MONOGRAPH

The FTC
Cigarette Test

O

Method for
Determining
Tar, Nicotine,
and Carbon
Monoxide Yields
of US. Cigarettes

Report of the NCI Expert Committee

NATIONAL INSTITUTES OF HEALTH

National Cancer Institute

1

TheFTC
Cigarette Test

O

Method tor
Determining
Tar, Nicotine,
and Carbon
Monoxide Yields
of US. Cigarettes

i

!

Report of the NCI Expert Committee
Mis.

j I I

bldg 10, 10 CENTER ; I
( BETHESDA, MD

U.S. DEPARTMENT OF HEAlTH
AND HUMAN SERVICES

Public Health Service
National Institutes of Health

dr

r

1

Foreword

In response to the emerging scientific evidence that cigarette smoking
posed a significant health risk to the user, in the early 1950's the major
cigarette manufacturers began widespread promotion of filtered cigarettes
to reassure smokers that, regardless of whatever unhealthy constituents
were in cigarette smoke, filters were a "scientific" breakthrough.

Advertisements for Viceroy's "health guard filter" stated, "DENTISTS
ADVISE — Smoke VICEROYS — ^The Nicotine and Tars Trapped by The Viceroy
Filter CAN NEVER STAIN YOUR TEETH!" and "Leading N.Y. Doctor Tells
His Patients What to Smoke — Filtered Cigarette Smoke Is Better For Health.
The Nicotine and Tars Trapped . . . Cannot Reach Mouth, Throat Or Lungs."
Chesterfield was "Best for you — low in nicotine, highest in quality," while
L&M's were "Just What the Doctor Ordered." Lorillard Tobacco Company
stressed its science-based Kent micronite filter (the original micronite filter
was made of asbestos) and claimed it removed seven times more tar and
nicotine than any other cigarette, which "put Kent in a class all by itself
where health protection is concerned." Of course, we know today that not
only were these claims patently false, but the cigarette companies knew it.

In the early 1950's the Federal Trade Commission (FTC) challenged a
variety of health claims made for cigarettes in their advertising, including
claims about tar and nicotine. In 1955 FTC published advertising guidelines
that, among other things, prohibited claims by cigarette manufacturers
that a particular brand of cigarettes was low in tar and nicotine or lower
than other brands, when it had not been established by competent scientific
proof that the claim was true and the difference was significant. Cigarette
manufactures, however, continued to advertise tar numbers. In the absence
of a standardized test methodology, this resulted in what is referred to as a
"tar derby" — a multitude of inconsistent, noncomparable claims that did not
give consumers a meaningful opportunity to assess the relative tar delivery
of competing brands. The tar derby ended in 1960 when discussions with
FTC culminated in an industry agreement to refrain from tar and nicotine
advertising.

In 1966, however, the U.S. Public Health Service (PHS) prepared a
technical report on "tar" and nicotine that concluded, "The preponderance
of scientific evidence strongly suggests that the lower the 'tar' and nicotine
content of cigarette smoke, the less harmful would be the effect." In reaching
this conclusion, the report noted the clear relationship between dose of
cigarette smoke received by the smoker and disease risk. Regardless of how
dose was calculated — by number of cigarettes smoked per day, age of
initiation, total number of years one smoked, or depth of inhalation,
mortality rates among smokers increased. When smokers quit smoking,
their risk was reduced in proportion to the length of time off cigarettes.

iii

Smoking and Tobacco Control Monograph No. 7

Subsequent to the PHS statement, FTC reversed its decision banning tar
and nicotine claims in advertising and established a standardized testing ^

protocol for assessing tar and nicotine yields. Today that protocol is widely i

known as the FTC test method. In 1980 the protocol was broadened to i

include measurement of the carbon monoxide yields of cigarettes as well. i

The initial protocol adopted by FTC was largely based on the work of |

U.S. Department of Agriculture chemist C.L. Ogg, as published in the Journal '■

of the Association of Official Agricultural Chemists in 1964. It appears, however, j
that this protocol was based on one person's observations about how people j

smoked. i

1

Much the same protocol had been proposed by American Tobacco |

Company researchers in 1936. Writing in the July issue of Industrial and ^

Engineering Chemistry, J.A. Bradford and colleagues noted, "The present i

writer's arbitrarily selected rate is a 35-cc puff of 2-second duration taken I

once a minute." i

However, cigarettes consumed at that time were vastly different from I

those manufactured and marketed later. In fact, tar and nicotine levels j

began to decline during the 1950's, concurrent with the mass marketing of
filter cigarettes. Market share of filter cigarettes increased from almost zero |
in 1950 (0.6 percent of the market) to 50 percent by decade's end. Total |

cigarette sales, which had begun to decline after the first public statements |

about the hazards of smoking in the early 1950's, rebounded to new highs. !

i

Although filter efficiency may have contributed to some of the reduction I
in tar/nicotine yields in the 1950's, the decline resulted mostly from less I

tobacco being used to make filtered as opposed to unfiltered cigarettes.

However, during the 1960's and 1970's major cigarette design changes
resulted in significantly lower machine-measured cigarette yields. The j

changes included increased use of ventilated tobacco rods and filters, I

use of more porous cigarette papers, and increased use of expanded and |

reconstituted tobacco. Concurrent with these modifications in cigarette i

design, cigarette manufacturers increasingly made use of additives in |

manufacturing. Today about 600 different compounds are routinely j

added to domestic cigarette brands, yet no routine testing is performed ,

to determine whether these compounds pose any additional health risk j

to the smoker when they are burned in a cigarette. [

U.S. market share of cigarettes yielding 15 mg tar or less went from I

3.6 percent in 1970 to 44.8 percent by 1980. The sales-weighted average tar
and nicotine yields of all U.S. cigarettes are now approximately 12 mg tar and i.
0.9 mg nicotine. By comparison, sales-weighted yields in the early 1950's |

were 35 mg tar and 2.5 mg nicotine. j

As consumption of low-yield cigarettes began to proliferate, the public
health community became concerned that these products were not what
they seemed, increasingly, scientific studies documented that smokers who
switched to these low-yield products smoked them differently, thus negating

IV

Foreword

the reason many of them changed in the first place — to lower their health
risk.

The U.S. Congress also voiced its concern in 1978 when it enacted the
Health Services and Centers Act. Section 403 of that legislation directed the
U.S. Department of Health and Human Services (DHHS) to conduct a "study
or studies of (1) the relative health risks associated with smoking cigarettes
of varying levels of tar, nicotine, and carbon monoxide; and (2) the health
risks associated with smoking cigarettes containing any substances commonly
added to commercially manufactured cigarettes." The Secretary of the
Department of Health and Human Services addressed this issue as part of
the 1981 Surgeon General's report. The Health Consequences of Smoking: The
Changing Cigarette. The overall conclusion of that report was clear: "There
is no safe cigarette and no safe level of consumption." Although the report
did note that smoking cigarettes with lower yields of tar and nicotine
reduces the risk of lung cancer to some extent, the benefits are minimal
in comparison with giving up cigarettes entirely. Evidence relating to heart
disease, other cancers, or chronic obstructive lung disease was not sufficient
to permit conclusions to be drawn. As to the accuracy of the FTC test
method, the report stated: "The 'tar' and nicotine yields obtained by present
testing methods do not correspond to the dosages that the individual smokers
receive: In some cases they may seriously underestimate these dosages."

Growing numbers of questions were raised about the accuracy of the
FTC test protocol to measure tar, nicotine, and carbon monoxide levels
from low-yield cigarettes — questions raised not just by the public health
community but also within the tobacco industry. Competitors complained
to FTC that Brown and Williamson's (B&W) Barclay brand cigarette did not
test accurately with the FTC test method. They argued that the brand was
designed with unique air ventilation channels that caused it to test low on
the FTC method. The ventilation channels, which remained open when
Barclays were smoked on the FTC machine, were rendered inoperable
when a human being smoked the cigarettes. In April 1983 FTC announced
that its testing method understated values for constituents in Barclay
cigarettes, and as a result, until new testing methods were developed, FTC
would no longer report an official rating for Barclay cigarettes. Later, FTC
took similar steps with respect to other B&W cigarette varieties that used
a filter design similar to Barclay's.

Eventually FTC closed its cigarette testing laboratory, in part because of
insufficient expertise within the agency to carry out an increasingly complex
and costly testing program. Since 1987, constituent levels for domestic
cigarette brands have been determined for the manufacturers by the Tobacco
Institute Testing Laboratory with oversight by FTC. The Tobacco Institute
serves as a trade organization as well as the information and lobbying arm
of the tobacco industry.

In June 1994 the Chairman of the House Subcommittee on Health and
the Environment wrote the Director of the National Cancer Institute (NCI),
asking him to convene a meeting of experts to ". . . review and make

V

Smoking and Tobacco Control Monograph No. 7

recommendations on the accuracy and appropriateness of the Federal Trade
Commission's method for determining the relative 'tar' and nicotine content
of cigarettes." A similar request was received from the FTC Chairman asking
that NCI convene a consensus conference on the topic and outlining several
areas it wished to be considered.

On December 5 and 6, 1994, a meeting of the NCI ad hoc expert
committee was convened under the aegis of the President's Cancer Panel to
examine this issue. The committee consisted of 1 1 individuals from diverse
scientific backgrounds and experience. The committee had the benefit of
excellent presentations from 14 experts whose professional careers were not
only involved in research on smoking, but who have been active contributors
to this field of scientific inquiry. Two of the individual participants were
cigarette industry scientists, who participated in all discussions.

From the outset of the committee's deliberations, it was clear that the
intent of the meeting was not to redesign the FTC testing protocol but, rather,
to examine the protocol and make suggestions for improvement, if warranted.
To provide a framework for discussion, the committee was asked to consider
three basic questions:

1. Does the evidence presented clearly demonstrate that
changes are needed in the current FTC protocol for
measuring tar, nicotine, and carbon monoxide? If yes,
what changes are required?

2. Should constituents other than tar, nicotine, and carbon
monoxide be added to the protocol?

3. Does the FTC protocol provide information useful to
smokers in making decisions about their health?

1. The committee reached the following conclusions with respect to the first

question.

A. The smoking of cigarettes with lower machine-measured yields has a
small effect in reducing the risk of cancer caused by smoking, no effect
on the risk of cardiovascular diseases, and an uncertain effect on the
risk of pulmonary disease. A reduction in machine-measured tar yield
from 15 mg tar to 1 mg tar does not reduce relative risk from 15 to 1.

B. The FTC test protocol was based on cursory observations of human
smoking behavior. Actual human smoking behavior is characterized
by wide variations in smoking patterns, which result in wide variations
in tar and nicotine exposure. Smokers who switch to lower tar and
nicotine cigarettes frequently change their smoking behavior, which
may negate potential health benefits.

C. Accordingly, the committee recommends the following changes to
the I'FC protocol:

VI

Foreword

1. This system should also measure and publish information on the
range of tar, nicotine, and carbon monoxide yields that most
smokers should expect from each cigarette sold in the United
States.

2. This information should be clearly communicated to smokers.

3. A simple graphic representation should be provided with each pack
of cigarettes sold in the United States and in all advertisements.

The representation should not imply a one-to-one relationship
between measurements and disease risk.

4. The system must be accompanied by public education to make
smokers aware that individual exposure depends on how the
cigarette is smoked and that the benefits of switching to lower
yield cigarettes are small compared with quitting.

D. There should be Federal oversight of cigarette testing, but such testing
should continue to be performed by the tobacco industry and at
industry expense.

E. The questions involved in the purpose, methodology, and utility
of the FTC protocol are complex medical and scientific issues that
require ongoing involvement of Federal health agencies, including
the National Institutes of Health, the Food and Drug Administration,
and the Centers for Disease Control and Prevention.

F. The system should be reexamined at least every 5 years to evaluate
whether the protocol is maintaining its utility to the smoker.

G. When a cigarette manufacturer makes significant changes in cigarette
design that affect yields, it should notify the appropriate Federal
agency.

II. With regard to the second question, the committee recommends that to
avoid confusing smokers, no smoke constituents other than tar, nicotine,
and carbon monoxide be measured and published at the present time.
Smokers should be informed of the presence of other hazardous smoke
constituents with each package and with all advertisements. These
constituents should be classified by toxic effects.

III. In considering the third question, the committee reached the following
conclusions:

A. Information from the testing system is useless to smokers unless
they have ready access to it. The information from the testing system
should be made available to all smokers, including those who smoke
generic brands and other brands not widely advertised.

B. Brand names and brand classifications such as "light" and "ultralight"
represent health claims and should be regulated and accompanied, in
fair balance, with an appropriate disclaimer.

vii

Smoking and Tobacco Control Monograph No. 7

C. The available data suggest that smokers misunderstand the FTC test
data. This underscores the need for an extensive public education
effort.

As Chairman of the President's Cancer Panel under whose aegis this
meeting was convened, 1 would like to express here my admiration and deep
appreciation to the members of the NCI ad hoc committee and its expert
consultants for a job well done. In transmitting this report to both the
U.S. Congress and the Federal Trade Commission, it is my sincere hope
that the recommendations contained herein will receive the serious and
thoughtful consideration they deserve.

Harold P. Freeman, M.D.

Chairman, President's Cancer Panel

A^cknowledgments

The FTC Cigarette Test Method for Determining Tar, Nicotine, and Carbon
Monoxide Yields ofU.S. Cigarettes: Report of the NCI Expert Committee was
developed under the general editorship of the Smoking and Tobacco Control
Program (STCP), National Cancer Institute (NCI), Donald R. Shopland,
Coordinator.

In organizing the December 5-6, 1994, meeting of the NCI Ad Hoc
Committee of the President's Cancer Panel on the FTC Test Method for
Determining Tar, Nicotine, and Carbon Monoxide Levels in Cigarettes, NCI
had the expert advice and assistance of many individuals both in and out of
Government service. In particular, the Coordinator and STCP staff members
would like to acknowledge the following individuals who served as part of an
informal planning group for the conference:

Judith Wilkenfeld, Esq.

Food and Drug Administration

Rockville, MD

Jack E. Henningfield, Ph.D.

Addiction Research Center

National Institute on Drug Abuse

Baltimore, MD

Michael P. Eriksen, Sc.D.

Centers for Disease Control and Prevention

Atlanta, GA

Shira D. Modell, Esq.

Federal Trade Commission

Washington, DC

Special recognition is due John M. Pinney and Joseph G. Gitchell,
Pinney Associates, Bethesda, MD, for their help with overall conference
organization and planning. Mr. Pinney also served as facilitator for all
consensus deliberations by the expert panel.

We would like to express our sincere appreciation to
the following memhers of the NCI Ad Hoc Committee of
the President's Cancer Panel.

Chairman,

President's Cancer Panel

Executive Secretary,
President's Cancer Panel

Harold P. Freeman, M.D.
Director of Surgery
Harlem Hospital Center
New York, W

Maureen O. Wilson, Ph.D.
National Cancer Institute
Bethesda, MD

IX

Smoking and Tobacco Control Monograph No. 7

Panelists Fred Bock, Ph.D.

Miami, FL

Dorothy K. Hatsukami, Ph.D.

Associate Professor of Psychiatry
University of Minnesota
Minneapolis, MN

Sandra Headen, Ph.D.

Social/Community Psychologist
University of North Carolina School of
Public Health
Chapel Hill, NC

Dietrich Hoffmann, Ph.D.

Associate Director and Chief
Division of Environmental Carcinogenesis
American Health Foundation
Valhalla, NY

John R. Hughes, M.D.

Professor

Departments of Psychiatry, Psychology,
and Family Practice
University of Vermont
Burlington, VT

Diana Petitti, M.D., M.P.H.

Director

Division of Research and Evaluation
Kaiser Permanente
Pasadena, CA

William S. Rickert, Ph.D.

President
Labstat, Inc.

Kitchener, Ontario
CANADA

Saul Shiftman, Ph.D.

Professor

Department of Psychology
University of Pittsburgh
Pittsburgh, PA

Maxine L. Stitzer, Ph.D.

Professor

Department of Psychiatry and Behavioral
Sciences

Johns Hopkins University School of Medicine
l-'rancis Scott Key Medical Center
Baltimore, MD

X

Acknowledgments

Invited Speakers

Ray Woosley, M.D., Ph.D.

Chair

Department of Clinical Pharmacology
Georgetown University
Washington, DC

Neal L. Benowitz, M.D.

Professor of Medicine and Chief
Division of Clinical Pharmacology and
Experimental Therapeutics
University of California, San Francisco
San Francisco, CA

Joel B. Cohen, Ph.D.

Distinguished Service Professor and Director
Center for Consumer Research
University of Florida
Gainesville, FL

Gary A. Giovino, Ph.D., M.S.

Chief

Epidemiology Branch

Centers for Disease Control and Prevention
Atlanta, GA

Michael R. Guerin, Ph.D.

Section Head, Organic Chemistry
Oak Ridge National Laboratory
Oak Ridge, TN

Jeffrey E. Harris, M.D., Ph.D
Massachusetts General Hospital
Associate Professor
Department of Economics
Massachusetts Institute of Technology
Cambridge, MA

Jack E. Henningfield, Ph.D.

Chief

Clinical Pharmacology Branch
Addiction Research Center
National Institute on Drug Abuse
Baltimore, MD

Dietrich Hoffmann, Ph.D.

Associate Director and Chief
Division of Environmental Carcinogenesis
American Health Foundation
Valhalla, NY

XI

Smoking and Tobacco Control Monograph No. 7

Lynn T. Kozlowski, Ph.D.

Professor and Head
Department of Biobehavioral Health
Pennsylvania State University
University Park, PA

C. Lee Peeler, Esq.

Associate Director
Division of Advertising Practices
Federal Trade Commission
Washington, DC

Harold C. Pillsbury, Jr.

Rockville, MD

Jonathan M. Samet, M.D., M.S.

Chairman

Department of Epidemiology
Johns Hopkins University

School of Hygiene and Public Health
Baltimore, MD

James P. Zacny, Ph.D.

Assistant Professor

Department of Anesthesia and Critical Care
University of Chicago
Chicago, IL

Tobacco Industry J. Donald deBethizy, Ph.D.

Representatives Vice President, Product Evaluation

R.J. Reynolds Tobacco Company
Bowman Gray Technical Center
Winston-Salem, NC

David E. Townsend, Ph.D.

Principal Scientist
R.J. Reynolds Tobacco Company
Bowman Gray Technical Center
Winston-Salem, NC

The Coordinator and STCP staff members also gratefully
acknowledge the authors who made this monograph possible.
Attributions for those chapters with authors follow.

Chapter 1. Cigarette Testing and the C. Lee Peeler, Esq.

Federal Trade Commission: Federal Trade Commission

A Historical Overview Washington, DC

A cknowledgmen ts

Chapter 2.

Chapter 3.

Chapter 4.

Review of the Federal Trade Harold C. Pillsbury, Jr.
Commission Method for Rockville, MD

Determining Cigarette Tar
and Nicotine Yield

Changes in Cigarette Design Dietrich Hoffmann, Ph.D.
and Composition Over Time American Health Foundation

and How They Influence the Valhalla, NY
Yields of Smoke Constituents

Mirjana V. Djordjevic, Ph.D.
American Health Foundation
Valhalla, NY

Klaus D. Brunnemann, M.S.
American Health Foundation
Valhalla, NY

Attitudes, Knowledge, and
Beliefs About Low-Yield
Cigarettes Among
Adolescents and Adults

Gary A. Giovino, Ph.D., M.S.
Centers for Disease Control and
Prevention
Atlanta, GA

Scott L. Tomar, D.M.D., Dr.P.H.
Centers for Disease Control and
Prevention
Atlanta, GA

Murli N. Reddy, M.S.

Centers for Disease Control and
Prevention
Atlanta, GA

John P. Peddicord, M.S.

Centers for Disease Control and
Prevention
Atlanta, GA

Bao-Ping Zhu, Ph.D., M.B.B.S., M.S.
Centers for Disease Control and
Prevention
Atlanta, GA

Luis G. Escobedo, M.D., M.P.H.
Centers for Disease Control and
Prevention
Atlanta, GA

Michael P. Eriksen, Sc.D.

Centers for Disease Control and
Prevention
Atlanta, GA

xiii

Smoking and Tobacco Control Monograph No. 7

Chapter 5.

Cigarette Smoke Components
and Disease: Cigarette
Smoke Is More Than a
Triad of Tar, Nicotine,
and Carhon Monoxide

Jeffrey E. Harris, M.D., Ph.D.
Massachusetts General Hospital
Massachusetts Institute of Technology
Cambridge, MA

Chapter 6.

The Changing Cigarette and
Disease Risk: Current Status
of the Evidence

Jonathan M. Samet, M.D., M.S.

Johns Hopkins University

School of Hygiene and Public Health
Baltimore, MD

Chapter 7.

Biomarkers of Cigarette
Smoking

Neal L. Benowitz, M.D.

University of California, San Francisco
San Francisco, CA

Chapter 8.

Pharmacology and Markers:
Nicotine Pharmacology and
Addictive Effects

Jack E. Henningfield, Ph.D.
Addiction Research Center
National Institute on Drug Abuse
Baltimore, MD

Leslie M. Schuh, Ph.D.

Wayne State University
Detroit, MI

Chapter 9.

Consumer/Smoker
Perceptions of Federal Trade
Commission Tar Ratings

Joel B. Cohen, Ph.D.
University of Florida
Gainesville, EL

Chapter 10.

Sensitivity of the Federal
Trade Commission Test
Method to Analytical
Parameters

Michael R. Guerin, Ph.D.

Oak Ridge National Laboratory
Oak Ridge, TN

Chapter 11.

Human Smoking Patterns

James P. Zacny, Ph.D.

University of Chicago
Chicago, IL

Maxine L. Stitzer, Ph.D.

Johns Hopkins University School
of Medicine

Francis Scott Key Medical Center
Baltimore, MD

Chapter 12.

Compensation for Nicotine
by Smokers of Lower Yield
Cigarettes

Lynn T. Kozlowski, Ph.D.
Pennsylvania State University
University Park, PA

Janine L. Pillitteri
Pennsylvania State University
University Park, PA

Acknowledgments

Chapter 13.

Cigarette Design
Technologies Reduce Smoke
Yield and Expand Consumer
Choices: The Role and Utility
of the FTC Test Method

David E. Townsend, Ph.D.

RJ. Reynolds Tobacco Company
Bowman Gray Technical Center
Winston-Salem, NC

SECTION IV. Overview of 1980 to 1994
Research Related to the
Standard Federal Trade
Commission Test Method
for Cigarettes

Michael D. Mueller, M.S.P.H.
R.O.W. Sciences, Inc.
Rockville, MD

Finally, the Coordinator and STCP staff members would like to
acknowledge the significant contributions of the following staff
members of R.O.W. Sciences, Inc., Rockville, MD, and MasiMax
Resources, Inc., Rockville, MD, who provided technical and
editorial assistance in the preparation of this monograph. In
particular, we would like to acknowledge the contribution of
Richard H. Amacher, M.S., who served as Project Manager from
September 1989 through August 1995 for the contract under which
this publication was produced. We would also like to thank
Marilyn M. Massey, M.P.H., MasiMax Resources, and Jacqueline M.
Tressler, R.O.W. Sciences, who currently serve as Project Manager
and Subcontract Manager, respectively, for the contract under
which this publication was produced, for their valuable
contribution during the Dual phases of the mongraph's
development. Special recognition is also due to Janies R. Libbey,
M.P.I.A., who served as Managing Editor for this publication, and
Traci Cherrier, who served as Senior Conference Specialist for the
NCI ad hoc committee meeting that resulted in this monograph.

R.O.W. Sciences, Inc.

Douglas Bishop, Art Director

Catherine W. Chapman, Executive Assistant
Rebecca A. Charton, Senior Librarian

Ruth E. Clark, Word Processing Specialist
Daria T. Donaldson, Proofreader

Catherine Hageman, Word Processing Supervisor
Stephen P. Luckabaugh, Information Specialist
Frances Nebesky, Senior Copyeditor
Sheila Proudman, Technical Writer/Resource Coordinator

Esther M. Roberts, Administrative Assistant
Donna Selig, Copyeditor/Proofreader

XV

Smoking and Tobacco Control Monograph No. 7

Keith W. Stanger, Graphics Services Coordinator

Donna Cay Tharpe, Quality Control Proofreader

Sonia Van Putten, Word Processing Specialist

MasiMax Resources, Inc.

Rebecca H. Razavi, Copyeditor/Proofreader

Aida M. Teymouri, Administrative Secretary
ABOUT THE MONOGRAPH

This volume is the seventh in the series of Smoking and Tobacco Control
monographs published by the National Cancer Institute since 1991. The
monographs were specifically established by NCI to provide an authoritative
source of information about issues important to those individuals and
institutions involved in smoking and tobacco use control.

This report was compiled in response to a request to the National Cancer
Institute by the then Chairman of the Subcommittee on Health and the
Environment, U.S. House of Representatives, asking that a scientific panel of
experts be convened to review and make recommendations on the accuracy
and appropriateness of the Federal Trade Commission's test method for
assessing constituent yields for cigarettes on the U.S. market. The NCI
received a similar but more detailed letter from the Chairman of the Federal
Trade Commission in which the Commission outlined several areas for the
NCI ad hoc committee to consider (see page xix).

The Coordinator of NCI's Smoking and Tobacco Control Program, who
was given overall responsibility for the project, established a small informal
advisory group consisting of individuals from the FTC and various PHS
agencies to help organize the conference, suggest committee members, and
plan the agenda.

The NCI Ad Hoc Committee of the President's Cancer Panel on the
FTC Test Method for Determining Tar, Nicotine, and Carbon Monoxide
Levels in Cigarettes was convened December 5-6, 1994, in Bethesda, MD.
Harold P. Freeman, M.D., Chairman of the President's Cancer Panel, also
chaired these proceedings.

However, prior to the December conference the 1 1 members of the NCI
ad hoc committee (these individuals are identified in the "Acknowledgments"
to the monograph) were provided several resource materials in support of
their deliberations. These resources included copies of the 1981 Surgeon
General's report The Health Consequences of Smoking: The Changing Cigarette.

A Report of the Surgeon General, a detailed bibliography of the relevant
worldwide scientific literature, and a copy of an NCI-commissioned White
Paper titled "Overview of 1980 to 1994 Research Related to the Standard
Federal 'Trade Commission Test Method for Cigarettes." The White Paper,
which is published as Section IV of this monograph, represents a noncritical

Acknowledgments

summary of those research findings published since the 1981 Surgeon
General's report. Full copies of all articles were made available on demand
to members of the NCI ad hoc committee by NCI's information science
contractor, R.O.W. Sciences, Inc., of Rockville, MD.

The December 5-6, 1994, conference was organized similar to a consensus
conference. Prior to the formal opening of the conference, the committee
was asked to consider the three questions laid out on page vi of the
"Foreword."

On the first day, subject matter experts were invited to make formal,
structured presentations before the NCI ad hoc committee. (See
"Acknowledgments" for list of speakers.) The 13 individual chapters
published in Section 1 of this monograph are based on these presentations.
Each presentation was approximately 30 minutes in length, followed by a
question-and-answer session. Both members of the NCI ad hoc committee
and invited speakers fully participated in these discussions. During the
second day of deliberations, committee members and invited speakers
participated in a more open-ended discussion, with the goal of reaching
consensus on the three questions.

Open discussions ended midday December 6. Members of the NCI ad
hoc committee then met to finalize their recommendations and findings;
these were presented to the public during a press conference midafternoon
December 6. The FTC Cigarette Test Method for Determining Tar, Nicotine, and
Carbon Monoxide Yields of US. Cigarettes: Report of the NCI Ad Hoc Committee
is the culmination of that effort.

Individuals wishing to receive a copy of the audiotapes of the December
meeting may order these directly from Caset Associates at (703) 352-0091.

The cost per set is $75. Those individuals interested in receiving
a copy of the written transcript should contact Mr. Donald R. Shopland,
National Cancer Institute, Executive Plaza North, Room 241, 6130 Executive
Boulevard, Bethesda, MD 20892-7337.

xvii

Acknowledgments

ONE HUMORED ThiAO CONGRESS

HENffY A. WAXk^iN. CAllfOANlA. CHAIRMAN

MIKE STHAA. OKLAHOMA
RON WVOEN. OREGON

THOMAS J. eULEY. JR. VlMClNU
MICHAEL BILIRAKIS. FLORIDA
AA£ji McMillan. NOimi Carolina
J OCNNI6 HASTERT. tUJNOlS
FRED URTON. MICHIGAN
SILL RAXON NEW YORK
SCOTT KLUG. WISCONSIN
GANT A FRANKS. CONNECTICLTY
JAMES C GREENWOOD. RCNNSYLVANIA
CARLOS J. MOORHEAD CAUFORNIA
(EX OFfICIO}

BILL RICHARDSON. NEW MEXICO
JOHN BRrANT. TEXAS
J. ROT ROWtANO. GEORGIA
EDOLRHUS TOWNS. NEW YORK
GERRY L STUODS. MASSACHUSETTS
JIM 8LATTERV KANSAS
JIM COORCR TENNESSEE
FRANK RALLONC. JR.. NEW JERSEY
CRAIG A WASHINGTON. TEXAS
SHERROO BROWN. OHIO
MIKE KREIOLER. WASHINGTON
JOHN D DINOCLL MICMIGAM
(EX OfMClOl

RALRH M HALL. TEXAS

U.S. HOUSE OF REPRESENTATIVES
COMMITTEE ON ENERGY AND COMMERCE

SUBCOMMITTEE ON HEALTH AND THE ENVIRONMENT

2415 RAYBURN HOUSE OFFICE BUILDING
WASHINGTON. DC 20615-6118

KAREN NELSON. STAFF DIRECTOR

PHONE (202) 226-4962

June 7, 1994

Dr. Samuel Broder
Director

National Cancer Institute
National Institutes of Health
Building 31
Room 11A48
9000 Rockville Pike
Bethesda, Maryland 20892

Dear Dr. Broder:

I am writing to request that the National Cancer Institute sponsor a scientific
conference which would review and make recommendations on the accuracy and
appropriateness of the Federal Trade Commission’s method for determining the relative
"tar" and nicotine content of cigarettes. As you know, there is growing concern over the
current testing method because many public health and addiction experts believe it may
mislead smokers about the relative safety of a low tar, low nicotine product.

It has been suggested that a major reason for reliance upon the FTC test
procedure is to allow consumers the option of reducing their risk of disease by smoking a
brand deemed low in "tar" and nicotine. Consumer preference for low tar and nicotine
rated cigarettes accelerated during the 1970’s when NCI supported research strongly
suggested that such cigarettes offered the consumer a reduced risk of lung cancer. The
shift in consumer demand to these newer low yield cigarettes was quite rapid. In 1972
less than 2 percent of all cigarettes sold in the U.S. had a tar yield of less than 15 mg.
However, the major cigarette manufacturers were quick to use the FTC tar and nicotine
numtiers in their advertising and by the end of the decade 40 percent of all cigarettes
sold were under 15 mg. During the 1980’s considerable doubt was expressed by many
public health officials as to whether the tar and nicotine yields of cigarettes based on a
protocol developed in the 1950’s accurately reflect actual exposure and health risk levels
when smoking today’s cigarettes. Today approximately 60 percent of all brands are
considered low-tar.

XIX

Smoking and Tobacco Control Monograph No. 7

The NCI can provide an invaluable public service in sponsoring a scientific forum
to address these issues and formulate alternative recommendations. It would be
particularly helpful if a conference on this matter, perhaps in collaboration with the
National Institute on Drug Abuse and the Federal Trade Commission, could be convened
by October 1994.

Your consideration of this request is greatly appreciated. Please do not hesitate
to contact me or Ripley Forbes of the Subcommittee staff if we can answer any questions
or provide assistance in developing a conference agenda. I look forward to hearing firom
you.

With every good wish, I am

Sincerely,

HENRY A. WAXMAN
Chairman, Subcommittee on
Health and the Enviroiunent

XX

Contents

Page

Foreword iii

Acknowledgments ix

SECTION I. Invited Papers

Chapter 1. Cigarette Testing and the Federal Trade Commission:

A Historical Overview 1

Question-and-Answer Session 8

References 8

Chapter 2. Review of the Federal Trade Commission Method for

Determining Cigarette Tar and Nicotine Yield 9

Question-and-Answer Session 12

References 14

Chapter 3. Changes in Cigarette Design and Composition
Over Time and How They Influence the Yields

of Smoke Constituents 15

Introduction 15

Background 15

Changes in Cigarette Design and Composition 18

Summary 31

Question-and-Answer Session 33

References 35

Acknowledgments 37

Chapter 4. Attitudes, Knowledge, and Beliefs About Low-Yield

Cigarettes Among Adolescents and Adults 39

Introduction 39

Monitoring National Data 42

Survey Findings 43

Health Beliefs and Switching 47

Quitting Smoking 49

Discussion 51

XXI

Smoking and Tobacco Control Monograph No. 7

Question-and-Answer Session 53

References 56

Chapter 5. Cigarette Smoke Components and Disease:

Cigarette Smoke Is More Than a Triad of Tar,

Nicotine, and Carbon Monoxide 59

Introduction 59

Mainstream Vs. Sidestream Cigarette Smoke 59

Major Health Effects of Cigarette Smoke 60

Cancer 61

Noncancerous Lung Diseases 62

Atherosclerotic Cardiovascular Diseases 63

Cigarette Smoking and Human Reproduction 65

Absolute Risk Vs. Relative Risk 66

Tar, Nicotine, Carbon Monoxide, and Other

Smoke Constituents 66

Smoke Constituents, Cigarette-Related Disease,

and Modified Labeling of Cigarettes 67

Question-and-Answer Session 70

References 72

Acknowledgment 75

Chapter 6. The Changing Cigarette and Disease Risk:

Current Status of the Evidence 77

Introduction 77

Epidemiologic Evidence on the Changing Cigarette

and Disease Risks 79

Conclusions 86

Question-and-Answer Session 87

References 91

Chapter 7. Biomarkers of Cigarette Smoking 93

Introduction 93

Use of Various Biomarkers 93

Nicotine Absorption From Cigarettes 93

XXII

Contents

Nicotine Content of Tobacco Vs. FTC Yield 95

Quantitating Nicotine Intake in Smokers 95

Cotinine Levels and Nicotine Intake 98

Ultralow-Yield Cigarettes 101

Nicotine Intake and Machine-Determined Yield 102

Carbon Monoxide and FTC Yield 104

Tar-to-Nicotine Ratio 104

Conclusions 106

Question-and-Answer Session 108

References 110

Acknowledgments Ill

Chapter 8. Pharmacology and Markers: Nicotine

Pharmacology and Addictive Effects 113

Introduction 113

Cigarette Smoking as Drug Dependence 113

Nicotine Delivery Systems 115

Nicotine's Effects 116

Smoking and Nicotine Dose 119

A Proposal for More Meaningful Cigarette Labeling 122

Question-and-Answer Session 122

References 124

Chapter 9. Consumer/Smoker Perceptions of Federal

Trade Commission Tar Ratings 127

Introduction 127

Tar Level of Cigarettes 127

Knowledge of Advertised Tar Numbers 128

Smokers' Interpretations of Tar Numbers 130

Smokers' Use of Advertised Tar Numbers 132

Conclusions 132

Question-and-Answer Session 132

Reference

134

Smoking and Tobacco Control Monograph No. 7

Chapter 10. Sensitivity of the Federal Trade Commission

Test Method to Analytical Parameters 135

Introduction 135

Standard Machine Smoking 135

Influence of Smoking Parameters 137

Influence of Human Smoking Practices 141

Conclusions ' 149

Question-and- Answer Session 149

References 149

Acknowledgments 150

Chapter 11. Human Smoking Patterns 151

Introduction 151

How Do Humans Smoke? 151

Which Human Smoking Behaviors Determine

Smoke Exposure? 152

Are Human Smoking Patterns Dynamic or Static? 153

Does the FTC Method Accurately Reflect Human

Smoking Patterns? 154

Summary 156

Question-and-Answer Session 157

References 159

Chapter 12. Compensation for Nicotine by Smokers of

Lower Yield Cigarettes 161

Background 161

Cigarette Brand Switching in Experimental Research 161

Research 162

One Repeated-Measures Study of Self-Selected

Brand Switching 165

Smokers Can Get High Yields From the Lowest of the

Ix)w-Yield Cigarettes: More on the Issue of Vent Blocking 165

Graphic Information on Far and Nicotine Yields:

1 he Color-Matching 3'echnique 166

XXIV

Contents

Summary 168

Question-and-Answer Session 168

References 171

Chapter 13. Cigarette Design Technologies Reduce Smoke
Yield and Expand Consumer Choices: The Role
and Utility of the FTC Test Method 173

Background 173

Cigarette Design and Changes in the Cigarette Market 174

Utility of the FTC Test Method 176

Question-and-Answer Session 180

References 191

SECTION II. Transcript of Second-Day Discussion 193

SECTION III. Recommendations and Findings 239

SECTION FV. Overview of 1980 to 1994 Research Related

to the Standard Federal Trade Commission Test

Method for Cigarettes 249

Introduction 249

Parameters of the FTC Test Method and Current

Smoking Patterns 250

Impact of Changing Parameters of the FTC Test Method

on Absolute Yields of a Cigarette Brand and Relative Yields

of Different Brands 252

Tar and Nicotine Yield by the FTC Test Method and

Amounts Delivered to Smoker 253

Relative Yields of Different Brands by the FTC Test

Method and Amount of Nicotine Absorbed by Smokers 257

Yield by the FTC Test Method and Absorption of Nicotine

in Switchers 259

Yields by the FTC Test Method and Other Constituents

Using FTC Puff Profile 270

Proposals To Change the FTC Test Method 273

References 273

XXV

Chapter 1

Cigarette Xesting and the Federal Xrade
Commission: A Historical Overview^

C. Lee Peeler

Cigarette manufacturers began advertising their products' tar and
nicotine content before there was a standardized procedure for testing
cigarette output. In 1955, after a series of cases challenging a variety of
claims made for cigarettes (including tar and nicotine claims),^ the Federal
Trade Commission (Commission or FTC) published cigarette advertising
guides. Among other things, the guides prohibited claims that a particular
brand of cigarettes was low in tar and nicotine or lower than other brands
"when it has not been established by competent scientific proof . . . that the
claim is true, and if true, that such difference or differences are significant"
(Federal Trade Commission, 1988a).

However, cigarette manufacturers continued to advertise tar numbers.

In the absence of a standardized testing methodology, their claims resulted
in what is often referred to as the "tar derby" — a multitude of inconsistent,
noncomparable claims that did not give consumers a meaningful opportunity
to assess the relative tar delivery of competing brands. The tar derby ended
in 1960, when discussions with the Commission culminated in an agreement
by the industry to refrain from tar and nicotine advertising (Federal Trade
Commission, 1988b).

In 1964 the first Surgeon General's report on the health risks of smoking
concluded that cigarette smoking was a cause of lung cancer in men (U.S.
Department of Health and Human Services, 1964). In 1966 the Public
Health Service stated that "The preponderance of scientific evidence strongly
suggests that the lower the tar and nicotine content of cigarette smoke, the
less harmful would be the effect" (U.S. Department of Health and Human
Services, 1981, p. v).

It was in this environment that the Commission initiated two major
steps in 1966 to encourage cigarette manufacturers to provide consumers
with comparative information about their products' tar and nicotine yields.

' These remarks are the views of the staff of the Bureau of Consumer Protection. They do not necessarily
represent the view of the Commission or any individual commissioner.

^See, e.g., R.J. Reynolds Tobacco Co. v. FTC, 192 F.2d 535 7th Cir. (1951) (claims that Camel does not impair the
physical condition of athletes and aids digestion); American Tobacco Co., 47 F.T.C. 1393 (1951) (Lucky Strike
cigarettes advertised as less irritating to the throat than competing brands and containing less tar than four
other leading brands); P. Lorillard Co., 46 F.T.C. 735 (1950) (Old Gold cigarettes advertised as lowest of seven
leading brands in nicotine and throat irritating tars, and Beech-Nut cigarettes as providing "definite defense
against throat irritation"). See also, e.g., Leighton Tobacco Co., 46 F.T.C. 1230 (1950) (Phantom cigarettes
represented as causing no irritation of any kind).

1

Smoking and Tobacco Control Monograph No. 7

First, it ended the ban on tar and nicotine advertising by announcing that
factual statements of the tar and nicotine content of mainstream cigarette
smoke could be made if they were supported by tests conducted in accordance
with the so-called "Cambridge Filter method" and if they were not
accompanied by claims about reduced health hazards (Federal Trade
Commission, 1988a). Second, it authorized establishment of a laboratory to
analyze cigarette smoke and invited public comment on what modifications,
if any, should be made to the Cambridge Filter method for purposes of the
laboratory's procedures and how the test results should be expressed {Federal
Register, 1966). The modified Cambridge Filter method ultimately adopted
by the Commission is often referred to as the "FTC method."

By mid-1967 the laboratory was ready to begin testing cigarettes
{Federal Register, 1967).^ The Commission agreed, pursuant to Senator
Warren Magnuson's request,'* to report the test results to Congress
periodically, a process that continues today.

From the outset, the testing was intended to obtain uniform,
standardized data about the tar and nicotine yield of mainstream cigarette
smoke, not to replicate actual human smoking. The Commission recognized
that individual smoking behavior was just that — too individual to gauge
what a hypothetical "average" smoker would get from any particular cigarette:
"No two human smokers smoke in the same way. No individual smoker
always smokes in the same fashion" (Federal Trade Commission, 1967). The
purpose of the testing was "not to determine the amount of 'tar' and nicotine
inhaled by any human smoker, but rather to determine the amount of tar
and nicotine generated when a cigarette is smoked by machine in accordance
with the prescribed method" (Federal Trade Commission, 1967). Indeed, the
Cambridge Filter method did not attempt to duplicate an "average" smoker
but was "an amalgam of many choices" (Federal Trade Commission, 1967).
Because no test could accurately duplicate human smoking, the Commission
believed that the most important thing was to make certain the results
presented to the public were based on a reasonable, standardized method
and could be presented to consumers in an understandable manner.

The Commission next attempted to increase consumer awareness of the
ratings produced by its laboratory. In 1970 it proposed a trade regulation rule
that would have required disclosure of tar and nicotine ratings in all cigarette

’ For the first dozen years of its existence, the laboratory tested only for tar and nicotine. In 1980 the protocol
was modified to add testing for carbon monoxide.

^ Expressing the opinion held at that time by many people in the Federal Government, Senator Magnuson
stated that "By encouraging smokers to switch to low tar/nicotine cigarettes, we can contribute meaningfully
to the physical health of our nation. Publication of the Commission's testing results is one important facet

I he Commission expressed its views concerning dissemination of tar and nicotine figures in an
Odober 1967 letter to the National Association of Broadcasters; "The Commission favors giving smokers
as much information about the risks involved in smoking as is possible and to that end favors mandatory
disclosure of tar and nicotine content, as measured by a standard test."

2

Chapter 1

advertising {Federal Register, 1970). The rulemaking was suspended
indefinitely a short time later, when five of the major cigarette manufacturers
and three small companies agreed voluntarily among themselves to include
the ratings produced by the Commission's protocol in their advertisements.
That agreement, modified to reflect the discontinuance of the Commission's
laboratory, remains in effect today.^

There are a number of ways to lower a cigarette's tar and nicotine rating,
including adding filters that literally trap some of the constituents of the
tobacco smoke before they reach the machine, wrapping the tobacco plug in
paper that burns relatively quickly, and placing ventilation holes around the
circumference of the filter so that when a smoker or smoking machine puffs
on the cigarette, air is drawn into the filter and the resulting diluted mixture
of air and smoke yields lower tar and nicotine ratings than an undiluted puff
of smoke would yield. The last technique is often referred to as "aeration."

These types of changes in cigarette technology have focused the
Commission's attention on its protocol on two separate occasions since 1970.
In both cases, the Commission solicited public comments on certain aspects
of the FTC method. However, in neither instance did the information
received by the Commission form a sufficient basis for changing the protocol,
even though the limitations on the predictiveness of the FTC method caused
by compensatory smoking were clearly recognized by the mid-1980's.
("Compensatory behavior" is the tendency of consumers to offset the benefits
of a positive change in their behavior by making a second, negative change.
For example, a smoker who switches to a brand with lower tar and nicotine
ratings might smoke more cigarettes each day or smoke each one more
intensively, that is, inhale more deeply and/or take more puffs per cigarette.)
Following is a review of the two events referred to above.

Aeration first became an issue for the Commission in 1977, when
Lorillard, Inc., suggested that the depth to which cigarettes were inserted in
the Commission's smoking machine be decreased when the standard depth
would block some of a cigarette's ventilation holes, thereby impairing its
filtration system and resulting in higher ratings than if the holes were open.
The Commission solicited public comments on this question and also on
whether the insertion depth should be decreased beyond the point where
consumers cover the cigarette with their fingers or lips {Federal Register, 1977).

Of the seven cigarette companies that commented, only Lorillard
supported varying the standard insertion depth. However, none of the
responders addressed the question of whether the new insertion depth
would be more consistent with actual smoking practices. After reviewing

^ The American Tobacco Company did not sign the voluntary agreement, but similar disclosures have been
contained in its advertisements, pursuant to a 1971 consent agreement with the Commission. [In re American
Brands, Inc., 79 F.T.C. 255 (1971).]

3

Smoking and Tobacco Control Monograph No. 7

the comments, the Commission noted that the development of cigarettes
with ventilation holes near the tip had complicated the comparability of its
tar and nicotine ratings,^ but "that a change in the insertion depth would
cause a lack of continuity with previous test results" {Federal Register, 1978,
pp. 11856, 11857). The Commission decided not to modify the protocol "in
the absence of information indicating that a new insertion depth would be
more consistent with the manner in which smokers insert cigarettes in actual
use" {Federal Register, 1978, p. 11857).

Another controversy concerning the test method arose in the early 1980's
and involved the Brown & Williamson Tobacco Corporation's (B&W) Barclay
cigarette, which was designed with a channel ventilation system rather than
air holes.^ Competitors claimed that Barclay, which had received an official
FTC rating of 1 mg tar in 1981, did not test accurately on the FTC smoking
machine because the channels remained open during testing but were
rendered inoperable in practice. After careful consideration, the Commission
determined that its present test method did not accurately measure Barclay's
tar, nicotine, and carbon monoxide. It revoked the 1-mg rating, estimating
that Barclay should be rated between 3 and 7 mg of tar (based on testing by
independent consultants) and invited comments on a number of issues
relating to possible modification of its testing method, including using new
cigarette holders on the smoking machine that would simulate the reduction
in ventilation that occurred when people smoked Barclay {Federal Register,
1983). The Commission asked which modifications would yield the most
appropriate results for all cigarettes and whether modification of the cigarette
testing method would result in unintended consequences and affect possible
innovation in cigarettes design {Federal Register, 1983).

The Commission also took this opportunity to reiterate that its ratings
were relative; that the amount of tar, nicotine, and carbon monoxide any
particular cigarette delivered depended on how it was smoked; and that
in the case of ventilated filter cigarettes, delivery would be increased if
ventilation holes were blocked {Federal Register, 1983). It then invited

* Quoting its 1967 statement that the purpose of testing was not to determine the amount of constituents
inhaled by a human smoker but to determine the amount generated when a cigarette was smoked by a
machine in accordance with a prescribed protocol (see above), the Commission noted that:

The point of this statement was that the 1-TC's "tar" and nicotine values represented valid
standards for making comparisons among different cigarettes. Thus, if the consumer smoked
each different cigarette the same way, he would inhale "tar" and nicotine in amounts propor-
tional to the relative values of the FI'C figures. A person who smoked a 10 mg "tar" cigarette
would ingest half the "tar" he would by smoking a 20 mg "tar" cigarette providing he smoked
the same way. The development of cigarettes with ventilation areas within 1 1 mm of the tip
has complicated this simple relationship. (Federal Register, 1978, p. 11856)

' In conventional aerated cigarettes, air and smiAe mixed together as they passed through the filter. Outside air
drawn into liarclay's channels, however, went directly into the smoker's mouth before first mixing with any
smoke; dilution was supposed to occur in the mouth, not in the filter. Competitors alleged that because the
exit holes for the channels were close to the smoker's lips, they were crushed or covered by lips, thus reducing
dilution.

4

Chapter 1

comments on a wide range of issues concerning compensatory smoking
behavior:

Should the Commission further examine the implications for its
testing program of the issues raised by compensatory smoking
behavior, including hole blocking, when consumers smoke
lower "tar" cigarettes? What is the evidence that smokers use
higher "tar" cigarettes differently than lower "tar" cigarettes?

What is the evidence regarding the extent of hole blocking
by smokers of different ventilated filter cigarettes? Are there
problems regarding compensatory smoking behavior which are
significant enough to warrant further exploration of changes
in the method, beyond those necessitated by the Commission's
findings concerning Barclay? What lines of inquiry would
generate the most useful information if such an examination
is undertaken? For example, should the Commission explore
a system of categories or "bands" of "tar" content rather than
specific numerical estimates? Also, should consumers be
advised that the cigarettes' actual "tar" delivery depends on
how it is smoked? {Federal Register, 1983)

Shortly after the initial comment period closed,® a Federal district
court issued an opinion in the Commission's action against B&W over
advertisements that continued to describe Barclay as a 1-mg tar cigarette,
despite the Commission's revocation of Barclay's 1-mg rating [FTC v.

Brown & Williamson Tobacco Corp., 580 F. Supp. 981 (D.D.C. 1983), affd in
part, remanded in part, 778 F.2d 35 (D.C. Cir. 1985)]. During that litigation,
B&W contended that "recent scientific evidence demonstrates that the
FTC system is so flawed that it is itself deceptive" [580 F. Supp. at 984].^

The court recognized that compensatory smoking behavior complicated
the ratings question but rejected B&W's contention that the system provided
no benefit to consumers:

The FTC system attempts only to determine how much relative
tar and nicotine a smoker would get in his mouth were he to
smoke two cigarettes in the same manner. B&W has utterly
failed to show that the system does not do this. Nor has it
shown that a better method for determining the relative health

* Comments responsive to the April 13, 1983, Federal Register notice were originally due by June 30, 1983.

On June 4, 1984, however, the Commission reopened the comment period because certain information
that was relevant to the questions addressed in that notice, but had been previously under a court-ordered
seal, was now publicly available (Federal Register, 1984).

’ B&W argued that all cigarettes were subject to compensatory smoking behavior and thus all tar numbers
were "soft." The Commission acknowledged that low-yield cigarettes were subject to substantial variations
in actual smoker intake but contended that Barclay tested differently on the machine from other cigarettes.
The Commission's position was that the tar ratings provided a rough comparative scale; that is, a 1-mg
cigarette should be comparable to all other 1-mg cigarettes, if all are smoked in an identical manner.

5

Smoking and Tobacco Control Monograph No. 7

hazards of the many different varieties of cigarettes on the
market is currently feasible [580 F. Supp. at 985].

The comments ultimately submitted in response to the Commission's
questions about compensatory smoking reflected sharply disparate views.

On the one hand, the American Heart Association (AHA), American Lung
Association (ALA), and American Cancer Society (ACS) identified problems
with the existing methodology, expressed concern over the impact of
compensatory smoking behavior, and suggested extensive research to
improve the current testing and reporting procedures.^®

On the other hand, Philip Morris, R.J. Reynolds Tobacco Company,
and American Brands asserted that compensatory smoking behavior was
not relevant to the testing methodology and that devising a protocol that
accounted for compensatory smoking would require establishing a profile of
the average smoker, something the Commission had previously declined to
do because of the impossibility of accounting for all the relevant variables.
Lorillard stated that data on compensatory smoking were very limited and
therefore recommended that the existing system be kept intact. Liggett &
Myers suggested that perhaps all cigarette testing should be abolished because
smoking behavior could seriously affect tar and nicotine yields and smokers
could not be taught to change their behavior.

In response to the Commission's question about possible implementation
of a "banding" system for its tar and nicotine ratings, B&W (which had
just had Barclay's rating revoked) argued that the current system caused
manufacturers to emphasize small differences that might not exist, given
the realities of compensatory smoking, and that it should be replaced with
a system that would group products into high-tar, medium-tar, low-tar, and
ultralow-tar "bands." Philip Morris and American Brands argued that banding
would lead to a concentration of brands at the upper limit of each category
(in contrast to the existing system, which encouraged reductions across the
board). American Brands also contended that banding would confuse
consumers, whereas Philip Morris noted that it would substitute the
Government's judgment about the significance of differences in tar ratings
for that of the individual consumer.

the ALA stated that given the reality of compensatory smoking, lov^-tar cigarettes might not be as safe as
some consumers were being led to believe and that the Commission's testing and reporting procedures were
contributing to questionable advertisements for "safe" cigarettes. The ACS stated that the Commission's test
method should be modified to reflect current understanding of compensatory smoking behavior. The AHA
expressed its view that the Commission's testing and reporting procedures fostered the belief among
consumers that low-tar cigarettes were safer than high-tar brands. However, epidemiological evidence
showing a correlation between the risk of coronary heart disease and the number of cigarettes smoked
per day, but not a reduced rate of such disease among low-tar smokers, suggested that smokers of those
cigarettes might be engaging in compensatory smoking.

I he ALA and ACS recommended that research be conducted to determine how actual intake of tar and
other smoke constituents by smokers related to the Fl'C's ratings; following completion of this research,
the Commission should test each cigarette under a range of conditions replicating actual smoking behavior
and report those results with a warning that individual yield depends on individual smoking patterns.

6

Chapter 1

In short, there was no clear consensus as to specific action the
Commission could (or should) take to eliminate the limitations of the
test method. At the same time, abandoning the testing system without
instituting another method of tar testing would have been premature
because then-current epidemiological evidence suggested that there had
been a reduction in lung cancer deaths that might be attributable to declines
in average tar levels that had occurred since the 1950's (U.S. Department
of Health and Human Services, 1981).^^ Accordingly, at that time the
Commission made no changes to its cigarette test method to address
compensatory smoking.

In early 1987 the Commission decided to close its cigarette testing
laboratory. The Commission found that closing the laboratory was necessary
for several reasons, chiefly because the cost of the laboratory was significant
and the Commission would have had to commit significant additional funds
to continue its operation. The Commission also was persuaded that the
same information could be obtained from other sources and that other
means were available to verify the accuracy of industry testing results. In fact,
the Commission's operation of a testing system for the industry at taxpayer
expense was highly unusual. The common scenario is for the industry to
conduct its own testing under Government-specified testing protocols.

Since 1987 the Tobacco Institute Testing Laboratory (TITL) has continued
to test most cigarettes, using the Commission's approved methodology; the
companies report the results to the Commission pursuant to a compulsory
request, and the Commission publishes the results. TITL keeps the
Commission informed of proposed changes in the testing procedure and
solicits Commission approval for all significant changes. TITL's work is
regularly monitored by the Commission's contractor, Harold Pillsbury, Jr.

(this volume), who has virtually unrestricted access to the laboratory and
makes unannounced visits to inspect it and check the testing process.

Mr. Pillsbury also checks the data for consistency from run to run and from
year to year. Most industry members also have testing facilities; however, the
numbers published by the Commission are primarily TITL numbers. (Generic
and private label brands, as well as new cigarettes and cigarettes that are not
widely available, are not tested by TITL.)

Since the closing of its laboratory, the Commission has continued to
review advertising for today's low- and ultralow-yield cigarettes for deceptive
claims. In January 1995 the Commission approved a consent agreement with
the American Tobacco Company, settling charges over advertisements that
allegedly misused the Commission's tar and nicotine ratings by stating that
consumers would get less tar by smoking 10 packs of Carlton brand cigarettes

” In 1954 the tar yield of the sales-weighted average cigarette was 37 mg (U.S. Department of Health and Human
Services, 1981). By 1981 cigarettes yielding 15 mg of tar or less had 56 percent of the domestic market (Federal
Trade Commission, 1984).

7

Smoking and Tobacco Control Monograph No. 7

(which are rated as having 1 mg of tar per cigarette) than by smoking a single
pack of certain other brands of cigarettes (rated as having more than 10 mg
of tar per cigarette).

The Commission's desire to ensure that smokers have accurate and useful
information about their cigarettes led to its request for the conference, whose
reports are contained in this monograph.

QUESTION-AND-ANSWER SESSION

Mr. Peeler conducted a question-and-answer session simultaneously with
Mr. Pillsbury; see page 12.

REFERENCES

Federal Register. 31(215): 14278, November 4, 1966.

Federal Register. 32(147): 11178, August 1, 1967.

Federal Register. 35(154): 12671, August 8, 1970.

Federal Register. 42(79): 21155, April 25, 1977.

Federal Register. 43(56): 11856 and 11857, March 22,

1978.

Federal Register. 48(72): 15953-15955, April 13, 1983.

Federal Register. 49(108): 23120-23121, June 4, 1984.

Federal Trade Commission. "Statement of

Considerations." Press release. August 1, 1967, p. 2.

Federal Trade Commission. Report to Congress for 1981
Pursuant to the Federal Cigarette Labeling and
Advertising Act. Washington, DC: Federal Trade
Commission, 1984, p. 30 (Table 11).

Federal Trade Commission. Trade Regulation Reporter.

Vol. 6. Chicago: Commerce Clearing House, Inc.,

1988a, 1 39,012.70.

Federal Trade Commission. Trade Regulation Reporter.
Vol. 3. Chicago: Commerce Clearing House, Inc.,
1988b, 1 7853.51 at 11730.

U.S. Department of Health, Education, and Welfare.
Smoking and Health, Report of the Advisory Committee
to the Surgeon General of the Public Health Service.

PHS Publication No. 1103. Rockville, MD: U.S.
Department of Health and Human Services, Public
Health Service, 1964.

U.S. Department of Health and Human Services. The
Health Consequences of Smoking: The Changing
Cigarette. A Report of the Surgeon General. DHHS
Publication No. (PHS) 81-50156. Rockville, MD:
U.S. Department of Health and Human Services,
Public Health Service, Office on Smoking and
Health, 1981.

8

Chapter 2

Review of the Federal Xrade Gommission
IVlethod for Determining Cigarette Xar and
Nicotine Yield

Harold C. Pillsbury, Jr.^

The "Federal Trade Commission (Commission or FTC) method" is the
methodology that the Commission adopted almost 30 years ago for testing
cigarettes. This methodology is still used today by the Tobacco Institute
Testing Laboratory (TITL), with some minor modifications. The FTC method
determines the relative yield of individual cigarettes by smoking them in a
standardized fashion, according to a predetermined protocol, on a smoking
machine. The FTC test method was based on the "Cambridge Filter method"
developed by Ogg (1964), which called for 2-second, 35-mL puffs to be taken
until a 23-mm butt length remained on the cigarette. More about how these
parameters were selected is presented below.

For the testing procedure, as implemented initially by the FTC's cigarette
testing laboratory and currently by TITL, cigarettes are collected by an
independent firm that purchases two packages of each cigarette variety^
in each of 50 locations throughout the United States. (If some varieties
or brands are not available in certain locations, additional packs will be
purchased in locations where they are available.) They are mailed to the
testing laboratory; the postmark serves as verification that they were
purchased in different locations. Individual cigarettes to be tested are
selected on a random basis, two from each pack. Before being smoked,
the cigarettes are "conditioned" by being placed on storage trays in a room
maintained at 75 °F and 60 percent relative humidity for not less than
24 hours.

The machine used in the Commission's laboratory had 20 "ports"
(openings); the smoking machine currently used by TITL also has 20 ports.
Each opening is fitted with a filter holder, into which a cigarette is inserted
for smoking, and a filter pad, on which particulate matter from the cigarette
smoke is collected. Gases pass through the pad and are collected in specially
designed plastic bags.

' These remarks are the views of the staff of the Bureau of Consumer Protection. They do not necessarily
represent the view of the Commission or any individual commissioner.

^ A particular brand of cigarettes may have more than a dozen varieties, depending on whether it is available
in different lengths, in regular and menthol flavors, in hard and soft packaging, and in regular, light, and
ultralight versions. For example, the Commission's 1994 tar and nicotine report lists 20 varieties of Marlboro.

9

Smoking and Tobacco Control Monograph No. 7

The machines are calibrated to take one puff of 2-second duration and
35-mL volume every minute. Cigarettes are smoked to a butt length of
23 mm or the length of the overwrap plus 3 mm, whichever is longer.

When the cigarette has been smoked down to the prescribed length, it
burns through a string that has been placed on that mark; this causes a
microswitch to be flipped, which in turn disconnects that particular port
of the smoking machine. (Although this seems like a fairly unsophisticated
way of terminating the test, more sophisticated methods — such as infrared
detectors and thermal sensors — have been tried and rejected over the years.)

Five cigarettes of each variety are smoked, one at a time, using the same
filter holder.^ (A total of 100 cigarettes of each variety are smoked to get the
official tar, nicotine, and carbon monoxide ratings.) After the smoke from
those five cigarettes has been filtered through each filter pad, the holder is
removed and weighed. The difference between the weight of the holder
before and after the smoking process divided by the number of cigarettes
smoked is the total particulate matter collected from the cigarette smoke.

The filter pad is then extracted with a solvent,^ and the moisture content
is determined by injecting a measured amount of the extract into a gas
chromatograph and comparing the resulting peak against the standard curve.
Ratings for the three constituents reported by the Commission are then
determined as follows:

• Nicotine: As with moisture, a specified amount of the extract from the
filter pad is injected into a gas chromatograph, and the resulting peak
is compared against the standard curve.^

• Carbon monoxide: The gas collected in the plastic bag is passed
through an infrared detector to determine carbon monoxide levels.

• Tar: Tar level is determined by subtracting water and nicotine levels
from total particulate matter.

Tar and carbon monoxide figures are rounded up or down to the nearest
milligram, while nicotine figures are rounded to the nearest 10th of a
milligram. Varieties with tar and carbon monoxide results below 0.5 mg
per cigarette or nicotine results below 0.05 mg are reported as <0.5 mg or
<0.05 mg, respectively, because the FTC test method is not sensitive enough
to report these components at lower levels.

Although the ratings are based on 100 cigarettes, at least 150 (and
preferably 200) cigarettes of each variety are needed for the test to ensure

‘ To make certain that the machine is working properly, at least 4 of the 20 ports are reserved on each run for
"monitor" cigarettes — cigarettes with known yields for tar, nicotine, and carbon monoxide.

* the solution contains extractant and internal standards: 2-propanol containing 1 mg anethole per mL as
an internal standard for nicotine and 20 mg ethanol per ml, as an internal standard for water.

* Ultraviolet spectroscopy was used to determine nicotine until 1980, when it was replaced by gas
chromatography.

10

Chapter 2

that 100 are successfully smoked. Common technical problems that
can cause a filter pad to be discarded include lighting failures and port
leaks. During the last year of the FTC laboratory's operation, fewer than
300 varieties of cigarettes were tested, and the testing cycle (which included
curing, marking, and smoking the cigarettes, etc.) lasted approximately
12 months. There were 933 cigarette varieties rated by the TITL in the
Commission's 1994 report.

The author once had the opportunity to ask Dr. Ogg (who worked as a
tobacco chemist for the U.S. Department of Agriculture) how he came up
with the specific parameters of his protocol. He said that he had based them
on observations of how people smoke under different conditions. He had
spent a lot of time watching people smoke (at the office, on the street, etc.),
sometimes timing them with a stopwatch. His observations told him that
people smoked differently under different conditions. For example, someone
deep in thought might take only one or two puffs before the cigarette burned
out, whereas someone who seemed extremely nervous might puff constantly.
In short, there was no such thing as an "average" smoker and no way to
derive a set of testing parameters that would replicate actual human smoking,
so Dr. Ogg had to select parameters that seemed reasonable in light of his
observations.^ Dr. Ogg also collected cigarette butts from ash trays in hotels,
restaurants, and offices and measured how long they were; the resulting
average length became the butt length called for by his protocol.

When the Commission adopted a slightly modified version of the
Cambridge Filter method in 1967 for use in its newly opened cigarette testing
laboratory, it was the author's opinion that the Commission's procedures (as
implemented on the 20-port smoking machine selected by the Commission)
were clearly superior to all other methods currently in use at that time. The
FTC method had its limitations, most significantly that the information it
generated would not tell any individual smoker how much tar and nicotine
he or she would get from a particular brand of cigarette. However, there was
simply no way to get that information, and the FTC method did provide a
smoker with accurate comparative information about the relative amounts
of tar and nicotine delivered by various cigarettes when they were smoked
in precisely the same manner. In addition, it provided a uniform analytical
procedure that could be replicated in different laboratories simultaneously
and in the same laboratory over time; therefore, not only could many brands
of cigarettes be compared with each other at any time, but long-term pictures
of tar and nicotine levels over the years also were possible.

® During the December 5-6, 1994, National Cancer Institute conference, it was learned that a protocol using
the same parameters for the testing of cigarettes had been proposed by The American Tobacco Company
researchers many years before Dr. Ogg published his article (Bradford et al., 1936) ("arbitrarily" selecting a
2-second, 35-mL puff once a minute, although another researcher who had studied human smoking habits
used a 40-mL puff).

11

Smoking and Tobacco Control Monograph No. 7

QUESTION-AND-ANSWER SESSION

DR. HARRIS: I was curious about the very last statement on the tape: The
results are sent to the cigarette manufacturers who, in turn, report the
numbers to the Federal Trade Commission?

MR. PILLSBURY: Yes. We get the tar and nicotine data directly from the
cigarette manufacturers so that we can hold them responsible if there is
anything wrong with the numbers.

DR. HARRIS: To your knowledge, do the numbers reported under the
compulsory process by the manufacturers ever deviate from those that are
measured in the Tobacco Institute laboratory?

MR. PILLSBURY: The only thing I can tell you is that they are checked.

DR. STITZER: Could you remind us how the original Cambridge Filter
method was altered when the FTC method was developed?

MR. PILLSBURY: The original smoking machine was a four-port smoker that
used a column of water to draw from the cigarettes. When this new machine
came out, the filter pads and the holders were pretty much the same. The
only thing that has been changed is that the machine has been modified so
that carbon monoxide can be analyzed at the same time that the cigarettes
are being smoked.

DR. STITZER: So, there wasn't a puffing protocol that went along with the
original method?

MR. PEELER: We published, at the time that we adopted the method, a fairly
detailed protocol for how the test was supposed to be done. I suppose the
question is, did that protocol that we published differ from the original
method in the parameters that were required?

MR. PILLSBURY: No. They were pretty much the same as in the original
method.

DR. RICKERT: How much of a difference would you have to have in tar yields
between two brands before they would be considered to be different in the
statistical sense?

MR. PEELER: We publish the numbers and try to have a large enough sample
so that there are differences in those numbers. But the question of whether
there is a significant difference in those numbers is what we need to know
from you.

DR. RICKERT: What I am referring to is that on the tables in the UK there
is a footnote that reads, "Ignore differences in 2 mg in tar and CO," and I was
wondering whether that is the same sort of position that we have here?

MR. PILLSBURY: Fhe only thing that is done is they are rounded. Five and
above are rounded up; four and down are rounded down. We make no
criteria as to whether one with 14 mg is better for you than one with 15 mg.
We are just publishing the ratings of the cigarettes as they fall.

12

Chapter 2

MS. WILKENFELD: I think the answer is that, at least originally, we used to
publish the table with a standard deviation and that therefore there was a
significance between each degree of tar yield. We do not have confidence
in yields below .5, and that is announced in the report.

DR. PETITTI: About how long does it take to finish puffing one cigarette, and
what is the difference in the time that it might take to puff a cigarette that is
a very-high-tar cigarette vs. a cigarette that is very low tar?

MR. PILLSBURY: The difference in the length of time it takes to smoke a
cigarette is primarily a factor of how long the cigarette is, how tight the
tobacco is packed, how hard it is, and how much gas flows through the
cigarettes. Most of the cigarettes take approximately 10 minutes to smoke.
We have had longer cigarettes that have gone up to 12 to 13 minutes.

DR. PETITTI: Could you give me a range of the shortest vs. the longest?

Is it 5 minutes vs. 15, or is it 9 minutes vs. 12?

MR. PILLSBURY: Any range I would have to give you right now would be a
guess, because 1 haven't followed the range that closely. But 1 believe that
probably the shortest cigarette we have ever had is probably around 6 or
7 puffs per cigarette, and the longest one ran almost 15 puffs, but that was
a very long cigarette.

DR. BENOWITZ: Could you explain the rationale for the parameters that are
used in the current method? How did you arrive at the present protocol?

MR. PEELER: Let me ask Mr. Pillsbury to address what Dr. Ogg's rationale
was in the documents because he actually had an opportunity to discuss that
with Dr. Ogg. 1 think that if you look at the documents that the Commission
published at the time of the adoption of the testing methodology in 1967,
the Commission is fairly clear that, whatever Dr. Ogg's rationales were, it
did not believe it could replicate average smoking conditions. And so it was
picking parameters that were essentially fairly arbitrary.

MR. PILLSBURY: When we first started the lab, I talked to Dr. Ogg to quite
some extent on this topic. He had actually gone out there with a stopwatch
in his pocket and ridden the trains, and watched people in meetings and
so forth, and tried to get some feeling for how they were smoking. He came
back rather confused, because it seemed as though everybody smoked
differently: from the fellow who got on the train and looked at his
newspaper and lit his cigarette and never took another puff on it until it
burned down to the man who was sitting down arguing with somebody,
smoking like mad. So, he came up with what he considered a fairly average
way of smoking, so that you didn't get a big long firebox on the end of
the cigarette and you kept it burning.

As far as the butt length is concerned, they went out and picked up
cigarettes from ash trays in hotels and restaurants and so forth and did actual
measurements on those. And the best butt length that they could come up
with was 23, or the overwrap plus 3.

13

Smoking and Tobacco Control Monograph No. 7

MR. PEELER: Again, by the time the Commission adopted the methodology
in 1967, the Commission was very clear that it was not trying to establish
average smoking parameters.

DR. BOCK: 1 think that it goes back to the 1938 paper by the American
Tobacco Company group. 1 talked with Bradford and Harlan in Richmond
in 1953, and they, again, had gone to parties and watched what their friends
were doing. They were the same parameters, 1 believe, and it was based on
a group of probably upper-middle-income-level Richmondites.

DR. SHIFFMAN: You mentioned that the original FTC action on this was
under the FTC's general authority to prevent deceptive advertising. Now,
at the moment, you are also reporting the results of these tests to Congress.
Has there been any evolution in the FTC's authority in this area, or is it
still under this broad mandate?

MR. PEELER: No. The FTC's involvement in this issue continues to be under
its authority to regulate deceptive or unsubstantiated claims in advertising.
And, in the case of tar and nicotine testing in particular, there are two
variations: (1) We do have a voluntary agreement from the industry to
include this information in their advertising, and (2) we have had this
longstanding practice of sending the reports of this testing to Congress,
which was originally established in response to requests from the Commerce
Committee. But the only legal authority that we have in this area is our
authority to require claims in advertising to be truthful and to be
substantiated.

DR. COHEN: 1 want to return to the point of the statistical significance of the
yields. 1 think that is a very central question for the record. 1 would just like
to point out that there are three different sources of variance here that ought
to be considered: (1) variance due to product characteristics, such as product
design features; (2) variance due to individual smoking characteristics; and
(3) variance due to testing methodology.

Each of those sources of variance can be estimated separately, and it
may be very important later on, as the panel does its work, to consider the
implications of variance in each of those three separately.

REFERENCES

Bradford, J. A., Harlan, W.R., Hanmcr, H.R. Nature of
cigarette smoke. Technique of experimental
smoking. Industrial and Engineering (Chemistry 28(7):
836-839, 1936.

Ogg, C.L. Determination of particulate matter and
alkaloids (as nicotine) in cigarette smoke, foumal of
the Association of Official Agricultural Chemists 47:
356, 1964.

14

Chapter 3

Changes in Cigarette Design and Composition
Over Xime and How Xhey Influence the
Yields of Smoke Constituents

Dietrich Hoffmann, Mirjana V. Djordjevic, and Klaus D. Brunnemann

INTRODUCTION Since the first epidemiological reports on the association of

cigarette smoking with lung cancer, the composition of tobacco blends
and the makeup of commercial cigarettes in the United States as well as in
Western Europe have undergone major changes. Measured on the basis of
standardized machine smoking conditions, the sales-weighted average tar
and nicotine deliveries in U.S. cigarette smoke have decreased from 38 mg
and 2.7 mg, respectively, in 1954 to 12 mg and 0.95 mg, respectively, in
1993. The lower emissions have been primarily accomplished by using
efficient filter tips and highly porous cigarette paper and by changing the
composition of the tobacco blend. The latter includes the incorporation
of reconstituted and expanded tobaccos into the blend. Concurrent with
the reduction of tar and nicotine in the smokestream, there also occurred
a reduction of carbon monoxide, phenols, and carcinogenic polynuclear
aromatic hydrocarbons (PAHs). These reductions were partially tied to an
increase in the nitrate content of the tobacco blend used for U.S. cigarettes.
The addition of nitrate was initially targeted at decreasing the smoke yields
of PAHs; however, that this also would cause a gradual increase of the
carcinogenic, tobacco-specific N-nitrosamines (TSNAs) was not recognized
until there was awareness of those compounds as smoke constituents in
the 1970's.

These observations were based on measurements of yields from
cigarettes that were smoked under standardized laboratory conditions,
initially established in 1936, and adopted by the U.S. Federal Trade
Commission (FTC) in 1969. These conditions do not reflect the smoking
patterns of the smokers of filter cigarettes, who currently account for the
consumption of 97 percent of all cigarettes produced in the United States.

The current filter cigarette smoker tends to smoke more intensely and to
inhale more deeply. Thus, the actual exposure to toxic and tumorigenic
agents in the inhaled smoke of filter cigarettes is not necessarily in line
with the machine smoking data.

BACKGROUND In 1950 epidemiological studies reported that lung cancer was

particularly prevalent among cigarette smokers (Wynder and Graham, 1950;
Doll and Hill, 1950). These observations in the United States and the United
Kingdom were confirmed by the Royal College of Physicians (1962) and by
the U.S. Surgeon General in 1964 (U.S. Department of Health, Education, and
Welfare, 1964). These reports and the emerging knowledge of the presence

15

Smoking and Tobacco Control Monograph No. 7

of carcinogens and tumor promoters in cigarette smoke led to a gradual
change in the design and composition of commercial cigarettes in North
America, Western Europe, and other developed countries (Hoffmann and
Hoffmann, 1994a; Jarvis and Russell, 1985). The modifications were intended
to reduce both the toxicity and the carcinogenic potential of the cigarette
smoke. Although research on the changing cigarette was pursued in several
countries, this chapter deals primarily with the developments relating to
U.S. cigarettes between 1954 and 1993.

At the basis of all analytical assessments of smoke composition lies the
standardization of machine smoking methods, first suggested for empirical
cigarette smoking in Europe (Pfyl, 1933; Pyriki, 1934). In the United States,
Bradford and colleagues (1936) developed a procedure for cigarette smoking
on the basis of "arbitrarily selected" parameters of a 35-mL puff volume,
a 2-second puff duration, and one puff per minute. The only goal of this
method was to offer a means for comparing the smoke yields of various
types of cigarettes; there was no intent to simulate human smoking patterns.
The influences on smoke yields and composition that are exerted by the
overall physical characteristics of a cigarette — including its length and the
butt length to which it is smoked, its circumference, whether it is filtered
or nonfiltered, and the effects of the puff volume, puff frequency, and puff
duration; the type and cut of tobacco used as a filler; the properties of the
wrapper; and the mode of precipitation of the condensate — were described
in many research papers during the 1960's (Wynder and Hoffmann, 1967).

For regulatory purposes, Pillsbury and colleagues (1969) adapted in principle
the method of Bradford and coworkers (1936) and made some refinements
to establish what became known as the FTC method; the smoking parameters
were still a 35-mL puff volume, a 2-second puff duration, and a 1-puff-per-
minute frequency. What was new was the definition of the butt length to
which a cigarette was to be smoked. Butt lengths were set to be 23 mm for
plain cigarettes and length of the filter plus overwrap with an additional
3 mm for filter cigarettes. CORESTA, the International Organization for
Research on Tobacco, developed a comparable method that is widely used
in most of the developed countries (CORESTA, 1991-1993).

This chapter describes the analytical data obtained with the FTC method,
although many studies (Russell, 1980; Herning et al., 1981; Kozlowski et al.,
1982; Fagerstrdm, 1982; Haley et al., 1985; Byrd et al., 1994) have shown
that the standardized machine smoking method does not reflect the smoking
habits of consumers of filter cigarettes. This is especially so for filter cigarettes
with low and ultralow smoke yields, because smokers of such cigarettes
tend to inhale more deeply and draw puffs more frequently to satisfy a
physiologically conditioned need for nicotine (U.S. Department of Health
and Human Services, 1988).

Figure 1 presents the sales-weighted average tar and nicotine deliveries
of all U.S. domestic brands for the years 1954 through 1993 (Hoffmann and
Hoffmann, 1994a). Fhis figure also shows the major changes in the makeup
of U.S. cigarettes, such as the introduction of filter tips, porous cigarette

16

Chapter 3

Figure 1

Sales-weighted average tar and nicotine deliveries, 1 954-1 993

All Domestic Brands on U.S. Market

Key: RT = reconstituted tobacco; F = filter; ET = expanded tobacco.
Source: Hoffmann and Hoffmann, 1994a.

paper, reconstituted tobacco, filter tip ventilation, and use of expanded
tobacco. Similar developments occurred in most industrialized countries,
albeit at a somewhat slower pace and about 5 to 10 years after the
introduction of these changes in the United States (Hoffmann and
Hoffmann, 1994b; Jarvis and Russell, 1985; U.S. Department of Health

17

Nicotine (mg)

Smoking and Tobacco Control Monograph No. 7

and Human Services, 1988). Jarvis and Russell (1985) first observed for
English cigarettes that the smoke delivery of nicotine was not reduced to
the same extent as that of the tar. During the past 10 to 15 years, the same
observation was made for U.S. cigarettes. Figure 1 does not reflect the
gradual change in the tobacco blend of U.S. cigarettes with regard to an
increase of the hurley tobacco share from about 35.9 percent in 1950 to
46.5 percent in 1982; the remainder of the tobacco blend consists primarily
of bright tobacco with about 5 to 8 percent oriental tobacco and 1 percent
Maryland tobacco (Grise, 1984).

CHANGES IN
CIGARETTE
DESIGN AND
COMPOSITION

Cigarettes With
Filter Tips

Since 1955 the U.S. sales-weighted average smoke yields have
declined from 38 mg tar and 2.7 mg nicotine to 12 mg and 0.95 mg,
respectively (Figure 1). A major reason for the decrease in smoke
yields is the wide acceptance of filter cigarettes. Their use steadily
increased in America from 0.56 percent of all cigarettes smoked in
1950 to 19 percent in 1955, 51 percent in 1960, 82 percent in 1970,
92 percent in 1980, and more than 97 percent since 1993 (Figure 2)
(Hoffmann and Hoffmann, 1994b; U.S. Department of Agriculture, 1993).
Most filter tips (15 to 35 mm) are made of cellulose acetate; only a low
percentage of cigarettes are made with composite filters of cellulose acetate
with charcoal. Since about 1968, increasing proportions of the cellulose
acetate filter tips are perforated with one or more lines of tiny holes placed
near the middle of the filter tow. Today up to 50 percent of all cigarette
filter tips in the United States have various degrees of perforations. The
conventional filter cigarettes are acceptable to consumers with a maximal
draw resistance of up to about 130 mm water column (Kiefer and Touey,
1967). The filters reduce primarily the smoke yields of particulate matter
and thus the nonvolatile smoke constituents. The efficiency of cellulose
acetate filters for total particulate matter (TPM) removal can be increased by
reducing the diameter of the filaments without increasing the draw resistance
(Table la) or by using a longer filter tip (Table lb). In the mainstream smoke
of the U.S. blended cigarette with a pH below 6.3 to 6.5, more than 90 percent
of the nicotine is present in the particulate matter as a salt with organic acids
(Kiefer and Touey, 1967; Brunnemann and Hoffmann, 1974).

Conventional cellulose acetate has the capability to selectively reduce
some of the volatile and semivolatile compounds in the smokestream,
especially when the filter is treated with certain plasticizers, such as glycerol
triacetate. Some of the volatile smoke constituents that are ciliatoxic agents,
such as acrolein, are removed selectively, even beyond the reduction of
I'PM, by retention on such treated filter tips. Phenols and cresols, a group
of semivolatiles, also are removed selectively up to 80 to 85 percent, as are
the highly carcinogenic dialkylnitrosamines, of which up to 75 percent can
be retained on cellulose acetate filters (George and Keith, 1967; Brunnemann
and Hoffmann, 1977).

Filter tips with perforations allow dilution of the smoke with air.
Moreover, drawing puffs through perforated filter cigarettes reduces the
velocity of the air drawn through the burning cone. As a result, less of the

18

Chapter 3

Figure 2

Percentage of all U.S. cigarettes with filter tips

i

I

inner core of the burning cone is depleted of oxygen, and thus the levels
of carbon monoxide, hydrogen cyanide, and some other volatiles are
selectively reduced in the smoke of cigarettes with perforated filter tips
(Figure 3) (National Cancer Institute, 1977). Furthermore, the lower velocity
of the generated smoke increases the efficiency of the filter. However, the
tumorigenicity of the resulting tar does not change compared with that of
the tar of a conventional, nonperforated cellulose acetate filter cigarette
(National Cancer Institute, 1977). In principle, the smoke of a cigarette
can be diluted to an unlimited degree by air; however, the consumers'
nonacceptance of these cigarettes is the limiting factor.

The use of charcoal particles in one of two or three sections of a filter
tip, or sprayed onto the cellulose acetate, also offers the opportunity to
selectively reduce certain volatile smoke constituents, such as the ciliatoxic
hydrogen cyanide, acetaldehyde, and acrolein (National Cancer Institute,

1977; Tiggelbeck, 1968). However, replacing one section of the filter tip
with charcoal also leads to less reduction of TPM than can be achieved with

19

Smoking and Tobacco Control Monograph No. 7

Table 1 a

Effect of filament diameter on filter efficiency^

Approximate

Pressure Drop

Tar Removed

Filament Diameter (p)

(mm of H2O)

(percent)

22

55.7

30

20

55.7

33

17

53.1

36

14

55.7

38

12.6

53.1

43

Table 1 b

Effect of filter length on efficiency**

Filter Length

Pressure Drop

Tar Removed

(mm)

(mm of H2O)

(percent)

15

42

26.2

20

57

33.3

25

71

39.7

30

85

45.5

35

99

50.8

® Cellulose acetate, 1 7 mm in length, 25-mm circumference.
“ Cellulose acetate, 24.6-mm circumference.

Key: n = micron (W^ meter); Hfi = water.

Source: Kiefer and Touey, 1967.

a filter tip of the same length but made entirely of cellulose acetate (Figure 4)
(Brunnemann et al., 1990). Charcoal-containing filter tips are efficient in
selectively reducing certain volatile aromatic hydrocarbons, such as benzene
and toluene, from the smoke of the early puffs; yet, they release these
hydrocarbons during the later puffs (Brunnemann et al., 1990).

I'oday, more than 70 percent of all cigarettes sold in Japan have
charcoal-containing filter tips (Wynder and Hoffmann, 1994). Only a few
percent of the cigarettes sold in the United States have such filters. Although
more Japanese men smoke comparable numbers of cigarettes per day than
American men do and the smoke yields per cigarette in Japan are similar
to those in the United States, Japanese men have a significantly lower lung
cancer incidence rate (Wynder and Hoffmann, 1994; Wynder et al., 1992).
Among other factors, the lower yields of ciliatoxins, such as acrolein and
hydrogen cyanide, in the smoke of cigarettes with charcoal filter tips may
be partly responsible for the lower lung cancer rate in Japan.

20

Chapter 3

Figure 3

Regression lines for all the investigated smoke components

10 20 30 40 50 60 70

Dilution (percent)

Key: CO = carbon monoxide; tot. aid. = total volatile aldehydes; NO = nitrogen oxide; H^O = water;

00^ = carbon dioxide; HON = hydrogen cyanide.

Source: Norman, 1974.

21

Smoking and Tobacco Control Monograph No. 7

Figure 4

Filtration of smoke constituents

Viscose

15-mm Single Filters

Acetate

Paper

100

I 75

■D

a 50

c

o

25

F C A P N

Paper/Carbon

F C A P N
Compound

Acetate/Carbon
(e)

F C A P N

Bonded Carbon

Acetate + Paper

7.5- + 7.5-mm Dual Filters

Acetate + Paper/Carbon Acetate I + Paper/Carbon

F C A P N

Acetate + Bonded Carbon

(d)

F C A P N
Compound

F C A P N

Acetate -t- Acetate/Carbon Viscose + Bonded Carbon

Compound

Key: F = formaldehyde; C = hydrogen cyanide; A = acrolein; P = total phenols; N = nicotine.
Source: Williamson et al., 1965.

22

Chapter 3

Cigarette Paper With increasing permeability, porous cigarette papers significantly
reduce tar, carbon monoxide, and nitrogen oxides but not low-molecular-
weight gas phase components in the smokestream. Perforated cigarette
paper also significantly reduces hydrogen cyanide, whereas nicotine
reduction is less (National Cancer Institute, 1977) (Figure 5). In a recent
study it was found that porous cigarette paper reduces not only smoke yields
of carbon monoxide and tar but also of volatile nitrosamines, TSNAs, and
benzo(fl)pyrene (BaP) (Brunnemann et al., 1994). However, the reduction

Figure 5

Percentage change in smoke yield and composition with perforated, 0.5 percent
citrate paper

Key: NO = nitrogen oxide; CO = carbon monoxide.
Source: Owens, 1978.

23

Smoking and Tobacco Control Monograph No. 7

of TSNAs and BaP is not selective. On a gram-to-gram basis, the tars obtained
from cigarettes with high-porosity paper still have the same tumorigenic
activity as does the tar from control cigarettes that have conventional
cigarette paper (National Cancer Institute, 1977).

Reconstituted Reconstituted tobacco (RT) was first used after World War 11 as a

Tobacco binder for cigars and until the beginning of the 1960's on a limited

scale for cigarettes (Halter and Ito, 1979). The interest in RT grew with
the observation that cigarettes made exclusively from RT delivered lower
smoke yields of tar, phenols, and BaP. On a gram-to-gram basis, this tar had
significantly lower tumorigenicity on mouse skin and in the respiratory tract
of hamsters (Wynder and Hoffmann, 1965). In 1974 the Research Institute
of the German Cigarette Industry reported that forced exposure of Syrian
golden hamsters to the smoke of cigarettes filled exclusively with RT gave
significantly lower tumor incidence in the upper respiratory tract of the
animals than treatment with the smoke of a blended cigarette containing
only lamina of bright, hurley, and oriental tobacco (Dontenwill, 1974).

Reconstituted tobacco, or homogenized sheet tobacco as it is sometimes
called, is a paperlike sheet approaching the thickness of tobacco laminae.

RT is made from tobacco dust, fines, and particles from ribs and stems;
various additives may be incorporated. The process for making RT can be
divided into four general classes. The first two relate to the papermaking
process; the third involves a slurry; and the fourth is based on the
preparation of a tobacco paste with rollers using water or low-boiling
solvents. For the papermaking process, a mixture of fines, midribs, and
sometimes tobacco stems is broken up and extracted with water. The extract
is concentrated by evaporation. The insoluble residue is macerated further,
and the resulting material is formed into a paperlike web on a papermaking
machine. The web is dried and then impregnated with the concentrated
extract; this web is then further dried and cut. The shredded material is
added to the tobacco blend. Because the water extract of the tobacco
contains nicotine and this extract is added in concentrated form to the
tobacco web, this process has been considered a "nicotine-enriching process."
In one papermaking process, cellulose fiber is added to increase the filling
power and stability of the resulting RT.

In making RT by the slurry process, dry tobacco materials are finely
divided and often mixed with small amounts of adhesive, then suspended
in water. The resulting slurry is placed on a metallic band on which it is
dried. I'he resulting sheet is shredded and added to the tobacco blend. In
the rolling process, only small amounts of water are added to the mixture of
tobacco fines, dust, and finely powdered ribs; this paste is placed onto rollers
with different speeds, resulting in a sheet with limited filling power and
tensile strength.

The potential to produce RT in various forms with different densities
and filling powers and thereby to modify the tumorigenicity of tars and
whole smoke encouraged the National Cancer Institute (NCI) in the 1970's
to explore the use of various types of R F for recommendations of a less

24

Chapter 3

hazardous cigarette. The results documented that RT, especially RT
resulting from the paper process with cellulose fiber as an additive, offered
an opportunity to significantly reduce the cigarette smoke yields of tar,
nicotine, phenols, and PAHs, as well as the tumorigenicity of the resulting
tar. The most encouraging results were achieved with RT resulting from
the paper process using only tobacco stems (Table 2).

Today, most blended U.S. cigarettes contain 20 to 30 percent RT, which
is also now widely used in Europe, Canada, and Japan.

Puffed, Expanded, In the early 1970's a new tobacco preparation was introduced for
and Freeze-Dried the blended cigarette, that of "puffed," "expanded," or "freeze-
Tobaccos dried" tobacco. Using these materials, less tobacco is required

to fill a cigarette. The principle is to expand the tobacco cell walls by quick
evaporation of water and other vaporizable agents. This causes a rapid
pressure increase in the cells by heat and/or the reduction of external
pressure.

Table 3 summarizes the smoke yields of experimental cigarettes made
exclusively from puffed, expanded, or freeze-dried tobaccos. The smoke data
are compared with those from the smoke of the control cigarette. The tars
from the smoke of cigarettes made from expanded and freeze-dried tobaccos
were significantly less tumorigenic than tar from the control cigarettes
(National Cancer Institute, 1980).

Table 2

Smoke yields of cigarettes made from reconstituted tobacco (RT) by paper
processes and from control cigarettes

Components

RT

Stems Only

RT

Blend

Control

Weight (mg)

1,011.0

1 ,060.0

1 ,226.0

Tar (mg)

11.3

11.7

25.9

Nicotine (mg)

0.2

0.7

1.7

Carbon Monoxide (mg)

11.9

11.8

16.1

NO, (ixg)

586.0

343.0

367.0

Hydrocyanic acid (pg)

73.5

81.9

201.0

Acetaldehyde (pg)

1 ,027.0

948.0

1 ,065.0

Acrolein (pg)

99.0

105.0

109.0

Benz(a)anthracene (ng)

13.1

9.8

46.3

Benzo(a)pyrene (ng)

8.9

7.4

27.8

Key: NO^ = N (>95 percent) + NO^ (<5 percent).
Source: National Cancer Institute, 1976a and 1976b.

25

Smoking and Tobacco Control Monograph No. 7

Table 3

Smoke analysis of cigarettes made from puffed, expanded, and freeze-dried tobaccos
and from control cigarettes

Smoke

Component

Puffed

Tobacco

Expanded

Tobacco

Freeze-Dried

Tobacco

Control

Carbon Monoxide (mg)

9.33

11.80

12.30

18.00

Nitrogen Oxides (^g)

247.00

293.00

235.00

269.00

Hydrogen Cyanide (pg)

199.00

287.00

234.00

413.00

Formaldehyde (pg)

20.70

21.70

33.40

31.70

Acetaldehyde (pg)

814.00

720.00

968.00

986.00

Acrolein (pg)

105.00

87.70

92.40

128.00

Tar (mg)

15.60

18.20

16.30

36.70

Nicotine (mg)

0.78

0.74

0.82

2.61

Benz(a)anthracene (ng)

13.70

11.80

15.30

37.10

Benzo(a)pyrene (ng)

11.80

8.20

9.20

28.70

Source: National Cancer Institute, 1976b.

The use of puffed, expanded, or freeze-dried tobacco, together with the
use of filter tips and reconstituted tobaccos, has had a major impact on the
amounts of leaf tobacco needed per average U.S. cigarette. In about 1950
1,230 mg of leaf tobacco were required for one cigarette, whereas only
785 mg were needed in 1982 (Grise, 1984).

Physical As the length of a cigarette increases, there is more opportunity for air

Parameters to enter through the paper and for certain gaseous components, for

of Cigarettes example, carbon monoxide and hydrogen cyanide, to diffuse out of the
paper into the environment. Assuming that all other factors remain
Length the same and only the length of the cigarette increases, there will be

a higher smoke yield of tar and nicotine because more tobacco is burned
(Moore and Bock, 1968). In the past, it was claimed that tobacco absorbs
only slightly less of the smoke particulates than a cellulose acetate filter tip
(Dobrowsky, 1960). This may have been true in the early 1960's, but modern
cellulose acetate filter tips are more efficient in retaining smoke particulates
than the tobacco column of a cigarette.

Circumference With the packing density remaining constant, a decrease in

circumference of a cigarette reduces the amount of tobacco available for
burning. As a result, tar and nicotine yields in the smokestream are reduced
ri able 4) as are the yields of carbon monoxide and several other volatile
smoke constituents (DeHardeleben et al., 1978).

26

Chapter 3

Table 4

Effect of cigarette circumference on tar and nicotine in mainstream smoke

Circumference (mm)

Tar

Delivery (mg)

Nicotine

26

23.3

1.56

25

21.5

1.46

24

19.9

1.35

23

18.2

1.21

Source: DeBardeleben et al., 1978.

Tobacco Cut Studies have shown that modifying tobacco from fine to coarse cut
causes the number of puffs per cigarette to increase (DeBardeleben et al.,
1978). In general, cigarettes that are filled with a more coarsely cut tobacco
burn less efficiently than those made with fine-cut tobacco. One report,
comparing the smoke of cigarettes filled with coarse-cut tobacco (1.27 mm)
with smoke from cigarettes made with fine-cut tobacco (0.42 mm),
showed only slight differences in smoke yields (Spears, 1974). However,
a comparison of tars from cigarettes with given tobacco cut at rates of
20, 30, or 50 cuts per inch (1.27, 0.85, and 0.51 mm, respectively) showed
in a bioassay that the finer the cut of the tobacco, the lower the
tumorigenicity of the resulting tar (Wynder and Hoffmann, 1965).

Packing Density Increasing the mass of the tobacco in a cigarette — increasing the

packing density — causes yields of tar and nicotine in the smoke to rise.
However, packing more than 1.0 g of tobacco into an 85-mm cigarette causes
the yields of tar and nicotine in the smoke to decrease, most likely because
of increased retention by the tobacco acting as a filter (Figure 6).

Tobacco Pesticides Since 1969 the use of chlorinated pesticides has been banned

in the cultivation of tobacco in the United States. As a result, 1,1,1-trichloro-
2-(4,4'-dichlorodiphenyl)ethane (DDT) and l,l,-dichloro-2-2(4,4'-
dichlorodiphenyl)ethane (DDD) in tobacco and in cigarette smoke have
drastically decreased. In the tobacco of a cigarette made in 1965, 13.4 ppm
DDT and 20.2 ppm DDD were measured, and in the tobacco of the leading
cigarette brand made in 1993, only 0.02 ppm DDT and 0.013 ppm DDD
were detected, a decrease of more than 98 percent (Djordjevic et al., 1995).
The small amounts of residual DDT and DDD in more recently produced
cigarettes appear to originate from imported tobaccos used for blended
cigarettes.

It was reported in 1981 that U.S. tobacco contains 250 ppb of the
carcinogenic N-nitrosodiethanolamine (NDELA). This nitrosamine is formed
by N-nitrosation of the secondary amine diethanolamine during tobacco

27

Smoking and Tobacco Control Monograph No. 7

Figure 6

Effect of cigarette weight/packing density on particulate matter

O

t2

0)

(0

o

o

+

ra

O)

E

o>

C

0)

k_

0)

C7)

o

c

O

O

z

C7)

E

Key: TPM = total particulate matter.
Source: DeBardeleben et al., 1978.

28

Chapter 3

processing. The major source of diethanolamine in tobacco in 1981 was
the sucker growth inhibitor MH-30, which is the diethanolamine salt of
maleic hydrazide (Brunnemann and Hoffmann, 1981). Because of the ban
on MH-30 for tobacco treatment, NDELA levels have decreased to less than
100 ppb in cigarette tobacco (Brunnemann and Hoffmann, 1991). The
remaining NDELA may be at least partially due to the contamination with
diethanolamine from packaging materials.

Several pesticides are still being used on tobacco; these include
insecticides, fumigants, and insect growth regulators (Benezet, 1989).

There is only limited knowledge about the residues of these agents on
cigarette tobacco and about their role during smoking.

Additives In April 1994, the major U.S. cigarette companies released a list of

599 additives used in the manufacture of cigarettes (Tobacco Reporter
Staff, 1994). Little is known about the fate of such additives during the
smoking of cigarettes. An exception is menthol, which amounts to less
than 2.5 mg in U.S. mentholated cigarettes (Perfetti and Gordin, 1985).
Menthol is not carcinogenic in rodents (National Cancer Institute, 1979),
nor does this readily volatilized compound give rise to measurable amounts
of carcinogenic hydrocarbons, including BaP, during the smoking of
cigarettes (Jenkins et al., 1970).

The list of additives also contains inorganic salts, such as ammonium
and potassium carbonates, and bicarbonates. These additives possibly
increase the pH of cigarette smoke. Beyond pH 6.0, cigarette smoke contains
increasing amounts of unprotonated nicotine; with smoke pH at 6.9, about
10 percent of the nicotine is present in the smoke in free form; at pH 7.85
this rises to 50 percent (Brunnemann and Hoffmann, 1974). The free
nicotine is present predominantly in the vapor phase of the smoke and is
more quickly absorbed through the oral mucosa than nicotine in salt form
(Armitage and Turner, 1970). Data are urgently needed for examining the
change in pH of the smoke of cigarettes with additives.

Although most additives that are used as flavor-enhancing agents are
sprayed onto tobacco in milligram amounts and may therefore generate at
most microgram amounts of toxic or tumorigenic agents in the smoke, it is
nevertheless important to document the fate of such compounds when they
are added to cigarettes, cigars, or pipe tobacco.

Tobacco Blend Most U.S. cigarettes manufactured worldwide are blended cigarettes.
The composition of the tobacco blend has a major influence on the pH,
toxicity, and tumorigenicity of the smoke. Many tobacco lines are available,
including about 60 species and about 1,000 different tobacco varieties (Tso,
1972). The wealth of this source permits the manipulation of the tobacco
plant and its components and leads to selective use of those portions of
the plant that enhance or reduce specific agents in the smoke. This is then
reflected in the toxicity and/or carcinogenicity of the smoke. For example,
there are flue-cured tobacco lines that contain 0.2 to 4.75 percent nicotine
and hurley lines with 0.3 to 4.58 percent nicotine (Chaplin, 1975).

I

29

tAllNi

Smoking and Tobacco Control Monograph No. 7

Furthermore, flue-cured tobacco leaves harvested from the lowest stalk
position contain 0.08 to 0.65 percent nicotine, whereas those from the
highest positions contain between 0.13 and 4.18 percent nicotine (Tso,

1977). The resulting smoke differs widely in its concentration of toxic and
tumorigenic agents (Hoffmann and Hoffmann, 1994a). Another example
is the BaP content of the smoke generated from leaves harvested from the
lowest stalk position, which ranges between 14.9 and 18.2 ng per cigarette,
contrasted with BaP in the smoke from the leaves of the highest stalk
position, which ranges between 23.2 and 35.2 ng per cigarette (Rathkamp
et al., 1973).

The first comparative study of the smoke of cigarettes made exclusively
from bright, oriental, hurley, and Maryland tobacco was published by
Wynder and Hoffmann (1963). The BaP levels in the smoke per cigarette
(without filter tip) were 53, 44, 24, and 18 ng, respectively. The tars from the
smoke of cigarettes made with bright and oriental tobaccos were significantly
more tumorigenic than the tars from hurley and Maryland tobaccos (Wynder
and Hoffmann, 1963). A large-scale study by NCI confirmed the observation
that the smoke of hurley tobacco is lower in BaP and other carcinogenic
agents than the smoke of bright tobacco and that the tar has less tumorigenic
activity than the tar from bright tobacco (National Cancer Institute, 1980).

During the past three decades, the nitrate content of the U.S. cigarette
blend increased from 0.3 to 0.5 percent to 0.6 to 1.35 percent (U.S.
Department of Health and Human Services, 1981; Fischer et al., 1990).

During smoking, the nitrates in tobacco give rise to nitrogen oxides that
scavenge C,H-radicals and thereby inhibit the pyrosyn thesis of carcinogenic
PAHs; at the same time, nitrogen oxides are involved in the formation of
nitrosamines from secondary and tertiary amines in tobacco (Rathkamp and
Hoffmann, 1970; Hoffmann et al., 1994). The result is that today the smoke
of the U.S. blended cigarette has lower concentrations of PAHs but higher
concentrations of N-nitrosamines than the smoke of the U.S. blended
cigarette three decades ago. Figure 7 shows the decrease per cigarette of
BaP from 50 ng in 1965 to 20 ng in 1992 and the concomitant increase of
the levels of the organ-specific lung carcinogen 4-(methylnitrosamino)-l-
(3-pyridyl)-l-butanone (NNK) from 110 ng in the late 1970's to 176 ng in
1992. These data pertain to the smoke of a leading United States nonfilter
cigarette. NNK is formed from nicotine during tobacco processing and
smoking (Hoffmann and Hoffmann, 1994a). In laboratory animals,
carcinogenic PAHs induce primarily squamous cell carcinoma, whereas
NNK elicits mainly adenocarcinoma in the peripheral lung. One major
reason for the steep ascent of lung adenocarcinoma incidence in cigarette
smokers in the United States compared with the more modest rise of
squamous cell carcinoma may lie in the more intense smoking of the low-
nicotine cigarette. The deeper inhalation of the smoke from these cigarettes
has led to higher yields of NNK and lower yields of BaP in the smoke of the
more recent cigarettes. Fhis modification has created a different profile of
smoke carcinogens that is likely reflected in the changed tumor morphology
that has emerged since the 1960's (Wynder and Hoffmann, 1994).

30

Chapter 3

Figure 7

BaP and NNK In mainstream smoke of a leading U.S. nonfilter cigarette, 1959-1992

200

150

100

50

1955 1960 1965 1970

1975

Year

1980 1985 1990

0

1995

Source: Hoffmann and Hoffmann, 1994a.

SUMMARY Table 5 indicates the potential roles that filter tips, perforated filter tips,
cigarette paper, reconstituted tobacco, expanded tobacco, and an increase
of the share of bright and hurley tobacco in the cigarette blend have in
affecting the smoke yields of selected toxic and tumorigenic agents. These
observations have largely been taken into account with respect to the
manufacture of blended U.S. filter cigarettes, which accounted for 97 percent
of all cigarettes sold on the U.S. market in 1993. The result is a cigarette
that delivers smoke with generally lower toxicity and tumorigenicity than
products that were smoked 40 years ago. However, all the measurements
on which this evaluation are based were obtained by standardized machine
smoking with parameters that are not in line with the real practices of men
and women who smoke the modern, low-yield, filter-tipped cigarettes
(Russell, 1980; Herning et al., 1981; Kozlowski et al., 1982; Fagerstrom, 1982;
Haley et al., 1985; Byrd et al., 1994). Is it thus safe to say that the modern
cigarette is really less harmful?

31

NNK (ng)

Smoking and Tobacco Control Monograph No. 7

Table 5

Changes in cigarette design and composition:

Effects on smoke yields of selected toxic agents

Smoke

Compound

Filter

Perforated

Filter

Cigarette

Paper

Reconstituted

Tobacco

Expanded

Tobacco

Bright

Tobacco

Burley

Tobacco

Tar

a

e

a

a

a

b

a

Nicotine

a

e

c

a

a

c

c

pH

NC

NC

NC

NC

NC

d

b

CO

c

a

NC

a

a

b

d

HCN

Volatile

NC

a

NC

a

a

c

c

Aldehydes

Volatile

NC

a

NC

a

a

b

a

Nitrosamines

e

e

NC

a

a

6

b

Phenol

e

e

NC

a

a

b

a

PAHs

a

e

NC

a

a

b

a

TSNAs

a

e

NC

f

f

e

b

® Significant decrease.

“ Trend for increase.

‘ Can increase, can decrease.

“ Trend for decrease.

® More than a 50-percent decrease.

' Unknown.

Key: CO = carbon monoxide: HCN = hydrocyanic acid; PAHs = polynuclear aromatic hydrocarbons;
TSNAs = tobacco-specific N-nitrosamines; NC = no significant change.

How can the human risk from cigarette smoking truly be assessed?
Should we not above all remember that the only way to prevent smoking-
related diseases is abstinence from tobacco? Meanwhile, millions of smokers
in the United States and worldwide continue to smoke cigarettes and to use
other forms of tobacco because of their dependence on nicotine. Smoking
cessation efforts have had success for many but are not likely to stem the tide
of an enormous epidemic of smoking-related diseases that will be seen in the
coming decades in those parts of the world that have hardly begun to tally
the incidence and mortality from tobacco-related illness.

In the United States, we have today several sensitive techniques that can
assist in determining uptake and even an individual's capacity for activating
vs. detoxifying xenobiotics, such as the toxins and carcinogens from tobacco
smoke (Bryant et al., 1988; Santella et al., 1992; Melikian et al., 1993; Hecht
et al., 1994), but these so|)histicated methods of risk assessment are research
tools that for now do little to guide the consumer. One may agree with the
content of an editorial published in the New York l imes (1989) that read:
"Obviously, no smoking is better than smoking, but the best should not be

32

Chapter 3

the enemy of the good. There is a strong social case for encouraging
manufacturers to develop safer cigarettes that will sell." If we take this
premise as a realistic approach to the tobacco and illness dilemma in our
Nation, how can our regulatory agencies effectively protect the consumer
and on what type of measurement should risk assessment from cigarette
smoking be based? This is the question to be resolved. The authors hope
that presentation of some historical background will assist with this aim.

QUESTION-AND-ANSWER SESSION

DR. HENNINGFIELD: Dr. Hoffmann, the influence of some parameters,
such as increasing puff quantity, would be pretty obvious for their impact;
you would take in more smoke. But what about the factor of changing the
intensity of a puff? For example, the FTC method uses 35 mL over 2 seconds,
or say about 18 mL per second. What would be the impact of tripling the
intensity by going to, say, 60 mL per 1 second?

DR. HOFFMANN: This has been done by various groups, including
Dr. Benowitz, Dr. Auston, and Dr. Ogg. All have shown that when you
smoke more intensely (I think one report makes up to four or five puffs
per minute, with puff volumes up to 55 mL), you obviously increase the
smoke yields for cigarette smoke; based on epidemiological observations,
but you inhale deeper.

Now, this is reflected in the yield of nicotine respectively as one of its
major metabolites. And in fact, R.J. Reynolds Tobacco Company has recently
shown a very low yielding cigarette. They determined 90 percent of all
metabolites, and I think the results are in here. They have shown that with
the very low yielding cigarettes, the smoker inhales more than one would
expect from machine smoking data, based on the nicotine metabolites.

Machine smoking data may be all right for the cigarette without a filter
tip, but based on all these studies (I think there are eight all total), the
smoker of a low yielding cigarette inhales deeper and takes more puffs,
smokes more intensely.

DR. RICKERT: Dr. Hoffmann, I think you were intimately involved in the
NCTs less hazardous cigarette program a number of years ago. Why was
that program abandoned?

DR. HOFFMANN: The timing was not right — I do not know the details.

I work in the laboratory, and that is outside the field. It was purely politics.

DR. HARRIS: Dr. Hoffmann, you presented trends in some cigarette smoke
components over time. What do you know, if anything, about gross
characteristics of cigarette smoke, such as the trends in the pH of American
cigarette smoke or in the oxidation reduction potential of smoke?

DR. HOFFMANN: The pH has increased slightly; it is slightly higher in
filtered cigarettes, in perforated filter cigarettes, and in RT.

33

VAIIM

Smoking and Tobacco Control Monograph No. 7

There has been a slight increase in unprotonated nicotine, but it is a
minor difference, because it is still a blended cigarette. If you smoke a
French cigarette, which are the black or burley type cigarettes, they have
a pH of 7.5, or 40 percent of the nicotine is unprotonated; whereas, in our
blended U.S. cigarettes, less than 5 percent is unprotonated. As an English
study by Turner and others has shown, when you have unprotonated
nicotine, most of it has a quicker result to the mucous membrane, especially
of the oral cavity. In other words, when you have unprotonated nicotine,
not in salt form but in free base, most of it is in the water phase, and
therefore it is absorbed more quickly by the surface of the bronchial
epithelium or the oral cavity.

Therefore, you would rarely see a Frenchman taking as deep inhalations
as a smoker of a blended cigarette with an active filter tip. You watch a
Paris cab driver and you will see that they never inhale; he just dangles the
cigarette on the side of his mouth, because he would get a tremendous
nicotine kick if he inhaled.

DR. HARRIS: Does the protonation state of nicotine, whether it is protonated
or free base, affect the measurement method of nicotine as currently used by
the FTC?

DR. HOFFMANN: No, the pH is not measured. 1 do not see the need because
so far, in our U.S. blended cigarettes, there are no major differences. That
may change, but at present, it is not.

DR. HUGHES: 1 noticed over time that the tar and nicotine yields have
changed somewhat. What is your opinion about how feasible it is, using
existing techniques, to change that ratio?

DR. HOFFMANN: The first study was performed in the United Kingdom
by Russell. It demonstrated that the ratio of tar to nicotine, which was
originally 100 to 6, has changed to 100 to 10. We see this in low yielding
cigarettes. In other words, the nicotine is not reduced to the same extent
that the tar is reduced.

DR. HUGHES: And how feasible would it be for the manufacturers to
deliberately change that ratio at this point?

DR. HOFFMANN: I'hey can do it easily by changing the tobacco variety,
which is high in nicotine. We have heard about genetic engineering for
a tobacco variety that is very high in nicotine. So that is possible. 1 mean,
the manufacturer has everything in his hand to have high nicotine and
low tar or vice versa.

In fact, for a brief time, there was a cigarette on the market that was free
of nicotine. I he nicotine was extracted from the tobacco with supercritical
fluid transaction, and the tobacco was then used for cigarettes. So, the
tobacco industry has a whole spectrum from high- to low-nicotine yield.

Fhat depends on what the consumer requests.

Chapter 3

REFERENCES

Armitage, A.K., Turner, D.M. Absorption of nicotine
in cigarette and cigar smoke through the oral
mucosa. Nature 226: 1231-1232, 1970.

Benezet, H.J. Chemical control of pests in stored
tobacco. Recent Advances in Tobacco Science 15: 1-25,
1989.

Bradford, J.A., Harlan, W.R., Hanmer, H.R. Nature of
cigaret smoke. Technic of experimental smoking.
Industrial and Engineering Chemistry 28(7): 836-839,
1936.

Brunnemann, K.D., Hoffmann, D. The pH of tobacco
smoke. Food and Cosmetics Toxicology 12: 115-124,
1974.

Brunnemann, K.D., Hoffmann, D. Assessment of
carcinogenic, volatile N-nitrosamines in tobacco
and in mainstream and sidestream smoke of
cigarettes. Cancer Research 37: 3218-3222, 1977.

Brunnemann, K.D., Hoffmann, D. Assessment of the
carcinogenic N-nitrosodiethanolamine in tobacco
products and tobacco smoke. Carcinogenesis 2:
1123-1127, 1981.

Brunnemann, K.D., Hoffmann, D. Decreased
concentrations of N-nitrosodiethanolamine and
N-nitrosomorpholine in commercial tobacco
products. Journal of Agricultural and Food Chemistry
39: 207-208, 1991.

Bmnnemann, K.D., Hoffmann, D., Gairola, C.G., Lee,
B.C. Low ignition propensity cigarettes: Smoke
analysis for carcinogens and testing for mutagenic
activity of the smoke particulate matter. Food and
Chemical Toxicology 32: 917-922, 1994.

Bmnnemann, K.D., Kagan, M.R., Cox, J.E., Hoffmann,
D. Analysis of 1,3-butadiene and other selected gas-
phase components in cigarette mainstream and
sidestream smoke by gas chromatography-mass
selective detection. Carcinogenesis 11(10): 1863-
1868, 1990.

Bryant, M.S., Vineis, P., Skipper, P.L., Tannenbaum,
S.R. Hemoglobin adducts of aromatic amines:
Associations with smoking status and type of
tobacco. Proceedings of the National Academy of
Sciences of the United States of America 85(24): 9788-
9791, 1988.

Byrd, G.D., Robinson, J.H., Caldwell, W.S., deBethizy,
J.D. Nicotine uptake and metabolism in smokers.
Recent Advances in Tobacco Science 20: 5-31, 1994.

Chaplin, J.F. Genetic influence on chemical
constituents of tobacco leaf and smoke. Beitrage
zur Tabakforschung 8: 233-240, 1975.

CORESTA. Recommended Method No. 22. Routine
analytical cigarette-smoking machine
specifications, definitions and standard conditions.
CORESTA Information Bulletin 124-140, 1991-1993.

DeBardeleben, M.Z., Claflin, W.E., Gannon, W.F. Role
of cigarette physical characteristics on smoke
composition. Recent Advances in Tobacco Science 4:
85-111, 1978.

Djordjevic, M.V., Fan, J., Ferguson, S., Hoffmann, D.
Self-regulation of smoking intensity. Smoke yields
of the low-nicotine, low-"tar" cigarettes.
Carcinogenesis 16(9): 2015-2021, 1995.

Dobrowsky, A. The absorption of tobacco smoke: How
far is the cigarette its own filter? Tobacco Science 4:
126-129, 1960.

Doll, R., Hill, A.B. Smoking and carcinoma of the
lung. Preliminary report. British Medical Journal 2:
739-748, 1950.

Dontenwill, W.P. Tumorigenic effect of chronic
cigarette smoke inhalation on Syrian golden
hamsters. In: Experimental Lung Cancer.
Carcinogenesis and Bioassays, E. Karbe and J.F. Park
(Editors). New York: Springer-Verlag, 1974, pp. 331-
359.

Fagerstrbm, K.O. Effects of a nicotine-enriched
cigarette on nicotine titration, daily cigarette
consumption, and levels of carbon monoxide,
cotinine, and nicotine. Psychopharmacology 77(2):
164-167, 1982.

Fischer, S., Spiegelhalder, R., Preussmann, R. Tobacco-
specific nitrosamines in European and U.S.
cigarettes. Archiv fur Geschwulstforschung 60: 169-
177, 1990.

George, T.W., Keith, C.H. The selective filtration of
tobacco smoke. In: Tobacco and Tobacco Smoke.
Studies in Experimental Tobacco Carcinogenesis, E.L.
Wynder and D. Hoffmann (Editors). New York:
Academic Press, 1967, pp. 577-622.

Grise, V.N. Market growth of reduced tar cigarettes.
Recent Advances in Tobacco Science 10: 4-14, 1984.

Haley, N.J., Sepkovic, D.W., Hoffmann, D., Wynder,
E.L. Cigarette smoking as a risk factor for
cardiovascular disease. VI. Compensation with
nicotine availability as a single variable. Clinical
Pharmacology and Therapeutics 38: 164-170, 1985.

Halter, H.M., Ito, T.l. Effect of tobacco reconstitution
and expansion processes on smoke composition.
Recent Advances in Tobacco Science 4: 113-132, 1979.

Hecht, S.S., Carmella, S.G., Foiles, P.G., Murphy, S.E.
Biomarkers for human uptake and metabolic
activation of tobacco-specific nitrosamines. Cancer
Research 54(7 Suppl): 1912s-1917s, 1994.

Heming, R.I., Jones, R.T., Bachman, J., Mines, A.H.
Puff volume increases when low-nicotine cigarettes
are smoked. British Medical Journal (Clinical Research
MJ 283(6285): 187-189, 1981.

35

Smoking and Tobacco Control Monograph No. 7

Hoffmann, D., Bmnnemann, K.D., Piokopczyk, B.,
Djordjevic, M.V. Tobacco-specific N-nitrosamines
and Areca-derived N-nitrosamines: Chemistry,
biochemistry, carcinogenicity and relevance to
humans. Journal of Toxicology and Environmental
Health 41: 1-52, 1994.

Hoffmann, D., Hoffmann, 1. Tobacco consumption and
lung cancer. In: Lung Cancer, H.H. Hansen (Editor).
Boston: Kluwer Academic Publications, 1994a, pp. 1-
42.

Hoffmann, D., Hoffmann, 1. Recent developments in
smoking-related cancer research. Journal of Smoking
Related Diseases 5: 77-93, 1994b.

Jarvis, M.J., Russell, M.A. Tar and nicotine yields of U.K.
cigarettes 1972-1983. Sales-weighted estimates from
non-industry sources. British Journal of Addiction
80(4): 429-434, 1985.

Jenkins, R.W.J., Newman, R.H., Chavis, M.K. Cigarette
smoke formation studies II. Smoke distribution and
mainstream pyrolytic composition of added *‘‘C-
menthol (U). Beitrage zur Tabakforschung 5: 299-301,
1970.

Kiefer, J.E., Touey, G.P. Filtration of tobacco smoke
particles. In: Tobacco and Tobacco Smoke. Studies in
Experimental Tobacco Carcinogenesis, E.L. Wynder and
D. Hoffmann (Editors). New York: Academic Press,
1967, pp. 545-575.

Kozlowsid, L.T., Rickert, W.S., Pope, M.A., Robinson,
J.C., Frecker, R.C. Estimating the yields to smokers of
tar, nicotine, and carbon monoxide from the "lowest
yield" ventilated filter cigarette. British Journal of
Addiction 77: 159-165, 1982.

Melikian, A.A., Prahalad, A.K., Hoffmann, D. Urinary
trans,trans-muconic acid as an indicator of exposure
to benzene in cigarette smokers. Cancer Epidemiology,
Biomarkers and Prevention 3: 47-51, 1993.

Moore, G.E., Bock, F.G. Tar and nicotine levels of
American cigarettes. National Cancer Institute
Monograph 28: 89-94, 1968.

National Cancer Institute. Bioassays ofdl-Menthol for
Possible Carcinogenicity. Vol. 98. Carcinogenesis
Technical Report Series. Bethesda, MD: National
Intitules of Health, 1979, pp. 1-123.

National Cancer Institute, Smoking and Health
Program. Toward Less Hazardous Cigarettes. The First
Set of Experimental Cigarettes. DHEW Publication No.
(NIH) 76-905. Bethesda, MD: U.S. Department of
Health, Fxlucation, and Welfare, Public Health
Service, National Institutes of Health, 1976a.

National Cancer Institute, Smoking and Health

Program. Toward Less Hazardous (Cigarettes. The Second
Set of Experimental Cigarettes. DHEW Publication No.
(Nil!) 76-1111, Bethesda, MD: U.S. Department of
Health, Ixlucation, and Welfare, Public Health
Service, National Institutes of Health, 1976b.

National Cancer Institute, Smoking and Health
Program. Toward Less Hazardous Cigarettes. The
Third Set of Experimental Cigarettes. DHEW
Publication No. (NIH) 77-1280. Bethesda, MD:

U.S. Department of Health, Education, and
Welfare, Public Health Service, National Institutes
of Health, 1977.

National Cancer Institute, Smoking and Health
Program. Toward Less Hazardous Cigarettes. The
Fourth Set of Experimental Cigarettes. Bethesda, MD:
U.S. Department of Health, Education, and
Welfare, Public Health Service, National Institutes
of Health, 1980.

New York Times. Safer cigarettes (editorial). March 3,
1989. p. A38.

Norman, V. The effect of perforated tipping paper on
the yields of various smoke components. Beitrage
zur Tabakforschung 7: 282-287, 1974.

Owens, W.F., Jr. Effect of cigarette paper on smoke
yield and composition. Recent Advances in Tobacco
Science 4: 3-24, 1978.

Perfetti, T.A., Gordin, H.H. Just noticeable
differences. Studies of mentholated cigarette
products. Tobacco Science 29: 57-66, 1985.

Pfyl, B. Zur Bestimmung des Nikotins II. Mitteilung.
Zeitschrifi fur Lebensmittel-Untersuchung und-
Forschung 66: 501-510, 1933.

Pillsbury, H.C., Bright, C.C., O'Connor, K.J., Irish,
F.W. Tar and nicotine in cigarette smoke. Journal-
Association of Official Analytical Chemists 52: 458-
462, 1969.

l^riki, C. Verteilung von Nikotin im Rauch von
Zigaretten. Chemiker-Zeitung 58: 279-280, 1934.

Rathkamp, G., Hoffmann, D. Inhibition of the
pyrosynthesis of several selective smoke
constituents. Beitrage zur Tabakforschung 5: 302-
.306, 1970.

Rathkamp, G., Tso, T.C., Hoffmann, D. Smoke
analysis of cigarettes made from bright tobaccos
differing in variety and stalk position. Beitrage zur
Tabakforschung 7: 179-189, 1973.

Royal College of Physicians. Smoking and Health.
London: Pitman, 1962.

Russell, M.A. 11. The case for medium-nicotine, low-
tar, low-carbon monoxide cigarettes. Banbury
Report s: 297-310, 1980.

Santella, R.M., Grinberg-Gunes, R.A., Young, T.L.,
Dickey, C., Singh, V.N., Wang, L.W., Perera, F.P.
Cigarette smoking related polycyclic aromatic
hydrocarbon-1 )NA adducts in peripheral
mononuclear cells. Carcinogenesis 13: 2041-2045,
1992.

Spears, A.W. Effect of manufacturing variables on
cigarette smoke composition. CORESTA Bulletin
d'lnformation 65-78, 1974.

Chapter 3

Tiggelbeck, D. Comments on selective cigarette-smoke
filtration. National Cancer Institute Monograph 28:
249-258, 1968.

Tobacco Reporter Staff. Safe cigarette flavorings not
a risk, says independent panel. Tobacco Reporter
I 121(7): 32-39, 1994.

I Tso, T.C. The potential for producing safer cigarette
tobacco. Agriculture Science Review 10: 1-10, 1972.

Tso, T.C. Simple conelation and multiple regression
among leaf characteristics, smoke components, and
biological responses of bright tobaccos. Agricultural
Research Service Technical Bulletin 1551: 1-135, 1977.

U.S. Department of Agriculture. Percent of total
cigarette production which are filter-tipped
cigarettes. Foreign Tobacco 8: 40-41, 1993.

I U.S. Department of Agriculture, Economic Research
I Service. Tobacco Situation and Outlook Report. TBS-
i 229. Washington, DC: U.S. Department of

I Agriculture, 1994.

U.S. Department of Health, Education, and Welfare.
Smoking and Health. Report of the Advisory Committee
to the Surgeon General of the Public Health Service.
Publication No. 1103. Rockville, MD: U.S.
Department of Health, Education, and Welfare,
Public Health Service, 1964.

U.S. Department of Health and Human Services. The
Health Consequences of Smoking: The Changing
Cigarette. A Report of the Surgeon General. DHHS
Publication No. (PHS) 81-50156. Rockville, MD:

U.S. Department of Health and Human Services,
Public Health Service, Office on Smoking and
Health, 1981.

U.S. Department of Health and Human Services. The
Health Consequences of Smoking: Nicotine Addiction.

A Report of the Surgeon General, 1988. DHHS
Publication No. (CDC) 88-8406. Rockville, MD: U.S.
Department of Health and Human Services, Public
Health Service, Centers for Disease Control, Center
for Health Promotion and Education, Office on
Smoking and Health, 1988.

Williamson, J.T., Graham, J.F., Allman, D.R. The
modification of cigarette smoke filters. Beitrage
zur Tabakforschung 3: 233-242, 1965.

Wynder, E.L., Graham, E.A. Tobacco smoking as a
possible etiologic factor in bronchiogenic
carcinoma. A study of six hundred and eighty-four
proved cases. Journal of the American Medical
Association 143: 329-336, 1950.

Wynder, E.L., Hoffmann, D. Ein experimenteller
Beitrag zur Tabakrauchkanzerogenese. Deutsche
Medizinische Wochenschrift 88: 623-628, 1963.

Wynder, E.L., Hoffmann, D. Reduction of the
tumorigenicity of cigarette smoke. Journal of the
American Medical Association 192: 88-94, 1965.

Wynder, E.L., Hoffmann, D. (Editors). Tobacco and
Tobacco Smoke. Studies in Experimental Tobacco
Carcinogenesis. New York: Academic Press, 1967.

Wynder, E.L., Hoffmann, D. Smoking and lung
cancer: Scientific challenges and opportunities.
Cancer Research 54: 5284-5295, 1994.

Wynder, E.L., Taioli, E., Fujita, Y. Ecologic study of
lung cancer risk factors in the U.S. and Japan with
special reference to smoking and diet. Japanese
Journal of Cancer Clinics 83: 418-423, 1992.

ACKNOWLEDGMENTS The authors greatly appreciate the editorial assistance of

Use Hoffmann and Jennifer Johnting. The authors' work cited in this article
was supported by NCI grant CA-29580.

Chapter 4

I A^ttitudes, Knowledge, and Beliefs About
j Low-Yield Cigarettes Among Adolescents and
^ Adults

ii

Gary A. Giovino, Scott L. Tomar, Murli N. Reddy, John P. Peddicord,
Bao-Ping Zhu, Luis G. Escobedo, and Michael P. Eriksen

I

, INTRODUCTION Per capita consumption of cigarettes in the United States increased
1 rapidly from 1900 to 1963 (Miller, 1981; U.S. Department of Health and

I Human Services, 1989); however, since the January 1964 release of the

I first Surgeon General's report on smoking (U.S. Department of Health,

Education, and Welfare, 1964), cigarette consumption has been declining
(Miller, 1981; U.S. Department of Agriculture, 1987 and 1994). In 1994 per
capita consumption was about the same as during World War II (Miller, 1981;
U.S. Department of Agriculture, 1994). However, the prevalence of smoking
I was slightly higher in the 1940's (Centers for Disease Control and Prevention,

I 1994a; U.S. Department of Health and Human Services, 1988), indicating

that smokers in the 1990's consumed more cigarettes per day than did
smokers in the 1940's (Harris, 1994; U.S. Department of Health and
Human Services, 1980).

Falls in per capita consumption of cigarettes seem linked to health
concerns. For example, in the early 1950's, scientific and popular articles
led to increasing concern about smoking-related cancers. American and
British studies provided a scientific foundation for the mounting health
concerns (Doll and Hill, 1950 and 1952; Levin et al., 1950; Wynder and
Graham, 1950). Articles such as "Cancer by the Carton," published in
the Reader's Digest (Norr, 1952), also carried the message to many people
(U.S. Department of Health and Human Services, 1989).

One apparent result of these early health communications was the
marked increase in the consumption of filter-tipped cigarettes. In the 1940's
few people smoked those varieties (U.S. Department of Agriculture, 1962),
but by 1992 about 97 percent of cigarettes sold had filters (Figure 1) (Federal
Trade Commission, 1994). Switching to filtered cigarettes was promoted by
slogans such as "Kent with the micronite filter is smoked by more scientists
and educators than any other cigarette" (Anonymous, 1985).

The release of the first Surgeon General's report on smoking was a
major turning point in public perception of the health threat of tobacco
(U.S. Department of Health, Education, and Welfare, 1964; U.S. Department
of Health and Human Services, 1989). In response, cigarette companies
began introducing cigarettes in the 1960's and early 1970's that yielded.

39

Figure 1

Domestic market share of all filter-tipped cigarettes and those filter-tipped cigarettes
yielding si 5 mg tar: United States, 1946-1992

Smoking and Tobacco Control Monograph No. 7

}U33Jad

o

o

in

o

LO

in

CM

o

(0

0)

}uaojdd

40

Sources: U.S. Department of Agriculture, 1962; Federal Trade Commission, 1994.

j Chapter 4

by the Federal Trade Commission (FTC) method (Pillsbury et al., 1969),

15 mg or less tar (Federal Trade Commission, 1994; Slade, 1989; Warner,
1985). By 1992 these so-called milder cigarettes had captured about
69 percent of the market (Federal Trade Commission, 1994).

The lower tar cigarettes were accompanied by advertisements such
as the following:

Vantage is changing a lot of my feelings about smoking. I like
to smoke, and what I like is a cigarette that is not limited on
taste. But 1 am not living in an ivory tower. 1 hear the things
being said about high tar smoking as well as the next guy. So,

1 started looking for a low tar smoke that had some honest-to-
goodness taste (Anonymous, 1977).

It is believed that the Vantage advertisements targeted "intelligent"
smokers (Pollay, 1990).

Since 1974, FTC has collected data on advertising and promotion of
cigarettes yielding 15 mg or less tar (Figure 2) (Federal Trade Commission,
1994). As pointed out by Davis (1987), for many years the proportion of
advertising and promotional expenditures for lower tar cigarettes exceeded

Figure 2

Domestic market share and proportion of total advertising and promotional expenditures
related to cigarettes yielding si 5 mg tar, by year: United States, 1975-1992

Source: Federal Trade Commission, 1994.

Smoking and Tobacco Control Monograph No. 7

market share, suggesting an attempt to increase market share. As shown in
Figure 2, the two proportions are converging. In 1992 lower tar cigarettes
accounted for 69 percent of market share and 71 percent of advertising
expenditures (Federal Trade Commission, 1994).

One major purpose of the marketing of these varieties of cigarettes
appears to have been to alleviate smokers' health concerns (Pollay, 1990;
Warner 1985). The advertisements seem to have achieved a large part of
their goal. In 1993 a Gallup Organization poll posed the following question:
"Besides selling the product, what message do you think cigarette advertising
is trying to get across when it uses terms like low tar, low nicotine, or low
yield?" (Gallup Organization, Inc., 1993, pp. 22). Fifty-eight percent of
respondents (56 percent of smokers and 60 percent of nonsmokers) answered
that the message indicates a positive health benefit, that is, that the brand
is safer, healthier, less harmful, not as bad for you, or less cancerous (Gallup
Organization, Inc., 1993).

MONITORING Three national surveys helped shed light on the patterns in j

NATIONAL DATA attitudes, knowledge, and beliefs about low-yield cigarettes: i

the 1986 Adult Use of Tobacco Survey (AUTS), the 1987 National Health
Interview Survey (NHIS) Cancer Control Supplement, and the 1993 Teenage
Attitudes and Practices Survey (TAPS). The 1986 AUTS was a national
telephone survey of approximately 13,000 Americans ages 17 years and '

older (Pierce et al., 1990). The nationally representative sample of the 1987 f
NHIS included about 22,000 Americans ages 18 years and older who were j
interviewed primarily in their homes (Schoenborn and Boyd, 1989). The I
1993 TAPS sample included about 13,000 people 10 to 22 years of age !

who were contacted via telephone or in their homes (Centers for Disease j
Control and Prevention, 1994b). The 1993 TAPS included a cross-sectional |
component of persons 10 to 15 years of age in 1993 and a followup j

component of a cohort of persons first interviewed in 1989 who were 15 !

to 22 years old in 1993 (Centers for Disease Control and Prevention, 1994b).

There are difficulties in using the 1993 TAPS data to make prevalence |
estimates. Some participants lost to followup were more likely to be
smokers in 1989, a phenomenon that would be likely to decrease the overall !|
prevalence estimate (Centers for Disease Control and Prevention, 1994b). j
The data used for this report are not used to generate smoking prevalence
estimates; rather, they look at characteristics of persons who reported that '
they were currently smoking. !i

I

'Fhe 1986 AU'FS and the 1987 NHIS questions used to determine tar :
levels assessed items such as brand name, filter vs. nonfilter, pack hardness, f
cigarette length, mentholation, and if the cigarette was regular, light, or
ultralight. Fhe tar level assigned is based on responses to the questions using
FTC tables (Federal Trade Commission, 1985). The tar categories used for i
this report are (1 ) less than or equal to 6 mg, (2) 7 to 15 mg, and (3) 16 mg
or more. ( The actual cutpoints used here are 6.99 mg and 15.99 mg.) ^

S

42

Chapter 4

SURVEY FINDINGS The percentage distribution of tar yield of the usual brand of

cigarettes smoked among current smokers by sex and age is
Use of Low-Tar shown in Figure 3. Female smokers were more likely to smoke

or Light Cigarettes lower tar yield brands than men. Smokers 18 to 24 years of age
were less likely to use the lower tar brands than smokers ages 25 to 44 or 45
to 64. These patterns were similar to those found by the AUTS for both
current and former smokers.

With regard to race and ethnicity (Figure 4), white Americans who
smoked in 1987 were more likely to smoke lower tar and nicotine
cigarettes (76.8 percent) than Hispanics (67.8 percent) or black Americans
(52.4 percent). Education is a strong correlate of smoking cigarette brands
with 15 mg or less tar (Figure 4). Beginning with persons who have
completed 9 to 1 1 years of education, as education increased, smokers
were more likely to smoke low-tar brands.

In the 1993 TAPS, adolescents and young adults who smoked and
usually bought their own cigarettes were asked what brands they smoked.
Furthermore, they were asked, "Is the brand you smoke regular, light, or

Figure 3

Prevalence (by percent) of current smokers' use of cigarette brands^ with 5 mg tar,
by sex and age: Ages 18 and older, United States, 1987

■ s6 mg tar □ 7-1 5 mg tar

c

0)

o

0>

Q.

80-

60-

40-

20-

Male Female

Sex

1 8-24 25-44 45-64

Age (years)

65-f

78.8

® Self-reported usual brand.

Source: National Center for Health Statistics, 1987.

43

Percent

Smoking and Tobacco Control Monograph No. 7

Figure 4

Prevalence (by percent) of current smokers' use of cigarette brands® with ^15 mg tar,
by race and education: Ages 18 and older, United States, 1987

Race/Ethnicity Education (years completed) |

" Self-reported usual brand. I

Source: National Center for Health Statistics, 1987.

ultralight?" Of note, "light" and "ultralight" are terms used in advertising
and may not correlate precisely with tar and nicotine levels (Davis et al.,

1990). However, these are terms that smokers frequently use in describing i
the brands they smoke. |

I'here are two key findings from the TAPS data. First, among 10- to
18-year-olds and 19- to 22-year-olds, females were more likely than males ^
to smoke light and ultralight cigarettes (Figure 5). Very few males smoked '
ultralight cigarettes. Second, the proportion of males and females using these
brands increased with age. This pattern among young persons (increasing ^
use of light and ultralight brands with increasing age) is reflected in both jj
the 1 987 NHIS and the 1993 FAPS.

),

The 1993 TAPS race and ethnicity findings are similar to those detected I,
by the NHIS: White youth were most likely to smoke light cigarettes j

(52.6 percent), followed by Hispanic youth (44.5 percent), with much j

smaller pro[)ortions of black youth (15 percent) reporting use of these brands !
(Figure 6). Anecdotal evidence also indicates that African-American youth j
begin with higher tar cigarettes (Galkq) International Institute, 1992).

I

44

«

I

I Chapter 4

;i

I

' Figure 5

! Prevalence (by percent) of use of light and ultralight cigarettes among current smokers/
by sex and age: Ages 10 to 22, United States, 1993

Male Female Male Female

1 0- to 1 8-Years-Old 1 9- to 22-Years-Old

® Who usually buy their own cigarettes.

Source: Centers for Disease Control and Prevention, 1993.

In the 1993 TAPS, students were asked to rate how well they were doing *

in school; the categories used here were less than average, average, better '

than average, and much better than average. The percentage of young
smokers who smoked light or ultralight cigarettes increased with level of
performance in school: from 30 percent for those who performed less than
average to 66 percent for those who performed much better than average
(Figure 6).

Brand Switching Brand switching is one measure of the perceived health risk

associated with lower tar yield cigarettes. The 1986 AUTS asked the following
question of current smokers: "Thinking of your entire smoking history,
have you ever switched from one cigarette to another, just to reduce the
amount of tar and nicotine?" Former smokers were asked, "Did you ever
switch from one type of cigarette to another just to reduce the amount
of tar and nicotine?" Approximately 38 percent of current smokers and
26 percent of former smokers answered "Yes."

The 1987 NHIS asked current smokers, "Have you ever switched to a low
tar and nicotine cigarette just to reduce your health risk?" About 44 percent
of current smokers answered that they had switched for that reason. As

45

Smoking and Tobacco Control Monograph No. 7

Figure 6

Prevalence (by percent) of use of light and ultralight cigarettes among current smokers,^
by race/ethnicity and school performance: Ages 10 to 22 years. United States, 1993

■ Ultralights □ Lights

66.4

White Black Hispanic < Average Average > Average » Average

Race/Ethnicity School Performance

* Who usually buy their own cigarettes.

Source: Centers for Disease Control and Prevention, 1993.

shown in Figures 7 and 8, there are clear trends and differences by sex, age,
race/ethnicity, and education. Figure 7 shows that females (48.4 percent)
were more likely to switch than males (39.4 percent). Smokers in the 25-to-
44 and 45-to-64 age groups were most likely to have switched to lower yield
brands (45.2 and 45.9 percent, respectively), followed by smokers older
than age 64 (41.3 percent) and those 18 to 24 years (36.4 percent). Figure 8
shows that whites (47 percent) were more likely to switch than Hispanics
(30.9 percent) or African-Americans (30.8 percent), and the more educated
were more likely to switch than the less educated.

Smokers of low-tar yield varieties were more likely to have switched.

That is, among smokers consuming brands yielding 6 mg or less tar,

74 percent of current smokers in the 1986 AUTS had ever switched compared
with 19 percent of smokers consuming cigarettes yielding 16 mg or more
tar. Fhese patterns were similar for both former smokers (as reported by
the AUTS) and current smokers (as reported by the NlllS).

Persons who switched brands were more likely to smoke low-tar yield
brands. For example, according to the 1986 AUTS, 22 percent of switchers
smoked brands yielding 6 mg or less tar compared with 5 percent of people

46

Chapter 4

Figure 7

Percentage of current smokers who have ever switched brands,^ by sex and age:
Ages 18 and older, United States, 1987

60 r

50 — 48.4

Male Female 18-24 25-44 45-64 65+

Sex Age (years)

® To lower tar/nicotine brands to reduce their health risks.

Source: National Center for Health Statistics, 1987.

who had never switched. This suggests that many smokers switch to lower
tar brands rather than starting with those brands.

HEALTH BELIEFS Survey data on health beliefs shed light on possible factors that
AND SWITCHING may drive or influence smokers' switching to lower tar cigarette
brands. The surveys indicate that current smokers of lower tar brands and
persons who had switched brands were more likely to acknowledge health
risks than those who smoked higher tar brands or who had not switched
brands. Figures 9 and 10 illustrate this relationship between tar yield of the
smoker's brand and beliefs that smoking is related to cancer and emphysema.

It is worth pointing out that the majority of smokers of high-tar
cigarettes, as well as smokers who have never switched, acknowledged the
health risks of smoking (Figure 10). However, there is an inverse gradient
for both variables.

Similarly, concerns about health risks decrease as tar yields rise (Table 1).
Among smokers who switched brands, 85 percent stated that they were

47

Smoking and Tobacco Control Monograph No. 7

Figure 8

Percentage of current smokers who have ever switched brands/ by race/ethnicity and
education: Ages 18 and older. United States, 1987

White Black Hispanic 0-8 9-11 12 13-15 16+

Race/Ethnicity Education (years completed)

* To lower tar/nicotine brands to reduce their health risks.
Source: National Center for Health Statistics, 1987.

concerned about the health effects of smoking compared with 70 percent of
those who had never switched. Furthermore, people in the lower tar yield
categories and those who switched were more likely to respond that their
health had been affected by their smoking, and they were more likely to
report that a doctor had advised them to quit.

Moreover, people who smoke low-tar cigarettes and those who switched
were more likely to acknowledge that some brands are more hazardous than
others (Table 1). Smokers of low-tar brands were more likely to state that
their brand is less hazardous compared with smokers of higher tar brands.
Among switchers, 33 percent believed that their brand is less hazardous than
other brands. For smokers who had never switched, only 16 percent held
this belief.

In the 1993 TAPS, adolescents and young adults who smoked light and
ultralight cigarettes were asked why they smoked those brands. Four reasons
were most commonly cited: Thirty-three percent of respondents said that

48

Chapter 4

Figure 9

Percentage of current smokers who believe that low-tar cigarettes pose reduced cancer risk,
by tar yield and history of switching: Ages 18 and older, United States, 1987

Tar Yield (mg) History of Switching

Source: National Center for Health Statistics, 1987.

they smoked light or ultralight cigarettes because they taste better, 29 percent
said they are less irritating, 21 percent said they thought these cigarettes
were healthier than other brands, and 19 percent said they "just liked them."

QUITTING SMOKING The surveys revealed some interesting trends with regard to
quitting. In the 1987 NHIS, participants were asked to identify techniques
they had used in their efforts to quit smoking. Among participants who had
switched brands, 38 percent said they had ever switched to lower tar and
nicotine cigarette brands as a quitting strategy; 62 percent switched for other
reasons (Table 2). Switchers were more likely to have tried these quitting
strategies, with the exception of quitting cold turkey, than smokers who had
never switched. This suggests that switchers were seeking help with quitting.
In addition, those who smoked lower tar cigarettes were slightly more likely
to have sought help during previous quit attempts than were persons who
smoked higher tar cigarettes.

However, the data from the 1986 AUTS indicate that the prevalence of
cessation increases with increasing tar yield (Figure 11). That is, ever-smokers
who smoked higher tar yield brands were more likely to have quit than
people who smoked lower tar brands. Respondents who had never switched
were more likely to have quit smoking than switchers.

49

Smoking and Tobacco Control Monograph No. 7

Figure 10

Percentage of current smokers who believe that cigarette smoking is related to emphysema,
by tar yield^ and history of switching*’: Ages 18 and older. United States, 1987

100 r

s6 mg 7-15 mg alBrng Ever Never

Tar Yield (mg) History of Switching

* Of current brand.

“ Ever switched to lower tar/nicotine brands to reduce health risks.
Source: National Center for Health Statistics, 1987.

Table 1

Health concerns and beliefs of current smokers, by tar yield and history of switching^,
by percent: Ages 17 and older. United States, 1986

History of

Tar Yield (mg) Switching

<6

7-15

>16

Ever

Never

Concerned About Health Effects

84

79

68

85

70

Some Brands More Hazardous Than Others

60

46

39

54

40

Their Brand Is Less Hazardous Than Others

48

26

12

33

16

• Ever switched to reduce tar and nicotine.
Source: Centers for Disease Control, 1986.

f

^ Chapter 4

i

Table 2

Quit strategies ever used by current smokers, by tar yield and history of switching®, by percent:
Ages 18 and older. United States, 1987

Tar Yield (mg)

History of
Switching

<6

7-15

>16

Ever

Never

Switch to Low Tar

37

22

18

38

6

Special Filters

14

9

8

13

4

Gradual Reduction

39

34

36

42

27

Nicotine Gum

16

10

10

12

8

The Great American Smokeout

10

9

8

12

6

Cold Turkey

86

84

82

82

85

Book/Pamphlet

9

9

7

10

5

Relatives/Friends

18

18

18

20

13

I ^ Switching to lower tar and nicotine brand to reduce health risks.
Source: National Center for Health Statistics, 1987.

Among persons who had ever been regular smokers, those who smoked
low-tar cigarettes and those who switched to lower tar brands were more
likely to have made a recent effort to quit smoking and relapsed and were
less likely to be former smokers (data not shown). Among smokers who had
never tried to quit, smokers of low-tar cigarettes and those who switched to
low-tar cigarettes were more likely to have considered quitting (data not
shown).

DISCUSSION These data seem to reflect an interplay of the forces of motivation to

quit and nicotine dependence (Russell, 1981). Smokers of lower tar cigarettes
appear to be especially interested in quitting and are more actively seeking
help than smokers of higher tar cigarettes. Perhaps when lower tar smokers
were unsuccessful in their attempts to quit, they switched to a lower tar brand
to allay their fears about the health consequences of continuing to smoke.

The tacit health claims associated with advertisements of the lower tar brands
may have allayed smokers' health concerns (Davis, 1987). Because of the
cross-sectional nature of the data, however, further research on the topic is
warranted.

Not all switching is a step toward quitting. Three of every five smokers
who had ever switched to lower tar and nicotine brands did not do so as a
quitting strategy. Both low-tar cigarette smokers and ever-switchers were
more likely, compared respectively with high-tar smokers and persons who
had never switched brands, to (1) acknowledge the dangers of smoking.

51

Smoking and Tobacco Control Monograph No. 7

Figure 1 1

Prevalence of cessation among ever-smokers, by tar yield and history of switching®:
Ages 17 and older, United States, 1986

Tar Yield (mg) History of Switching

^ Ever switched to reduce tar/nicotine.

Source: Centers for Disease Control, 1986.

(2) say that their health has been affected, (3) be concerned about health
effects, and (4) believe that their cigarettes are safer.

The data on prevalence of cessation are especially intriguing, given
that low-tar cigarette smokers and ever-switchers are better educated and
it is known that persons with more years of education are less likely to be
smokers and more likely to have quit (Centers for Disease Control and
Prevention, 1994a; Giovino et al., 1994). These data and the Pollay (1990)
observation that the tobacco industry seems to be targeting lower tar yield
cigarettes toward more highly educated smokers deserve consideration. I'he
innovation of quitting smoking, which started among persons with more
education, may have been replaced by the innovation of switching to lower
tar brands (Rogers, 1983; U.S. Department of Health and Human Services,
1989).

52

Chapter 4

As stated by Samet (this volume), available evidence indicates that
smoking lower tar cigarettes only minimally reduces smokers' health risks.
The reduced prevalence of cessation among smokers who have switched
brands and smokers of low-yield cigarettes, coupled with beliefs among some
in the public that these cigarettes are safer, suggest that low-yield cigarettes
'I have kept many smokers smoking who otherwise might have quit. The net

1 effect of the introduction and mass marketing of these brands, then, may

j have been and may continue to be an increased number of smoking-

attributable deaths.

j QUESTION-AND- ANSWER SESSION

I

! DR. SHIFFMAN: I don't know if you have these data, but 1 am wondering,

! when smokers in these surveys make a deliberate switch, do you have a sense

j of how big a jump they make in the FTC tar and nicotine values?

1

DR. GIOVINO: We have begun to look at the issue of the penultimate brand
! vs. the current or the last brand. The reason I can't give you a direct answer

is because we looked at it as a function of whether or not they smoke more
now or less now. And 1 will have to check this, but I think it was about
.2 mg nicotine.

DR. HOFFMANN; We know now that nicotine is one major reason that
people smoke or chew tobacco. Therefore, you could have classified your
groupings according to nicotine, which I would have done, because that is
why people smoke; it is not for the tar.

DR. GIOVINO: 1 think the analysis clearly could be done both ways, and
I understand your reasoning. The reason that I felt comfortable with tar is
because it is based on perceptions. A lot of this is based on perceptions of
health risks.

My guess is that they are so highly correlated that the analysis would
find very similar findings, and if the committee would like me to do that,
we can certainly do that.

DR. PETITTI: This is a pretty technical question, but your last slide had a
conclusion that low-tar smokers are less likely to be former smokers and
switchers are less likely to be former smokers. 1 presume those are age
I adjusted?

DR. GIOVINO: We did age-specific analyses. We did not have time to do
age-adjusted analyses. We used three age categories: 17 to 34 years, 35 to
64 years, and 65 plus years. For switching, the relationship held in every
category; for low tar, it held in every category except the 17 to 34 category.

DR. BENOWITZ: I wonder if you have any information on smoking of the
really ultralows, like 1 mg and below, because there is some evidence that
the yields from those are really fundamentally different, and I will be talking
about that later. But do you Imow anything about the characteristics of
those smokers?

53

Smoking and Tobacco Control Monograph No. 7

DR. GIOVINO: The numbers in those categories became very small. You
know, at 6 mg or less, it was 10 or 12 percent. At 1 mg, the numbers would
have been ....

DR. BENOWITZ: So, no one is smoking them.

DR. GIOVINO: Very small numbers, yes.

DR. FREEMAN: Do you have any guess or reason why young black males
in particular are smoking so much less today, since it is obviously not a
function of education. Do you have any sense of why that is happening?

DR. GIOVINO: What Dr. Freeman is referring to are the trends in the High
School Senior data, in National Health Interview Survey data among people
18 to 24, in the National Household Survey on Drug Abuse data, the Youth
Risk Behavior Survey data, the TAPS data, and others, that show that African-
American youth are much less likely to smoke than white youth.

1 will take 2 minutes, because it is an interesting study. It is not a school
dropout effect, because when we look at dropouts, white kids who have
dropped out are much more likely to smoke than African-American kids.
Also, regardless of race and ethnicity, all kids who drop out are more likely
to smoke.

We don't believe that it is because they have switched to other drugs.

We have looked at Monitoring the Future data, and it does not look like
cigarette smoking has been replaced by an increased use of alcohol and
other drugs.

There are some data to suggest that differential misclassification may
explain some of the difference. There was a paper by Karl Bauman in the
American Journal of Public Health that showed that African-American youth
may be a little more likely to differentially underreport in a household
survey. Household surveys pose the most serious concerns about
confidentiality, unless serious steps are taken to protect confidentiality.

We see lower smoking rates among blacks in school surveys, where
there is greater privacy. And even in Bauman's household survey, mean
validated tobacco use was three times higher in white youth than in African-
American youth.

To answer your question in more detail, variables like discretionary
income, parental education, importance of religion, and how well they do
in school do not explain it. In other words, the trends seem to be down in
African-Americans more than white youth in just about all the subcategories
that we have carved out.

There are ex[)lanations, and some were presented in the 1994 Surgeon
General's report: There have been changes in attitudes about smoking,
and the attitudes held by African-American youth changed in a much more
health-promoting direction than the attitudes among white youth. There

54

Chapter 4

appears to have been some sort of social climate change, such that cigarette
smoking does not appear to be as socially acceptable among African-
Americans; there are certainly some reports of grassroots involvement at
the church and other levels.

There also appears to be a differential concern about the potential
weight-controlling effects of cigarettes, with African-American youth being
less obsessed with slimness than white youth.

It is a very intriguing phenomenon and one that we have examined in
detail.

DR. FREEMAN: Is this reflected in the 18- to 24-year-old group?

DR. GIOVINO: The prevalence trends have definitely translated into the
18- to 24-year-old age group, and even in the 25 to 29 age group. African-
Americans start smoking about a year later in life, but the differences we
are seeing are not enough, and we are definitely seeing translation into the
young adult population.

DR. STITZER: One more question on the youth. Your data seem to
contradict the popular wisdom that youth begin with light cigarettes.

I wondered if there were any data suggesting that they do play some role
in initiation or original experimental use?

DR. GIOVINO: Some of that dogma, if I understand it right, is that it might
have influenced young girls starting because they were less irritating, and
that seems to be part of the scenario. Young girls are more likely to have
used the lights or the ultralights, to the extent that the cross-sectional data
can tell us exactly.

I find myself thinking this, and again, this is hypothesis generation:

You see a lot more ads for regular cigarettes than you do for light cigarettes,
especially if you think about Marlboros, Camels, Newports, etc. Regardless
of the reason, it is possible that they start on the regulars, that the thought
of quitting occurs to them, they have difficulty quitting and the thought is,
"Well, I have got to do something here, so maybe I will switch." It is a
hypothesis.

DR. KOZLOWSKI: A number of years ago, Fred Silverstein, Scott Feldon, and
I published a paper in the Journal of Health and Social Behavior on the role of
low-yield cigarettes and the recruitment to smoking, particularly in women,
we found, in a school sample.

And you have to think that there were some young women who were
particularly sensitive to the effects of smoking. Not all were. In other words,
a small percentage of the market were under great social pressure to take up
smoking, and the low-yield cigarette, smoked without vent blocking and so
on, provided a nice trial-sized dose. So, it helped some people, but it was not
across the board.

55

Smoking and Tobacco Control Monograph No. 7

REFERENCES

Anonymous. Health claims made for filter cigarettes.
New York State Journal of Medicine 85(7); 317, 1985.

Anonymous. Vantage is changing a lot of my feelings
about smoking. Time Magazine, November 7, 1977.
pp. 86-87.

Centers for Disease Control and Prevention. Adult
Use of Tobacco Survey. Public use data tape, 1986.

Centers for Disease Control and Prevention. Teenage
Attitudes and Practices Survey. Public use data tape,
1993.

Centers for Disease Control and Prevention. Cigarette
smoking among adults — United States, 1993.
MMWR. Morbidity and Mortality Weekly Report 43:
925-930, 1994a.

Centers for Disease Control and Prevention. Changes
in cigarette brand preferences of adolescent
smokers — United States, 1989-1993. MMWR.
Morbidity and Mortality Weekly Report 43: 577-581,
1994b.

Davis, R.M. Current trends in cigarette advertising and
marketing. New England Journal of Medicine 316: 725-
732, 1987.

Davis, R.M., Healy, P., Hawk, S.A. Information on tar
and nicotine yields on cigarette packages. American
Journal of Public Health 80: 551-553, 1990.

Doll, R., Hill, A.B. Smoking and carcinoma of the lung.
British Medical Journal 2: 740-748, 1950.

Doll, R., Hill, A.B. A study of the aetiology of
carcinoma of the lung. British Medical Journal 2:
1271-1286, 1952.

Federal Trade Commission. "Tar," Nicotine and Carbon
Monoxide of the Smoke of 207 Varieties of Domestic
Cigarettes. Washington, DC: Federal Trade
Commission, 1985.

Federal Trade Commission. Federal Trade Commission
Report to Congress for 1 992. Washington, DC: Federal
Trade Commission, 1994.

Gallup International Institute. Teen-age Attitudes and
Behavior Concerning Tobacco. Report of the Focus
Groups. Princeton, NJ: The George H. Gallup
International Institute, 1992.

Gallup Organization, Inc. The Public's Attitudes Toward
Cigarette Advertising and Cigarette Tax Increase.
Princeton, NJ: The Gallup Organization, Inc., 1993,
pp. 22-24.

Gioviiio, G.A., Schooley, M.W., Zhu, B.P., Chrismon,
J.H., lomar, S.l,., Peddicord, J.P., Merritt, R.K.,
Husten, C.G., Friksen, M.P. Surveillance for selected
tobacco-use behaviors — United States, 19(KJ-1994.
MMWR. Morbidity and Mortality Weekly Report 43(3):
1-43, 1994,

Harris, j . A. Working Model for Predicting the

Consumption and Revenue Impacts of Large Increases iti
the U.S. Federal Cigarette Excise Fax. Working |)aj)er
no. 4803. (Cambridge, MA; National Bureau of
Fxonornic Research, 1994.

Levin, M.L., Goldstein, H., Gerhardt, P.R. Cancer and
tobacco smoking; A preliminary report. Journal of
the American Medical Association 143: 336-338, 1950.

Miller, R. U.S. cigarette consumption, 1900 to date. In:
Tobacco Yearbook, W. Harr (Editor). Bowling Green,
KY: Cockrel Corporation, 1981, p. 53.

National Center for Health Statistics. National Health
Interview Survey Cancer Control Supplement.
Public use data tape, 1987.

Norr, R. Cancer by the carton. Reader's Digest
December: 7-8, 1952.

Pierce, J.P., Hatziandreu, E.H., Flyer, P., Hull,].,

Maklan, D., Morganstein, D., Schreiber, G. Report of
the 1986 Adult Use of Tobacco Survey. OM 90-2004.
Rockville, MD: U.S. Department of Health and
Human Services, Public Health Service, Centers for
Disease Control, Office on Smoking and Health,
1990.

Pillsbury, H.C., Bright, C.C., O'Connor, K.J., Irish,

F.W. Tar and nicotine in cigarette smoke. Journal
of Official Analytical Chemists 52: 458-462, 1969.

Pollay, R.W. "The Functions and Management of
Cigarette Advertising." Unpublished manuscript.
University of British Columbia, 1990. 20 pp.

Rogers, E.M. (Editor). Diffusion of Innovations. New
York: Free Press, 1983.

Russell, M.A. Dependence and motivation to stop
smoking. In: Smoking and Arterial Disease, R.M.
Greenhalgh (Editor). Bath, United Kingdom;

Pitman Medical, 1981, pp. 285-288

Schoenborn, C.A., Boyd, G. Smoking and Other Tobacco
Use: United States, 1 987. Vital and Health Statistics.
Series 10: Data From the National Health Interview
Survey (No. 169). DHHS Publication No. (PHS) 89-
1597. Rockville, MD; U.S. Department of Health
and Human Services, Public Health Service, Centers
for Disease Control, National Center for Health
Statistics, 1989.

Slade, j. The tobacco epidemic: Lessons from history.
Journal of Psychoactive Dntgs 21 : 281-291, 1989.

U.S. De[)artment of Agriculture. Tobacco Situation.
rs 99. Washington, DC: U.S. Department of
Agriculture, Economic Research Service, 1962.

U.S. Department of Agriculture. Tobacco Situation and
Outlook Report. TS 200. Washington, DC: U.S.
Department of Agriculture, Commodity Economics
Division, Economic Research Service, 1987.

U.S. Department of Agriculture. I'obacco Situation and
Outlook Rejwrt. TS 226. Washington, DC: U.S.
Department of Agriculture, Commodity Economics
Division, Economic Research Service, 1994.

U.S. Department of Health and Human Services. The
Health Consequences of Smoking for Women. A Rejwrt
of the Surgeon General. Rockville, MI): U.S.
Dejrartment of Health and Human Services, Public
Health .Service, Office of the A.ssistant .Secretary for
Health, Office on Smoking and Health, 1980.

56

Chapter 4

U.S. Department of Health and Human Services. The
Health Consequences of Smoking: Nicotine Addiction. A
Report of the Surgeon General, 1988. DHHS
Publication No. (CDC) 88-8406. Rockville, MD: U.S.
Department of Health and Human Services, Public
Health Service, Centers for Disease Control, Center
for Health Promotion and Education, Office on
Smoking and Health, 1988.

U.S. Department of Health and Human Services.
Reducing the Health Consequences of Smoking: 25
Years of Progress. A Report of the Surgeon General,
1989. DHHS Publication No. (CDC) 89-8411.
Rockville, MD: U.S. Department of Health and
Human Services, Public Health Service, Centers for
Disease Control, Center for Chronic Disease
Prevention and Health Promotion, Office on
Smoking and Health, 1989.

U.S. Department of Health, Education, and Welfare.
Smoking and Health: A Report of the Advisory
Committee to the Surgeon General of the Public Health
Service. PHS Publication No. 1103. Rockville, MD:
U.S. Department of Health, Education, and
Welfare, Public Health Service, 1964.

Warner, K.E. Tobacco industry response to public
health concern: A content analysis of cigarette ads.
Health Education Quarterly 12: 115-127, 1985.

Wynder, E.L., Graham, E.A. Tobacco smoking as a
possible etiological factor in bronchogenic
carcinoma. A study of six hundred and eighty-four
proved cases. Journal of the American Medical
Association 143: 329-336, 1950.

KS

w

r

c

57

■

1

1

■t

I

Chapter 5

Cigarette Smoke Components and Disease:
Cigarette Smoke Is IVIore Xhan a Xriad of Xar,
Nicotine, and Carbon IVIonoxide

Jeffrey E. Harris

INTRODUCTION Cigarette smoke is a complex mixture of chemicals. Some smoke
components, such as carbon monoxide (CO), hydrogen cyanide (HCN), and
nitrogen oxides, are gases. Others, such as formaldehyde, acrolein, benzene,
and certain N-nitrosamines, are volatile chemicals contained in the liquid-
vapor portion of the smoke aerosol. Still others, such as nicotine, phenol,
polyaromatic hydrocarbons (PAHs), and certain tobacco-specific nitrosamines
(TSNAs), are contained in the submicron-sized solid particles that are
suspended in cigarette smoke.

In view of this chemical complexity, cigarette smoke has multiple,
highly diverse effects on human health. It is not unexpected that multiple
chemicals in cigarette smoke can contribute to any single adverse health
effect.

Thus, HCN may affect the human respiratory system by its toxic effects
on the cilia that line the respiratory tract. At the same time, HCN may cross
the placenta and have toxic effects on the growing fetus. In addition, HCN
also may cause nerve damage in cigarette smokers with optic neuropathy
(Costagliola et al., 1989). Although the PAHs and TSNAs in the particulate
phase of cigarette smoke are known carcinogens, catechols and phenols in
the particulate phase also are considered carcinogens or tumor promoters.
Benzene and formaldehyde in the liquid-vapor portion of the smoke also
may be carcinogenic.

Aside from specific chemical constituents, certain physical-chemical
properties of smoke may participate in disease processes. Thus, the pH of
the smoke may affect the site and degree of nicotine absorption as well as the
smoker's depth of inhalation. The oxidation-reduction state of the smoke can
be important because oxidants influence the maturing of cholesterol-laden
plaques in the coronary arteries and other blood vessels. In short, cigarette
smoke is far more than a triad of tar, nicotine, and carbon monoxide. This
fact needs to be considered carefully in any discussion of the adequacy of
current cigarette testing methods or current cigarette labeling practices.

MAINSTREAM VS. Both smokers and nonsmokers can incur adverse health
SIDESTREAM effects from the smoke of burning cigarettes. Smokers inhale

CIGARETTE SMOKE mostly mainstream (MS) smoke, which is drawn through

the burning tobacco column and filter tip and exits through the mouthpiece
of the cigarette. Nonsmokers inhale mostly sidestream (SS) smoke, which is

59

Smoking and Tobacco Control Monograph No. 7

emitted into the surrounding air between puffs from the end of the smolder-
ing cigarette. Sidestream smoke is the major source of environmental tobacco
smoke (ETS).

Although SS and MS smoke have qualitatively similar chemical
compositions, the respective quantities of individual smoke constituents
can be quite different (U.S. Department of Health and Human Services,

1987 and 1989). For example, in studies of nonfilter cigarettes smoked
by machines, the yield of CO in undiluted SS smoke was 2.5- to 4.7-fold
that of MS smoke, whereas the corresponding SS/MS ratio for N-
nitrosodimethylamine (NDMA), an animal carcinogen, was 0.2
(U.S. Department of Health and Human Services, 1989). In one compilation
of toxic and tumorigenic agents in cigarette smoke, the SS/MS ratio ranged
from 0.03 to 130 (Hoffmann and Hecht, 1990). In another study, the
concentration of the carcinogen 4-aminobiphenyl in undiluted SS smoke
was 32-fold that of MS smoke. The SS smoke from so-called reduced-yield
cigarettes does not necessarily have reduced emissions of toxic and
carcinogenic chemicals (Adams et al., 1987; Rando et al., 1992).

Whereas exposure to SS smoke depends on the distance from the
burning cigarette and conditions of ventilation, the higher concentrations
of certain toxic and carcinogenic chemicals in SS smoke result in measurable
levels of these chemicals in nonsmokers exposed to ETS. For example,
nonsmokers exposed to relatively high concentrations of SS smoke have
detectable urinary levels of the metabolites of the tobacco-specific
nitrosamine 4-(methylnitrosamine)-l-(3-pyridil)-l-butanone (NNK)

(Hecht et al., 1993). Young children exposed to ETS via their smoking
mothers have detectable levels of PAH-albumin adducts in their blood
(Crawford et al., 1994).

Exposures to specific chemical agents in ETS can in turn produce
pathological effects in humans and in animal models. The CO in SS smoke
reduces the blood's ability to deliver oxygen to the heart, an effect that is
especially important in patients with coronary heart disease (CHD) (Sheps et
al., 1990). Secondhand cigarette smoke activates blood platelets, which in
turn play a role in the development of atherosclerotic plaques in CHD
(Glantz and Parmley, 1995).

I he remainder of this chapter focuses on the chemical components of
MS smoke and their health effects on cigarette smokers; however, the
components of SS smoke and their health effects on nonsmokers cannot
be ignored.

MAJOR HEALTH The major health effects of cigarette smoke include:

cancer;

noncancerous lung diseases;

atherosclerotic diseases of the heart and blood vessels; and
toxicity to the human reproductive system.

EFFECTS OF

CIGARE n E SMOKE .

60

Chapter 5

CANCER

Other health effects of cigarette smoke, such as retardation of healing of
peptic ulcers and interaction with certain therapeutic drugs, are not
considered in detail here.

The epidemiologic evidence on the degree (if any) to which filter-tipped
and low-tar cigarettes have reduced the risks of smoking-related diseases are
reviewed by Samet (this volume).

The psychoactive drug in cigarette smoke is nicotine. Cigarette smoking
is a highly controlled form of self-administration of this drug. Nicotine use
is self-reinforcing. Attempts to stop smoking lead to craving, withdrawal
symptoms, and high rates of relapse (U.S. Department of Health and Human
Services, 1988; Harris, 1993). The psychoactive effects of nicotine are
discussed in detail in chapters by Benowitz (this volume) and Henningfield
and Schuh (this volume).

Cigarette smoking causes cancers of the lung, esophagus, larynx, oral
cavity, bladder, and pancreas in male and female smokers. In fact, cigarette
smoking is the major cause of lung cancer in the United States, accounting
for 90 percent of cases in men and 79 percent in women (U.S. Department
of Health and Human Services, 1989). Smoking is also reported to increase
the risks of cancers of the kidney, liver, anus, penis, and uterine cervix as
well as several forms of acute leukemia (Garfinkel and Bofetta, 1990;

U.S. Department of Health and Human Services, 1982, 1989, and 1990).

Numerous epidemiological studies covering the experience of millions
of men and women over many years show that smokers' risks of developing
cancer increase with the number of cigarettes smoked daily, the lifetime
duration of smoking, and early age of starting smoking. Smoking cessation
gradually reduces cancer risk, although a persistent excess risk has been
observed even two decades after cessation (U.S. Department of Health and
Human Services, 1989 and 1990). Cigarette smoke interacts with other
causative agents, including alcohol, asbestos, radon daughters, certain viruses,
and certain workplace exposures, in the development of human cancers
(U.S. Department of Health and Human Services, 1982, 1989, and 1990).

Condensates collected from cigarette smoke cause mutations and damage
to DNA (deoxyribonucleic acid) in laboratory assays of mutagenesis (Cairola,
1982) as well as malignant transformation (in laboratory tests) of a chemical's
ability to induce malignant changes in mammalian cells. The most widely
used experimental system is the mouse skin bioassay, in which cancers are
induced by the repeated application of condensates of cigarette smoke to
the shaved skins of mice.

Humans naturally puff on cigarettes. The puffed smoke, in a volume
of about 30 to 70 mL, is temporarily retained in the smoker's mouth, after
which it may be inhaled deeply into the lungs. By contrast, some laboratory
animals breath by panting, and others are obligate nose breathers. Even with
installation of smoke through artificial airways, it can be quite difficult to
get the animals to inhale deeply, as human smokers do. Accordingly, the

61

Smoking and Tobacco Control Monograph No. 7

distribution and retention of smoke components in the respiratory systems
of laboratory animals may not mimic natural human smoking. Nevertheless,
long-term smoke inhalation regularly induces tumors of the larynx in Syrian
golden hamsters. Direct installation of cigarette tar into the airways of
laboratory animals causes lung cancers (Hoffmann and Hecht, 1990;

U.S. Department of Health and Human Services, 1982).

MS cigarette smoke contains more than three dozen distinct chemical
species considered to be tumorigenic in humans or animals (Hoffmann and
Hecht, 1990; U.S. Department of Health and Human Services, 1982 and
1989). Among the most prominent are PAHs such as benzo(fl)pyrene (BaP);
aza-arenes such as dibenzo-acridine; N-nitrosamines such as NDMA; aromatic
amines such as 4-aminobiphenyl; aldehydes such as formaldehyde; other
organics such as benzene; and certain inorganic compounds such as arsenic,
nickel, and chromium. Some of these chemicals alone are capable of
initiating tumors in laboratory animals; others can promote the development
of previously initiated cancers. Still others indicate direct human
epidemiological evidence of carcinogenicity.

Certain chemical components of smoke may contribute to specific
cancers. For example, TSNAs may contribute to cancers of the lung, larynx,
esophagus, and pancreas, whereas 4-aminobiphenyl and certain aryl amines
may contribute to cancer of the bladder (Vineis, 1991). Benzene in cigarette
smoke may play a role in smoking-induced leukemia (Melikian et al., 1993).

NONCANCEROUS Cigarette smoking is the main cause of chronic obstructive lung

LUNG DISEASES disease (COLD), also called chronic obstructive pulmonary disease
(U.S. Department of Health and Human Services, 1984a). Smoking accounts
for 84 percent of COLD deaths in men and 79 percent in women (U.S.
Department of Health and Human Services, 1989).

COLD is a slowly progressive illness that develops after repeated insults
to the lung over many years. In the early years after starting to smoke, an
individual may report no symptoms. However, even at this early stage
breathing tests can often detect abnormalities in the small terminal airways
of the lung (Beck et al., 1981; Seely et al., 1971; U.S. Department of Health
and Human Services, 1984a), and these abnormalities have been directly
observed in autopsy studies of young smokers who died suddenly
(Niewoehner et al., 1974). For smokers in their twenties, there is already a
dose-response relationship between the extent of abnormal lung tests and
the number of cigarettes smoked daily. In random population surveys, from
17 to 60 percent of adult smokers younger than age 55 have detectable small
airway dysfunction (U.S. Department of Health and Human Services, 1984a).

Over the course of an individual's two decades or more of smoking, a
constellation of chronic respiratory changes develops. I'hese chronic lung
injuries include (1 ) mucus hypersecretion with chronic cough and phlegm; ,
(2) airway thickening and narrowing, resulting in obstruction to airflow I
during expiration; and (3) emphysema, that is, abnormal dilation of the air i
spaces at the end of the res[)iratory tree, with destruction of the walls lining

62

Chapter 5

the air sacs, resulting in further airflow obstruction. These changes can
cause significant respiratory impairment, disability, and death. Although
individual patients vary in the relative contribution of these three changes,
those with clinically severe COLD typically have all three.

Although a minority of cigarette smokers will develop clinically
severe COLD, some chronic deterioration in lung structure or function is
demonstrable in most long-term smokers (U.S. Department of Health and
Human Services, 1984a). Some smokers show more chronic cough and
phlegm, others more airway obstruction. In general, breathing function
declines with the increase in a person's cumulative exposure to smoke,
measured in pack-years (Dockery et al., 1988).

Cigarette smoke produces pathological changes in the lungs of smokers
by a number of different mechanisms (U.S. Department of Health and
Human Services, 1990). Cigarette smoke is toxic to the cilia that line the
central breathing passages. These cilia, in combination with mucus
secretions, defend against deep inhalation of foreign material (U.S.
Department of Health and Human Services, 1984a). Smoking also induces
many abnormalities in the inflammatory and immune systems within the
lung (U.S. Department of Health and Human Services, 1985). In particular,
cigarette smoke causes inflammatory cells to produce an enzyme called
elastase, which in turn breaks down elastin, an important protein that lines
the elastic walls of the air sacs (Fera et al. 1986; U.S. Department of Health
and Human Services, 1984a). Moreover, oxidants present in cigarette smoke
can inactivate a separate protective enzyme called alpha^-antitrypsin, which
inhibits the destructive action of elastase Qanoff, 1985; U.S. Department of
Health and Human Services, 1984a).

Many organic and inorganic chemicals in the gaseous, volatile, and
particulate phases of cigarette smoke appear to contribute to smoke's
toxicity to the respiratory system, including hydrocarbons, aldehydes,
ketones, organic acids, phenols, cyanides, acrolein, and nitrogen oxides.

Some components contribute to the development of chronic mucus
hypersecretion in the central airways, whereas others play a greater role
in the production of small airway abnormalities and emphysematous injury
to the peripheral air sacs. Oxidizing agents in smoke inhibit the enzymes
that defend against the destruction of lung elastin (U.S. Department of
Health and Human Services, 1984a).

ATHEROSCLEROTIC Cigarette smoking is a major contributing cause to CHD, stroke,

CARDIOVASCULAR and other atherosclerotic diseases of the circulatory system (U.S.

DISEASES Department of Health and Human Services, 1984b and 1989).

Atherosclerosis is a chronic disease that can affect the arterial blood
vessels in virtually every part of the human body. The most important
form of atherosclerosis in the United States is coronary atherosclerosis.

Its manifestations, which include angina, heart attack, heart failure, and
sudden death, are described by the inclusive term coronary heart disease.
Atherosclerosis involving the arteries supplying the brain is a form of

63

Smoking and Tobacco Control Monograph No. 7

cerebrovascular disease. Atherosclerosis involving the arteries to the limbs is
called peripheral vascular disease (PVD).

In numerous epidemiologic studies of millions of people, cigarette
smokers have been found to have higher rates of heart attack, sudden death,
and other manifestations of CHD. They also have higher rates of stroke, PVD,
and other atherosclerotic lesions (U.S. Department of Health and Human
Services, 1984b and 1989; U.S. Department of Health, Education, and Welfare,
1979). In the Cancer Prevention Study II (CPS-II) of more than 1 million
people followed from 1982 through 1986, men currently smoking had a
94-percent greater risk of CHD than lifelong nonsmokers, whereas women
currently smoking had a 78-percent greater risk. In smokers younger than
age 65, men had a 181-percent greater risk and women a 200-percent greater
risk (U.S. Department of Health and Human Services, 1989).

Cigarette smoking is sometimes called an independent risk factor for
CHD because smokers' CHD rates are found to be higher even when other
risk factors such as gender, blood pressure, and cholesterol level are taken into
account. It is sometimes called a modifiable risk factor because one can
reduce or stop smoking. Although smoking obviously cannot be a cause of
CHD in someone who never smoked, it can be an important contributor to
CHD in a smoker. Among 548,000 deaths from CHD in the United States in
1985, an estimated 115,000 would not have occurred but for the presence of
cigarette smoking (U.S. Department of Health and Human Services, 1989).

Cigarette smoke appears to enhance the atherosclerotic process by several
different mechanisms (U.S. Department of Health and Human Services, 1990;
Glantz and Parmley, 1995). Cigarette smoking affects cholesterol metabolism.
Smokers repeatedly have been observed to have lower levels of the protective
high-density lipoprotein (HDL) cholesterol (Willett et al., 1983), and smoking
cessation raises HDL cholesterol (Rabkin, 1984). In animal models, cigarette
smoke can damage the inner lining of blood vessels, thus enhancing the
transfer of low-density lipoprotein (LDL) cholesterol particles across the
arterial wall and into the developing cholesterol-laden plaque (Krupski et al.,
1987; Zimmerman and McGeachie, 1987; Penn et al., 1994). Cigarette
smoking also can affect the blood clotting system, including the adherence
of blood platelets to the lining of arterial blood vessels (Pittilo et al., 1984;

U.S. Department of Health and Human Services, 1984b; Burghuber et al.,

1986) and the formation of blood clots that block a narrowed artery. Acrolein
in cigarette smoke may be partly responsible for its platelet-adhering effects
(Selley et al., 1990). Cigarette smoke also can cause spasm of the coronary
arteries.

Many chemical components of cigarette smoke have been implicated in
the development of atherosclerotic disease. Nicotine, the major psychoactive
component of smoke, causes powerful changes in heart rate and blood
circulation. Nicotine appears to cause injury to the arterial lining (Krupski et
al., 1987). Carbon monoxide in cigarette smoke binds to the hemoglobin in
red blood cells, thereby reducing the oxygen-carrying capacity of the blood
(Sheps et al., 1990). PAHs, such as 7,12-dimethylbenz(«,/;)anthracene and

64

i

Chapter 5

BaP, have been found to accelerate the development of atherosclerosis in
animal models; this suggests that cell injury and cell proliferation (or
hyperplasia) may contribute to the development of the growing plaque
(Glantz and Parmley, 1991). Hydrogen cyanide, nitrogen oxides, and
chemical components of cigarette tar also have been implicated. Free
radicals in cigarette smoke, which are highly reactive oxygen products,
are damaging to the heart muscle cells (Church and Pryor, 1985).

CIGARETTE SMOKING Cigarette smoking adversely affects sexual and reproductive

AND HUMAN function in women in a number of different ways. Cigarette

REPRODUCTION smoking appears to impair female fertility (Baird and Wilcox,

1985; Daling et al., 1987; Mattison, 1982; U.S. Department of Health and
Human Services, 1980). Among the possible mechanisms are direct toxicity
to eggs, interference with motility in the female reproductive tract, and
alterations in immunity that predispose female smokers to infections that
block the Fallopian tubes (Chow et al., 1988).

Maternal cigarette smoking has serious adverse effects on the outcome
of pregnancy. These include retarded fetal growth; low birth weight;
spontaneous abortion; certain complications of pregnancy, labor, and
delivery, such as bleeding during pregnancy and prolonged premature
rupture of membranes; and infant death (U.S. Department of Health and
Human Services, 1980, 1989, and 1990; U.S. Department of Health,
Education, and Welfare, 1979). Direct nicotine toxicity has been suggested
as a mechanism for spontaneous abortion (U.S. Department of Health and
Human Services, 1990). Although a smoking-induced reduction in maternal
weight gain contributes to fetal growth retardation (U.S. Department of
Health and Human Services, 1980; Werler et al., 1985), the evidence points
to oxygen starvation of the fetus and placenta as important factors. Carbon
monoxide in cigarette smoke can cross the placenta and bind to the
hemoglobin in fetal blood. Smoking causes constriction of the umbilical
arteries, impairing placental blood flow. Nicotine, which also crosses the
placenta, can have a number of toxic effects on the fetus (U.S. Department
of Health and Human Services, 1980). The carcinogen 4-aminobiphenyl
crosses the placenta in a mother who smokes and adducts with the
hemoglobin in the fetus' blood (Coghlin et al., 1991). Cyanide, another
component of cigarette smoke, also has been implicated.

Women currently smoking enter nonsurgical menopause about 1 to
2 years earlier than nonsmokers (U.S. Department of Health and Human
Services, 1990). Heavy smokers experience an even earlier menopause than
light smokers. This effect has important consequences for women's health,
because the rates of osteoporosis and atherosclerotic cardiovascular diseases
increase after menopause. One proposed mechanism for early menopause is
that PAHs in smoke are directly toxic to ovarian follicles (Mattison, 1980).

Cigarette smoking also may affect male reproductive performance. In
several studies, men who report impotence (i.e., the inability to maintain an
erection sufficient for intercourse) were more likely to be cigarette smokers.
This association between smoking and impotence is particularly common

65

Smoking and Tobacco Control Monograph No. 7

among men who have high blood pressure or diabetes and appears to be
a consequence of increased atherosclerotic disease in the blood vessels
supplying the genitalia rather than an effect on sexual drive.

ABSOLUTE RISK Human epidemiology can be used to estimate quantitatively the
VS. RELATIVE risk of specific diseases to human smokers. For example, in the
RISK CPS-11 study of smoking practices and mortality rates among

1.2 million U.S. adults followed from 1982 through 1986, about 0.8 percent
of current male smokers ages 65 or older died of lung cancer each year (U.S.
Department of Health and Human Services, 1989), whereas the comparable
annual lung cancer death rate was about 0.04 percent among men ages 65 or
older who never smoked. These quantitative risk estimates are often termed
"absolute risks." That the continuing smokers' risk of lung cancer was
twentyfold that of nonsmokers is an expression of "relative risk."

Estimating relative risks from analyses of chemical composition of
different cigarettes is far more complicated. For example, the smoke from
cigarette A might contain 0.05 mg of BaP, a known carcinogen, whereas
the smoke from cigarette B might contain 0.02 mg of BaP. To estimate
human lung cancer risks from these data alone would require a number of
assumptions relating the dose of BaP to the incidence lung cancer in humans.
Whereas cigarette A had 2.5-fold as much BaP as cigarette B, it cannot be
concluded automatically that the relative risk of getting lung cancer for
those smoking cigarette A is 2.5-fold greater than those smoking cigarette B.
The relative concentrations of benz(a)anthracene, another carcinogen in the
PAH group, might be higher or lower.

Toxicity studies in nonhuman species also can give estimates of
relative risk, but applying these estimates directly to humans requires
caution. The fact that the smoke from cigarette C might produce twice as
many revertants as cigarette D in a particular strain of the Ames salmonella
assay is an indicator that C contains higher concentrations of certain
mutagens. Likewise, if cigarette E produced three times as many tumors as
cigarette F in a mouse skin carcinogenesis assay, we can conclude that
cigarette E contains higher concentrations of certain carcinogens, including
tumor initiators and tumor promotors (DuMouchel and Harris, 1983).

TAR, NICOTINE, Some studies (e.g., Adams et al., 1987) suggest that the yields

CARBON MONOXIDE, of most toxic agents in cigarette smoke are correlated with
AND OTHER SMOKE their tar, nicotine, and CO deliveries. Still other studies

CONSTITUENTS show the correlation to be weak at best. Kaiserman and

Rickert (1992) found a 0.89 correlation between the declared tar level and
the BaP delivery of 35 brands of Canadian cigarettes. However, for 16-mg
tar brands, the measured BaP ranged from 15 to 28 ng per cigarette. Fischer
and colleagues (1991) found no correlation between tar delivery and the
concentration of certain TSNAs in 170 Imropean cigarettes.

I he lack of a perfect correlation between tar values and specific chemical
yields is not simply an artifact of measurement error. As Hoffmann and
colleagues (this volume) report, there are many alternative methods to reduce

66

Chapter 5

cigarette smoke constituents, including various filter designs, changes in
paper porosity, mixing of tobacco species, and the use of reconstituted
tobacco sheets and expanded tobacco. However, all these methods do not
reduce every smoke constituent uniformly. For example, perforated filter
tips selectively reduce the volatile and gaseous components of cigarette
smoke, whereas reconstituted tobacco sheets reduce BaP and tar but not
acrolein or acetaldehyde. Likewise, as reported by Hoffmann and coworkers
(this volume), the increased hurley tobacco content (and with it, the nitrate
content) of at least one marketed cigarette resulted in an increase in the
delivery of NNK, a tobacco-specific nitrosamine, over the course of three
decades.

In a study of cigarette brands sold in the United Kingdom from 1983
through 1990, Phillips and Waller (1991, p. 469) concluded that, "with the
exception of nitrogen monoxide, which is strongly dependent upon the
type of tobacco, and the delivery of some phenols and PAHs, which may
be affected to a minor extent by the design of cigarette," the three routinely
monitored smoke components (tar, nicotine, and CO) provided "an adequate
guide" to the yields of the other chemical entities examined. However, as
the foregoing review of cigarette smoke constituents and disease suggests,
the exceptions may prove the rule. It would be unscientific to claim that
the absolute human risk or even the relative risk of a particular brand of
cigarettes is lower merely because, on average, everything but TSNAs, phenols,
and PAHs seems to be lower. With phenols and related flavorant compounds
implicated in smoke-induced chromosomal damage Qansson et al., 1988),
it would seem that, at minimum, biological testing would be warranted.

As discussed elsewhere in this volume, the yields of nicotine and carbon
monoxide are significantly influenced by the smoker's style or "topography"
of smoking, including number of puffs, interval between puffs, velocity and
volume of each puff, depth of draw, length of cigarette smoked, depth of
inhalation into the lungs, and other factors. It is possible that these
differences in smoking topography might selectively influence the yields
of some smoke chemicals more than others. Fischer and colleagues (1989)
found that TSNA yields depended on the total volume of smoke inhaled
by the smoker and that total smoke volume was increased for smokers
of low- and medium-tar cigarettes. Studies of smokers' exposure to specific
carcinogenic compounds (e.g., by measurement of PAH adducts to DNA) do
not always show a relationship between exposure and self-reported smoking
intensity (Santella et al., 1992).

SMOKE CONSTITUENTS, Henningfield and colleagues (1994) recently proposed

CIGARETTE-RELATED modified labeling of cigarettes. Their proposed new

DISEASE, AND MODIFIED new cigarette label included a warning statement;

LABELING OF CIGARETTES categorization of nicotine yield; nicotine content; tar,
nicotine, and CO deliveries (average and maximal); harmful additives; and
information about factors affecting nicotine delivery. The use of a nicotine-
yield category was intended to replace such marketing terms as "light" and
"ultralight." These authors noted, "An additional strategy that could be used

67

Smoking and Tobacco Control Monograph No. 7

to assist consumers in making informed decisions would be to fully
disclose the tobacco smoke constituents of potential health significance,
analogous to harmful constituent disclosure of foods" (Henningfield et al.,
1994, pp. 312-313).

The new nutritional labels mandated by the Food and Drug
Administration (FDA) on all packaged foods contain information on a
wide array of vitamins, minerals, cholesterol, total fat, and saturated fat.
These labels reflect the product's characteristics. They make no pretense
that any two individuals will eat breakfast cereal in the same way. Nor
do they imply that each and every consumer will understand or want to
understand each and every entry on the label.

In the same way, the author has designed a "mock" cigarette label
(Figure 1) to indicate what such an FDA-style label for cigarettes might look
like. This is a sample and is not intended to reflect any current brand on the
market. The opening box gives an explanation as well as a warning about
the ways in which a smoker can obtain higher yields by changing his or her
style of smoking. Then some "basic cigarette facts" would be included, such
as length, type of filter, and weight of tobacco. In addition to data on the
range of yields of tar (total particulates less nicotine) and nicotine, the label
would show the range of yields of important smoke chemicals.

The concept of full disclosure of cigarette characteristics is entirely
consistent with the current Federal Trade Commission (FTC) method. In
fact, the current FTC measurements of tar, nicotine, and CO are included
in the proposed mock label. In addition, as we move to an era where both
short- and long-term biological testing have become commonplace in
industry, one might imagine a rating system based on the Ames test, skin
painting, and other studies. Illustrative results for such biological testing
are included in the mock label.

One might object that such detailed disclosure of cigarette characteristics
will confuse the smoker. Such an assertion is unscientific and unfair. To
publish a label that discloses, for example, the tobacco-specific nitrosamine
contents of a particular brand of cigarettes is no more confusing or
complicated than printing a label that discloses the riboflavin and potassium
yields of a particular brand of breakfast cereal. It would be remarkable
to discover cereal manufacturers or consumer advocates arguing that the
vitamin contents or trace metal levels of cereals should be withheld from
consumers because vitamin E and zinc levels might correlate — at least
roughly — with dietary fiber contents.

Iiid

I'o a limited degree, researchers have studied consumers' responses to
advertised tar and nicotine ratings of cigarettes. But there are no data — at
least in the public domain — on the possible effects of providing consumers
with additional cigarette-specific information of the type considered in the
mock label.

68

i Chapters

Figure 1

1 "Mock" cigarette label

; Our tar/nicotine label has changed! The Food and Drug Administration now requires each pack of
1 HARRIS Ultras to display the deliveries of the most important chemicals in your cigarette smoke. Your
i own smoke intake of these chemicals may vary from low to high depending on the size of your puffs,
the number of puffs per minute, the depth of your draw, and how far you smoke your cigarette down to
• the filter overwrap. For a factsheet about the new cigarette label, write to: New Cigarette Label

Factsheet, P.O. Box 7551, Brookline, MA 02146.

1 Basic Cigarette Facts

1 Cigarette Length

100 mm

i Cigarette Diameter

8 mm

1 Length of Filter Plus Plugwrap

20 mm

: Total Cigarette Weight

1 .20 gm

! Tobacco Weight

0.90 gm

i Type of Filter

Cellulose Acetate

Design of Filter

Perforated

U.S. -Grown Tobacco

55%

Cigarettes Per Pack

20

Delivery Per Cigarette

Low

High

i Nicotine

1.0 mg

1.5 mg

; Carbon Monoxide

14 mg

20 mg

Carcinogenic PAHs

0.1 pg

0.2 pg

Tobacco-Specific Nitrosamines

0.3 pg

2.0 pg

Hydrogen Cyanide

0.4 mg

0.6 mg

Acrolein

60 pg

140 pg

Formaldehyde

20 pg

100 pg

Nitrogen Oxides

0.1 mg

0.4 mg

Catechols

200 pg

400 pg

Phenols

45 pg

60 pg

Nickel

0.1 pg

0.2 pg

Total Particulates Less Nicotine

6 mg

16 mg

Redox Potential of Smoke

160 mV

240 mV

1 pH of Whole Smoke

5.8

6.0

' Biological Test Results

: Ames Salmonella

+

ii Tracheal Installation

++

1 Mammalian Cell Transformation

+

' Syrian Golden Hamster Inhalation

+

i! Mouse Skin Carcinogenesis

++

1 Antielastase Test

++++

INGREDIENTS: Domestic flue-cured, Burley, and oriental leaf tobaccos; flavorants (including menthol

in HARRIS ULTRAGREENS), and humectants, including diethylene glycol. Citric
acid added to cigarette paper. Residues of maleic hydrazide (a suckercide used in

1 tobacco growing) less than 1 part per million.

1

1 WARNING: Keep out of reach of children!

Key: PAHs = polyaromatic hydrocarbons.

1

69

Smoking and Tobacco Control Monograph No. 7

In any case, smokers constitute only the demand side of the cigarette
market. On the supply side are a handful of cigarette manufacturers who,
so far as is known, go to considerable lengths to determine the detailed
characteristics of competitors' products. From time to time, a cigarette
manufacturer will disclose the level of a particular chemical in a particular
brand. One classic example is the claim by one manufacturer, in the early
1960's, that a particular brand delivered smoke with reduced phenol, an
announcement that coincided with scientific reports that the phenol in
cigarette smoke inhibited the cilia lining in the respiratory tract. However,
without systematic and complete disclosure requirements, such "competition"
will remain haphazard at best. In 1989 the tar content was listed on only
14 percent and the nicotine content on only 11 percent of U.S. cigarette
packages (Davis et al., 1990).

Enhanced and complete disclosure of cigarette characteristics by a
standardized label would create a basis for more effective competition among
manufacturers. If Hoffmann and colleagues' (this volume) data are
generalizable, then the growing trend toward use of hurley tobaccos in
American cigarettes might have resulted in increased deliveries of TSNAs,
even as other smoke constituents have declined. Without specific disclosure
of tobacco-specific nitrosamines, it is unclear how this deleterious trend
would be reversed or even detected. As economists know, competition among
manufacturers over a specific brand characteristic, such as a cigarette's TSNA
delivery, does not require that the average smoker — or even most smokers —
know what a "nitrosamine" is.

QUESTION-AND-ANSWER SESSION

DR. HOFFMANN; We go through stages in all research; so we go through
stages in tobacco research. The first stage was to identify those agents that,
in the laboratory animal, cause disease. The second stage we are in now is a
biomarker stage. This gives us 4-aminobiphenol, bihemoglobin, and tobacco-
specific nitrosamines.

I do think we are now in a better position to judge the relationship between
smoke components and disease. One should not forget that we have now
moved past the stage of biomarkers where it is solely the identification of
agents, and I do think one should not have such a negative outlook.

DR. HARRIS: Yes. I did not include as possible endpoints by which to
compare individual cigarettes the possibility that these components may
be found bound to the hemoglobin and red cells, or circulating proteins,
or albumen in the blood. It is a fact that certain biomarkers, certain
hydrocarbons, 4-aminobicarbons, and other compounds, have been now
found bound to blood proteins or other compounds, not only among those
who smoke cigarettes, but recently, in the Journal of the National Cancer
Institute, among those who are expo.sed passively to cigarette smoke.

Whether those can be used for a com[)arative analysis of different cigarettes
I do not know. Hut 1 would, in order at least to be provocative or speculative.

Chapter 5

state that we will soon be entering an era when we can make comparisons
among cigarettes by more than merely standards of chemical constituents.

DR. KOZLOWSKl: I imagine that the HARRIS ULTRA is an ultralow-tar
cigarette, and it is perforated. And so, like other current 1-mg tar cigarettes,
it might be 80 or 90 percent perforated, so you get air dilution of between
80 and 90 percent. 1 am disappointed that the label, in talking about
compensatory smoking factors, does not mention the issues of vent blocking
and staying away from the vents, because in fact, increased puff numbers
would have a relatively small effect, if this 90-percent dilution factor was
not eliminated.

DR. HARRIS: In designing that mock label, I did not attempt to be
scientifically precise as to absolutely everything one would put in about
proper directions for use or factors that might affect yield. Actually, your
yields could vary, and then afterward put in a perforated tip. If you will
look carefully, it was a 6-mg tar cigarette but with very high nicotine —

1 mg of nicotine.

One could argue, however, that if one is to continue to publish what
basically are the FTC data, expanded possibly to include a high or low range,
or to include other constituents, that there ought to be something about
directions for appropriate use.

DR. HEADEN: Dr. Harris, in proposing a possible design for a cigarette label,

I would like to know your opinion of who you think the audience is for that
label. Is it the tobacco industry and other regulators who might possibly be
able to interpret the information that you have? Is it the consumer, who
might smoke that cigarette, many of whom have lower educational levels,
or would it be both?

DR. HARRIS: I think it would have to be considerate of everyone and,
although it may sound as if I would add the results of additional constituents
just to satisfy some intellectually rarified audience, I raise the question, why
do we put such a large array of constituents on our ordinary food supply?
Some people might argue that we should not insult consumers by assuming
that they cannot pick and choose, to understand or use meaningfully some
forms of information rather than others. At this point, unless there is a
solid confirmation that all those constituents of smoke, or characteristics
of cigarettes, are simply summarized by the amount of tar, one wonders
whether the lack of that information is at least deceiving some people.

That is the best I can do to answer that.

DR. SONDIK: Dr. Harris, the label is intriguing, and I would not want to
spend too much time on it, but a couple of points might differentiate it
from the FDA label, which I happen to believe is one of the major advances
in nutrition. The FDA label is designed to aid people in developing their
diets. Their diet consists of all types of food, and the idea is to integrate
all of these things together, which is the idea behind putting all of these
different measurements on it, not just a single measurement, such as
calories, for example, or total fat.

71

Smoking and Tobacco Control Monograph No. 7

The second thing is, that label must have education along with it. And
that is part of the program in NCI, the Department of Agriculture, the FDA,
and others who are involved in a very intense education program — to try to
be sure that the public knows how to interpret a label like this. So, in a
sense, that label is more complex and aimed at a variety of types of decisions.

I would think that a label such as you proposed would be aimed at
perhaps a single decision, which is whether or not this is a useful thing for
me to do, trading off whatever my immediate gain might be, and pleasure,
vs. long-term health effects. Is there a way of getting that onto the label?

DR. HARRIS: 1 do not know. It also has occurred to me that once more
dimensions to cigarettes are specifically disclosed, that would be the basis of
further competition among cigarette manufacturers. So, the manufacturers
would then be seeing not only whether a cigarette is low or high in tar,
nicotine, and carbon monoxide, but in other specific components, too.

That means that while the consumer does not specifically choose among
high- or low-benzo(u)pyrene cigarettes, the disclosure of such contents
provides an incentive for manufacturers to try to reduce that component.
This is the same way that the disclosure of saturated fat contents in certain
breakfast cereals or other foods, even without consumer knowledge,
provides an incentive for some manufacturers to try to reduce that content.
Nevertheless, it provides some incentive on the supply side, not just the
demand side.

DR. BENOWITZ: I know that you are not intending to be totally
comprehensive about your mock insert, or label, and I think it is worthwhile
keeping in mind the parallel with foods. If you are talking about limiting
intake, there really are only two contents that we know about that might
limit intake. One is the amount of the tobacco in the cigarette, which you
did put down, and the other is the amount of nicotine contained in there,
which is something that people do not often think about. But the amount
of nicotine in tobacco limits what a person can get. And the intake of
nicotine is not necessarily correlated at all with the yield.

So, I think that when we think about any sort of labeling for content,
the nicotine content, which is the maximum available dose one could get,
should really be a part of it.

DR. HARRIS: 1 noticed that that was in your original proposal, and 1 am not
an expert on the degree to which nicotine content is very limiting for how
many smokers. 1 would rather defer that to a later discussion, as to how
irn[)ortant that is.

RIJERtNCES

Adams, J.l)., O'Mara-Adams, K.J., Mofimaim, I). Toxic
and carcinogenic agents in undiluted mainstream
smoke and sidestream smoke ol different tyjX'S of
cigarettes. C«rdr/uxf«t's/s 8(S): 72^J-7 M, 1987.

Ifaird, D.l)., Wilcox, A.J. Cigarette smoking assrxiated
witfi delayed conception, loumul of the Amerutin
MeJiial AsuKiution 251: 2979-2981, 1985.

Heck, (i.J., Doyle, C.A., .Schacliter, K.N. .Smoking and
lung function. Arfieriiun Review of Respimtory Disease
i2A: 149-15.5, 1981.

Hurgluiber, ().(]., I’un/.engrul)er, C., Sin/inger, II.,
Haber, I’., Silberl)auer, K. I'latelel sensitivity to
prostacyclin in smokers and non-smokers. Chest
90(1): .'14-.f8, 1986.

72

Chapter 5

Chow, W.-H., Daling, J.R., Weiss, N.S., Voigt, L.F.
Maternal cigarette smoking and tubal pregnancy.
Obstetrics and Gynecology 7: 167-170, 1988.

Church, D.F., Pryor, W.A. Free-radical chemistry of
cigarette smoke and its toxicological implications.
Environmental Health Perspectives 64: 111-126, 1985.

Coghlin, J., Gann, P.H., Hammond, S.K., Skipper, P.L.,
Taghizadeh, K., Paul, M., Tannenbaum, S.R.
4-Aminobiphenyl hemoglobin adducts in fetuses
exposed to the tobacco smoke carcinogen in utero.
Journal of the National Cancer Institute 83(4): 274-
280, 1991.

Costagliola, C., Rinaldi, M., Giacoia, A., Rosolia, S.,
Cotticelli, C., Rinaldi, E. Red cell glutathione as a
marker of tobacco smoke-induced optic
neuropathy. Experimental Eye Research 48(4): 583-
586, 1989.

Crawford, F.G., Mayer, J., Santella, R.M., Cooper,

T.B., Ottman, R., Tsai, W.Y., Simon-Cereijido, G.,
Wang, M., Tang, D., Perera, F.P. Biomarkers of
environmental tobacco smoke in preschool
children and their mothers. Journal of the National
Cancer Institute 86{18): 1398-1402, 1994.

Daling, J., Weiss, N., Spadoni, L., Moore, D.E., Voigt,

L. Cigarette smoking and primary tubal infertility.
In: Smoking and Reproductive Health, M.J. Rosenberg
(Editor). Littleton, MA: PSG Publishing, 1987, pp.
40-46.

Davis, R.M., Healy, P., Hawk, S.A. Information on tar
and nicotine yields on cigarette packages. American
Journal of Public Health 80(5): 551-553, 1990.

Dockery, D.W., Speizer, F.E., Ferris, B.G., Jr., Ware,
J.H., Louis, T.A., Spiro, A. III. Cumulative and
reversible effects of lifetime smoking on simple
tests of lung function in adults. American Review
of Respiratory Disease 137(2): 286-292, 1988.

DuMouchel, W.H., Harris, J.E. Bayes methods of
combining the results of cancer studies in humans
and other species. Journal of the American Statistical
Association 78: 293-308, 1983.

Fera, T., Abboud, R.T., Richter, A., Johal, S.S. Acute
effect of smoking on elastaselike activity and
immunologic neutrophil elastase levels in
broncheolar lavage fluid. American Review of
Respiratory Disease 133: 568-573, 1986.

Fischer, S., Spiegelhalder, B., Preussmann, R. Influence
of smoking parameters on the delivery of tobacco-
specific nitrosamines in cigarette smoke — -a
contribution to relative risk evaluation.
Carcinogenesis 10: 1059-1066, 1989.

Fischer, S., Spiegelhalder, B., Preussmann, R. Tobacco-
specific nitrosamines in commercial cigarettes:
Possibilities for reducing exposure. lARC Scientific
Publications 105: 489-492, 1991.

Gairola, C.G. Genetic effects of fresh cigarette smoke
in Saccharomyces cerevisiae. Mutation Research 102:
123-136, 1982.

Garfinkel, L., Bofetta, P. Association between smoking
and leukemia in two American Cancer Society
prospective studies. Cancer 65: 2356-2360, 1990.

Glantz, S.A., Parmley, W.W. Passive smoking and
heart disease: Epidemiology, physiology, and
biochemistry. Circulation 83: 1-12, 1991.

Glantz, S.A., Parmley, W.W. Passive smoking and
heart disease: Mechanisms and risk. Journal of the
American Medical Association 273: 1047-1053, 1995.

Harris, J.E. Smoking and nothingness. In: Deadly
Choices: Coping with Health Risks in Everyday Life.
New York: Basic Books, 1993, pp. 151-177.

Hecht, S.S., Carmella, S.G., Murphy, S.E., Akerkar, S.,
Bmnnemann, K.D., Hoffmann, D. A tobacco-
specific lung carcinogen in the urine of men
exposed to cigarette smoke. New England Journal
of Medicine 329(21): 1543-1546, 1993.

Henningfield, J.E., Kozlowski, L.T., Benowitz, N.L.

A proposal to develop meaningful labeling for
cigarettes. Journal of the American Medical Association
272(4): 312-314, 1994.

Hoffmann, D., Hecht, S.S. Advances in tobacco
carcinogenesis. Handbook of Experimental
Pharmacology, Vol. 94, No. I, C.S. Cooper and P.L.
Grover (Editors). Berlin: Springer-Verlag, 1990.

Janoff, A. Elastases and emphysema: Current

assessment of the protease-antiprotease hypothesis.
American Review of Respiratory Disease 132: 41 7-433,
1985.

Jansson, T., Curvall, M., Hedin, A., Enzell, C.R. In
vitro studies of the biological effects of cigarette
smoke condensate. III. Induction of SCE by some
phenolic and related constituents derived from
cigarette smoke. A study of structure-activity
relationships. Mutation Research 206(1): 17-24,

1988.

Kaiserman, M.J., Rickert, W.S. Carcinogens in tobacco
smoke: Benzo(a)pyrene from Canadian cigarettes
and cigarette tobaccos. American Journal of Public
Health 82: 1023-1026, 1992.

Krupski, W.C., Olive, G.C., Weber, C.A., Rapp, J.H.
Comparative effects of hypertension and nicotine
on injury-induced myointimal thickening. Surgery
102: 409-415, 1987.

Mattison, D.R. Morphology of oocyte and follicle
destruction by polycyclic aromatic hydrocarbons in
mice. Toxicology and Applied Pharmacology 53: 249-
259, 1980.

Mattison, D.R. The effects of smoking on fertility
from gametogenesis to implantation. Environmental
Research 28: 410-433, 1982.

Melikian, A.A., Prahalad, A.K., Hoffmann, D. Urinary
trans, truns-muconic acid as an indicator of
exposure to benzene in cigarette smokers. Cancer
Epidemiology, Biomarkers and Prevention 2(1): 47-51,
1993.

73

Smoking and Tobacco Control Monograph No. 7

Niewoehner, D.E., Kleinerman, J., Rice, D.B. Pathologic
changes in the peripheral airways of young cigarette
smokers. New England journal of Medicine 291: 755-
758, 1974.

Penn, A., Chen, L.C., Snyder, C.A. Inhalation of steady-
state sidestream smoke from one cigarette promotes
atherosclerotic plaque development. Ciradation 90:
1363-1367, 1994.

Phillips, G.F., Waller, R.E. Yields of tar and other smoke
components from UK cigarettes. Food and Chemical
Toxicology 29(7): 469-474, 1991.

Pittilo, R.M., Clarke, J.M., Harris, D., Mackie, I.J.,

Rowles, P.M., Machin, S.J., Woolf, N. Cigarette
smoking and platelet adhesion. British Journal of
Haematology 58(4): 627-632, 1984.

Rabkin, S.W. Effect of cigarette smoking cessation on
risk factors for coronary atherosclerosis: a control
clinical trial. Atherosclerosis 53(2): 173-184, 1984.

Rando, R.J., Menon, P.K., Poovey, H.G., Lehrer, S.B.
Assessment of multiple markers of environmental
tobacco smoke (ETS) in controlled, steady-state
atmospheres in a dynamic test chamber. American
Industrial Hygiene Association journal 53(1 1): 699-704,
1992.

Santella, R.M., Grinberg-Funes, R.A., Young, T.L.,
Dickey, C., Singh, V.N., Wang, L.W. Perera, F.P.
Cigarette smoking related polycyclic aromatic
hydrocarbon-DNA adducts in f>eripheral
mononuclear cells. Carcinogenesis 13(1 1): 2041-
2045, 1992.

Seely, J.E., Zuskin, E., Bouhys, A. Cigarette smoking:
Objective evidence for lung damage in teenagers.
Science 172: 741-743, 1971.

Selley, M.L., Bartlett, M.R., McGuiness, J.A., Ardlie,

N.G. F!.ffects of acrolein on human platelet
aggregation. Chemico-Biological Interactions 76(1):
101-109, 1990.

Sheps, D.S., Herbst, M.C., Hinderliter, A.L., Adams,

K.F., Ekelund, L.G., O'Neil, J.J., Goldstein, G.M.,
Bromberg, P.A., Dalton, J.l.., Pallenger, M.N., Davis,
S.M., Koch, G.G. Production of arrythmias by
elevated carboxyhemoglobin in patients with
coronary artery disease. Annals of Internal Medicine
113(5): 343-351, 1990.

U.S. Dejjartment of Health and Human Services. The
Health Consequences of Smoking for Women. A Report of
the Surgeon General. Rockville, Ml): U.S. De()artment
of Health and Human Services, Public Health
Service, Office of the Assistant Secretary for Health,
Office on Smoking and Health, 1980.

U.S Department of Health and tfumaii Services. I'he
Health (Consequences of Smoking: Cancer. A Rejiort of the
Surgeon (leneral. DHHS Publication No. (PHS) 82-
S0179 Kixkville, Ml): U.S. Department of Health
and Human Services, Public Health Service, Office on
Smoking and ffealtfi, 1982.

U.S. Department of Health and Human Services. The
Health Consequences of Smoking: Chronic Obstnictive
Lung Disease. A Report of the Surgeon General. DHHS
Publication No. (PHS) 84-50205. Rockville, MD:

U.S. Department of Health and Human Services,
Public Health Service, Office on Smoking and
Health, 1984a.

U.S. Department of Health and Human Services. The
Health Consequences of Smoking: Cardiovascidar
Disease. A Report of the Surgeon General. DHHS
Publication No. (PHS) 84-50204. Rockville, MD:

U.S. Department of Health and Human Services,
Public Health Service, Office on Smoking and
Health, 1984b.

U.S. Department of Health and Human Services. The
Health Consequences of Smoking: Cancer and Chronic
Lung Disease in the Workplace. A Report of the Surgeon
General. DHHS Publication No. (PHS) 85-50207.
Rockville, MD: U.S. Department of Health and
Human Services, Public Health Service, Office on
Smoking and Health, 1985.

U.S. Department of Health and Human Services. The
Health Consequences of Involuntary Smoking. A Report
of the Surgeon General. DHHS Publication No. (CDC)
87-8398. Rockville, MD: U.S. Department of Health
and Human Services, Public Health Service, Centers
for Disease Control, Center for Health Promotion
and Education, Office on Smoking and Health,
1987.

U.S. Department of Health and Human Services. The
Health Consequences of Smoking: Nicotine Addiction.

A Report of the Surgeon General, 1988. DHHS
l\iblication No. (CDC) 88-8406. Rockville, MD:

U.S. Department of Health and Human Services,
Public Health Service, Centers for Disease Control,
Center for 1 lealth Promotion and Education, Office
on Smoking and Health, 1988.

U.S. Department of Health and Human Services.
Reducing the Health Conseipcences of Smoking: 25
Years of Progress. A Report of the Surgeon General,

1989. Dims Publication No. (CDC) 89-8411.
Rockville, MD: U.S. Department of Health and
Human Services, Public Health Service, Centers
for Disease C’.ontrol, Center for c:hronic Disease
Prevention and Health Promotion, Office on
Smoking and Health, 1989.

U.S. De|)artment of Health and Human Services. The
Health Benefits of Smoking (Cessation. A Rejmrtofthe
Surgeon General, 1990. DHHS Publication No. (C^DC)
90-8416. Rockville, Ml): U.S. De[)artment of Health
and Human Services, Public Health Service, Centers
lor Disease Control, CX-nter tor (Ihronic Disease
Prevention and Health Promotion, Office on
Smoking and Health, 1990.

74

Chapter 5

U.S. Department of Health, Education, and Welfare.
Smoking and Health. A Report of the Surgeon General.
DHEW Publication No. (PHS) 79-50066. Rockville,
MD: U.S. Department of Health, Education, and
Welfare, Public Health Service, Office of the Assistant
Secretary for Health, Office on Smoking and Health,
1979.

Vineis, P. Black (air-cured) and blond (flue-cured)
tobacco and cancer risk. 1: Bladder cancer. European
Journal of Cancer 27 {\\y. 1491-1493, 1991.

Werler, M.M., Pober, B.R., Holmes, L.B. Smoking and
pregnancy. Teratology 32: 473-481, 1985.

Willett, W., Hennekens, C.H., Castelli, W., Rosner, B.,
Evans, D., Taylor, J., Kass, E.H. Effects of cigarette
smoking on fasting triglyceride, total cholesterol,
and HDL-cholesterol in women. American Heart
Journal 105(3): 417-421, 1983.

Zimmerman, M., McGeachie, J. The effect of nicotine
on aortic endothelium: A comparative study.
Atherosclerosis 63: 33-41, 1987.

ACKNOWLEDGMENT Financial support from the National Cancer Institute is
acknowledged, but the contents of this manuscript are the author's sole
responsibility.

75

V

Chapter 6

Xhe Changing Cigarette and Disease Risk:
Current Status of the Evidence

Jonathan M. Samet

INTRODUCTION Since the early 1950's when filter tip cigarettes were first widely
introduced, the cigarette has evolved continually through modifications
intended to reduce yields of tar and nicotine (U.S. Department of Health and
Human Services, 1991). Following the introduction of the filter tip cigarette,
sales-weighted averages of tar and nicotine deliveries show a temporal trend
of declining yield, which continues to the present (Figure 1). In the face
of continued modifications of the cigarette and the seemingly associated
changes in exposure of smokers to cigarette smoke components, questions
have been raised concerning the implications of the changing cigarette for
disease risks in smokers.

Only epidemiologic studies can provide information on modification of
the risks of smoking as the cigarette has evolved, and only epidemiologic data

Figure 1

Tar and nicotine content of U.S. cigarettes, sales-weighted average basis, 1957-1987

1955 1960 1965 1970 1975 1980 1985 1990

Year

Source: U.S. Department of Health and Human Services, 1989.

P

77

Sales-Weighted Nicotine (mg)

Smoking and Tobacco Control Monograph No. 7

can measure the risks of cigarettes under the "natural" circumstances of use.
However, the dynamic nature of the exposure (Figure 1) challenges the
epidemiologic researcher to classify accurately the pattern of cigarette use
when changes are made that may not be indexed by tar and nicotine yields
measured with a smoking machine.

In considering the health implications of the changing cigarette,
the concepts of exposure and dose are fundamental. Exposure has been
defined by the National Research Council (1991) as the amount of material
potentially available for interaction with a human, that is, material in contact
with a person at a boundary, whether that boundary be the skin, lung, or the
alimentary tract. On the other hand, dose is the amount of material that
enters the organism. Dose may be further classified as the internal dose
(i.e., the amount of material deposited) or as the biologically effective dose
(i.e., the amount of material delivered to some biologically relevant site).
Changes in the cigarette can be interpreted as potentially leading to changes
in exposure; the health consequences of changing exposure vary with any
resultant changes in dose of components of cigarette smoke that cause disease.

The physiological functioning of the lung is also relevant to
understanding the linkages in changes in the cigarette to changes in
exposure and dose. The lung is a complex organ with several different
"compartments," including the upper airway that extends from the nose
and mouth to the larynx; the airways of the lung itself, which include the
trachea, bronchi, and bronchioles; and the parenchyma of the lung, which
includes the interstitium and the airspaces, or alveoli. The lung behaves as
a filter that absorbs and deposits gaseous and particulate components of
smoke throughout its surfaces during the act of smoking. The sites and
extent of deposition of inhaled mainstream components vary, depending
on solubility and other characteristics of gas phase components and the
sizes of the particles. Cigarette smoke is a dynamic mixture in the respiratory
tract, changing with humidification of the mixture, growth of particles,
and changing composition as components are selectively removed by the
filtration process (U.S. Department of Health and Human Services, 1984).

These physiological considerations imply that there is no simple relationship,
linear or nonlinear, between reported tar and nicotine yields — a measure
of exposure — and biologically effective doses of toxic smoke components
delivered to the sites of injury in the respiratory tract.

The measures of cigarette smoking used in epidemiologic research on
smoking and health can be classified as estimating either exposure or dose.

I he most widely used measures, for example, information on cigarette
smoking (duration of smoking, number of cigarettes smoked per day, and
type or brand of cigarettes smoked), are exposure measures. Biomarkers that
can be inter{)reted as indicators of dose include levels of carboxyhemoglobin, 1
nicotine, and cotinine (U.S. Department of Health and Human Services, i

1990). Thus, for ei)idemiological i)urposes, researchers use exposure measures, ^
typically obtained by questionnaire, and dose measures, based on biomarkers. i
l or exam[)le, cigarettes smoked j)er day is an exposure measure, whereas f

78

Chapter 6

pack-years (packs smoked per day multiplied by the number of years smoked)
is a cumulative exposure measure. An estimate of kilograms of tar deposited
in the lung is an absorbed dose measure; nanograms of benzo(a)pyrene, for
example, reaching basal cells might be considered a biologically effective dose
for carcinogenesis. New markers take dose measures to the molecular level
(Vineis and Caporaso, 1995).

To assess the consequences of changes in the cigarette, it is necessary to
have information on how changes in tar and nicotine yield, as assessed by
the Federal Trade Commission (FTC) method, affect dose measures, extending
to the molecular level. Any new approach to testing cigarette yields should
be designed to be informative both as an exposure measure and as an
indicator of biologically relevant doses of cigarette smoke components.

EPIDEMIOLOGIC
EVIDENCE ON
THE CHANGING
CIGARETTE AND
DISEASE RISKS

Overview

Epidemiologic evidence is available on the effect of the changing
cigarette on all-cause mortality and on three major categories of
disease caused by cigarette smoking: lung and other cancers,
nonmalignant respiratory diseases, and cardiovascular disease
(CVD). The Health Consequences of Smoking: The Changing Cigarette.
A Report of the Surgeon General (U.S. Department of Health and
Human Services, 1981) addressed the changing cigarette, covering
the relevant toxicologic and epidemiologic evidence. This chapter considers
key epidemiologic publications since that report but does not provide a
systematic overview of the many studies on the changing cigarette.

The principal study designs that have been used to address the health
consequences of the changing cigarette are the ecological study, a descriptive
approach conducted at the group level, and cohort and case-control studies,
analytic approaches conducted at the individual level. Cross-sectional studies
have proven informative in investigating nonmalignant respiratory diseases.
The ecological approach is exemplified by a comparison of temporal changes
in rates of smoking-related diseases with patterns of consumption of various
types of cigarettes. The American Cancer Society (ACS) studies of large
groups of volunteer participants are cohort studies; the participants were
enrolled, information about smoking was obtained on enrollment and
periodically thereafter, the population was followed over time, and mortality
was ascertained. Some of the earliest evidence on smoking and lung cancer
was obtained in the classic case-control studies conducted by Doll and Hill
(1950) and Wynder and Graham (1950). In these studies, the smoking habits
of patients hospitalized with lung cancer were compared with the smoking
habits of control patients having another disease.

Evidence from epidemiologic studies has well-known strengths and
limitations (Rothman, 1986). Epidemiologic research has had a central role
in characterizing the consequences of the changing cigarette because it
supplies direct information on the consequences of varying tar and nicotine
yield products. Thus, the findings inherently consider compensatory
changes in inhalation patterns or in numbers of cigarettes smoked and
provide the evidence needed to answer the question of immediate public

79

Smoking and Tobacco Control Monograph No. 7

health relevance: whether disease risk varies with cigarette tar and nicotine
yield as determined by the FTC method.

Exposure misclassification is a potential threat to the validity of studies I
of the changing cigarette. Typically, the exposure of smokers to cigarettes I
of varying tar and nicotine yields is estimated based on information on |

brands and types of cigarettes smoked. However, smokers may not be able ;

to provide a fully accurate history of brands used throughout their lifetimes; ,

therefore, estimates of tar and nicotine yield are potentially subject to error
(U.S. Department of Health and Human Services, 1990). The consequences
of misclassification include biased estimates of the effect of tar and nicotine
yields and reduction of statistical power. Additional methodological concerns I
include the possibility of selection bias if smokers affected by symptoms
or disease tend to switch to lower yield products; another concern is
confounding by other aspects of lifestyle if smokers of lower yield cigarettes j
differ substantially in lifestyle characteristics from those smoking higher yield ;
products. However, the research challenge of studying the consequences of
the changing cigarette is no different from the challenge posed by other j
complex mixtures of inhaled agents, and epidemiologic research has the j
advantage of integrating the effects of the mixture, even though individual j
components may be interacting in ways that are difficult to characterize. |

I

Lung Cancer The ACS's Cancer Prevention Study I (CPS-I) provided early evidence j
on the risks of lower tar and nicotine cigarettes (Hammond et al., 1976). j

CPS-1 included about 1 million volunteers who were followed from 1960 to i

1972. Mortality was examined by three categories of tar intake — high, jj

medium, and low. For all causes of mortality and for lung cancer mortality, jj.
the standardized mortality ratios declined as estimated tar or nicotine intake jiii
declined (Table 1). The findings were similar for males (Table 1) and for jl

females (data not shown). However, comparison with mortality in never- i|

smokers shows that smokers of even the lowest tar and nicotine products ‘j
nonetheless had substantially higher mortality rates. :

Other studies have had similar findings for lung cancer. Wynder and jf
colleagues at the American Health Foundation have conducted an ongoing !
case-control study of smoking and lung cancer that provides information on '
cigarette type and lung cancer risk over decades since the 1950's. Reports i
from this study have consistently shown that smokers of lower tar products, I
indexed in a variety of ways, have reduced lung cancer risk (Wynder et al., :
1970; Wynder and Rabat, 1988). For example, in a recent report based on
cases from the late 1970's and early 1980's, risks were examined separately ■; !
for persons with squamous cell and sniall cell carcinomas of the lung j'

(Kreyberg 1) and adenocarcinoma of the lung (Kreyberg 11) (Wynder and
Rabat, 1988). .Smoking was classified as 100 percent filter, 100 percent j

nonfilter, or intermediate, by number of switchers from nonfilter to filter.

For smokers of filter cigarettes only, risks were approximately 10 to 30 percent
less than those of smokers of nonfilters only ( Fable 2). I

I'

;| Chapter 6

Table 1

Standardized mortality ratios for men in Cancer Prevention Study I for total
mortality, lung cancer, and coronary heart disease (CHD) by tar and nicotine
intake

Tar and Nicotine Intake

Deaths

High"

Medium‘s

Low*^

Total Deaths

1960-1966

1.00

0.90

0.88

1967-1972

1.00

0.98

0.81

Lung Cancer

1960-1966

1.00

0.96

0.83

1967-1972

1.00

0.94

0.79

CHD

1960-1966

1.00

0.91

0.93

1967-1972

1.00

1.03

0.82

® High = 2.0 to 2.7 mg nicotine and 25.8 to 35. 7 mg tar.
Medium = intermediate.

Low = <1.2 mg nicotine and tar generally <17.6 mg.
Source: Hammond et al., 1976.

Table 2

Adjusted odds ratios and 95-percent confidence intervals for males in the
American Health Foundation case-control study, by level of filter smoking

Pattern of Smoking

Tumor Type

Kreyberg 1

Kreyberg II

Odds

Ratio

95%

Confidence

Interval

Odds

Ratio

95%

Confidence

Interval

Nonfilter Only

1.00

1.00

—

Switchers (1-9 years)

0.83

0.59-1.17

0.96

0.61 - 1.51

Switchers (10-t- years)

0.66

0.49 - 0.90

0.79

0.53 - 1.18

Filter Only

0.69

0.37 - 1 .27

0.87

0.43 - 1 .54

Source: Wynder and Kabat, 1988.

A multicenter case-control study conducted in Europe during the late
1970's also provided information on cigarette type and lung cancer risk
(Lubin et al., 1984). In this study, risk for lung cancer increased progressively
! in both males and females as the proportion of filter use declined from

I 100 percent. Findings were similar in a case-control study that was

conducted in New Mexico from 1980 through 1983, although a linear

81

Smoking and Tobacco Control Monograph No. 7

dose-response relationship between lung cancer risk and the extent of filter
cigarette smoking was not observed (Pathak et al., 1986). Other recent case-
control studies have provided comparable results (Wilcox et al., 1988;
Kaufman et al., 1989).

Temporal patterns of lung cancer rates also have been interpreted as
indicating lower lung cancer risks among smokers of lower tar and nicotine
cigarettes. It has been suggested that the recent decline in lung cancer
mortality rates among younger males may reflect changes in the cigarette
(World Health Organization, 1986). This downturn has been observed in
the United States and other countries (Gilliland and Samet, 1994).

Nonmalignant Cigarette smoking has diverse effects on the structure and function of

Respiratory the lung and is a cause of chronic bronchitis and chronic obstructive

Diseases pulmonary disease (COPD) (U.S. Department of Health and Human

Services, 1984). The persistent obstruction to airflow in the lung that is the
hallmark of COPD reflects underlying changes in the small airways of the
lung and emphysema, which is the permanent destruction of the air spaces
of the lung. Chronic bronchitis, a condition of chronic sputum production,
reflects hyperplasia of the lining of the airways of the lung and mucous gland
proliferation. Compared with nonsmokers, smokers have a greater frequency
of cough and production of phlegm, manifestations of the inflammation of
the lung and increased mucus production secondary to smoking, and
wheezing; smokers also have lower lung function.

A significant number of adults in the United States have COPD, which
now causes more than 60,000 deaths annually (U.S. Department of Health
and Human Services, 1984). The natural history of this disorder has been
described through longitudinal investigations that have monitored lung
function over time in smokers and nonsmokers (U.S. Department of Health
and Human Services, 1984; Sherman et al., 1993). In nonsmokers, lung
function increases through late adolescence and early adulthood, maintains
a plateau across the third and fourth decades, and then begins to decline.

In smokers, the decline begins at a younger age and tends to be steeper.

The rate of decline increases with the number of cigarettes smoked per day
but varies widely among smokers. With continued smoking, those with
more rapid rates of decline eventually deteriorate to a level of lung function
associated with impairment, and COPD is diagnosed. Although cessation
earlier in the evolution of the disease is followed by return of the rate of
decline to that of nonsmokers (U.S. Dej)artment of Health and Human
Services, 1990), smoking cessation at this point in the natural history of
the disease is not followed by im[)rovement in lung function.

l indings have been reported that provide insights concerning tar and
nicotine yields and res[)iratory sym[)toms and lung function level. Auerbach
and c(41eagues (1979) quantitated smoking-related changes in the lungs of
men having autopsies at a Veterans Administration hospital in New Jersey. In
a rigorously investigated series of autopsied lungs, these investigators showed
that smokers from a period during which cigarettes had com{)aratively high
yields (A tar and nicotine (1955 to 1960) had more changes in the airways at

82

Chapter 6

various smoking levels compared with smokers from a later period (1970 to
1977). They interpreted this temporal pattern as indicating that cigarettes
with lower tar and nicotine yields had less effect on lungs than did higher
yield cigarettes.

A number of studies have shown that smokers of lower yield cigarettes
have comparatively lower rates of respiratory symptoms. Respiratory
questionnaire data collected in the late 1970's from approximately
6,000 Pennsylvania women are illustrative (Schenker et al., 1982). The brand
of cigarettes currently smoked was determined and used with FTC tar yield
information to classify the smokers by tar exposure. Tar yield was positively
associated with cough and phlegm but not with wheezing or shortness of
breath. For cough and phlegm, there were consistent exposure-response
relationships with an approximate doubling of symptom frequency from
the lowest to the highest exposure category (Table 3). The findings of other
studies are similar. For example, a large study of civil servants in the United
Kingdom, the Whitehall study, showed that the percentage of smokers
reporting phlegm increased with tar yield within each stratum of cigarettes
smoked per day, even the lowest (Higenbottam et al., 1980).

Table 3

Absolute and relative risks of chronic cough and chronic phlegm in Pennsylvania
women by smoking status, cigarettes smoked per day (CPD), and tar yield of
current brand

Chronic Cough

Chronic Phlegm

Smoker

Classification

Risk

Relative Risk

Risk

Relative Risk

Never-Smokers

0.038

1.00

0.033

1.00

Ex-Smokers

0.056

1.46

0.052

1.58

Current Smokers
1-14 CPD

7 mg tar

0.073

1.92

0.067

2.04

1 5 mg tar

0.103

2.71

0.085

2.56

22 mg tar

0.137

3.61

0.103

3.12

15-24 CPD

7 mg tar

0.136

3.58

0.155

4.67

1 5 mg tar

0.185

4.87

0.190

5.74

22 mg tar

0.240

6.32

0.226

6.82

25+ CPD

7 mg tar

0.273

7.18

0.234

7.05

1 5 mg tar

0.353

9.29

0.281

8.48

22 mg tar

0.430

11.32

0.327

9.87

Source: Schenker et al., 1982.

83

irmktNSi

Smoking and Tobacco Control Monograph No. 7

Respiratory morbidity also has been investigated. Follovymp of outpatient
visits by enrollees in a Kaiser-Permanente group over 1 year showed that there
was a reduced risk for pneumonia and influenza but not other respiratory
conditions, associated with use of low-tar and -nicotine products (Petitti and
Friedman, 1985a). However, in comparison with nonsmokers, smokers using
low-tar and -nicotine cigarettes had an increased risk for pneumonia,
influenza, and COPD.

Not all studies show less disease associated with lower yield cigarettes.

One recent study from Finland found that symptom levels in young smokers
who were just initiating smoking did not depend greatly on tar yield (Rimpela
and Teperi, 1989). In this 6-year followup study, the youths were surveyed
on several occasions, and the relationship between tar yield and symptom
onset was determined. There was little evidence of less symptom occurrence
in the new smokers using low-tar cigarettes in comparison with those
smoking higher tar cigarettes. Moreover, symptoms were far more frequent
in the smokers of low-tar cigarettes in comparison with nonsmokers. In a
randomized trial in the United Kingdom, lower tar cigarettes were not
associated with either lower symptom frequency or higher level of ventilatory
function, as assessed by measuring the peak expiratory flow rate (Withey et
al., 1992a and 1992b). The investigators monitored urinary nicotine
metabolites and concluded that compensation led to comparable levels
across the trial period.

The evidence does not suggest a relationship between tar yield and lung
function level. For example, in the Whitehall study (Higenbottam et al.,
1980), there was no cross-sectional relationship between tar yield and level
of the forced expiratory volume in 1 second. In the Normative Aging Study
(Sparrow et al., 1983), a longitudinal study of U.S. veterans, tar yield of the
usual brand of cigarettes smoked was not associated with decline of forced
expiratory volume in 1 second.

C:ardiuva.vcular Harris (this volume) discusses mechanisms by which cigarette

Disea.se smoking causes CVD. Through some of these mechanisms, cigarette

smoking is anticipated to increase the incidence of new cases (i.e., to cause
more disease), whereas other mechanisms are anticipated to exacerbate the
status of those who already had disease (U.S. Department of Health and
Human Services, 1990). Thus, factors promoting atherogenesis would
increase incidence, whereas factors such as sympathomimetic stimulation
by nicotine or impairment of oxygen delivery by carbon monoxide might
be expected to have more immediate effects and contribute to morbidity
and mortality among those with coronary artery disease.

Strong evidence does not exist for either lower incidence or less morbidity
from coronary heart disease (CUD) among smokers of lower yield cigarettes.

In the American Cancer Society's Cd’S-1 study (Hammond et al., 1976),
smokers (A lower tar |)roducts did have lower mortality from heart disease
('fable 1 ). On the other hand, two case-control studies carried out during

84

Chapter 6

the 1980's, one involving men (Kaufman et al., 1989) and the other
involving women (Palmer et al., 1989), did not show evidence of reduced
risk for smokers smoking lower nicotine products. Both studies included
persons with a first and nonfatal myocardial infarction. In the 1980-1981
study of men younger than 54, neither nicotine nor carbon monoxide yields
of current brand were associated with risk of myocardial infarction (Table 4).
From 1985 to 1988, a similar case-control study of women as old as 65 with
nonfatal myocardial infarction also showed no relationship between nicotine
or carbon monoxide yields of current brand of cigarettes and risk of
myocardial infarction (Table 5).

The study of Kaiser-Permanente enrollees also supplied relevant
information (Petitti and Friedman, 1985b). Hospitalization for a variety
of cardiovascular outcomes was assessed in relation to type of cigarettes
smoked, after adjusting for other predictors. Using a multivariate regression
model, the investigators found relatively small increases in risk for
hospitalization as tar yield increased.

Table 4

Relative adjusted risk of myocardial infarction in men by nicotine and carbon
monoxide yield of cigarettes smoked

Smoker Status

Relative Risk

95%

Confidence

Interval

Never-Smoker

1.0

2.5 - 6.7

Current Smoker
Nicotine yield (mg)

< 0.8

3.8

2.3 - 6.5

0.8-0.9

4.1

2.5 - 6.7

1. 0-1.1

3.4

2.2 - 5.3

1.2-1 .4

2.4

1.5 -3.8

^1.5

3.2

1.9 -5.6

Carbon monoxide yield (mg)

< 10

3.5

1.9 -6.6

10-14

4.4

2.6 - 7.5

15-17

3.2

2.1 -5.0

18

2.9

1.8 -4.5

a 19

3.3

1.8 -6.0

Source: Kaufman et al., 1983.

85

Smoking and Tobacco Control Monograph No. 7

Table 5

'Relative adjusted risk of myocardial infarction in women by nicotine and carbon
monoxide yield of cigarettes smoked

Smoker Status

Relative Risk

95%

Confidence

Interval

Never-Smoker

1.0

Current Smoker
Nicotine yield (mg)

< 0.40

4.7

2.8 -8.0

0.40-0.63

3.3

2.3 - 4.8

0.64-0.75

3.2

2.2 - 4.5

0.75-1.00

4.7

3.4 - 6.5

1.01-1.06

3.6

2.6 - 5.0

1 .07-1 .29

5.1

3.4 - 7.5

2:1.30

4.2

2.4 - 7.2

Carbon monoxide yield (mg)

< 4.8

4.9

2.9 - 8.2

4.8-9.1

4.4

2.4 - 4.9

9.2-11.1

3.8

2.7 - 5.4

11.2-14.4

3.8

2.7 - 5.2

14.5-15.0

4.1

2.9 - 5.7

15.1-18.0

4.2

2.9 - 6.2

> 18.0

4.8

2.8 -8.1

Source; Palmer et al., 1989.

CONCLUSIONS The Health Consequences of Smoking: The Changing Cigarette: A Report

of the Surgeon General (U.S. Department of Health and Human Services, 1981)
offered conclusions on these three major classes of disease. Do these
conclusions remain tenable in light of more recent evidence?

With regard to cancer, the report concluded that:

Today's filter-tii)ped, lower 'tar' and nicotine cigarettes produce
lower rates of lung cancer than do their higher 'tar' and
nicotine predecessors. Nonetheless, smokers of lower 'tar' and
nicotine cigarettes have much higher lung cancer incidence and
mortality than do nonsmokers (U.S. Department of Health and
Human Services, 1981, p. 18).

The more recent case-control evidence remains consistent with the first
compofient of this conclusion.

With regard to COIM), the report concluded that it was unknown
whether risk was lower for smokers of low-tar and -nicotine cigarettes

86

Chapter 6

compared with risk for smokers of higher tar and nicotine cigarettes. There is
no consistent evidence that risk for this disease is associated with the tar and
nicotine yield of the cigarettes smoked.

For CVD, the 1981 conclusion remains appropriate: . . . the
overall changes in the composition of cigarettes that have
occurred during the last 10 to 15 years have not produced a clearly
demonstrated effect on cardiovascular disease, and some studies
suggest that a decreased risk of CHD may not have occurred
(U.S. Department of Health and Human Services, 1981, p. 125).

Our research needs have changed little from the agenda set out
in that report 15 years ago. The report called for further surveillance
of the characteristics of smoke in relation to the type of cigarettes,
further characterization of compensatory changes in smoking, better
understanding of doses of tobacco smoke components delivered to the
lung, and additional epidemiologic research. Ongoing characterization of
the health consequences of the changing cigarette should be implemented
and maintained through cohort studies such as CPS-I or case-control
methods. New biomarkers of exposure and dose should be applied to
better understand the relationships of FTC tar and nicotine yields with
biologically effective doses of smoke components.

QUESTION-AND-ANSWER SESSION

DR. PETITTI: It actually does amaze me that the conclusions of this report
are the same as they were in 1981. It also amazes me how little information
has developed in this field over the past 14 years.

I wanted you to comment on an issue that was, I think, not particularly
well addressed in the 1981 report and has troubled me about the
epidemiological data. It has to do with the tendency to examine the
risk of lung cancer in strata defined by number of cigarettes smoked per
day. When you define smoking by number of cigarettes smoked per day,
you do take into account compensation by inhalation and amount smoked,
but you don't take into account any kind of compensation that might
occur because of a tendency to smoke an increased number of cigarettes
per day and smoking a lower yield brand. That would suggest that in order
to take that into account in the epidemiology, you would have to move
people to a different category of number of cigarettes smoked per day.

Do you think that epidemiology can address this issue, and how do you
think that places limitations on the first conclusion related to lung cancer,
particularly?

DR. SAMET: It is a good question and I think much of the discussion
about smokers' behavior that will follow will get at just how complex the
physiology is and how difficult it is to make these determinations in the
laboratory.

87

Smoking and Tobacco Control Monograph No. 7

Then, if you think about trying to develop approaches that might be
used in epidemiological studies, based around questionnaires to try to
develop tools that would provide a better measure of dose, which I think
is what you are calling for, it becomes very difficult.

You know, using some of our nested approaches, one might begin to use
biomarkers within studies, within cohort studies, probably particularly, to
sort this out. But 1 think you are pointing to a significant limitation of
approaching this question in large population studies.

DR. BENOWITZ: The biggest effect was clearly in the lung cancer data, and
the lung cancers occurred as a result of cigarettes smoked a long time ago.

Is there any evidence that there is any difference in risk if you looked at
modern or filtered cigarettes?

DR. SAMET: Let me see if 1 can rephrase the question. Are you asking, has
there been an attempt to assess whether some estimate of tar dose, or tar
received, is a better predictor of lung cancer risk than simply proportion of
filter use?

DR. BENOWITZ: Yes. What 1 am wondering is, is there any relevance to
the data when people were mostly smoking nonfiltered cigarettes to today's
cigarette market, where they are filtered? Can the whole thing be done just
by adding a filter?

DR. SAMET: Probably the right answer to the question is: 1 do not know.

But if we think we could begin to use the information from studies of
smokers of old nonfiltered products, through smokers of newer products,
to try and define some kind of an exposure-response relationship, then
I suppose it could be done. But 1 think that, if we were to do that, it would
be subject to a great deal of uncertainty.

DR. HARRIS: 1 noticed that one of the studies omitted from your review
was the second American Cancer Study, CPS-11, which followed people from
1982 to 1986. 1 am wondering if anyone knows whether that study will be
analyzed in terms of the yield or type of cigarette and health outcomes.

DR. SAMET: There has already been a paper describing the demographics of
tobacco use in that study and predictors of tar yield by various demographic
predictors. 1 would anticipate seeing such an analysis eventually.

DR. WOOSLEY: We have already heard this morning how the marketing and
the promotion of the low-tar and -nicotine cigarettes have been toward the
more highly educated portion of the population. We have already seen how
they responded to that by switching. We have already heard how they have
expressed greater concern for their overall health.

I have a serious concern. Do you feel the data have adequately addressed
the possibility that you are looking at a subset of the population who have
done something else to modify their health risks and, therefore, have looked
at a selected population with decreased negative outcomes because of these

88

Chapter 6

other factors, and that we really have not seen any influence of the cigarettes
themselves?

DR. SAMET: 1 referred to that set of concerns under the rubric of selection
bias. That is, people may select themselves to products based on either
their response to what they were smoking or other characteristics that are
relevant — an argument in epidemiology called confounding.

I think you are right; these are concerns. 1 think, on the other hand, in
many of the studies there have been attempts to "adjust," to the extent one
can, for such differences in the characteristics of those using different types
of products. As you look across the consistency of the evidence in different
populations with different approaches to controlling for such factors, and
different study designs, a consistency emerges, 1 think at least for lung
cancer, that would suggest some modest reduction of risk for those using
the lower delivery products.

Could there be some element of residual bias in there? 1 certainly
could not exclude it. But when we weigh the evidence in an attempt
to understand those other factors, the socioeconomic indices and other
measures in different studies would support that conclusion.

DR. BENOWITZ: I think that is the most important issue that we have to
address here today. If labeling something low-tar and -nicotine implies
improved health compared with higher tar and nicotine, I think those
confounders have the most impact on that decision.

DR. HOFFMANN: With regard to Dr. Benowitz' question to me, it is rather
interesting to see that multiple studies have shown that the increase in
adenocarcinoma today is much higher than previously, because the nature
has changed. So, to me, this has something to do with the cigarette. You get
more adenocarcinoma in the peripheral lung than in former times; it is a
ratio of 20 to 1 squamous cells, and today you have 1 to 1. So, I think at
least the type of lung cancer that appears today has something to do with
the change in cigarettes.

DR. SAMET: But certainly the histologic distribution of lung cancers has
changed and 1 agree; we would like to know why.

DR. DEBETHIZY: Your data about the relationship between nicotine and
cardiovascular disease are curious to me, because most of the data in the
literature show that people who smoke low-yielding cigarettes actually
absorb less nicotine. Could you comment on the fact that you do not see
any dose-response relationship there?

DR. SAMET: I am not sure how you would like me to comment. 1 am
describing the findings of a case-control study that describes how risks of
nonfatal myocardial infarction varied with the level of nicotine or carbon
monoxide intake, as estimated by what brand was being smoked at the
time of the infarct.

89

Smoking and Tobacco Control Monograph No. 7

These are not biomarker data, so there is no inference in these particular
subjects as to what the level of nicotine or carbon monoxide may have
been. The question is, again, looking at the yield or brand as an estimate of
exposure, there was simply no relationship observed in these observational
studies.

DR. HUGHES: In most of these studies, the control group is labeled
nonsmokers. Is that usually never-smokers?

DR. SAMET: In most of the studies that are labeled nonsmokers, that is
a never-smoker group. You basically will see two contrasts: vs. never-
smokers or, in some of the studies, the contrast has been made between
sort of the lower exposure group vs. the higher exposure group.

DR. HUGHES: The reason 1 asked that is, it seems to me that using controls
of ex-smokers would be important for two reasons. One, it would be a
control for the confounds that Dr. Woosley mentioned earlier. Second,
all your studies have to do with switching cigarettes. None of them has
to do with the alternative of either quitting or switching to a low-nicotine
cigarette. Are there data to inform the consumer of the question, how much
do 1 want to improve my health by quitting, vs. how much do 1 improve my
health by switching to a low-tar cigarette?

DR. SAMET: Certainly, there are abundant data on how risks of diseases vary
following cessation. 1 do not want to complicate this, and it was the subject
of the 1990 Surgeon General's report. These risks vary in complex ways for
different diseases, depending on the age at which the smoker stopped
smoking and the duration of successful abstinence from smoking.

So, it is somewhat difficult to capture a single number that describes the
risk in ex-smokers. It has to be done in a far more complex way. But, on
the other hand, there are data sets, like the American Cancer Society data
sets, that would allow one to describe how risks change following smoking
cessation, for example. And it would be possible to derive some quantitative
contrast between what might happen to smokers of different ages, different
prior smoking histories, with switching products vs. cessation.

DR. RICKERT: On your emphysema slide, the one that dealt with the
changes in lung function, there was a label that said, "never smoked and
not susceptible to the effects of tobacco smoke." Do you have any idea
what proportion of the population of smokers fell into the category "not
susceptible"?

DR. SAME'!’: Such numbers are not readily available. 1 think most people
who work in this field would guess that with regard to COED, perhaps 20 to
25 percent of continued smokers seemed to fall into this grouj) of ra[)id lung
function decline.

DR. RICKER T: Are there any |)ostulated mechanisms why smokers should
be in that grou[)?

90

Chapter 6

DR. SAMET: There are many postulated mechanisms, some of which
Dr. Harris already surveyed. They are essentially mechanisms having to
do with the balance between factors in the lung that injure it and those
that protect it, and how that balance may be shifted in individual smokers,
either by virtue of genetics or aspects of smoking, toward destruction rather
than susceptibility. It is the subject of a great deal of research.

DR. HEADEN: The next Surgeon General's report will be on smoking and
tobacco use among ethnic minorities. 1 want to remind the group that some
smoking patterns among ethnic minorities, particularly African-Americans,
differ substantially from smoking patterns of whites. For example, African-
Americans have extremely low daily rates of smoking, but they smoke very
high tar and nicotine cigarettes. Thus, it suggests that perhaps we need some
new data, oversampling for African-Americans and perhaps other ethnic
groups, particularly males, to find out what the relationships would be for
these subgroups.

DR. SAMET: 1 would certainly agree.

REFERENCES

Auerbach, O., Hammond, C., Garfinkel, L. Changes in
bronchial in relation to cigarette smoking, 1955-
1960 vs. 1970-1977. New England Journal of Medicine
300: 381-386, 1979.

Doll, R., Hill, A.B. Smoking and carcinoma of the
lung. British Medical Journal 2: 739-748, 1950.

Gilliland, F.D., Samet, J.M. Incidence and mortality
for lung cancer; Geographic, histologic, and
diagnostic trends. Cancer Surveys 19: 175-195, 1994.

Hammond, E.C., Garfinkel, L., Seidman, H., Lew, E.A.
Tar and nicotine content of cigarette smoke in
relation to death rates. Environmental Research
12(3); 263-274, 1976.

Higenbottam, T., Clark, T.J., Shipley, M.J., Rose, G.
Lung function and symptoms of cigarette smokers
related to tar yield and number of cigarettes
smoked. The Lancet 1(8165): 409-411, 1980.

Kaufman, D.W., Helmrich, S.P., Rosenberg, L.,
Miettinen, O.S., Shapiro, S. Nicotine and carbon
monoxide content of cigarette smoke and the risk
of myocardial infarction in young men. New
England Journal of Medicine 308(8): 409-413, 1983.

Kaufman, D.W., Palmer, J.R., Rosenberg, L., Stolley,

P., Warshaver, E., Shapiro, S. Tar content of
cigarettes in relation to lung cancer. American
Journal of Epidemiology 129(4): 703-711, 1989.

Lubin, J.H., Blot, W.J., Berrino, F., Flamant, R., Gillis,
C.R., Kunze, M., Schmahl, D., Visco, G. Patterns of
lung cancer risk according to type of cigarette
smoked. International Journal of Cancer 33(5): 569-
576, 1984.

National Research Council, Committee on Advances
in Assessing Human Exposure to Airborne
Pollutants. Human Exposure Assessment for Airborne
Pollutants. Washington, DC: National Academy
Press, 1991.

Palmer, J.R., Rosenberg, L., Shapiro, S. "Low yield"
cigarettes and the risk of nonfatal myocardial
infarction in women. New England Journal of
Medicine 320: 1569-1573, 1989.

Pathak, D.R., Samet, J.M. , Humble, C.G., Skipper, B.J.
Determinants of lung cancer risk in cigarette
smokers in New Mexico. Journal of the National
Cancer Institute 76(4); 597-604, 1986.

Petitti, D.B., Friedman, G.D. Respiratory morbidity in
smokers of low- and high-yield cigarettes. Preventive
Medicine 14: 217-225, 1985a.

Petitti, D.B., Friedman, G.D. Cardiovascular and other
diseases in smokers of low yield cigarettes. Journal of
Chronic Disease 38: 581-588, 1985b.

Rimpela, A., Teperi, J. Respiratory symptoms and low
tar cigarette smoking — a longitudinal study on
young people. Scandinavian Journal of Social
Medicine 17: 151-156, 1989.

Rothman, K.J. Modem Epidemiology. Boston, MA;

Little, Brown and Company, 1986.

Schenker, M.B., Samet, J.M., Speizer, F.E. Effect of
cigarette tar content and smoking habits on
respiratory symptoms in women. American Review
of Respiratory Disease 125: 684-690, 1982.

Sherman, C.B., Tollerud, D.J., Heffner, L.J., Speizer,

F.E., Weiss, S.T. Airway responsiveness in young
black and white women. American Review of
Respiratory Disease 148(1): 98-102, 1993.

91

Smoking and Tobacco Control Monograph No. 7

Sparrow, D., Stefos, T., Bosse, R., Weiss, S.T. The
relationship of tar content to decline in pulmonary
function in cigarette smokers. American Review of
Respiratory Disease 127(1); 56-58, 1983.

U.S. Department of Health and Human Services. The
Health Consequences of Smoking: The Changing
Cigarette. A Report of the Surgeon General. DHHS
Publication No. (PHS) 81-50156. Rockville, MD:

U.S. Department of Health and Human Services,
Public Health Service, Centers for Disease Control,
Office on Smoking and Health, 1981.

U.S. Department of Health and Human Services. The
Health Consequences of Smoking: Chronic Obstnictive
Lung Disease. A Report of the Surgeon General. DHHS
Publication No. (PHS) 84-50205. Rockville, MD:

U.S. Department of Health and Human Services,
Public Health Service, Centers for Disease Control,
Office on Smoking and Health, 1984.

U.S. Department of Health and Human Services.
Reducing the Health Consequences of Smoking: 25
Years of Progress. A Report of the Surgeon General.
DHHS Publication No. (CDC) 89-8411. Rockville,
MD: U.S. Department of Health and Human
Services, Public Health Service, Centers for Disease
Control, Centers for Chronic Disease Prevention
and Health Promotion, Office on Smoking and
Health, 1989.

U.S. Department of Health and Human Services. The
Health Benefits of Smoking Cessation. A Report of the
Surgeon General. DHHS Publication No. (CDC) 90-
8416. Rockville, MD: U.S. Department of Health
and Human Services, Public Health Service, Centers
for Disease Control, Centers for Chronic Disease
Prevention and Health Promotion, Office on
Smoking and Health, 1990.

U.S. Department of Health and Human Services.
Strategies To Control Tobacco Use In the United States:
a blueprint for public health action in the 1990's, D.M.
Bums, J.M. Samet, and E.R. Gritz (Editors). NIH
I>ublication No. 92-3316. Rockville, MD: U.S.
Department of Health and Human Services, Public
Health Service, National Institutes of Health,
National Cancer Institute, 1991.

Vineis, P., Caporaso, N. Tobacco and cancer:
Epidemiology and the laboratory. Environmental
Heakh Perspectives 103(2): 156-160, 1995.

Wilcox, H.B., Schoenberg, J.B., Mason, T.J., Bill, J.S.,
Stemhagen, A. Smoking and lung cancer: Risk as a
function of cigarette tar content. Preventive Medicine
17(3): 263-272, 1988.

Withey, C.H., Papacosta, A.O., Swan, A.V., Fitzsimons,
B.A., Burney, P.G., Colley, J.R., Holland, W.W.
Respiratory effects of lowering tar and nicotine
levels of cigarettes smoked by young male middle
tar smokers. 1. Design of a randomised controlled
trial. Journal of Epidemiology and Community Health
46(3): 274-280, 1992a.

Withey, C.H., Papacosta, A.O., Swan, A.V., Fitzsimons,
B.A., Ellard, G.A., Burney, D.G., Colley, J.R.,

Holland, W.W. Respiratory effects of lowering tar
and nicotine levels of cigarettes smoked by young
male middle tar smokers. 11. Results of a randomised
controlled trial. Journal of Epidemiology and
Community Health 46(3): 281-285, 1992b.

World Health Organization. Tobacco Smoking. lARC
Monographs on the Evaluation of the Carcinogenic Risk
of Chemicals to Humans. Lyon, France: International
Agency for Research on Cancer, 1986.

Wynder, E.L., Graham, E.A. Tobacco smoking as a
possible etiologic factor in bronchiogenic
carcinoma. A study of six hundred and eighty-four
proved cases. Journal of the American Medical
Association 143; 329-346, 1950.

Wynder, E.L., Rabat, G.C. The effect of low-yield
cigarette smoking on lung cancer risk. Cancer 62:
1223-1230, 1988.

Wynder, E.L., Mabuchi, K., Beattie, E.J. The

epidemiology of lung cancer. Recent trends. Journal
of the American Medical Association 213: 2221-2228,
1970.

92

Chapter 7

Biomarkers of Cigarette Smoking

Neal L. Benowitz

INTRODUCTION This chapter addresses the following question: To what extent do
smoking-machine-derived tar, nicotine, and carbon monoxide ratings of
cigarettes predict how much of those substances smokers actually absorb
into their bodies?

Two issues need to be clarified. First is the difference between delivery
and content: What a cigarette delivers to the smoker is not the same as what
is present in the cigarette tobacco. Second is the issue of compensation vs.
regulation or titration: Kozlowski and Pillitterri (this volume) focus on
compensation — the individual's smoking behavioral response to a change
in a cigarette brand; this chapter focuses on cigarettes that people have
self-selected to smoke. Whether behavioral adjustment to nicotine yields
indicates regulation or titration or compensation is not important. What is
important is the relationship between what people choose to smoke and
their intake of various tobacco-derived toxins.

USE OF The biomarkers most widely used to quantitate exposure to tobacco

VARIOUS smoke include nicotine, its metabolite cotinine, carbon monoxide,

BIOMARKERS and with less success, thiocyanate. Recent investigation has focused
on various hemoglobin and DNA (deoxyribonucleic acid) adducts and
excretion of nitrosamines in the urine. These latter measures represent
important future directions, but there are inadequate data in large enough
populations to make conclusions about the relationship between these
measures and U.S. Federal Trade Commission (FTC) yields. The use of
mutagenic activity of the urine is discussed to address the utility of the
tar-to-nicotine ratio that is computed from the "FTC method" in predicting
relative human exposure to tar and nicotine. This is an important
consideration in estimating human risks from different types of cigarettes.

NICOTINE Nicotine is rapidly absorbed from cigarettes. It enters arterial
ABSORPTION circulation first, then venous circulation; nicotine levels then fall
FROM relatively quickly as it is redistributed from the bloodstream to various

CIGARETTES body tissues. Subsequently, nicotine levels fall off with an elimination
half-life of about 2 hours (Benowitz, 1988).

The intake of nicotine from a single cigarette can be approximated by
measuring the nicotine blood concentration profile after a person smokes
a single cigarette. The area under the plasma concentration-time curve is
a reflection of systemic dose. The 24-hour nicotine consumption also can
be estimated. Volunteer smokers have been studied smoking cigarettes on a
research ward, where blood levels could be sampled frequently. Blood levels
rise with smoking in the morning, more or less plateau through the latter
part of the day, and then fall overnight (Benowitz and Jacob, 1984a).

Carbon monoxide levels also build up during the day, plateau, and then

Smoking and Tobacco Control Monograph No. 7

fall overnight. By sampling blood periodically throughout the day for
measurement of nicotine levels, it is possible to estimate daily exposure to
nicotine in human smokers.

The metabolic disposition of nicotine in humans has been determined
based on urine-recovery studies plus infusion studies of nicotine and cotinine
(Figure 1) (Benowitz et al., 1994). On average, about 70 to 80 percent of
nicotine is converted to cotinine, which is the main proximate metabolite.
Most studies of nicotine intake from cigarettes producing different yields
have used cotinine as the marker of nicotine intake. Cotinine is extensively
metabolized, primarily to hflMS-3'-hydroxycotinine. Nicotine, cotinine,
and hydroxycotinine also are conjugated as glucuronides. In the urine, a
relatively small amount of cotinine is excreted unchanged compared with
the total amount generated. However, in general, urine cotinine is well
correlated with plasma cotinine so that urine cotinine can be used as a
surrogate for plasma cotinine concentration Qarvis et al., 1984). Saliva
cotinine also is highly correlated with plasma cotinine and has been used
in the same way.

Plasma cotinine levels fluctuate somewhat throughout the day. There
is about a 15-percent change in cotinine levels from morning to night.

Figure 1

Quantitative scheme of nicotine metabolism, based on average excretion of metabolites as
percentage of systemic dose during transdermal nicotine application

4.2%

13.0%

9.8%

4.4%

0.4%

12.6%

2.4%

2.0%

7.4%

I

33.6% :

Note: Compounds in ovals indicate excretion in urine, and associated numbers indicate percentage of systemic dose
of nicotine.

Source: BenowiU et al., 1994

94

Chapter 7

reflecting the approximately 16-hour half-life of cotinine (Benowitz et al.,
1983a). Because of the relatively small circadian variation, cotinine levels
can be measured at various times of the day, and this value can be used as
representative of the average daily cotinine level.

It is possible, by measuring all the metabolites in the urine, to account
for an average of 90 percent of the nicotine dose (Benowitz et al., 1994).

An approach to estimating nicotine consumption is to measure all the
metabolites in the urine and sum them up. At steady state (where the rates
of intake of drug and generation of metabolism are the same as rates of
elimination of drug and metabolites), this sum of all metabolites in a
24-hour urine excretion reflects the amount of nicotine that a person takes
in each day.

NICOTINE CONTENT As noted earlier, cigarette content is not the same as cigarette
OF TOBACCO VS. yield or delivery. Figure 2 shows data from a 1983 study

FTC YIELD (Benowitz et al., 1983b) that investigated the nicotine content

of tobacco. The nicotine concentration of tobacco averaged 1.6 percent.

There was no relationship between nicotine content in the whole tobacco
rod and the FTC-predicted nicotine yield. There was a significant inverse
relationship between the concentration of nicotine and the FTC nicotine
yield. Thus, the yield as measured by smoking machine gives no information
whatsoever about the content of nicotine or other potential toxins in the
tobacco. The content of nicotine in the tobacco simply represents the
ultimate limit of the nicotine dose. The FTC method provides no information
about the amount of nicotine that could be obtained from the tobacco if
a person smoked it in a way to optimize intake.

QUANTITATING There are four general methods for quantitating the intake of
NICOTINE INTAKE nicotine from tobacco: (1) In circadian fashion, measure blood
IN SMOKERS nicotine levels during cigarette smoking (Benowitz and Jacob,

1984a and 1984b). If the clearance of nicotine also is measured by
intravenous infusion of nicotine, blood levels during smoking can be
converted to an absolute daily dose of nicotine. (2) The same can be done
with blood level data after a person has smoked one or two cigarettes
(Benowitz et al., 1991). (3) Blood cotinine levels during ad libitum cigarette
smoking have been used widely to estimate nicotine intake, which is
discussed below. (4) Finally, as mentioned by Byrd and colleagues (1995),
measuring urine nicotine and metabolites during ad libitum smoking can be
used to estimate nicotine intake. These four ways can be used to address the
question of how much nicotine is being taken into the body from smoking.

Table 1 presents a summary of data on the dose per cigarette from
the first three methods. The first method was used to study 44 smokers,
measuring blood levels during 24 hours of smoking, at steady state (Benowitz
and Jacob, 1984a, 1984b, and 1985). The dose was estimated to be about
1 mg per cigarette, with a range of 0.37 to 1.60 mg per cigarette.

95

Smoking and Tobacco Control Monograph No. 7

Figure 2

Nicotine content of cigarettes as compared with FTC-determined values
(regression analysis)

Note: “Nicotine intake per cigarette (mg)” indicates total amount of nicotine in the length of cigarette
tobacco rod smoked in the standard FTC smoking machine assay.

Key: NS = not significant.

Source: Benowitz et al., 1983b.

The second method is based on studies of persons smoking one or two
cigarettes (Benowitz et al., 1994). This method produced an average dose of
2.3 mg per cigarette, with a range of 0.37 to 3.47 mg. The study paradigm
was one in which smokers were deprived overnight and given only one or
two cigarettes to smoke in the morning. These were the only cigarettes
allowed all day. The unusually high dose per cigarette most likely reflected

96

Chapter 7

Table 1

Summary of absolute bioavailability of nicotine from cigarette smoking studies

Systemic Dose
(mg/cigarette)

Standard

Method

N

Average

Deviation

Range

Reference

1

22

1.04

0.36

0.37-1.60

Benowitz and Jacob, 1 984a

11

1.00

0.15

0.87-1.48

Benowitz and Jacob, 1 984b

11

0.90

0.15

—

Benowitz and Jacob, 1 985

2

10

2.29

1.00

0.37-3.47

Benowitz et al., 1991

3

20

0.87

0.41

0.22-1 .92

Benowitz and Jacob, 1994

the smokers' anticipation of no more cigarettes becoming available that day.

This finding illustrates the tremendous range of nicotine intake a smoker

has when there is a need, or an anticipated need, for nicotine. The intake

of nicotine per cigarette in this study was double that typically consumed ,

from ad libitum daily smoking. Consistent with this observation was

another study in which subjects tripled their intake of nicotine per cigarette

by smoking more intensely when the number of cigarettes allowed to be

smoked per day was limited (Benowitz et al., 1986a). "

The third method, that is, measuring blood cotinine concentrations, J

resulted in an estimated dose of about 0.9 mg of nicotine per cigarette, with 1

a range of 0.22 to 1.92 mg per cigarette (Benowitz and Jacob, 1994). What is ; ,

the quantitative relationship between nicotine intake and yield? Figure 3 "j

shows nicotine intake data from volunteer smokers studied whose plasma
nicotine levels were measured while they smoked their usual brand of ''

cigarettes ad libitum while in a research ward (Benowitz and Jacob, 1984a).

There was no correlation between the FTC-measured nicotine yield and
study-measured intake of nicotine. The only yield that turned out to be
accurate was 1 mg, which is fortuitous because it represents the average
consumption. Also, most smokers of nonfiltered cigarettes took in less
nicotine than predicted from the FTC yield. People who smoked low-yield
cigarettes took in, on average, more nicotine than predicted by FTC yield.

It is possible that in the 1940's and 1950's, when people smoked cigarettes
with a nominal yield of 2.5 mg or higher of nicotine, they may in fact have
been undersmoking those cigarettes and taking in considerably less smoke
per cigarette than they do now. That behavior might explain the change in
lung cancer pathology over the years. That is, a change in depth of inhaling
and intensity of smoking may affect the location of the lung tumor.

97

Smoking and Tobacco Control Monograph No. 7

Figure 3

Regression analysis of relationship between nicotine intake per cigarette and
machine-determined nicotine yield

FTC Machine Nicotine Yield (mg)

Key: solid line = regression line; broken line = line of identity, which indicates points at which
measured nicotine intake per cigarette equals machine-determined nicotine yield;

NS = not significant.

Source: Benowitz and Jacob, 1984a.

COTININE What is the quantitative relationship between cotinine levels and
LEVELS AND the intake of nicotine from smoking? To address this question, dual
NICOTINE infusions of deuterium-labeled nicotine and cotinine were given to

IN'I AKE smokers (Benowitz and Jacob, 1994). From resultant blood level data,

it was possible to calculate the percentage of nicotine that is converted to
cotinine and the clearance of cotinine per se. From these [)arameters can be
derived a factor (K) that relates a given blood level of cotinine at steady state

98

Chapter 7

to a given daily intake of nicotine, a factor that averages 0.08. Thus, for a
typical smoker with a level of 300 ng/mL, nicotine intake is estimated to be
24 mg per day. Based on average cigarette consumption, that represents an
intake of about 1 mg of nicotine per cigarette. This K factor did not vary as a
function of whether a person was a smoker or nonsmoker, brands of cigarette
smoked, or gender. Thus, the author is aware of no bias in using this K factor
to estimate the dose of nicotine based on a plasma cotinine concentration.

Data from a study of 136 smokers entering a smoking cessation program
are shown in Figure 4 (Benowitz et al., 1983b). There was a weak relationship
between FTC yield and cotinine level. The slope of this relationship was
shallow and, in this study, not significant. From the lowest to the highest
yield of cigarettes, there was only a 5- to 10-percent change in cotinine level,
reflecting a 5- to 10-percent change in nicotine intake. There was a much
stronger correlation between cigarettes per day and cotinine level (or nicotine
intake). Thus, the greater the number of cigarettes a person smokes, the more
nicotine is taken in. This is important because some studies, such as that of
Rosa and colleagues (1992), purport to show a strong relationship between

I Figure 4

; Afternoon blood cotinine concentrations (Group 1) as compared by regression analysis with the

Note: These smokers’ values were so similar that plots of individual values overlapped. The total number of subjects
shown in Panel B is lower because data for a few subjects were incomplete. Morning blood cotinine
concentrations (Group 2, not shown) were on average slightly lower but had similar correlations with the number
of cigarettes (r = 0.45) and the FTC yield (r = 0.06).

I Key: NS = not significant; 0=7 observation; • = 2 observations; A = 3 observations; □ = 4 observations.
i Source: Benowitz et al., 1983b.

99

t

!

Smoking and Tobacco Control Monograph No. 7

predicted nicotine intake and cotinine levels. But nicotine intake was
calculated by Rosa and colleagues as the multiple of cigarettes per day times
FTC yield. The strength of the relationship between this hybrid parameter !
and FTC yield derives primarily from the cigarettes-per-day term rather than
from the FTC-yield term.

Figure 5 presents data by Gori and Lynch (1985) based on a population I

of more than 800 smokers recruited at shopping malls. These were not '

smokers who were trying to stop smoking. Plasma cotinine and nicotine j
concentrations were measured. The average cotinine concentration was j
about 300 ng/mL. Again, there was only a shallow slope in the relationship |
between FTC nicotine and cotinine level, with little difference in cotinine
level comparing the lowest and the highest FTC nicotine yields. I

Figure 5

Mean plasma nicotine and cotinine concentrations as a fraction of FTC nicotine
yield of cigarettes smoked: n = 865

FTC Nicotine (mg/cigarette)

Key: solid line = mean; broken line = 95-percent confidence intervals.
Source: Gori and Lynch, 1985.

Plasma Nicotine (ng/mL)

Chapter 7

ULTRALOW-YIELD Yields from the ultralow cigarette may differ from yields from
CIGARETTES other cigarettes. Figure 6 shows data from another study of

smokers of cigarettes of different 5delds compared by category of FTC nicotine
(Benowitz et al., 1986b). The high-yield category was 1.0 mg of nicotine or
higher; the low was 0.60 to 0.99 mg; the very low was 0.20 to 0.59 mg; and
the ultralow was less than 0.20 mg nicotine per cigarette by FTC method.

I

i

!

I

i

I

Figure 6

Expired carbon monoxide, plasma thiocyanate, blood nicotine, and cotinine
concentrations in 248 habitual smokers of cigarettes according to FTC yield

40

35

30

25
30

25

20

175
150
125

20
15
10

350
300
250
200
150

<0.20 0.20-0.59 0.60-0.99 al.OO

FTC Nicotine Yield (mg)

® Significant difference compared with other yields.

Key: ppm = parts per million.

Source: Benowitz et al., 1986b.

101

Smoking and Tobacco Control Monograph No. 7

The ultralow cigarettes are typically rated 1 mg of tar or less. Smokers of
ultralow-yield cigarettes smoked on average a few more cigarettes per day
than other smokers. This appears to be one way in which these smokers are
compensating for lower nicotine yields. Of note, carbon monoxide levels !
were similar for all yields. Thiocyanate, nicotine, and cotinine levels were the
same for smokers of cigarettes with nicotine yields of 0.20 mg and higher.

Only in the ultralow group was there any reduction in nicotine exposure,
about 30 percent. Thus, cotinine levels produced by smoking ultralow-yield
cigarettes, instead of averaging 300 ng/mL, averaged about 200 ng/mL.

Gori and Lynch (1983) presented similar findings in a larger group of smokers.
Smokers of the low-yield cigarette brand had the same mean cotinine levels
as smokers of all other cigarette brands. In contrast, smokers of ultralow-yield
cigarettes had lower cotinine levels, averaging about 200 ng/mL. Note that
200 ng/mL still corresponds to a daily intake of 16 mg of nicotine per day. i
If the FTC yield of 0.1 mg nicotine per cigarette were correct, one would '

need to smoke 160 cigarettes per day to achieve an intake of 16 mg. These i

smokers were not smoking 160 cigarettes per day. Thus, the FTC information
on the ultralow-yield cigarette does not provide meaningful quantitative >

information, although there may be a difference between the ultralow- and i
higher yield cigarettes. fi

I

NICOTINE INTAKE Table 2 provides a summary of several studies of nicotine

AND MACHINE- intake vs. machine-derived yields. These are studies that have i

DETERMINED YIELD examined the relationship between FTC machine yield and [

nicotine intake measured either by cotinine concentrations or by nicotine j
concentration. Rickert and Robinson (1981) reported plasma cotinine i

concentrations vs. machine nicotine yield and found no relationship.

Russell and coworkers (1980) studied 330 subjects and found a weak
relationship between plasma nicotine concentration and yield. Benowitz
and colleagues (1983b) studied 272 smokers interested in smoking cessation
and found no significant relationship between plasma cotinine and yield. i
Lbert and coworkers (1983) found a shallow relationship between plasma
nicotine and yield. Gori and Lynch (1985) found a very shallow slope with t
865 subjects but also found significant relationships because of the large j
number of subjects. In a study by Benowitz and colleagues (1986b), cotinine I
concentrations were virtually the same for any cigarette with a yield of *■

0.2 mg and more and were a third less for the ultralow cigarettes. In a study I
by Russell and coworkers (1986), the 392 smokers studied showed a shallow ^
relationship between cotinine level and nicotine yield. Rosa and colleagues I
(1992) found a shallow slope relating cotinine level vs. ITC yield, similar to '|

that of other studies. However, when Rosa and coworkers (1992) combined I
cigarettes per day times I'FC yield, they found a strong relationship, which ‘
they interpreted as supporting the utility of the machine test method. f

In 298 Hispanics, Coultas and coworkers (1993) showed findings similar
to those of the other studies.

The Byrd and colleagues study (1995) was the only study with different '
results; thirty-three volunteers were asked to [)rovide 24-hour urine sam|)les
in wfiich nicotine and metabolites were measured. I'he nicotine intake was

102

Chapter 7

Table 2

Studies of nicotine intake compared with machine nicotine yield

iStudy

Population

Nicotine
Yields (mg)

Results

IRickert and Robinson,
‘ 1981

84 during routine medical
exams

0.25-1 .3

PCOT vs. Mach-N
r = 0.08

I Russell et al., 1980

330 from smokers’ clinics
or research volunteers

0.5-3.5

PNIC vs. Mach-N
r = 0.21®

iBenowitz et al., 1983b

1

1

i

272 seeking smoking
cessation therapy

<0.1-1 .9

BCOTvs. FTC-N
r = 0.15 (n = 137)
r = 0.06 (n = 123)

1 Ebert et al., 1983

76 — mix of smoking cessation,
hospital employees,
ambulatory patients

0.1-1.5

PNIC vs. FTC-N
r = 0.25®

j Gori and Lynch, 1 985

1

865 recruited from shopping
malls; 10 or more cigarettes
per day

0.1-1.6

PNIC vs. FTC-N
r = 0.37®

PCOT vs. FTC-N
r = 0.23®

Benowitz et al., 1986b

248 seeking smoking
cessation (1 37 from
previous study)

0.1-1.9

BCOT values similar for
FTC-N 0.21 to > 1.0
BCOT 2/3 of others for
FTC-N < 0.20

Russell et al., 1986

392 from smokers’ clinics

BCOT vs. Mach-N
r = 0.13®

BNIC vs. Mach-N
r = 0.26®

Rosa etal., 1992

1

125 attending military
medical center

0.38-1.38

BCOT vs. Mach-N
r = 0.30

1 Coultas et al., 1993

!

298 from Hispanic
household survey

—

SCOT vs. FTC-N
r = 0.12

! Byrd et al., 1995

I

1

33 volunteers

0.13-1.3

Urine N + metabolites
vs. FTC-N
N/24hr: r = 0.68®
N/cig: r = 0.79®

^ p < 0.05.

j Key: PCOT = plasma cotinine concentration; Mach-N = smoking-machine-determined nicotine yield; PNIC = plasma
j nicotine concentration; BOOT = blood cotinine concentration; FTC-N = machine yield by FTC method;

I BNIC = blood nicotine concentration; SCOT = saliva cotinine concentration; N = nicotine.

103

r-aai

Smoking and Tobacco Control Monograph No. 7

taken on the basis of total recovery. This study found a strong relationship
between yield and nicotine intake per cigarette per day that was totally
different from any other study's findings. There are serious methodological
concerns that might affect these conclusions. First, there were only
33 subjects, and recruitment procedures were unclear. In contrast, data
already presented involving 2,000 individuals have shown a weak or no
relationship between cotinine and FTC nicotine yields, so there is a problem
of generalizability of the Byrd data. Second, an examination of particular
data in the 1-mg tar group results in an average intake of 9 mg nicotine.
However, the studies of Gori and Lynch (1985) and Benowitz and colleagues
(1986b) showed an average cotinine concentration of 200 ng/mL for large
groups of smokers of the same ultralow-yield cigarettes. A cotinine level
of 200 ng/mL would correspond to an average daily intake of 15 or 16 mg,
not the 9 mg reported by Byrd and colleagues (1995). Thus, even if only
one group is studied, the subjects are not representative of the much larger
numbers of subjects that have been studied by other investigators.

CARBON Gori and Lynch (1985) have provided data on carbon monoxide levels in
MONOXIDE a large group of smokers of cigarettes of different yields (Figure 7). Their
AND FTC study and other studies have found virtually no relationship between

YIELD carbon monoxide levels and FTC yields, even for the ultralow group.

Thus, FTC carbon monoxide yield appears to be of no value in predicting
human carbon monoxide exposure.

TAR-TO-NICOTINE The tar-to-nicotine ratio also must be considered. Some authors
RATIO have argued that even if there is only a small reduction of

nicotine, because the machine tar-to-nicotine ratios are lower for low-yield
cigarettes, there is a disproportionately greater overall health benefit due
to reduced tar exposure (Russell et al., 1986). The question is whether
machine-determined tar-to-nicotine ratios predict ratios of exposure in
human smokers.

The author attempted to examine this question by studying mutagenic
activity of urine by Ames test. This test involves culturing salmonella bacteria
that are unable to generate histidine and therefore cannot grow. However,
if the bacteria are mutated so that they can make histidine, they can grow.
Growth can be quantitated by the number of colonies on a culture plate,
and the number of colonies can be used as a measure of mutagenic activity
of chemicals that were added to the salmonella culture before incubation.

It is well known that the urine of cigarette smokers is mutagenic, presumably
reflecting exposure to chemicals found primarily in cigarette smoke tar
(Yamasaki and Ames, 1977).

Figure 8 shows urine mutagenicity data from one individual smoking
his or her own brand of cigarettes who switched to a Camel (1 mg nicotine)
cigarette, then switched to a I'rue (0.4 mg nicotine) cigarette, and followed
with a period of no smoking (al)stinence). Urine mutagenicity was fairly
stable for the individual, and there was no difference between smoking the
Camel and True cigarettes.

104

Chapter 7

Figure 7

Mean expired air carbon monoxide values as a function of FTC carbon monoxide yield of
cigarettes smoked

Key: solid line = mean; broken line = 95-percent confidence intervals.
Source: Gori and Lynch, 1985.

Figure 9 shows data from a crossover study of urine mutagenicity and
nicotine intake from smokers smoking their own brand, high-yield (Camel),
and low-yield cigarettes (True) (Benowitz et al., 1986b). The mutagenic
activity of the urine was lower for both Camel and True compared with
the usual brand, most likely because smokers did not like these other brands
of cigarettes as much as they liked their own. However, the mutagenicity
values for Camel and True were similar. The ratio of mutagenic activity
over the 24-hour period under the nicotine plasma concentration-time curve
(the latter being an estimate of daily nicotine intake) was used as a surrogate
for tar-to-nicotine ratio and was the same in all conditions, although the
machine-predicted ratios were 14.8 for smoker's own, 15.1 for Camel, and
11.5 for True. Thus, the in vivo tar-to-nicotine ratio did not correspond to
differences in the machine-determined tar-to-nicotine ratios for different
brands.

Figure 9 also shows similar data when switching to Carlton, which is an
ultralow-yield cigarette. There was a small difference in the ratio of mutagenic

105

Smoking and Tobacco Control Monograph No. 7

Figure 8

Urinary mutagenicity based on 24-hour urine collections in a habitual smoker
smoking his or her own brand. Camel (high-yield), and True (low-yield) cigarettes

Own Camel True Abstinence

Note: Mutagenic activity tends to be constant from day to day and falls to the DMSO control value
(similar to that of nonsmokers) rapidly after stopping smoking.

Key: DMSO = dimethyl sulphoxide.

Source: Benowitz, 1989.

activity to tar for Carlton compared with other cigarettes, but the difference
was not near the values of 13.5, 15.4, and 7.3, which were predicted by FTC
values.

rhe data of Rickert and Robinson (1981) shown in Table 3 explain the
discrepancy between measured and predicted tar-to-nicotine ratios. These
data, based on smoking machine tests, show that when cigarettes are smoked
more intensely, the tar-to-nicotine ratio of low-yield cigarettes increases
substantially. Fhus, when smokers compensate for low-yield cigarettes by
smoking them more intensely, the tar-to-nicotine ratio increases. I'herefore,
tar-to-nicotine ratios published by the FT’C method cannot be used to make
estimates of what the overall tar exposure will be for actual smokers.

CX)NCLUSIONS The suggestion that there is a meaningful quantitative relationship
between FTC-measured yields and actual intake is misleading. I'here do
appear to be differences in nicotine ex[)osure comparing high- vs. low-yield

I Chapter 7

II

Figure 9

Average urine mutagenicity and ratio of mutagenic activity to nicotine exposure for subjects in
high-low yield (Group 1) and high-ultralow yield (Group 2) studies

Group 1 Group 2

^Indicates significant difference (p < .05, repeated measures analysis of variance) compared with own brand.
“Indicates significant difference compared with high-yield cigarettes.

Note: Despite lower ratios of tar and nicotine based on FTC testing for low- and ultralow-yield cigarettes, ratios of
mutagenic activity to nicotine exposure were not different while subjects smoked high-, low-, or ultralow-yield
cigarettes. Also, bars indicate standard error of the mean.

Key: AUC^i^ = ratio of mutagenic activity to nicotine exposure; rev = revertant colonies: nic = nicotine; O = smoker's
own brand; H = high-yield cigarettes (Camel): L = low-yield cigarettes (T rue); UL = ultralow-yield cigarettes
(Cambridge or Carlton).

Source: Benowitz et al., 1986b.

Table 3

Influence of intensity of smoking on tar-to-nicotine ratio, based on smoking machine studies

Standard Yield Tar-to-Nicotine Ratio Under

(mg) Different Smoking Conditions

Group

N

Tar

Nicotine

Standard

Moderate

Intensive

1

4

< 2

<2

9.2

9.9

11.1®

II

10

2-5

0.2 -0.5

10.3

11.7®

12.2®

III

8

5-10

0.5 - 0.9

11.3

11.9

12.6®

IV

9

10-14

0.8 - 1.0

12.7

13.3®

12.4

V

5

14-17

0.9- 1.0

15.7

16.5®

14.7

“p <0.05 compared with standard smoking machine conditions.
Key: N = number of brands tested.

Source: Rickert et al., 1983.

107

Smoking and Tobacco Control Monograph No. 7

cigarettes, but the differences are small and not quantitatively proportional
to nominal yield. Tar and nicotine ratings are poor predictors of human
intake, except for those cigarettes that happen to be rated by smoking
machines as 1 mg nicotine per cigarette, in which case that rating
fortuitously fits the population average. Tobacco manufacturers have
stated that the FTC method was never intended to measure intake in any
individual. The author agrees. However, data for 2,000 people summarized
here indicate that the FTC method does not work for the general population
of smokers either.

In general, the FTC method underestimates human exposure. Smoking-
machine-derived tar-to-nicotine ratios, which have been used to argue the
benefit of switching to low-yield cigarettes, are not of value because these
ratios change with changes in smoking behavior. On the other hand,
because there is some relationship between yields and nicotine, and although
the slope of that relationship is shallow, it is not recommended that smokers
regress to smoking higher yield cigarettes.

QUESTION-AND-ANSWER SESSION

DR. RICKERT: When we first looked at this back in 1981, there was
absolutely no relationship between yields and uptake as measured by
cotinine. And then, as you follow the studies from 1981 through to 1994,
there seems to be a growing tendency toward an association of some sort,
and at the end you pointed out there was a shallow slope. Is this just a
spurious change with time, or do you feel this may be related to changes in
product characteristic because, obviously, the product that was smoked in
1981 is not the product that is smoked in 1994.

DR. BENOWITZ: Yes, when you look at the earlier studies, they show
basically the same slope that studies that were done in 1989 and 1990 are
showing. So, I think when we have a large enough population, we are
probably seeing it even back in the 1980's. Prior to that, 1 have no idea.

DR. BOCK: Dr. Benowitz, you were quoted as saying something to the effect
that compensation over the long term does not appear to be persistent. Is
that your opinion?

DR. BENOWITZ: Fhat statement was made in dealing with the question of
when people are shifting from one cigarette to a lower yield cigarette, will
there be permanent overcompensation? And the only studies that 1 found (1
think Dr. Kozlowski is going to talk about these) looked at carbon monoxide
levels and the amount of compensation. At least in one study, carbon
monoxide levels went up and then went down again.

But if you look at the issue of compensation broadly, how do you
interj)ret the fact that people smoke a cigarette with an FFC yield of .2 the
same as the one that has the FFC nicotine yield of 1 .5, and tliey have the
same nicotine cotinine levels? If you do not call it comj)ensation, you have
to think of something to call it.

108

Chapter 7

At some point in time, I do think people will be limited by how much
smoke they take into their lungs. I do not think it is relevant to modern
cigarettes as currently marketed, but it could be relevant to a low-nicotine
cigarette.

DR. BOCK: You made a distinction just now and said "overcompensation."
Is that what you intend to imply?

DR. BENOWITZ: Yes.

DR. BOCK: Because there is a little bit of a difference between
overcompensation and compensation.

DR. BENOWITZ: Yes, I know. It is a good point.

DR. HATSUKAMI: In the studies where you looked at the FTC yield and the
actual intake, have any of the studies differentiated people who actually
initiated with low-tar and low-nicotine cigarettes and those who switched?
Are there any differences in terms of slopes between those two groups of
people?

DR. BENOWITZ: I have never seen that, but obviously that is a very
interesting question in terms of initiation. The earlier data that we heard
from Dr. Giovino suggest that most low-yield cigarette smokers are people
who switched from higher yield, which I think is quite interesting. But I
do not know the percentage of people who start with low-yield cigarettes.

It would be a good question.

DR. DEBETHIZY: Would you say that, on average, the people who smoke
lower yielding cigarettes absorb less nicotine?

DR. BENOWITZ: Yes, but the slope is very shallow.

DR. DEBETHIZY: So, if people are smoking very low-yielding cigarettes, they
are absorbing less nicotine and the data do speak to that. So, compensation
is incomplete; there is not a flat line.

One of the studies that you pointed out up there said that people
absorbed, on average, 1 mg of nicotine from cigarettes. And I think that
it is important to point out that people who smoke lower yielding cigarettes
do absorb less nicotine.

DR. BENOWITZ: Yes, although it is unclear where the break is. Some of our
data have suggested that the break is actually with the very, very low-yield
cigarettes, rather than the cigarettes most people smoke. But I would accept
the fact that there is a shallow relationship. Understand, however, that you
are talking about a 10-percent variation in nicotine intake, going across
yields from 0.1 to 1.6. So, there is some reduction in nicotine intake per
cigarette on average, but it is very small.

DR. DEBETHIZY: The other point I want to make is, you do rightfully point
out that I will discuss a little later why our study, which is the Byrd study,
may be unique from the plasma cotinine studies, in the fact that it is done

109

Smoking and Tobacco Control Monograph No. 7

with a completely different method and what we think is a more precise
method.

DR. WOOSLEY: 1 think the point you make is that the predictive accuracy of
any yield numbers, except at the low end of the scale, is useless, and 1 think
that is the most important message that 1 got out of your presentation.

The other message 1 have gotten out of the presentation was that this
indicates that the differences in mortality that we saw earlier, which had the
potential to be confounded, are very likely to be confounded because of the
lack of a difference in exposure that your data indicate.

DR. BENOWITZ: 1 would certainly agree with that. On the coronary heart
disease data, for example, with the low-yield cigarettes, I think you can
virtually assume that their exposure to everything was substantially the same.
It is no surprise that there is no protective effect of smoking low-yield
cigarettes for heart disease.

DR. RICKERT: How relevant do you feel the absorption of nicotine is to the I
levels of other harmful constituents, such as benzo(a)pyrene and biphenyl,
which are probably more related to disease processes than nicotine per se?

DR. BENOWITZ; I think there is considerable variability, and one has to look i

at that issue. I don't think there are enough data to know for the range of !

products. I

REFERENCES

Benowitz, N.L. Pharmacologic aspects of cigarette
smoking and nicotine addiction. New England
Journal of Medicine 319: 1318-1330, 1988.

Benowitz, N.L. Dosimetric studies of compensatory
cigarette smoking. In: Independent Scientific
Committee on Smoking and Health: Nicotine, Smoking,
and the Low Tar Programme, London, November 18-
20, 1986, N. Wald and P. Froggatt (Editors). Oxford:
Oxford University Press, 1989, pp. 133-150.

Benowitz, N.L., Hall, S.H., Herning, R.I., Jacob, P. Ill,
Jones, R.T., Osman, A.-L. Smokers of low-yield
cigarettes do not consume less nicotine. New
England joumal of Medicine 309: 139-142, 1983b.

Benowitz, N.L., Jacob, P. III. Daily intake of nicotine
during cigarette smoking. Clinical Pharmacology and
Therapeutics 35: 499-504, 1984a.

Benowitz, N.L., Jacob, P. III. Nicotine and carbon
monoxide intake from high- and low-yield
cigarettes. Clinical Pharmacology and Therapeutics
36: 265-270, 1984b.

Benowitz, N.L., Jacob, P. 111. Nicotine renal excretion
rate influences nicotine intake during cigarette
smoking. Journal of Pharmacology and Exjierimental
Therapeutics 234: 153-155, 1985.

Benowitz, N.L., Jacob, P. III. Metabolism of nicotine to
cotinine studied by a dual stable isotope method.
(Uinical PharmuLology and Theraj)eutics 56(5): 483-
493, 1994.

Benowitz, N.L., Jacob, P. Ill, Denaro, C., Jenkins, R.

Stable isotope studies of nicotine kinetics and
bioavailability. Clinical Pharmacology and
Therapeutics 49: 270-277, 1991.

Benowitz, N.L., Jacob, P. Ill, Fong, L, Gupta, S. |

Nicotine metabolic profile in man: Comparison I
of cigarette smoking and transdermal nicotine.

Jourtjal of Pharmacology and Experimental i

Therajreutics 268: 296-303, 1994. |

Benowitz, N.L., Jacob, P. Ill, Kozlowski, L.T., Yu, L. j
Influence of smoking fewer cigarettes on exposure I
to tar, nicotine, and carbon monoxide. New j

England Journal of Medicine 315: 1310-1313, 1986a. |

Benowitz, N.L., Jacob, P. Ill, Yu, L., Falcott, R., Hall, |
S., Jones, R.'F. Reduced tar, nicotine, and carbon j
monoxide exposure while smoking ultralow- but ji
not low-yield cigarettes. Journal of the American
Medical Association 2S6: 241-246, 1986b.

Benowitz, N.L., Kuyt, F., Jacob, P. Ill, Jones, R.'F., >

Osman, A.-L. Cotinine disposition and effects.

Clinical Pharmacology and Therajreutics 34: 604-61 1,

198.3a. I

Byrd, G.D., Robinson, J. 11., (Caldwell, W.S., |

DeBethizy, J.D. (lomparison of measured and |

FTC-predicted nicotine uptake in smokers.
Psychopharmacology \22: 9S-U)3, 1995.

no

Chapter 7

Coultas, D.B., Stidley, C.A., Samet, J.M. Cigarette
yields of tar and nicotine and markers of exposure
to tobacco smoke. American Review of Respiratory
Disease 148: 435-440, 1993.

Ebert, R.V., McNabb, M.E., McCusker, K.T., Snow, S.L.
Amount of nicotine and carbon monoxide inhaled
by smokers of low-tar, low-nicotine cigarettes.
Journal of the American Medical Association 250:
2840-2842, 1983.

Gori, G.B., Lynch, C.J. Smoker intake from cigarettes
in the 1 mg Federal Trade Commission tar class.
Regulatory Toxicology and Pharmacology 3: 110-120,
1983.

Gori, G.B., Lynch, C.J. Analytical cigarette yields as
predictors of smoke bioavailability. Regulatory
Toxicology and Pharmacology 5: 314-326, 1985.

Jarvis, M., Tunstall-Pedoe, H., Feyerabend, C., Vesey,
C., Salloojee, Y. Biochemical markers of smoke
absorption and self reported exposure to passive
smoking. Journal of Epidemiology and Community
Health 38(4): 335-339, 1984.

Rickert, W.S., Robinson, J.C. Estimating the hazards of
less hazardous cigarettes. 11. Study of cigarette yields
of nicotine, carbon monoxide, and hydrogen
cyanide in relation to levels of cotinine,
carboxyhemoglobin, and thiocyanate in smokers.
Journal of Toxicology and Environmental Health 7:
391-403, 1981.

Rickert, W.S., Robinson, J.C., Young, J.C., Collishaw,
N.E., Bray, D.F. A comparison of the yields of tar,
nicotine, and carbon monoxide of 36 brands of
Canadian cigarettes tested under three conditions.
Preventive Medicine 12: 682-694, 1983.

Rosa, M., Pacific!, R., Altieri, L, Pichini, S., Ottaviani,
G., Zuccaro, P. How the steady-state cotinine
concentration in cigarette smokers is directly
related to nicotine intake. Clinical Pharmacology
and Therapeutics 52: 324-329, 1992.

Russell, M.A., Jarvis, M.J., Feyerabend, C., Saloojee, Y.
Reduction of tar, nicotine and carbon monoxide
intake in low tar smokers. Journal of Epidemiology
and Community Health 40(1): 80-85, 1986.

Russell, M.A.H., Jarvis, M., Iyer, R., Feyerabend, C.
Relation of nicotine yield of cigarettes to blood
nicotine concentrations in smokers. British Medical
Journal 5: 972-976, 1980.

Yamasaki, E., Ames, B.N. Concentration of mutagens
from urine by absorption with the nonpolar resin
XAD-2: Cigarette smokers have mutagenic urine.
Proceedings of the National Academy of Sciences of the
United States 74: 3555-3559, 1977.

ACKNOWLEDGMENTS This research was supported by grants DA-02277, CA-32389,
DA-01696, and RR-00083 from the National Institutes of Health. The author
thanks David Greene for editorial assistance.

Ill

Chapter 8

Pharmacology and IVIarkers: Nicotine
Pharmacology and A^ddictive Effects

Jack E. Henningfield and Leslie M. Schuh

INTRODUCTION Dosing characteristics of cigarette brands are estimated using

machines that smoke representative cigarettes from each brand according
to a protocol termed the Federal Trade Commission (FTC) method (Peeler,
this volume; Pillsbury, this volume). This technology and methodology
provide tar- and nicotine-dosing estimates of cigarettes that are misleading
to consumers and do not accurately predict what level of tar and nicotine
intake consumers will obtain by smoking a given brand of cigarettes
(Henningfield et al., 1994). An understanding of the dependence-producing
and other behavior-modifying effects of cigarette smoke is necessary to
understand why the FTC method is a poor predictor of the nicotine, tar,
and carbon monoxide levels people obtain from cigarettes. Cigarette
smoking behavior is influenced by nicotine dose, and smokers tend to
maintain nicotine intake within upper and lower boundaries (Kozlowski,
1989). In brief, nicotine produces dose-related tolerance, physical
dependence, and discriminative effects (i.e., effects that people can feel,
which modify mood and physiology), and smokers change their behavior
in response to these effects. Unlike human smokers, machines are not
nicotine dependent, nor do they modify their behavior based on the flavor
of the smoke.

The FTC method was developed in the 1960's to provide a relative
ranking of nicotine, tar, and carbon monoxide yields from various cigarettes
(Peeler, this volume; Pillsbury, this volume). This ranking has provided
consumers with the false sense that they can tell precisely the amount of
these substances they will obtain from a given cigarette. Since the 1960's
there have been many advances in the understanding of nicotine and
smoking behavior that can be useful in reforming this methodology. This
chapter provides an overview of relevant research, including (1) physiological
and behavioral pressures to sustain nicotine intake; (2) the relationship
between smoking and nicotine dose; (3) determinants of compensatory
behavior, including the role of nicotine and other factors, such as flavor;
and (4) measurement of smoking and nicotine intake.

CIGARETTE
SMOKING AS
DRUG DEPENDENCE

Addiction Severity

Several findings bear on the issue of the strength of dependence
on cigarettes. Although 70 to 90 percent of smokers are
interested in quitting, only one in three succeeds before age 65
(Fiore, 1992). There is good and bad news about coronary
bypass surgery and even a lung removal. The good news is
that these traumatic events are among the most powerful incentives to quit
smoking. If one intervenes with patients who undergo these procedures,
about one-half of them quit. However, the bad news is that the other half

113

Smoking and Tobacco Control Monograph No. 7

or more soon return to smoking (U.S. Department of Health and
Human Services, 1988). There are two lessons here. First, incentives and
motivation are important factors in the treatment of nicotine and other
drug dependencies. Second, incentives and motivation have limitations;
even the threat of death is not sufficient for half these smokers to stop
smoking.

This is a tenacious addiction in which, despite so many people wanting
and trying to quit, fewer than 1 in 10 has a 1-year success, and this means
that only 2 to 3 percent of smokers stop smoking each year (Fiore, 1992).
Indeed, as Kozlowski and colleagues (1989) show, more than half of heroin
and cocaine users and alcoholics rate smoking cigarettes as harder to give
up than these other drugs. Thus, there are strong biological pressures in
nicotine-dependent humans that do not exist in machines to sustain
addictive levels of nicotine intake.

Clinical As with dependence on other drugs, cigarette smoking tends to be a

Characteristics progressive, chronic, relapsing disorder (U.S. Department of Health
and Human Services, 1988). The most notable distinction between cigarette
smoking and other drug dependencies is that a much higher percentage of
people who start smoking escalate and graduate to dependent levels than
with other addictive drugs. About 1 in 10 smokers in this country is a low-
level smoker, termed a "chipper," who smokes 5 or fewer cigarettes per day
(U.S. Department of Health and Human Services, 1988); most of the rest
show evidence of dependence. This is in contrast to alcohol use, where
10 to 15 percent of alcohol drinkers are problem drinkers; the rest generally
drink in moderation and at times of their own choosing (U.S. Department
of Health and Human Services, 1988).

People do not start smoking a pack of cigarettes per day. They likely
would become ill at that level of nicotine intake. Rather, they start out
with low levels. Over months and years, most people progress to higher
and higher nicotine intake. They become tolerant; that is, nicotine loses
effectiveness with its continued presence in the body, and it is necessary to
increase the dose to maintain its effectiveness after repeated administrations.
Eventually, smokers do more than simply tolerate high nicotine doses; they
need continued nicotine to feel normal and function satisfactorily. At this
point, smokers may go to great lengths to continue smoking and sustain
their nicotine intake within upper and lower boundaries so that their intake !
does not fall low enough that they experience withdrawal symptoms or !
high enough to produce adverse effects (Kozlowski, 1989).

An important aspect of the chronic nature of tobacco dependence is
related to daily patterns of nicotine blood levels. When smokers wake up
in the morning, some residual nicotine remains in their blood from smoking
on the previous day. Blood concentrations rise as they smoke until, by
midafternoon, most smokers' intake equals metabolism and excretion, and
nicotine level stabilizes, l.evels fall rapidly overnight, and the cycle resumes
the next day. I'hus, blood concentrations never reach zero unless the person
quits smoking for more than a few days. Moreover, cotinine, an active

1 14

Chapter 8

nicotine metabolite, has a half-life of about 20 hours (Cummings and
Richard, 1988; Jarvis, 1989; Jarvis et al., 1987) and therefore persists in the
body even longer.

It is difficult to disrupt these patterns when people have access to
cigarettes. In a study by Benowitz and colleagues (1986a), people switched
from 30 to 5 cigarettes per day. Because they tended to smoke these
5 cigarettes much more intensely, they reduced carbon monoxide levels
by only one-half and nicotine levels by only about one-third. Thus, nicotine
intake remained high enough to sustain dependence.

After quitting smoking, most people relapse quickly, and about one-third
of the people who have quit smoking and remained abstinent for 1 year
relapse (Fiore, 1992). As with alcohol and heroin, most nicotine relapses
occur during the first 3 months of abstinence (Hunt et al., 1971). In fact, the
determinants of relapse (e.g., degree of dependence and negative emotional
states) and remission (e.g., substance-associated health problems and learning
to manage cravings) are also similar across these three classes of drug
dependence (U.S. Department of Health and Human Services, 1988).

Relapse to nicotine dependence has been studied in greater detail than
relapse to heroin, cocaine, and alcohol dependence. Data from a Mayo Clinic
study showed that, with minimal treatment intervention, one-quarter of the
people relapsed in 2 days and about one-half in the first week (Kottke et al.,
1989). More recent data on people who quit on their own showed that about
two-thirds relapse within 3 days (Hughes et al., 1992). The withdrawal
syndrome can be debilitating in its own right, but in the long run, its worst
health consequences may be that most efforts to quit smoking never survive
the withdrawal phase (Hughes et al., 1992), thereby dooming one-half of
persistent smokers to die prematurely because of their tobacco use (Peto et al.,
1994). Much of the benefit of current nicotine medications is providing
adequate nicotine replacement for that formerly provided by cigarettes to
help more people remain nonsmokers during the important first few weeks
of tobacco abstinence.

NICOTINE Tobacco products come in many different forms. All have toxicities and
DELIVERY dependence potential, and there is variation related to the type of tobacco
SYSTEMS product and route of administration. Although the focus here is on

cigarettes, at some point similar issues must be addressed with other tobacco
products that currently have no dosage labeling. For example, moist snuff
products vary widely in their nicotine-dosing capabilities, and there is
evidence that the variation is accomplished primarily by manipulation
of the pH level of the products by tobacco manufacturers (Henningfield
et al., 1995; Djordjevic et al., 1995), but neither tobacco companies nor
governmental agencies provide any form of nicotine dosage information
to consumers except in cigarette advertising.

The cigarette, which may be conceived of as a nicotine dispenser with
smoke as the vehicle, is the most toxic and dependence-producing form of
nicotine delivery. Nicotine is volatilized at the tip of a burning cigarette from

115

Smoking and Tobacco Control Monograph No. 7

which it is carried by particulate matter (tar droplets) deep into the lungs
with inspired air. The nearly 2,000 °F microblast at the cigarette's tip is also
the source of carbon monoxide and many other toxicologically significant
pyrolysis products. Nicotine is rapidly absorbed in the alveoli of the lungs,
concentrated in the pulmonary veins as a bolus, and pumped by the left
ventricle of the heart throughout the body. Absorption characteristics are
similar to those of gases, such as oxygen, that are exchanged in the lung
from inspired air to venous blood (Henningfield et al., 1993). Thus, smoke
inhalation produces arterial boli that may be 10 times more concentrated
than the levels measured in venous blood (Henningfield et al., 1990 and
1993).

Psychoactive effects have rapid onset and short duration, dissipating
within a few minutes. This short duration requires the user to self-administer
the drug repeatedly, perhaps taking hundreds of puffs per day. The cigarette
allows the smoker very fine, "fingertip," dose control. The powerful engulfing
sensory effects are also important in dependence. It is not just the drug but
the conditions and the cues that become associated with the drug that make
nicotine dependence so tenacious. Finally, the cigarette is a convenient,
portable system that permits easily repeated dosing.

Benowitz (this volume) reviewed the pharmacokinetics of various
nicotine delivery systems. Briefly, a cigarette produces a rapid spike of
nicotine in the arterial blood. Smokeless tobacco products are also rapid,
especially the higher pH tobacco products, and they require little practice
for the user to achieve high nicotine levels. Whereas the nicotine dose
obtained from a cigarette is largely determined by the behavior of the user,
the nicotine dose obtained from a "chew" of smokeless tobacco is largely
controlled by the product (Henningfield et al., 1995). In contrast to delivery
from tobacco products, delivery of nicotine from polacrilex (nicotine gum)
is slower and takes a great deal of practice and work to achieve even modest
nicotine plasma levels. Transdermal nicotine medications (patches) provide
slow absorption — so slow that users cannot reliably detect nicotine's effects.
Fhe speed of delivery is clearly an important determinant of addictive effects,
and the cigarette, like crack cocaine, provides an explosive dose of nicotine.

NK^OTINE'.S Nicotine is a fascinating psychoactive drug. It was used to help map the

EFFECT'S cholinergic nervous system early in the 20th century. Much of receptor
theory and many of the methods used to study competitive agonists and
antagonists were developed at the turn of the century using nicotine (Langley,
1905).

Nicotine has diverse effects, not only in the brain but also in the
adrenals at)d skeletal muscles. I'hese diverse effects may ex{)lain why a
smoker re[)orts that on some occasions cigarettes have relaxing effects and on
other occasions, stimulating effects. Lhis has been referred to as a [raradoxical
effect, but it is not paradoxical at all; other drugs generally referred to either
as sedatives or stimulants also produce both sedating and stimulating effects
((hhnan et al., 1990). Like tfie effects of these other drugs, nicotine's effects
are conqrlicated; tliey de[)end on tfie dose, tfie time since dosing, how the

Chapter 8

drug was administered, which responses are being measured, and other
factors (Henningfield and Keenan, 1993; Pomerleau and Rosecrans, 1989).

If people with histories of drug abuse are given nicotine, they like the
nicotine; that is, liking scale scores increase with greater doses within a
certain range of parameters (Henningfield et al., 1985). Among drug abusers,
similar findings are reported for morphine and amphetamines but not for
drugs that have little psychoactivity (Fischman and Mello, 1989). Such
psychoactive effects are predictive of addiction potential and are correlated
with the ability of a drug to serve as a reinforcer for animals and humans
(Griffiths et al., 1980). Nicotine is psychoactive in humans and is readily
discriminated by animals; several forms of nicotine delivery have been shown
to serve as reinforcers for animals and humans (U.S. Department of Health
and Human Services, 1988).

Physical The cellular and neurological changes that lead to tolerance

Dependence also lead to physical dependence so that when people abruptly

and Withdrawal discontinue tobacco use, withdrawal occurs (U.S. Department of
Health and Human Services, 1988). Withdrawal onset begins within a few
hours of the last cigarette; symptoms include decreased cognitive capabilities
and heart rate and increased dysphoria or depressed mood, insomnia,
craving, anxiety, irritability, restlessness, appetite, and tendency to smoke
(American Psychiatric Association, 1994; Hughes and Hatsukami, 1992).
Altered brain electrical potentials and hormonal output are generally opposite
in direction of those produced by acute nicotine administration, and
decrements in evoked electrical potentials of the brain indicate impaired
information processing capabilities (Pickworth et al., 1989; U.S. Department
of Health and Human Services, 1988).

Nicotine dependence seems to be mediated primarily by the activation
of nicotinic cholinergic receptors in the brain (U.S. Department of Health
and Human Services, 1988) and secondarily through the cascading effects
of nicotinic systems to modulate levels of hormones such as epinephrine
(adrenaline) and cortisol (Pomerleau and Pomerleau, 1984; U.S. Department
of Health and Human Services, 1988). The mesolimbic dopaminergic
reward system, which mediates the ability of cocaine to produce dependence,
also has been implicated in nicotine dependence (Corrigall, 1991; U.S.
Department of Health and Human Services, 1988). The cells of this system
are located in the ventral tegmental area of the midbrain. Axons project to
the limbic system — specifically, to the nucleus accumbens, olfactory tubercle,
nuclei of the stria terminalis, and parts of the amygdala. Behaviors followed
by such neural activation can become extremely persistent. Cortical effects
of nicotine administration include changes in local cerebral metabolism
(London and Morgan, 1993) and electroencephalogram results Qones, 1987).
Prominent endocrine effects include release of catecholamines, serotonin,
prolactin, growth hormone, arginine vasopressin, beta-endorphin, and
adrenocorticotropic hormone (Pomerleau and Pomerleau, 1984; U.S.
Department of Health and Human Services, 1988). These effects mediate
both the positive nicotine reinforcement sought by smokers and even

117

Smoking and Tobacco Control Monograph No. 7

animals (Corrigall, 1991; Henningfield and Goldberg, 1983; Pomerleau, 1992; ;
U.S. Department of Health and Human Services, 1988) and the negative
reinforcement of withdrawal symptoms that also fuel the compulsion to
smoke (Hughes and Hatsukami, 1992; Pomerleau and Pomerleau, 1984). i
Nicotine also produces increased expression of brain nicotinic receptors in
humans and animals (U.S. Department of Health and Human Services, 1988). |
Taken together, these physiologic effects confirm that nicotine exposure alters ■
the structure and function of the nervous system and leads to modification of (
behavior. Thus, there are physiological factors that drive smokers to sustain
continued nicotine intake across changing delivery systems.

.

Smokers may report that they feel impaired and distracted after only
a few hours of abstinence, and their performance on various cognitive and
psychomotor tasks can decline within approximately 4 hours (Heishman et
al., 1994). Symptoms are rapidly reversed with resumed smoking or nicotine
replacement, thus providing a potentially powerful source of reinforcement i
for continued smoking. The degree of reversal is generally proportional to j
the percentage of plasma nicotine that is replaced (Pickworth et al., 1989; i
U.S. Department of Health and Human Services, 1988). ■)

Data from a performance study indicated that when patients abstained
from cigarettes and used placebo gum, they made more errors and took ^
longer to complete a task than during their smoking baseline. When they j

were given 2 mg gum, their performance returned to baseline. With 4 mg j

gum, they did not do significantly better than at baseline, but 4 mg appeared j
to produce somewhat more reliable clinical effects than 2 mg (Snyder and 1
Henningfield, 1989). jj

The same pattern of effects occurs with theta power, a measure of brain I
function (Pickworth et al., 1989). This nicotine-withdrawal-induced deficit j
can be completely reversed with nicotine replacement. When other !

volunteers resumed smoking, electrocortical potentials recovered quickly !
in all volunteers. Interestingly, these people did not like the gum, and they i{;
were not trying to quit smoking. The lesson is that nicotine replacement |l'
can maintain physiological function and cognitive performance. The j|:'

conclusion relating to performance is not that nicotine makes the user
perform better, faster, or more intelligently but that nicotine deprivation
results in impairments that are quickly and dose-dependently reversed by ||
nicotine readministration (Heishman et al., 1994). ,|

I'he nicotine-withdrawal-induced decline in performance has practical |
ramifications in policy decisions. Currently, the Federal Aviation |

Administration is examining its policies on smoking by pilots in the ft

flight decks of commercial airlines. Because of the time course of nicotine \]
withdrawal, if smoking were eliminated in the flight deck, acutely deprived |
pilots might suffer withdrawal-induced performance declines on flights I

longer than approximately 4 hours. Thus, the nicotine withdrawal syndromej !
poses a potential safety hazard if it is not rationally addressed by appropriate , 1
strategies to detoxify pilots safely and treat their withdrawal symptoms with ^ i
nicotine replacement medications. ,

i' 1=

118

Chapter 8

The duration of the nicotine withdrawal syndrome varies across
individuals, but on average, the acute physical syndrome is worst during
the first month. Gross and Stitzer (1989) studied the time course of the
nicotine withdrawal syndrome in detail. In their study, people quit smoking
and received either active or placebo nicotine gum. People who received
active gum chewed an average of 6.9 pieces of 2 mg gum per day, which
provided less nicotine than they were obtaining by smoking cigarettes.

People given placebo gum gradually decreased their intake from 6.8 pieces
per day during the first week of treatment to 4.9 pieces per day by the
10th week. The nicotine gum substantially reduced withdrawal symptom
severity relative to that observed in placebo subjects.

Nicotine's Nicotine provides many effects that cigarette smokers may consider useful.

Beneficial These include weight control, mood control, and preventing withdrawal

Effects symptoms (U.S. Department of Health and Human Services, 1988). The

issue of whether nicotine would provide substantial cognitive enhancement
in healthy persons who had never been nicotine dependent is controversial.
In nonsmokers, nicotine administration can increase finger-tapping rate
and slightly (but significantly in some studies) attenuate the deterioration
in attention that occurs during protracted testing (Heishman et al., 1994).
However, complex cognitive performance may be impaired by nicotine in
cigarette smokers as well as nonsmokers (Heishman et al., 1994). On the
other hand, there is no question that nicotine intake restores withdrawal-
induced deficits (Snyder and Henningfield, 1989). Nicotine intake also may
provide some level of cognitive enhancement in persons who are cognitively
impaired by Alzheimer's disease (Heishman et al., 1994; Sahakian et al., 1989;
Newhouse and Hughes, 1991).

One of the Brown and Williamson Tobacco Corporation (B&W)
documents made available for the National Cancer Institute conference on
the FTC cigarette test method also supported the conclusion that nicotine's
central nervous system effects contribute to the strong motivation to use
tobacco products. The document concluded that

to understand smoking, just as any other behavior, it is necessary
to consider it as a process embedded within everyday life .... It
is apparent that nicotine largely underpins these contributions
through its role as a generator of central physiological arousal
effects which express themselves as changes in human
performance and psychological well being. (Brown and
Williamson, 1984)

SMOKING AND Nicotine dosage is an important factor in smoking behavior.

NICOTINE DOSE Currently available cigarettes allow people to fairly easily administer
the nicotine dose they need or desire (Henningfield et al., 1994). This was
true of a low-content cigarette, NEXT, that was marketed a few years ago and
removed from the market following poor sales, even though taste and draw
characteristics were similar to conventional cigarettes. With that cigarette,
the nicotine content was so low that no amount of compensatory puffing

119

Smoking and Tobacco Control Monograph No. 7

and inhaling could result in the extraction of substantial amounts of nicotine
(Butschky et al., 1995).

Compensatory Compensation is nicely described in the B&W documents (Brown and
Behavior Williamson, 1984) as "the tendency for a smoker to obtain similar

delivery, intake and uptake of smoke constituents on a daily basis from a
variety of products with different standard (machine-smoked) deliveries."

As the B&W researchers noted, if smokers are dependent, then the
nicotine they receive from cigarettes can be supplemented by other forms,
and this will reduce smoking. Likewise, cigarettes of different strengths are
smoked differently; that is, smokers given low-delivery cigarettes smoke
them more intensively and vice versa.

In fact, this is what has been found in many studies (U.S. Department
of Health and Human Services, 1988). Cigarette consumption increases
in response to reduced nicotine, and most compensation occurs at the
individual cigarette level, not by cigarettes per day. Whereas people given
cigarettes of lower nicotine yield also may smoke a few more cigarettes
per day, they smoke each of the cigarettes more intensely to obtain
proportionately more nicotine than the rating of nicotine yield would
suggest (Hill and Marquardt, 1980; Russell et al., 1980; Benowitz et al., 1983;
Robinson et al., 1983).

When people are given nicotine gum and their smoking is measured,
smoking decreases as the nicotine gum dose increases (Nemeth-Coslett and
Henningfield, 1986). When mecamylamine is administered to antagonize
nicotine's effects, people smoke more cigarettes, take more puffs per cigarette,
and take in more total smoke, as can be seen by increased carbon monoxide
level (Nemeth-Coslett et al., 1986; Rose et al., 1989). Taste and other sensory
factors are also important modulators of human smoking behavior (Butschky
et al., 1995; Rose and Behm, 1987; U.S. Department of Health and Human
Services, 1988).

This finding addresses why the nicotine dependence issue is relevant to
why the FTC method of measuring tobacco smoke constituents is seriously
flawed. Simply put, the FTC method uses machines that do not change their
behavior to self-administer a preferred nicotine dose or in response to the
taste of the smoke, as human smokers do. It may be an accurate predictor of
what smoking machines obtain under specifically programmed conditions,
but it is not an accurate predictor of what people get from cigarettes.

I'he dose-response relationship between FTC ratings and plasma nicotine
levels is weak, except at low doses (Russell et al., 1980 and 1986; Rickert and
Robinson, 1981; Benowitz et al., 1983 and 1986b; Robinson et al., 1983;

Gori and Lynch, 1985; Maron and Fortmann, 1987; Coultas et al., 1993).

I he relationship between cigarette dosage ratings and plasma nicotine levels
may be better in studies using research cigarettes where nicotine content

120

Chapter 8

varies. With other drugs, compensation can be diminished when the cost
of compensation increases. That is, if a drug becomes too costly in terms of
expense or physical difficulty in sustaining intake, users may not compensate
as effectively and will not administer as much of the drug as they did when
the cost was lower (U.S. Department of Health and Human Services, 1988;
Lemaire and Meisch, 1985; Bickel et al., 1993). Thus, if cigarettes have low
enough nicotine contents, smokers would be expected to adjust over time
to lower nicotine levels rather than spend the time and money necessary to
maintain constant dose intake. Conversely, most smokers probably would
not smoke 160 to 200 low-nicotine-content cigarettes per day to continue
to receive the intake that they previously obtained from conventional
cigarettes.

Measurement of The role of dependence is assumed by the authors and the tobacco

Smoking and industry to be important determinants of nicotine intake. Brown

Nicotine Intake and Williamson (1983) noted

the basic assumption is that nicotine, which is almost
certainly the key smoke component for satisfaction, is
fully released to the body system before exhalation takes
place. It is essential, therefore, to quantify the change
in chemical composition between inhaled and exhaled
smoke under different smoking conditions.

Cigarette dose determination is indeed complicated, and some may
suggest that it is so complex that use of the flawed FTC method might as
well continue simply because it has been used for nearly 30 years. However,
such a conclusion contradicts the enormous research advances made
over the past 30 years. This research can be used to devise a better method.
Furthermore, the complexity of dose determination is not unique to
cigarettes. The Food and Drug Administration (FDA) faces this issue
routinely whenever a manufacturer submits a new drug. Unless the drug
is injected into a vein, determination of dosing is complicated. If the drug
is delivered by an inhaler or oral capsule, many factors must be and are
considered so that consumers are provided with realistic estimates of what
they will get. In particular, they are provided with information relevant to
the maximal doses that they are likely to receive from a drug-delivering
product.

To provide accurate dosing information for drug delivery systems, FDA
uses different methods as indicated by the chemical and its delivery system;
moreover, verification of dosing estimates is accomplished in human
bioavailability testing studies because, in the final analysis, we care about
the dose that people receive, not the machine-derived dose. Also, if there
are factors that produce major changes in bioavailability, such as whether
the drug is taken with food or on an empty stomach, this can be indicated
in the labeling.

121

Smoking and Tobacco Control Monograph No. 7

A PROPOSAL
FOR MORE
MEANINGFUL
CIGARETTE
LABELING

One approach to more meaningful cigarette labeling is that described
by Henningfield and colleagues (1994). This approach was adapted
from that used by FDA to label food products with constituents of
health-related relevance. One issue that FDA addressed in food
labeling was serving size. In the case of cigarettes, research has
indicated the need for larger and more intense puffs from the machine to
more closely parallel smokers' behavior (U.S. Department of Health and
Human Services, 1988). A second issue for cigarette labeling is the need to
use biologically meaningful categories. For example, labels might specify
"no nicotine" or "low nicotine" instead of including numerical values that
imply that differences of a few percentage points have practical meaning and
provide the consumer with the illusion that she or he will obtain different
doses from different cigarettes. Similarly, terms such as "light" should be
banned altogether because they imply health benefits; these terms are
permitted with foods only if the food type provides a health benefit relative
to the conventional type of food in a given category. Actual nicotine
content of the cigarettes also should be provided to consumers because the
content determines the absolute limit of nicotine that could be extracted.

Nicotine delivery ratings also could be linked to other factors having
health effects, for example, tar. Thus, a low-nicotine-delivering cigarette
could not be labeled "low nicotine" unless it was also low in tar and carbon
monoxide delivery. A comparable situation in food labeling is that a label
may not use the phrase "fat free" if a product contains cholesterol. Finally,
nicotine yield estimates from standardized machine tests should be validated
with bioavailability testing, as is done with other drugs, because what is of
interest is the dose obtained by smokers.

This approach would not in itself solve the health problem posed by
tobacco use, but it would at least provide consumers with what they have
come to expect in the United States, namely, honest labeling that gives them
the information on which to make decisions about the products they use.
Three decades of research on cigarette smoking, nicotine dependence, and
measurement of tobacco constituent intake have provided the means to
give consumers such information.

QUESTION-AND ANSWER SESSION

DR. DEBETHIZY: Dr. Henningfield, you really did not speak very much
to the FTC method, but I think it is important to point out that the FTC
method was never intended to measure nicotine uptake.

1 also agree with you. I think we can do better in terms of measuring
nicotine uptake when we want to do that. I think the methods that have
been used in the past are estimates. 1 think the study that 1 will tell you
about a little bit later is a step in that direction, and I will be looking
forward to sharing that with you.

122

I would like to make a point about your proposal to measure content.

I have heard that a number of times today, and we have to remember that
people do not eat cigarettes; they smoke cigarettes. And there is no
indication that people obtain the amount of nicotine that is contained
in a cigarette.

DR. HENNINGFIELD: On content, I think that the most important thing
is bioavailability tests. Again, that is the gold standard: what people are
likely to get and generally under maximum conditions. The importance
of content, though, is that content limits the amount of nicotine that you
can get. If it is not there, you cannot get it.

DR. DEBETHIZY: I think the important thing is the FTC method is set up to
provide relative ranking, so that consumers can get an idea of what different
cigarettes will yield. It was not intended to measure uptake.

Now, if you want to measure uptake and evaluate the FTC method,
that is a different activity, and I think that we need to make sure that we
distinguish those two activities. One is to provide a relative ranking. The
FTC method has done an excellent job of that over the years.

DR. HENNINGFIELD: I am not addressing the method, but I think it is pretti-
clear that it has not done a good job of telling people what they will have in
their bloodstream. And that is what I am addressing: that what people get
in their bloodstream does not bear much relation to the FTC yields. So, I am
not sure how much use that has been.

DR. HARRIS: I see the dispute as distinguishing between an ordinal ranking
and a cardinal ranking. An ordinal ranking merely says one brand, to some
degree, delivers more or less nicotine than another; whereas, a cardinal
ranking would say, this brand delivers one-fifth as much or five times as
much.

And what I understand the dispute to be about is that the FTC ranking
actually may preserve an ordinal ranking in the roughest sense, but it does
not preserve the cardinal ranking. From what I can gather, a 10-percent
increase in FTC nicotine corresponds with, at most, about a 2-percent
increase in blood nicotine, roughly speaking, and that that is where the
problem lies.

DR. HENNINGFIELD: It is not even that good, because if the slope were
constant, you could maybe say there is an ordinal ranking. That still may
not tell you if it is meaningful if it was so trivial. But what Dr. Benowitz
showed was if there is a break.

In other words, at the ultralow end, those cigarettes are in a slightly
different category. From the data I have seen, it is not even a meaningful
ordinal ranking. It is a pretty flat ranking. The slope is, I would contend
until proven otherwise, biologically trivial.

Smoking and Tobacco Control Monograph No. 7

REFERENCES

American Psychiatric Association. Diagnostic and
Statistical Manual of Mental Disorders: Fourth Edition.
Washington, DC: American Psychiatric Press, 1994.

Benowitz, N.L., Hall, S., Herning, R.I., Jacob, P., Jones,
R.T., Osman, A.-L. Smokers of low-yield cigarettes
do not consume less nicotine. New England Journal
of Medicine 309: 139-142, 1983.

Benowitz, N.L., Jacob, P., Kozlowski, L.T., Yu, L.
Influence of smoking fewer cigarettes on exposure
to tar, nicotine, and carbon monoxide. New
England Journal of Medicine 31S: 1310-1313, 1986a.

Benowitz, N.L., Jacob, P., Yu, L, Talcott, R., Hall, S.,
Jones, R.T. Reduced tar, nicotine, and carbon
monoxide exposure while smoking ultralow- but
not low-yield cigarettes. Journal of the American
Medical Association 256: 241-246, 1986b.

Bickel, W.K., DeGrandpre, R.J., Higgins, S.T.
Behavioral economics: A novel approach to the
study of drug dependence. Dmg and Alcohol
Dependence 33: 173-192, 1993.

Brown and Williamson. "Summary." Tobacco
Corporation Research Conference, Rio de Janeiro,
Brazil, August 22-26, 1983.

Brown and Williamson. "Proceedings of the Smoking
Behaviour-Marketing Conference." Tobacco
Corporation Smoking Behaviour-Marketing
Conference, Montreal, Quebec, Canada, July 9-12,
1984.

Butschky, M.F., Bailey, D., Henningfield, J.E.,

Pickworth, W.B. Smoking without nicotine delivery
decreases withdrawal in 12-hour abstinent smokers.
Pharmacology, Biochemistry and Behavior 50(1): 91-
96, 1995.

Corrigall, W.A. Regulation of intravenous nicotine
self-administration — dopamine mechanisms. In:
Effects of Nicotine on Biological Systems, Advances in
Pharmacological Sciences, F. Adlkofer and K. Fhurau
(Editors). Boston: Berkhauser Verlag, 1991, pp. 423-
432.

(^ultas, D.B., Stidley, C.A., Samet, J.M. Cigarette
yields of tar and nicotine and markers of exposure
to tobacco smoke. American Review of Resjnratory
Disease 148: 435-440, 1993.

(.ummings, S.R., Richard, R.J. Optimum cutoff points
for biochemical validation of smoking status.
American Journal of Public I lealth 78: 574-575, 1988.

Djordjevic, M.V., Hoffmann, IX, Glynn, T., Connolly,
(j.N. U.S. commercial brands of moist snuff, 1994.
Assessment of nicrrtine, moisture, and [)ll. Tobacco
Control 4dx 62-66, 1995.

Fiore, M.C.. I rends in cigarette smoking in the United
States. I he ejridemiology of tobaccrr use. Medical
(Ainic s of North America 76: 289-303, 1 992.

Fischman, M.W., Mello, N.K. (Editors). Testing for Abuse
Liability ofDnigs in Humans. National Institute on
Drug Abuse Research Monograph 92. DHHS
Publication No. 89-1613. Washington, DC: Supt.
of Docs., U.S. Govt. Print. Off., 1989.

Gilman, A.G., Rail, T.W., Nies, A.S., Taylor, P. (Editors).
Goodman and Gilman's The Pharmacological Basis of
Therapeutics. New York: Pergamon Press, 1990.

Gori, G.B., Lynch, C.J. Analytical cigarette yields as
predictors of smoke bioavailability. Regidatory
Toxicology and Pharmacology 5: 314-326, 1985.

Griffiths, R.R., Bigelow, G.E., Henningfield, J.E.
Similarities in animal and human behavior. In:
Advances in Substance Abuse: Behavioral and Biological
Research, Vol. 1, N.K. Mello (Editor). Greenwich, CT:
JAI Press, 1980, pp. 1-90.

Gross, J., Stitzer, M.L. Nicotine replacement: Ten-week
effects on tobacco withdrawal symptoms.
Psychopharmacology 98: 334-341, 1989.

Heishman, S.J., Taylor, R.C., Henningfield, J.E.

Nicotine and smoking: A review of effects on human
performance. Experimental and Clinical Pharmacology
2: 345-395, 1994.

Henningfield, J.E., Goldberg, S.R. Nicotine as a
reinforcer in human subjects and laboratory
animals. Pharmacology, Biochemistry and Behavior
19: 989-992, 1983.

Henningfield, J.E., Keenan, R.M. Nicotine delivery
kinetics and abuse liability. Journal of Consulting and
Clinical Psychology 61: 1-8, 1993.

Henningfield, J.E., Kozlowski, L.T., Benowitz, N.L. A
proposal to develop meaningful labeling for
cigarettes. Journal of the American Medical Association
272: 312-314, 1994.

Henningfield, J.E., London, E.D., Benowitz, N.L.

Arterial-venous differences in plasma concentrations
of nicotine after cigarette smoking. Journal of the
American Medical Association 263: 2049-2050, 1990.

Henningfield, J.E., Miyasato, K., Jasinski, D.R. Abuse
liability and pharmacodynamic characteristics of
intravenous and inhaled nicotine. Journal of
Pharmacology and Experimental 'Therapeutics 234: 1-2,
1985.

Henningfield, J.E., Radzius, A., Cone, E.J. Estimation of
available nicotine content of six smokeless tobacco
(rroducts. Tobacco Control 4(1): 57-61, 1995.

Henningfield, J.IL, Stapleton, J.M. , Benowitz, N.L.,
Grayson, R.F., London, E.l). Higher levels of nicotine
in arterial than in venous blood alter cigarette
smoking. Drug and Alcohol Dependence 33: 23-29,
199.3.

Hill, P., Marquardt, 11. Plasma and urine changes after
smoking dillerent braruls of cigarettes. Clinical
Pharmacology and I'herapeutics 27: 652-658, 1980.

124

Chapter 8

Hughes, J.R., Gulliver, S.B., Fenwick, J.W., Valliere,
W.A., Cruser, K., Pepper, S., Shea, P., Solomon, L.J.,
Flynn, B.S. Smoking cessation among self-quitters.
Health Psychology 11(5): 331-334, 1992.

Hughes, J.R., Hatsukami, D.K. The nicotine
withdrawal syndrome: A brief review and update.
International Journal of Smoking Cessation 1: 21-26,
1992.

Hunt, W.A., Barnett, L.W., Branch, L.G. Relapse rates
in addiction programs. Journal of Clinical Psychology
27: 455-456, 1971.

Jarvis, M.J. Application of biochemical intake markers
to passive smoking measurement and risk
estimation. Mutation Research 222: 101-110, 1989.

Jarvis, M.J., Tunstall-Pedoe, H., Feyerabend, C., Vesey,
C., Saloojee, Y. Comparison of tests used to
distinguish smokers from nonsmokers. American
Journal of Public Health 77: 1435-1438, 1987.

Jones, R.T. Tobacco dependence. In: Psycho-
pharmacology: The Third Generation of Progress, H.Y.
Meltzer (Editor). New York: Raven Press, 1987,
pp. 1589-1595.

Kottke, T.E., Brekke, M.L., Solberg, L.I., Hughes, J.R. A
randomized trial to increase smoking intervention
by physicians: Doctors helping smokers. Round I.
Journal of the American Medical Association 261:
2101-2106, 1989.

Kozlowski, L.T. Reduction of tobacco health hazards
in continuing users: Individual behavioral and
public health approaches. Journal of Substance Abuse
1: 345-357, 1989.

Kozlowski, L.T., Wilkinson, A., Skinner, W., Kent, C.,
Frankin, T., Pope, M.A. Comparing tobacco
cigarette dependence with other drug
dependencies: Greater or equal “difficulty quitting"
and "urges to use" but less pleasure from cigarettes.
Journal of the American Medical Association 261: 898-
901, 1989.

Langley, J.N. On the reaction of cells and of nerve
endings to certain poisons, chiefly as regards the
reaction of striated muscle to nicotine and to
curari. Journal of Physiology 33: 374-413, 1905.

I Lemaire, G.A., Meisch, R.A. Oral dmg self-
administration in rhesus monkeys: Interactions
I between drug amount and fixed-ratio size. Journal
of the Experimental Analysis of Behavior 44: 377-389,
1985.

London, E.D., Morgan, M.J. Positron emission
tomographic studies on the acute effects of
psychoactive drugs on brain metabolism and
mood. In: Imaging Drug Action in the Brain, E.D.
London (Editor). Boca Raton, FL: CRC Press, 1993,
pp. 265-280.

I Maron, D.J., Fortmann, S.P. Nicotine yield and
measures of cigarette smoke exposure in a large
population: Are lower-yield cigarettes safer?
American Journal of Public Health 77: 546-549, 1987.

Nemeth-Coslett, R., Henningfield, J.E. Effects of
nicotine chewing gum on cigarette smoking and
subjective and physiologic effects. Clinical
Pharmacology and Therapeutics 39: 625-630, 1986.

Nemeth-Coslett, R., Henningfield, J.E., O'Keefe, M.K.,
Griffiths, R.R. Effects of mecamylamine on human
cigarette smoking and subjective ratings.
Psychopharmacology 88: 420-425, 1986.

Newhouse, P.A., Hughes, J.R. The role of nicotine and
nicohnic mechanisms in neuropsychiatric disease.
British Journal of Addiction 86: 521-526, 1991.

Peto, R., Lopez, A.D., Boreham, J., Thun, M., Heath,

C., Jr. Mortality From Smoking in Developed Countries,
1950-2000: Indirect Estimates From National Vital
Statistics. World Health Organization. Oxford:
Oxford University Press, 1994.

Pickworth, W.B., Heming, R.I., Henningfield, J.E.
Spontaneous EEG changes during tobacco
abstinence and nicotine substitution. Journal of
Pharmacology and Experimental Therapeutics 251:
976-982, 1989.

Pomerleau, O.F. Nicotine and the central nervous
system: Biobehavioral effects of cigarette smoking.
American Journal of Medicine 93(Suppl lA): 2S-7S,
1992.

Pomerleau, O.F., Pomerleau, C.S. Neuroregulators
and the reinforcement of smoking: Towards a
biobehavioral explanation. Neuroscience and
Biobehavioral Reviews 8: 503-513, 1984.

Pomerleau, O.F., Rosecrans, J. Neuroregulatory effects
of nicotine. Psychoneuroendocrinology 14: 407-423,
1989.

Rickert, W.S., Robinson, J.C. Estimating the hazards
of less hazardous cigarettes. 11. Study of cigarette
yields of nicotine, carbon monoxide, and hydrogen
cyanide in relation to levels of cotinine,
carboxyhemoglobin, and thiocyanate in smokers.
Journal of Toxicology and Environmental Health 7:
391-403, 1981.

Robinson, J.C. , Young, J.C., Rickert, W.S., Fey, G.,
Kozlowski, L.T. A comparative study of the amount
of smoke absorbed from low yield ('less hazardous')
cigarettes. Part 2: Invasive measures. British Journal
of Addiction 78: 79-87, 1983.

Rose, J.E., Behm, F. Refined cigarette smoke as a
method for reducing nicotine intake. Pharmacology,
Biochemistry and Behavior 28: 305-310, 1987.

Rose, J.E., Sampson, A., Levin, E.D., Henningfield, J.E.
Mecamylamine increases nicotine preference and
attenuates nicotine discrimination. Pharmacology,
Biochemistry and Behavior 32(4): 933-938, 1989.

Russell, M.A., Jarvis, M., Iyer, R., Feyerabend, C.
Relation of nicotine yield of cigarettes to blood
nicotine concentrations in smokers. British Medical
Journal 280(6219): 972-976, 1980.

Smoking and Tobacco Control Monograph No. 7

Russell, M.A., Jarvis, M J., Feyerabend, C., Saloojee, Y.
Reduction of tar, nicotine and carbon monoxide
intake in low tar smokers. Journal of Epidemiology
and Community Health 40(1): 80-85, 1986.

Sahakian, B., Jones, G., Levy, R., Gray, J., Warburton,
D. The effects of nicotine on attention,
information processing, and short-term memory in
patients with dementia of the Alzheimer's type.
British Journal of Psychiatry 154: 797-800, 1989.

Snyder, F.R., Henningfield, J.E. Effects of nicotine
administration following 12 h of tobacco
deprivation: Assessment on computerized
performance tasks. Psychopharmacology 97(1): 17-
22, 1989.

U.S. Department of Health and Human Services. The
Health Consequences of Smoking: Nicotine Addiction. A
Report of the Surgeon General, 1988. DHHS
Publication No. (GDC) 88-8406. Rockville, MD:

U.S. Department of Health and Human Services,
Public Health Service, Centers for Disease Control,
Center for Health Promotion and Education, Office
on Smoking and Health, 1988.

Chapter 9

Consumer/Smoker Perceptions of Federal
Trade Commission Tar Ratings

Joel B. Cohen

INTRODUCTION A telephone survey among a national probability sample of
1,005 adults (502 men and 503 women) 18 years of age and older was
conducted between November 17 and 20, 1994. Data were weighted by
age, sex, geographic region, and race so that each respondent was assigned
a single weight based on the relationship between the actual population
proportions of the listed characteristics and the comparable sample
proportions.

The author's estimate of every-day smoking (23 percent) matches current
assessments of adult U.S. smoking prevalence (22 percent). When every-day
and some-days smokers were combined, the current smoking percentage
(28.7 percent) was slightly higher than the Centers for Disease Control and
Prevention (CDC) (1994) comparable estimate of 26.5 percent for 1992. This
sample reported somewhat higher current smoking percentages for females
(29 percent) than did the 1992 CDC surveys (24.6 percent). Total smoking
reported by whites (29 percent) was slightly higher than in the 1992 CDC
surveys (27.2 percent), whereas total smoking reported by blacks in this
sample (27 percent) was virtually identical (27.8 percent). A high percentage
of those who report having attended but not graduated from college were
some-days smokers. When added to every-day smokers, this total was
substantially higher (36 percent) than that reported in the CDC surveys
(24 percent) and was closer to the CDC estimate for high school graduates
(31 percent). College graduates in this sample were also somewhat more
likely to smoke (19 percent compared with 15.5 percent reported in CDC
surveys). Age breakdowns were not entirely comparable among the surveys,
but the author's sample reported a higher incidence of smoking among
18- to 24-year-olds (32 percent compared with 26.4 percent).

TAR LEVEL OF Table 1 reports the tar levels of cigarettes last smoked, determined by

CIGARETTES asking the brand, size, and other characteristics of the cigarette. These
answers were compared with actual Federal Trade Commission (FTC) tar
ratings. In 15 percent of the cases, respondents could not provide sufficiently
detailed product information to make this comparison ("Cannot Determine"
respondents). These respondents were likely to come disproportionately from
lower tar categories. A four-category designation of tar levels was selected.

It allowed for somewhat greater differentiation among lower tar users, had
an equal number of rating scale points in each of the low-tar categories, and
was consistent with a recently proposed four-category nicotine and tar rating
system. Unweighted cell sizes for the five tar categories (including "Cannot
Determine") shown in Table 1 were small: 28, 75, 70, 116, and 48 for those
smoking cigarettes in the past 2 to 3 years.

127

Smoking and Tobacco Control Monograph No. 7

Table 1

Tar level (percent) of cigarette last smoked

Tar Levels (mg)

Smoker Classification
(weighted data) (N)

Very Low
1-5

Low

6-10

Medium

11-15

High

16+

Cannot

Determine

Current Smokers

Some-days smokers (56)

9

34

9

zo

Every-day smokers (232)

8

22

21

40

10

Recent (2 to 3 years) quitters (36)

11

11

25

28

25

Those Smoking in the Past 2 to 3 Years

All smokers (325)

9

22

19

35

14

Male (152)

5

24

13

42

17

Female (174)

12

21

25

29

12

White (268)

10

23

21

31

15

Black (28)

0

14

18

64

4

Hispanic (26)

4

15

4

58

19

18

High school or less education (1 07)

6

T5

21

41

At least some college education (1 46)

12

32

23

30

12

Smokers of regular size cigarettes (145)

5

28

11

40

16

Smokers of longer cigarettes (173)

12

19

27

32

1 1

Smokers of soft pack cigarettes (1 80)

13

17

22

33

16

Smokers of hard pack cigarettes (1 33)

3

29

17

41

9

Smokers of plain cigarettes (223)

9

24

18

3b

15

12

Smokers of menthol cigarettes (101)

8

20

23

38

Fifty-eight percent of current smokers smoked a cigarette with 15 mg or
less of tar, and 9 percent smoked a cigarette with 1 to 5 mg of tar. Recent I
quitters tended to come from relatively higher tar categories, consistent
with evidence suggesting that switching to the lowest tar cigarettes was a [

substitute for, rather than a stepping stone to, quitting. High-tar cigarette j]

use was more frequent among males, blacks, and Hispanics and decreased |

markedly with educational attainment.

KNOWLEDGE OF Those smoking cigarettes in the past 2 to 3 years were asked to tell j

ADVERTISED the interviewer the tar number of their most recently smoked ;

TAR NUMBERS cigarette. Seventy-nine percent indicated that they did not know.

rhis increased to about 90 percent for those having less than a high school ^
education, smokers ages 55 and older, and black smokers. Respondents
answering "do not know" then were asked to come as close as they could, <

and interviewers were to probe for their "best guess." Fifty-eight percent still j
reported not knowing. '

Initial responses were slightly more likely to be underestimates (9 percent)
than correct answers (defined as plus or minus 1 mg from the actual tar level) ; :

128

Chapter 9

or overestimates (6 percent in both of the latter two categories). When
probed responses were included in the analysis, there was a substantial
increase in responses that underestimated tar levels (from 9 to 20 percent);
there were only small changes in correct answers or overestimates. When
actual tar numbers were regressed against respondents' initial and probed
answers, the relationships were weak (r = .26 and .20, respectively).

Smokers of very-low-tar cigarettes had a much greater awareness of their
cigarettes' tar numbers. Thirty-nine percent of those who smoked 1- to 5-mg
tar cigarettes were correct initially, increasing to 50 percent with probing.
These figures stand in marked contrast to responses of smokers of cigarettes
with 6 to 10 mg tar, whose comparable percentages of correct responses were
4 and 9 percent, respectively.

To assess "knowledge in practice" (in addition to recall-based knowledge),
half the members of the sample were asked whether a 16-mg (or, for the other
half, a 5-mg) tar cigarette is lower in tar than most other cigarettes on the
market. The correct answers are "no" for the 16-mg tar cigarette and "yes"
for the 5-mg tar cigarette. Table 2 shows respondents' answers cross-tabulated
by the tar level of their most recently smoked cigarette. Whereas 35 percent
of the smokers of 1- to 5-mg tar cigarettes did not know that a 16-mg tar
cigarette was not lower in tar, between 55 and 66 percent of all other smokers
either did not know or gave incorrect responses to this question. For those

Table 2

Interpretation of Federal Trade Commission tar numbers corresponding to lower tar levels

Tar Levels (mg)

Interpretations

Very Low
1-5

Low

6-10

Medium

11-15

High

16+

Cannot

Determine

Believe That a 1 6-mg Tar
Cigarette Is Lower in Tar
Than Most Other Cigarettes
(N = 179)

% Correct

(14)

65

(36)

45

(40)

44

(64)

34

(25)

32

% Incorrect

0

10

10

16

12

% Do not know

35

45

46

50

56

Believe That a 5-mg Tar
Cigarette Is Lower in Tar
Than Most Other Cigarettes
(N = 158)

% Correct

(14)

15

(39)

34

(30)

44

(52)

27

(23)

25

% Incorrect

13

10

14

19

16

% Do not know

73

56

42

55

59

i

I

I

I

129

Smoking and Tobacco Control Monograph No. 7

smoking cigarettes having more than 5 mg of tar, between 56 and 74 percent
either did not know that a 5-mg tar cigarette was lower in tar than most other
cigarettes or said that it was not lower (with 10 to 20 percent incorrect).

SMOKERS' Two approaches were used to better understand how smokers

INTERPRETATIONS interpreted the advertised tar numbers. In the first, half the
OF TAR NUMBERS sample members were asked whether a pack-a-day smoker could
significantly lower his or her health risks due to smoking by switching from
a 20-mg tar cigarette to a 5-mg tar cigarette; for the other half, the switch
was to a 16-mg tar cigarette. In total, 56 percent of smokers thought that a
switch to a 5-mg tar cigarette would significantly lower health risks, whereas
28 percent thought that a switch to a 16-mg tar cigarette would significantly
lower health risks.

Table 3 cross-tabulates answers to these questions against the actual tar
levels of smokers' cigarettes. For the substantive shift to a 5-mg cigarette,
light-to-heavy tar cigarette smokers are evenly divided between believing
there would be a significant reduction in health risks and either believing
this would not be the case or being unsure about this. Whereas more than
60 percent of smokers did not think switching to a 16-mg tar cigarette would
lead to a significant reduction in health risks due to smoking, a sizable
proportion of light-to-heavy tar cigarette smokers either thought it would
or did not know.

Table 3

Inferences (percent) about health risks as a result of switching to lower tar cigarettes

Tar Levels (mg)

Inference

Very Low
1-5

Low

6-10

Medium

11-15

High

16+

Cannot

Determine

Switching From a 20-mg to a 5-mg Tar Cigarette
Would Significantly Reduce Health Risks

83

49

49

55

60

Switching From a 20-mg to a 5-mg Tar Cigarette
Would Not Significantly Reduce Health Risks

13

32

35

25

29

Do Not Know

4

19

15

20

12

Switching From a 20-mg to a 16-mg Tar Cigarette
Would Significantly Reduce Health Risks

18

35

28

25

33

Switching From a 20-mg to a 16-mg Tar Cigarette
Would Not Significantly Reduce Health Risks

68

61

61

61

37

Do Not Know

14

4

10

14

31

Relative Difference in Health Risks Between Those
Asked About Switching to a 5-mg and Those
Asked About Switching to a 16-mg Tar Cigarette

65

14

21

30

27

Chapter 9

The interpretation of data in Table 3 is complicated by almost certainly
differing beliefs of smokers in the four tar categories regarding the risks of
smoking a 20-mg tar cigarette and hence about the decrease in risk from
any reduction in tar level. Because the belief factor is likely to be a constant
in the two versions of this question, it is useful to examine the relative
reduction in health risk (i.e., the difference in benefits between switching
to the 5-mg tar alternative compared with the 16-mg alternative), shown in
the last row of Table 3. Once again, the evidence points to a clear difference
between smokers of cigarettes with 1 to 5 mg of tar and all other smokers.
These very-low-tar smokers believe that it takes a substantial reduction in
tar yields to significantly reduce health risk, whereas this belief does not
appear to be held by a substantial number of smokers in other categories.
Unfortunately, this belief also may support a judgment that a substantial
reduction in tar levels may be a reasonable substitute for quitting.

In the second approach, we examined smokers' understanding of the
distinction between tar yield and delivery, together with their willingness to
treat the numerical information as if it had ratio-scale properties rather than
merely ordinal properties. Many of those supporting the dissemination of
tar numbers have assumed that consumers would use these numbers in an
ordinal fashion, essentially as if they were simply rank-ordered data. Ordinal
scales do not possess the property that each numerical interval is of the same
magnitude (i.e., the difference between 1 and 2 being precisely equal to the
difference between 10 and 11). The FTC method may produce tar ratings
that have this interval scale property for tar yields, but it cannot be said to
do so for actual deliveries of tar because smokers' inhalation patterns seem
to vary as they move lower on the scale. A ratio scale has the further property
of having a genuine zero point so that it is proper to regard a scale score of
10 as being twice as high as a scale score of 5.

Respondents were asked to assume that a person switched from a 10-mg
tar cigarette to a 1-mg tar cigarette. Then the three statements shown in
Table 4 were read twice, and respondents were asked to decide which of these
came closest to their opinion. Primacy and recency effects were controlled
by rotating the order of the first and third statements. The first answer is the
correct choice, whereas the second answer suggests some reluctance to rely
on the absolute numerical values when thinking about such tradeoffs.

The general conclusion to be drawn from these data is that at least one-
quarter of smokers (i.e., those selecting the third interpretation) clearly have
been misled about the meaning of the tar yield numbers. Interestingly, this
increases to 44 percent for smokers of very-low-tar cigarettes, in line with
other evidence presented here; it also increases concern about the safety
reassurances that such very-low-tar cigarettes appear to provide.

131

Smoking and Tobacco Control Monograph No. 7

Table 4

Inferences (percent) about tradeoffs between tar deliveries and number of cigarettes smoked

Tar Levels (mg)

Inference (relative to a 10-mg tar cigarette)

Very

Low

1-5

Low

6-10

Medium

11-15

High

16+

Cannot

Determine

Person Probably Could Smoke More Than One,
but These Numbers Cannot Tell You How
Much Less Tar the Person Would Take in
From the 1-mg Tar Cigarette

28

33

31

40

39

Person Could Smoke More Than 1 or 2 but Less
Than 9 or 10 of the 1-mg Tar Cigarettes Without
Taking in More Tar

18

33

22

25

22

Person Could Smoke About 10 of the 1-mg Tar
Cigarettes Without Taking in More Tar

44

25

31

21

21

None of These/Do Not Know

10

10

16

14

18

SMOKERS' USE The final issue under study in this survey was whether smokers
OF ADVERTISED reported having used these tar numbers to make judgments about
TAR NUMBERS the relative safety of different brands of cigarettes. In answering

this question, only 14 percent of the sample indicated doing so. Once again,
the smokers of 1- to 5-mg tar cigarettes were different: 56 percent of them
reported using advertised tar numbers to make judgments about the relative
safety of various cigarettes.

CONCLUSIONS This study demonstrates inherent difficulties in using advertised tar
yield numbers to communicate meaningful information to consumers.

Most smokers do not seem to pay careful attention to the numerical values
per se, even to the extent of having a strong sense of the range of numerical
values. Smokers of cigarettes with low- to high-tar content had considerable
uncertainty about the health implications of switching to lower tar cigarettes.
However, very-low-tar numbers seem to have a strong appeal to a particular
group of smokers and may convey a message of absolute safety.

QUESTION-AND-ANSWER SESSION

DR. TOWNSEND: Dr. Cohen, can you tell me if you also asked the subjects
the category of cigarettes that they smoked; for example, was it regular or |
lights or ultralights? I

I

DR. COHEN: We did not ask them their perception of their cigarette. We I
asked them exactly what they smoked in terms of the size and whether their |
cigarettes were menthol or plain and hard pack or soft pack — but we did not 1
ask them their perception. i

I'.iZ

Chapter 9

DR. TOWNSEND: I am not speaking about their perception. I am speaking
about the advertising associated with the cigarettes that smokers purchase.

For example, if you go into a store to buy cigarettes, you can buy Winston
regulars, Winston Lights, or Winston Ultra Lights. And of course, those
relative categories are based on FTC tar numbers. So, you did not ask them
a question like that?

DR. COHEN: No, we didn't ask that specific question.

DR. TOWNSEND: My experience, in talking with a lot of consumers, is that
they do know, very clearly, the category of cigarettes that they are smoking,
even though some of them do not know the accurate numbers of the
cigarettes that they are smoking.

DR. COHEN: I would not disagree with you.

DR. TOWNSEND: And that ranking of categories is based on the FTC
numbers. So, I think that your conclusion that the numbers are useless,

I certainly do not agree with.

I think another example of that is in my recent purchase of a hot water
heater. 1 certainly used the energy efficiency ratings in making that choice.

I cannot tell you today what the energy efficiency rating actually is.

DR. COHEN: If we are here looking at the utility of the FTC tar numbers
in advertising, then I think it is fair to ask if people are taking away this
information. The assertion is made that this information has value to
people. I am examining that assertion.

Now, if you are saying, well, it is not the numbers they care about; it is
the categories, then you can present information that says they are done in
four categories.

DR. TOWNSEND: In addition to that, you also said that there is at least one
category where a high percentage of those smokers do look at those numbers
very carefully. So, I think your conclusion that the FTC numbers are useless
is certainly not true.

DR. COHEN: I do not know that I went that far; I stopped a little short of
that.

DR. TOWNSEND: And I think another very practical example of the utility
of the FTC relative numbers is, in fact, what has happened to the industry
over the past 40 years. We have reduced the level of tar delivery, by the
FTC method, from about 38 mg down to about 12 mg. I think Dr. Hoffmann
spoke to that very clearly.

What has happened to cause that dramatic a change is that people trade
off taste. The lower tar cigarettes generally have — or always have — fewer
taste characteristics. And people find that more acceptable. So, they are
making this tradeoff in the marketplace of taste and something else.

133

Smoking and Tobacco Control Monograph No. 7

DR. COHEN: I appreciate your position, though I think we are talking about
different issues here. I am talking about the utility of this information
presented in this form as numbers. I am not talking about the utility of
providing information about tar.

MS. WILKENFELD: 1 do want to add one thing to the mix to make it more
complicated. You said they had to tell you the name of the cigarette they
were smoking. And in order to get the actual tar number, they would have
had to report specifically about the category, for example, Marlboro and
Marlboro Lights. So, they may have reported correctly.

DR. KOZLOWSKI: I have found out that a lot of people in the United States
who do not smoke somehow have the impression that tar and nicotine
ratings are printed on the packs of cigarettes. They are in some places, for
example, in Canada, but not in the United States. I think the one notable
exception is the ultralow-tar cigarettes. You know, when you test Carlton
as low as 1 mg tar, they are right on the pack.

If it is not on packs, if a brand is not advertised, or a person does not see
an ad, how in the world would they know what the tar and nicotine yields
were?

DR. COHEN: I think there is a fundamental problem. I do not think the scale
has integrity. We had a scale that goes from 1 to 27. People don't care about j|
tar; they don't know what it is. They care about harmfulness; they care about ||
smoking risk. r

If you don't present information to people along a dimension that they
care about, they are not going to pay as much attention to it. And if you
don't present information to people in a way so that they know how to
use it, where the numbers have some meaningful quality, they will not pay
attention to it. Then people are not going to be able to do as much with it.

1 think there is a fundamental problem with providing information.

It may be the wrong information presented the wrong way. Other than
that, it is OK.

REFERENCE

Centers for Disease Control and Prevention. Cigarette
smoking among adults — United States, 1992, and
changes in the definition of current cigarette
smoking. MMWR. Murbiclity and Mortality Weekly
Report 4:i(\9): M2-.i46, 1994.

i:i4

] Chapter 10

Sensitivity of the Federal Trade Commission
Test ^4ethod to .Analytical Parameters

' Michael R. Guerin

I INTRODUCTION The Federal Trade Commission (FTC) test method for determining
j the tar, nicotine, and carbon monoxide yields of commercial cigarettes

I was designed to characterize and compare brands. Relevance to human

smoking was a consideration in choosing the test method, but the principal
objective was to select a method that provided the most accurate and
reproducible result. Relevance to human smoking was addressed by using
intermittent puffing and by choosing puff volume, puff duration, puff
frequency, and butt length based on observations of human smokers.
Accuracy and reproducibility were addressed by selecting a single set of
smoking conditions, demanding narrow tolerances for variation in the
conditions, and standardizing everything from cigarette selection, to the
smoking environment, to the laboratory analytical chemical methods.

Requirements associated with producing a standard method tend to
conflict with those associated with maximizing relevance to the human
situation. Bradford and colleagues (1936) recognized from the beginning
that humans smoke cigarettes in different and varying ways, but a
standardized procedure requires that variables be set and controlled.

For practical purposes, only one set of conditions could be selected.

At least two factors have led to an increased concern about the relevance
of the FTC test procedure. First, FTC results increasingly have been viewed
as a measure of human exposure and therefore health risk. The problem
is compounded by the assumption that even a small difference in FTC
results signifies a meaningful difference in human exposure. Second, a much
greater variety of cigarettes is available today. They range from nonfilter and
filter cigarettes similar to those available when the method was adopted, to
j increasingly popular products with very low FTC yields. Behavioral research

j has demonstrated that low-yield products are consistently smoked differently

than are higher yield products (Kozlowski et al., 1989).

This chapter reviews the nature of the FTC test procedure and the
influence of changes in its specifications on yields. Smoking parameters
likely to be different for humans from FTC machine smoking are emphasized.

STANDARD The quantities of tar, nicotine, carbon monoxide, and other constituents
i MACHINE in cigarette smoke are measured using smoking machines. One or more
j SMOKING cigarettes are smoked by a machine, the constituents of interest are collected
' in a suitable trap, and the contents of the trap are chemically analyzed,

j The quantity in the trap is divided by the number of cigarettes smoked to

j compute a yield (or delivery) per cigarette. In the case of the FTC procedure.

Smoking and Tobacco Control Monograph No. 7

the particle phase of the smoke is collected on glass fiber (Cambridge) filters,
and the gas phase (passing through the filter) is collected in gas sampling
bags. Carbon monoxide is measured in the gas sampling bags. The filter is
weighed to yield a measure of total particulate matter (TPM) and is analyzed
for nicotine and water content. Tar (or nicotine-free dry particulate matter)
is computed by subtracting the weights of nicotine and water from the
weight of TPM.

The principal reason for using smoking machines is to maximize the
reproducibility of results (DeBardeleben et al., 1991). This is particularly
important for quality control and product comparison and is essential for
interlaboratory comparisons. However, machine smoking is limited in that
it provides results accurate only for the specific set of smoking conditions
employed by the machine.

Smoking parameters used in the FTC procedure are based largely on
empirical observations of smokers reported by Bradford and colleagues
(1936). They suggested a nominal 35-mL ("mL" is used interchangeably in
the literature with "cc") puff volume of a 2-second duration taken once per
minute to a 23-mm butt length. Current FTC smoking conditions (Federal
Trade Commission, 1994) specify a puff volume of 35 ± 0.5 mL, a puff
duration of 2.0 ± 0.2 seconds, and a puff frequency of 1 per 60 ± 1 second.
Butt length is specified as 23 mm for nonfilter cigarettes and the length of
filter overwrap plus 3 mm for filtered cigarettes. The international standard
method (Thomsen, 1992) ISO 3308 currently uses the same conditions but
requires more stringent tolerances. Puff volume is 35 ± 0.25 mL, puff
duration is 2.0 ± 0.05 second, and puff frequency is 1 per 60 ± 0.5 second.

Machine-smoking parameters are only one of several conditions that
have been specified to constitute standard FTC testing. Other conditions
include the number and manner of selection of cigarettes to be tested,
cigarette conditioning (see below), the smoking environment, and the
methods and instrumentation used. FTC testing specifies the analysis of
100 cigarettes selected at random from two packages purchased at each of
50 geographical locations throughout the United States. Cigarettes must be
conditioned at 60 percent relative humidity and 24 °C for at least 48 hours
before smoking and must be smoked in a room maintained under the same
conditions. Smoking is performed using a Phipps and Bird 20-port linear
smoking machine, thus specifying by default that "restricted" rather than
"free" (butt end closed rather than open to the atmosphere between puffs)
smoking be performed. Finally, the air flow across the cigarettes must be
reproducible and controlled to control the rate at which the cigarette burns
between puffs.

The introduction of cigarettes with ventilated filters has made it
necessary to pay additional attention to the depth to which the cigarette is
inserted into the holder. The cigarette must be inserted sufficiently deep to
hold it firmly for the smoking j)rocess but not so deep as to occlude the
ventilation holes.

1 36

Chapter 1 0

j Standardization has produced a remarkably reproducible procedure

I given that the process involves the combustion of highly processed and

I packaged plant material. This is illustrated by the data shown in Table 1.

I Individual laboratories typically generate results with a precision of

i ±5 percent (relative standard deviation) or better for tar, nicotine, and

I carbon monoxide yields of high-tar products. Interlaboratory agreement

j is generally within 4 and 8 percent of the mean, depending on the

constituents and number of cigarettes considered. Precision and
interlaboratory agreement as a percentage of the mean are poorer for
very-low-delivery (e.g., 1 mg tar) products, but the absolute error is similar.
The procedure is sufficiently reproducible to allow rounding of FTC results
for tar and carbon monoxide to the nearest whole milligram based on a
] difference between brands of only 0.1 mg (0.4 mg or less reported as <1 mg

or below detection limit of the method, 0.5 mg or more rounded up to 1 mg,

I 1.04 mg rounded down to 1 mg, etc.). Results for nicotine are rounded to

the nearest tenth mg. Those with 0.05 mg or greater are rounded up,
whereas those with 0.04 mg or less are rounded down, as above.

INFLUENCE Each parameter specified in the FTC testing procedure influences
OF SMOKING the yields of tar, nicotine, and carbon monoxide. Restrictive
I PARAMETERS tolerances specified for acceptable puff volume, puff duration, and
so forth are required to allow comparison of similar products and to allow
interlaboratory comparability. Parameters such as cigarette conditioning
prior to smoking are specified to accommodate the realities of laboratory
measurements: in this case, that the cigarettes are likely to be analyzed after
long periods of cold storage. Minor variations in any of these parameters can
result in detectable differences in yields. Realistic (comparable with human
smoking practices) variations also can result in large differences in yields.

Darrall (1988) has reported a systematic study of the influence of smoking
parameters on yields of tar, nicotine, and carbon monoxide. Puff durations
of 1.6 seconds and 2.3 seconds produced essentially the same yields for very-
low-tar (^4 mg) cigarettes and almost indistinguishable yields for higher tar
products (Table 2). No clear trend toward increasing or decreasing yields
was noted. Changing puff volume from 35 to 40 mL produced a small but
generally consistent increase in tar and nicotine (Table 3). Low-tar products
I yielded 1 to 3 mg more tar and 0.1 to 0.3 mg more nicotine at 40-mL puff

j volumes than at 35-mL puff volumes. Higher tar products increased yield

by 2 to 5 mg of tar and 0.1 to 0.5 mg of nicotine. The increases, although
small, still may be larger than would be found using the standard FTC
method because the investigator in Darrall's study (1988) smoked at 2 puffs
per minute, thus increasing the number of puffs per cigarette. Larger changes
in puff volume produce larger changes in yields. Browne and colleagues
(1980) reported that particulate matter yield increased from 29 mg to 55 mg
for a U.S. blend experimental cigarette when the puff volume was changed
I from 17.5 mL to 50 mL under otherwise standard conditions. Carbon

monoxide yields were 9 mg and 20 mg for puff volumes of 17.5 mL and
: 50 mL, respectively.

137

Smoking and Tobacco Control Monograph No. 7 ■*

sc

<n

I

<

■D

C

(S

CN

I

<

O'

(N

</)

2

■o

c

o

■o

c

o

T5

3

(/)

^ O

ro

fO o

I- u

MH

0

"O

X

o

00

CD

o>

CO

o

O’

1—

in

CO

c

o

d

T-^

CD

CD

■r^

CD

CD

CD

CD

f—

+1

+1

■ft

1

+1

+1

+1

1

+1

+1

+1

1

o

c

■I-

O)

in

CO

CO

o

CO

O’

CO

V-»

o

co

O’

CO

in

CD

t”-'

c\i

c\i

cvi

(0

.a

CM

CM

CM

T —

T

T

”q)' O

0

^ Q

(B

O

lo “2

O) CO
*r^ *D

CO

h-

O’

O’

CM

CM

CO

y—

CO

O’

c

0

O

O

1 —

T—

O

O

o

o

o

o

o

o

C3) (D

c

CD

CD

CD

CD

CD

CD

CD

CD

CD

CD

CD

CD

o

+1

+1

+1

■H

■fl

+1

+1

+1

+1

+1

+1

+1

T3 +1

o

CM

CM

CD)

CM

CD

CD)

h-

CD)

r-

0) 0

Z

■O’

O’

1—

O’

CD)

O)

00

CD)

T—

CM

1—

1—

;c CD
^ CO

c\i

c\i

CM

cvi

CD

CD

CD

CD

CD

CD

CD

CD

i

0

>

<

CM

CD

CD

CM

00

in

■I-

CM

CM

CO

CD

■I-'

1-^

CD

CD

CD

CD

CD

CD

CD

CD

CD

CO

+1

+1

■H

■H

+1

+1

+1

+1

•fl

+1

+l

+1

h-

CD

o

in

CO

in

CD)

CD)

CD

CO

CO

O;

CD

CD

in

CD

in

c\i

c\i

CM

CO

1-^

T-^

iL

T-^

CO

CO

CO

CO

■>“

I—

tn

•c

o

CL

0)

E

3

■c

o

Q-

O

o

o

CM

CD ID

O

CM

CO CD

in o

CM

in

in

in in in

I- CM CO

1- CM CO

DC

CM

CM

<

CO CD

CO

<

in o

CM

CD in in in

CM CO

<d

.O)

Q

1C

c

o

c

r

f

Chapter 10

Table 2

Influence of puff duration on machine yields

Yield (mg/cigarette)®

Tar

Nicotine

Carbon Monoxide

Brand 1 .6 2.3

(FTC tar) Seconds Seconds

1 .6 2.3

Seconds Seconds

1.6

Seconds

2.3

Seconds

B (1 mg)

2

2

0.3

0.3

1

1

D (4 mg)

8

7

0.8

0.7

10

8

E (6 mg)

12

10

0.8

0.8

15

12

G (9 mg)

15

14

1.4

1.3

15

12

K (13 mg)

23

22

2.1

2.1

19

17

T (15 mg)

25

27

2.2

2.4

27

24

X (25 mg)

38

39

3.6

3.7

22

20

® 35-mL puff, 30-second interval.

Source: Darrall, 1988.

Table 3

Influence of puff volume

on

machine yields

Yield (mg/cigarette)^

Tar

Nicotine

Carbon Monoxide

Brand (FTC tar)

35 mL 40 mL

35 mL

40 mL

35 mL

40 mL

(B) KS-UM-V (1 mg)

2

3

0.3

0.4

1

1

(D) KS-EM-V (4 mg)

7

10

0.7

1.0

8

11

(G) KS-V (9 mg)

14

16

1.3

1.7

12

15

(H) Regular-V (12 mg)

21

21

0.9

1.0

25

25

(K) KS-V (13 mg)

22

24

2.1

2.2

17

17

(0) KS-NV (14 mg)

22

24

2.0

2.3

21

22

(P) KS-NV (14 mg)

24

28

1.6

2.4

18

24

(W) Regular-NF (16 mg)

27

26

2.1

2.4

16

14

P<) Regular-NF (25 mg)

39

44

3.7

4.2

20

21

® 2.3-second duration, 30-second frequency.

Key: KS = king size; UM = ultramild (< 4 mg tar); V = ventilated; EM = extra mild (4 to 7 mg tar); NV = nonventilated;
NF = nonfilter.

Source: Darrall, 1988.

139

Smoking and Tobacco Control Monograph No. 7

Puff frequency (Table 4) and filter ventilation (Table 5) were found to
have the greatest effect on yields. Decreasing the puff interval from 60 to
40 seconds increased the deliveries of tar, nicotine, and carbon monoxide by
20 to 50 percent on a per-cigarette basis. Using a puff interval of 30 seconds
increased deliveries by 40 to 90 percent. Blocking the ventilation system of
ventilated filter cigarettes has similar effects for products using a low degree
of ventilation and a much greater effect for highly ventilated products. This
is particularly important for very-low-delivery (e.g., si mg tar) products
because they typically use highly ventilated filters. Darrall (1988) reported
that complete blockage of the filter ventilation of a nominally 4.0-mg tar
product resulted in a tar yield of 10 mg (Table 5); nicotine increased from
0.5 to 0.8 mg, and carbon monoxide rose from 4 to 13 mg. Lower yield
products employ more highly ventilated filters than in Darrall's (1988)
example, and the influence of filter blockage would be expected to be
greater for such products.

The importance of filter ventilation to FTC testing is illustrated by the
results summarized in Figure 1. Nonfilter, filter (F), and ventilated filter (VF)
commercial cigarettes were smoked (see next section, "Influence of Human
Smoking Practices") under standard FTC conditions and again under standard
conditions but with 23 mm of the butt end taped (FTC+). All cigarettes.

Table 4

Influence of puff frequency on machine yields

Percentage Increase Over Standard Federal Trade Commission Method
40 Seconds 30 Seconds

Carbon Carbon

Puffs

Tar

Nicotine

Monoxide

Puffs

Tar

Nicotine

Monoxide

Regular (M-H)

28

31

29

26

62

60

49

38

Regular (M-H)

24

33

32

23

52

69

54

42

KS-UM-V (L)

47

55

24

43

90

154

47

67

KS-EM-V (L)

27

31

32

32

52

94

60

64

KS-EM-V (L)

30

21

19

27

69

84

48

79

KS (L)

38

19

19

24

76

54

38

43

KS-NV (L-M)

31

44

35

33

57

62

61

39

KS-NV (M)

26

26

16

16

60

60

32

32

IS-NV (L-M)

38

46

42

39

70

62

58

46

Key: M-H = middle to high tar (23 to 28 mg): KS = king size; UM = ultramild (<4 mg tar); V = ventilated; L = low tar (0 to
10 mg): EM = extra mild (4 to 7 mg tar); NV = nonventilated; L-M = low to middle tar (11 to 16 mg): M = middle tar
(17 to 22 mg); IS = international size.

Source: Darrall, 1988.

140

Chapter 10

Table 5

Influence of ventilation on machine yields

Yield per Cigarette (mg)

Percent Blockage

Brand

(ventilation)

Constituents

0

50

100

A (55%)

Tar

3.8

5.9

10.0

Nicotine

0.46

0.55

0.82

Carbon monoxide

3.8

6.0

12.7

B (35%)

Tar

9.2

10.6

12.8

Nicotine

0.90

0.90

0.98

Carbon monoxide

9.2

10.9

15.2

Source: Darrall, 1988.

including nonfilter, were taped. Nonfilter cigarettes and filter cigarettes
with little or no ventilation were seen to be only slightly affected by the
tape. Some effect would be expected for nonfilter cigarettes because taping
blocks air flow through the cigarette paper, but the changes observed were
barely statistically significant for the experimental design used. The effect
on ventilated filter cigarettes (VF-A to VF-F in Figure 1) was significant and
major. Products rated as FTC 1 mg tar yielded 5 mg or more of tar when
the ventilation was completely occluded. Products rated at 2 to 4 mg of
tar delivered up to 10 mg of tar. In the case of brand F-F, the substantial
increase in delivery when cigarettes were taped suggested that the filter
incorporated ventilation even though it was not obvious from visual
inspection. Trends for nicotine and carbon monoxide yields were generally
parallel to those for tar.

INFLUENCE OF Standardized machine smoking was developed to ensure that

HUMAN SMOKING differences in yields among cigarettes were caused by the nature
PRACTICES of the cigarettes and not by differences in the measurement

method. The FTC adopted standardized machine smoking to maximize
its ability to discriminate accurately among brands. The FTC test has been
successful for this purpose but is accurate only where cigarettes are smoked
as prescribed by the method.

The relevance of the FTC test parameters to human smoking practices
has been called into question as FTC ratings have increasingly been viewed
as a measure of human exposure. This concern is heightened by the
increasing popularity of low-tar and ultralow-tar products relying largely
on filter ventilation and by a better understanding of compensatory smoking
practices. Observation of more recent smoking practice showed that filter
ventilation was commonly compromised, puff volume was somewhat greater

141

Smoking and Tobacco Control Monograph No. 7

Figure 1

Tar yields using standard (FTC) smoking conditions and FTC smoking conditions
with tips taped (FTC+)

Cigarette Brand

Key: NF = nonfilter; F = fitter; VF = ventilated filter.

Source: Jenkins et al., 1982.

than the standard 35 mL, and a puff frequency of 2 to 3 per minute was more
common than was 1 per minute (U.S. Department of Health and Human
Services, 1988).

Jenkins and colleagues (1982) surveyed the influence of major changes
in smoking parameters on the yields of tar, nicotine, and carbon monoxide
by commercial cigarettes. Results are given in Fables 6 through 8.

142

Yield (mg/cigarette ± one standard deviation)

Chapter 10

2

C8

C

o

!/l

c

o

“O

c

o

(J

0£

c

o

O) L_

o

CD

CO

O)

o

0)

0) CO
TO CV

(U

>

<

o

"O ^

Q- r\i

o CO

i~ m

CO

o

(U

O cvj

If)

CO

o

0)

5

o

E

r^

^ o
^ u
“ra 5C

c

00

00

CO

p

ID

1—

CD

CO

CD

p

CVJ

p

CO

ID

T-'

cvj

cvj

1-^

CO

CO

cvj

CD

ID

CO

CO

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

3

n

CVJ

CO

ID

p

ID

CO

ID

ID

o

p

CO

00

00

CD

ID

CO

CD

ID

CO

y-t

CD

d

1 —

CO

CO

CO

CO

CVJ

CO

CVJ

CVJ

ID

ID

CO

CD

CD

_c

O)

o

00

CVJ

CD

O

CO

p

CO

00

CVJ

CD

CO

cvi

■1-^

CO

d

cvj

cvj

cvj

T-^

1-^

T-‘

cvj

cvj

CO

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+ 1

+1

+1

+1

3

n

00

CVJ

CD

ID

CD

p

p

CD

o

h-

CC)

cvi

id

d

d

CO

CO

y-t.

d

00

d

ID

ID

CO

ID

ID

CO

CO

CVJ

CD

CD

CD

_c

CO

CO

ID

00

p

CVJ

CD

ID

p

00

CO

p

1-^

T-^

o

d

d

d

CO

T-^

1-^

T“

CO

cvj

cvj

y —

CO

:t=

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

3

CL

CVJ

1^

00

CD

p

o

CD

CD

CVJ

CO

CVJ

CD

■iL

1^

00

cvj

■iL

CO

CO

ID

cvj

d

CD

CO

CO

CVJ

CO

CO

CO

CVJ

CO

c

(3)

CVJ

Oi

CO

ID

O

CD

CO

p

ID

CVJ

o

CD

1-^

CD

d

d

d

cvj

d

cvj

CO

T-^

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

0-

CO

o

T-

CVJ

h-

CVJ

p

p

P

ai

c6

r'i

CD

ID

CD

CO

y-^

CD

CVJ

CVJ

CVJ

c

lO

h-

CD

CVJ

CO

p

00

oo

p

CD

p

CD

CD

o

o

d

d

d

T-^

d

•r^

d

■r^

d

1-^

cvj

T-^

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

^

Q_

CO

o

a>

CD

CO

ID

r^

Vj"

CD

CD

o

ID

p

CO

CO

cvj

y-t.

d

d

CO

CD

CO

CD

00

cvj

tL

CD

CVJ

CVJ

CVJ

cz

CO

CO

1—

CVJ

CO

00

00

ID

CD

CO

h-

CD

o

CD

CD

d

d

d

d

d

d

d

d

d

T-^

d

d

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

CL

CO

h-

CO

p

p

1—

CD

00

CD

1—

CD

N-

CD

CD

d

00

CO

CO

cvj

d

cvj

d

1

1 —

0)

CO

CO

CD

o

CO

CO

CO

O

CD

p

p

c5

CD

cvj

cvj

<0

ra

CD

ID

CD

00

d

cr

CVJ

CVJ

CVJ

"O

c

<

LL

CD

uL

o

uL

Q

UJ

U.

<

liL

CD

LiL

O

CO

k_

1

U-

u!.

ljL

<

CQ

O

Q

LIJ

LL

1

LL

CQ

>

>

>

>

>

>

uL

uL

uL

u!.

uL

LL

z

z

2

i

f

143

® Below reporting limit.

^ Not determined.

Key: VF = ventilated filter; F = filter; NF = nonfilter.

Yield (mg/cigarette ± one standard deviation)

Smoking and Tobacco Control Monograph No. 7

o

High+
ip Taped
mL, 3 Se

CM

CM

00

CO

00

O

in

CM

00

o

CM

0)

:*=

3

+1

CM

CM

+1

+i

CO

+1

CO

+1

00

+1

CO

+1

o

+1

CD

CM

+1

O

+1

o

+1

o

o

+1

CM

CM
+ 1

CM

+1

in

CM

+1

CD

0.

O

O

CM

in

o

00

CM

CM

CD

o

in

r- lO

CO

c\i

cm’

f—

T—

c\i

cm’

c\i

c\i

T-

c\i

CM

CO

h-

O

<u

c

CO

CD

in

CO

T—

00

CO

CD

CO

in

CD

t—

CD

00

CO

O

o

■*“

T-

CO

o

O

CM

T-

CM

O) ^
^ 1

To

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

00

1—

CM

CM

CM

o

CD

T—

T~

CM

CD

oo

T—

05

E

o

T-

00

00

in

CM

1—

CO

00

CD

CM

CO

CD

CM

Q.

T—

CM

c\i

CM

■J—

T~

CM

CO

CO

m

CO

o

CD
0) CO
O)

o

in

O

in

CO

00

O)

CD

CM

CD

CD

CO

•«—

•*—

T*~

o

o

o

1—

o

CO

o

CM

O

T-

CM

CO

(0 ^

CO

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

(D —I

3

m

00

CM

CM

00

05

CM

CO

in

CD

CO

CD

CM

CD

< ^

CL

o

CO

CD

in

CM

CM

C35

CD

in

in

OO

CO

lO

CM

T—

T~

—

—

CVJ

CO

2

.i

Ol

o

u

c

o

(/)

c

_o

c:

o

o

Ofj

c

J2

o

E

</>

t "o

^ u

J3 4>

m 4;

144

o

0) c

oS|

CO

5

o

o B

h ™
Ll ce

■o c

in

h-

CO

00

CO

CD

CD

O

CD

CO

00

CO

+

05 C/J V_

9- CM ^

CO

CO

CO

p

P

p

p

p

P

p

p

p

p

o

^

+1

+1

+1

+1

+1

+1

ra +1

+1

+1

+1

+1

+1

+1

+1

1-

1— —I D

CM

CM

in

CM

1^

f —

O

T-*

OO

y—

CD

r-

CD

o

Ll

C5. F 0-

CO

CO

CO

CO

00

N-

CO

00

in

CO

CD

h- in

T~

'T~

in

CO

in

y—

CM

h-

CM

CM

CM

CO

in

•V—

CO

o

o

o

o

O

o

O

O

O

O

o

o

o

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

o

in

in

05

in

CD

CD

CD

T—

o

CO

CO

CM

T—

o

O

O

CD

CD

CD

l—

o

y—

05 c
C/5 ~

y~“

•d"

CO

O

in

in

CO

■d-

in

CO

00

o

o

o

o

o

o

t—

o

o

o

o

o

o

O

o

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

+1

E 0-

CO

o

in

T“

CM

CD

o

CO

in

y^

CD

05

o

d-

o

o

O

O

in

in

in

CM

CO

CD

h-

CM T-
tn CO CO

O)

o

o 1-

■M- CM T-

CM r^

in

CM CM CD

■M- in CO

TO

c

<

CD

LL

o

LL

Q

LU

LL

2

LL

LL

LL

LL

< CD

m

>

>

>

>

>

>

LL LL

Q UJ U_

c:

00

O

LL

■Q

II

fc

LL

O

f

2

Z

CQ

4

1

ventilated filter; F = filter; NF = nonfilter.

Yield (mg/cigarette ± one standard deviation)

Chapter 10

w

.Si

"><

o

c

o

E

c

o

re

u

c

o

(/>

c

o

o

u

C£

IS

o

E

O)

X

o

■§ w

o

(U

CO

o

(D

CO

TO

(D

>

<

O

"O ^

Q- c\j

O TO

i~ lo

CO

o

TO

O <>J

t

CO

$

o

E

00 o
-Si u
"re tC

d

r^

(N

O

00

P

CM

CM

•re-

CD

•re-

CM

p

CD

h-

re"

cd

CM

CM

CM

c\i

CM

ed

CM

CM

cd

d

in

To

+1

+1

+1

+1

+1

+1

+l

+l

+i

+l

+1

+l

+i

+1

+1

TO

in

CM

in

N-

P

CD

p

•re-

CM

p

■I—

00

•re-

p

in

0-

•re-

cd

r-^

00

cd

in

ed

<D

cd

in

ed

•re-

od

co

CO

CM

CM

CM

CM

CM

CO

CO

CM

CO

CM

CM

CM

CO

CM

c

I--

05

CD

h-

•re;

CD

r-^

CM

CM

CM

r-

P

■r-^

T“

•re-

1-^

d

1-^

CM

iri

cd

CM

evi

T*“

1-^

cd

CM

TJj

+1

+1

+1

+1

+1

+1

+l

+i

+l

+l

+1

+l

+1

+1

+1

TO

CL

00

o>

h-

O

CM

CO

00

00

■I—

•re-

1-

CD

CD

CO

p

cd

CM

od

CD

CD

od

in

CD

CD

ed

in

CM

in

d

T-t.

CO

CO

CM

CM

CM

CM

CM

CM

CO

CO

c

00

C3)

in

r^

CM

1-^

•re-

p

CD

p

CM

•re;

in

p

d

T-^

•r-^

d

d

d

c\i

T-^

CM

1-^

1-^

CM

d

1-^

In

M—

+1

+1

+1

+1

+1

+1

+i

+l

+l

+1

+i

+l

+1

+1

+l

TO

C\J

CM

CD

o

CO

•re-

CO

p

in

CD

P

CD

CD

0-

(d

CD

in

CM

ed

in

•re-

re-

cd

•re-

c\i

CVJ

CM

CM

CM

T“

CM

CM

c

in

cq

CM

•re-

re;

o

p

r--

p

P

00

•r-^

1 —

d

1-^

1-^

T--

d

•>—

1-

CM

T-

d

+1

+1

+1

+1

+1

+i

to

+i

+i

+1

+i

+i

+1

+1

+l

TO

CL

CO

00

h~

oo

00

CM

CD

p

■I—

■I—

oo

P

CD

C\j

00

od

d

d

od

CD

d

CM

CM

•re-

evi

T

cz

h-

in

CM

CO

p

CD

p

CO

p

p

p

00

p

d

d

d

d

d

•r^

d

•r-^

d

•r-^

d

T-^

+1

+1

+1

+1

+1

+i

+l

+l

+l

+1

+l

+l

+1

+1

+l

TO

CL

1-

o

CD

r-

CD

•re-

in

00

CD

p

CO

p

•re-

CM

CM

•iL

d

d

in

ed

od

od

CD

c\i

CO

t —

c

CO

in

CM

O

o

CD

lO

T—

in

in

in

CD

•re-

f-

CD

d

d

d

d

d

d

d

d

d

d

d

d

d

+1

+1

+1

+1

+1

+i

+i

+l

+i

+1

+i

+l

+1

+1

+l

CL

CD

00

C\J

00

h-

P

o

-t —

p

CD

CD

m

CD

in

p

CD

p

d

d

in

in

CD

cd

CM

CM

•re-

in

•re-

TO

o

O)

cq

p

p

p

m

CO

CD

•re-

in

r--

TO

O)

CM

•re-

T-^

•re-

in

CD

d

od

d

CM

CD

ed

X

T

■o

c

<

LL

CD

O

Q

LU

LL

<

LL

GQ

o

LL

TO
1

1

LL

LL

LL

LL

LL

<

m

O

Q

1

UJ

LL

LL

CQ

>

>

>

>

>

>

LL

LL

LL

LL

IL

IL

X

X

X

TD

cb

C

I

145

Key: VF = ventilated filter; F = filter; NF = nonfilter.

Smoking and Tobacco Control Monograph No. 7

Smoking parameters were chosen to represent less intense smoking (17-mL,
1-second puffs, once per minute), conditions considered to be "average"
smoking (45-mL, 2-second puffs, twice per minute) at the time, and extreme
("high") smoking (75-mL, 3-second puffs, three times per minute) conditions.
These results were compared with results generated using the standard FTC
conditions, FTC conditions with 23 mm of the butt end of each cigarette
taped to completely occlude tip ventilation (FTC-i-), and the extreme
conditions with the tips taped (high+). Yields under high+ conditions
were viewed as the maximum practical yields of the cigarettes.

The cigarettes chosen for analysis were selected by weight and pressure
drop (the differential pressure from end to end when air is drawn through a
cigarette at a rate of 1,050 mL per minute [equivalent to a 35-mL puff taken
over a 2-second period]) from two cartons purchased locally; they were
conditioned and smoked under FTC-specified environmental conditions.

The smoke was trapped and analyzed using FTC methods except that carbon
monoxide was determined using gas chromatography (Horton and Guerin,
1974) rather than nondispersive infrared spectroscopy. A single-port and
a linear four-port Filimatic smoking machine were used rather than the
standard 20-port machine, and one to six cigarettes were smoked per port
depending on the smoking conditions used. At least four ports of cigarettes
were smoked per brand or condition, but the precision of the results remained
2 to 3 times poorer than would be expected using the standard 20-port
protocol.

Results for tar deliveries are diagramed in Figure 2. Results for nicotine
and carbon monoxide generally parallel those for tar (although carbon
monoxide yields are more scattered and less systematically varied). Several
observations are apparent. First, the trend toward decreasing yields generally
parallels the decrease in FTC yields regardless of the conditions used for most
products. Second, products with barely detectable yields of tar measured by
the FTC method produce readily detectable quantities of tar when smoked
under reasonable conditions. Third, even the lowest FTC tar products can
yield 10 to 20 mg of tar under sufficiently aggressive smoking conditions.
Products with very low FTC tar yields that depend largely on filter ventilation
are those most subject to underestimation of practical yields by the FTC
method.

Several investigators have reported on the influence of more relevant
combinations of smoking conditions on the yields of tar and nicotine.

Fable 9 summarizes some of these observations for tar. Rickert and colleagues
(1983) reported that increasing the puff volume to 48 mL and decreasing
the puff interval to 44 seconds resulted in an increase of approximately
40 to 90 percent in the yield of tar over that found using standard ITC
conditions. Using the same conditions but also occluding 50 percent of the
filter ventilation resulted in an increase of from 70 to 500 percent depending
on the product. Percentage increase in yield tended to correlate inversely
with yield of ITC tar; that is, the lower the ITC yield, the greater the
{)ercentage increase.

146

Chapter 10

Figure 2

Influence of smoking parameters on constituent yield

(mg/cigarette)

Key: low = 1 7-mL puff volume, 1 -second duration, 1 puff/minute; FTC = 35-mL puff volume, 2-second duration, 1 puff/
minute: FTC+ = same as FTC plus butt end taped; average = 45-mL puff volume, 2-second duration, 2 puffs/
minute; high = 75-mL puff volume, 3-second duration, 3 puffs/minute; high-r = same as high plus butt end taped;
NF = nonfilter; F = filter; VF = ventilated filter.

Source: Jenkins et a!., 1982.

It

«

I

I

I

The Rickert and colleagues (1983), Darrall (1988), and Jenkins and
colleagues (1982) studies all considered the effect of increasing the puff
frequency from one per minute to two per minute. Puff volumes varied
from 40 to 48 mL across the studies, but the results were similar. Smoking
at two puffs per minute approximately doubled the tar yield for most
products tested.

Filter ventilation also has been considered. The Jenkins and colleagues
(1982) data illustrated that 100 percent ventilation blockage increased the
tar delivery by a factor of 10 to 20 for very low (< 1 mg) FTC tar products
if all other smoking parameters were kept constant. Kozlowski and
coworkers (1982) reported increases of a factor of 20 to 40 using conditions
of 100 percent blockage, a 47-mL puff volume, and a 44-second puff interval
for products rated as <1 mg FTC tar. The influence of ventilation blockage
was smaller but still important for products rated as 1 to 6 mg FTC tar (2 to
6 times increased delivery compared with that measured using the FTC
method without ventilation blockage), was readily detectable for products

147

Table 9

Influence of "more relevant" smoking parameters on tar yields

Smoking and Tobacco Control Monograph No.

03

03

■o

C

«

5

c

0>

E

' w

<u

Q.

X

LU

03

QC

CM
g 03

Lx

03

CM

00

O

O «

CM
g CJ3

Lx

^ "to
"5

03

03 CO

^ 03
03
_x
o

DC

03

03 CO
_ OO
^ 03
03
JX

o

DC

03

03 CO
^ 00
^ CJ3
03

o

DC

03

£2

03

}= ra

C33

p

03

E

CD

CD

c\i

m

CM

00

CD

03

CD

CJ3

CO

03

to

U3

CO

P 00 §

CM LO °

r^

CM

O

o

o

o

C33

CM

O

CM

CD

cd

03

cvi

1

cvi

cm'

CM

LO

o

o

oo‘

CD

c\i

r^

1

00

CD*

CO

Cvi

CO

CO

o

h-

CJ3

o

C33

O

1

cd

c\i

T*~

CM

00

CM

o

CO

o

1

o'

o

oo'

h-

CD

r-

CM

CO

cd

CM

co

CO

1

1

CM

c\i

1

CM

00

1

1

1

1

T —

o'

1

o

cvi

CO

c\i

CM

cvi

(O

CO

C33

CO

U3

c\i

cvi

CM

1

1

1

CO

-

'

'

1

'

1

1

1

Cvi

CM

00

1—

o

1^

U3

CM

cm'

CM

O

<33

h-

C33

C33

cvi

•>—

't —

1 —

T-^

1

1

oo

tT

cd'

CO

1

•«0;

CD*

1

1

CM

T-^

■r^

o

m

CO

CD

U3

1-

o

CM

CD

o

CM

A

03

c

g

c

o

O

«

c

03

E

k_

03

C3.

03

E

_D

O

>

3=

D

O

03

C/3

C

o

D

Q

it=

r3

o

03

(O

-X

o

o

CQ

c: iU

c

g

to

CL D_

D

0.

c

03

>

148

Key: NV = nonventilated.

Chapter 1 0

rated up to 10 mg FTC tar, and became insignificant for products rated as
15 mg FTC tar and higher. It is conceivable that a higher tar (e.g., > 10 mg)
product exists that incorporates a highly ventilated filter. Such a product
would be affected by ventilation blockage similar to the way lower tar
products are affected.

CONCLUSIONS The FTC procedure for measuring the tar, nicotine, and carbon
monoxide yields of cigarettes provides an accurate measure of yield for
cigarettes smoked in the specified manner. It serves the purpose of comparing
the yields of brands smoked under the same (specified) conditions. The utility
of the procedure for measuring human exposure is doubtful because it is
unlikely that all brands are smoked in the same way. This is especially the
case given the wide variety of products currently available. Results using
realistic combinations of puff volume, puff frequency, and filter ventilation
blockage suggest that human smoking conditions can produce from two times
(nonfilter and standard filter brands) to ten times (low-tar and very-low-tar
ventilated filter brands) the yields of tar that are measured by the FTC test.
Nicotine and carbon monoxide yields vary similarly.

The current FTC test procedure must continue to be used if there is a need
to compare current products with those of the past. New or additional sets of
smoking parameters must be adopted if a more accurate measure of human
exposure is desired.

QUESTION-AND-ANSWER SESSION

DR. RICKERT: There was a question I had asked earlier today and I wonder if
you could answer it. It looks as though you have some information about
interlaboratory variation, plus within-lab variation, plus variation over time.

If you measure, for example, a 12-mg cigarette — how different would another
brand have to be before you would be comfortable in calling it truly different?

DR. GUERIN: Certainly it would have to be more than 10 percent different. I
think that it is more like, at that range, about 2 mg different.

DR. RICKERT: So, you would say that, for example, 10 mg would be
considered different from one that was 14; but other than that, there would
be virtually no difference.

DR. GUERIN: Right.

REFERENCES

Bradford, J. A., Harlan, W.R., Hanmer, H.R. Nature of
cigaret smoke. Technic of experimental smoking.

Industrial and Engineering Chemistry 28(7): 836-839,

1936.

Browne, C.L., Keith, C.H., Allen, R.E. The effect of
filter ventilation on the yield and composition of
mainstream and sidestream smokes. Beitrage zur
Tabakforschung International 10: 81-90, 1980.

Darrall, K.G. Smoking machine parameters and
cigarette smoke yields. Science of the Total
Environment 74: 263-278, 1988.

DeBardeleben, M.Z., Wickham, J.E., Kuhn, W.F. The
determination of tar and nicotine in cigarette
smoke from an historical perspective. Recent
Advances in Tobacco Science 17: 115-149, 1991.

Federal Trade Commission. "Tar, Nicotine, and
Carbon Monoxide of the Smoke of 933 Varieties of
Domestic Cigarettes." Internal report prepared for
the National Cancer Institute. Washington, DC:
Federal Trade Commission, 1994.

149

Smoking and Tobacco Control Monograph No. 7

Horton, A.D., Guerin, M.R. Gas-solid

chromatographic determination of carbon
monoxide and carbon dioxide in cigarette smoke.
Journal - Association ofOfpdal Analytical Chemists
57: 1-7, 1974.

Jenkins, R.A., Pair, D.D., Guerin, M.R. "Deliveries of
Tar, Nicotine, and Carbon Monoxide of Selected
U.S. Commercial Cigarettes Smoked Under 'More
Relevant' Smoking Parameters." Oak Ridge
National Laboratory (ORNL) Project Topical Report
No. 120. Unpublished report, available from the
authors. Oak Ridge, TN, 1982, 13 pp.

Kozlowski, L.T., Heatherton, T.F., Frecker, R.C., Nolte,
H.E. Self-selected blocking of vents on low-yield
cigarettes. Pharmacology, Biochemistry and Behavior
33(4): 815-819, 1989.

Kozlowski, L.T., Rickert, W.S., Pope, M.A., Robinson,
J.C., Frecker, R.C. Estimating the yield to smokers
of tar, nicotine, and carbon monoxide for the
"lowest yield" ventilated filter cigarettes. British
Journal of Addiction 77: 159-165, 1982.

Rickert, W.S., Robinson, J.C., Young, J.C., Collishaw,
N.E., Bray, D.F. A comparison of the yields of tar,
nicotine, and carbon monoxide of 36 brands of
Canadian cigarettes tested under three conditions.
Preventive Medicine 12: 682-694, 1983.

Thomsen, H.V. International reference method for the
smoking of cigarettes. Recent Advances in Tobacco
Science 18: 69-94, 1992.

U.S. Department of Health and Human Services. The
Health Consequences of Smoking: Nicotine Addiction:

A Report of the Surgeon General, 1988. DHHS
Publication No. (CDC) 88-8406. Rockville, MD: U.S.
Department of Health and Human Services, Public
Health Service, Centers for Disease Control, Center
for Health Promotion and Education, Office on
Smoking and Health, 1988.

ACKNOWLEDGMENTS This is to thank Ms. Andi Palausky of our group for the
graphic summaries of data from the Jenkins and colleagues (1982) study.

This work was sponsored by the National Cancer Institute under Interagency
Agreement DOE No. 0485-FO53-A1 under U.S. Department of Energy
contract DE-AC05-840R21400 with Martin Marietta Energy Systems, Inc.

150

Chapter 1 1

Human Smoking Patterns

James P. Zacny and Maxine L. Stitzer

: INTRODUCTION It has been established that human exposure to tobacco smoke

constituents does not reflect package yield characteristics of cigarettes as
determined by Federal Trade Commission (FTC) smoking machine methods.

I This chapter describes some reasons for this discrepancy by examining

! features of human smoking behavior and how smoking behavior interacts

* with cigarette yield characteristics. The chapter is divided into four sections.

I The first section describes the topography of cigarette smoking; the second

I identifies the parameters of smoking topography that influence smoke

I exposure; the third shows that human smoking patterns are dynamic rather

than static; and the fourth draws conclusions about the relevance of the
FTC methodology to human smoking patterns.

HOW DO The first behavioral aspect of smoking involves holding the cigarette,
t HUMANS When smoking low-yield cigarettes (nicotine yield < 0.9 mg), smokers

' SMOKE? may knowingly or unknowingly block some or all the filter vents with

their fingers or lips. Blockage of these vents increases the density of
mainstream smoke that enters the mouth from the cigarette rod because S

the opportunity for air to be drawn into the smoke stream via the vents is "

reduced. Vent blocking essentially can turn a low-yield cigarette into a high-
yield cigarette. Over the past 10 years, Dr. Lynn Kozlowski has performed
a series of studies in which cigarette butts were assessed for vent blocking.

He obtained these butts from public access places such as shopping malls. j

From his butt analyses, he estimated the extent to which smokers in the
United States engage in vent blocking. In one study (Kozlowski et al., 1988), .

the incidence of partial or complete vent blocking of ultralow-yield cigarettes j,

(0.1 to 0.4 mg of nicotine) was 58 percent. In a more recent study, Kozlowski »

and colleagues (1994) collected butts of so-called "light" cigarettes (0.5 to
I 0.8 mg of nicotine yield) and found that 53 percent of the butts showed

I evidence of some degree of vent blocking. Vent blocking can be detected by

j looking at the filter stain: Cigarettes that are not vent blocked have a dark

stain in the middle of the filter toe with a visible white ring surrounding the
; stain (i.e., "bulls-eye" pattern); cigarettes that are vent blocked have filter

1 stains that encompass to varying degrees not only the middle of the filter

j toe but also the periphery.

! What are other features of smoking behavior? The smoker draws on

the cigarette, inhales the smoke into the lungs, then exhales. Drawing or
puffing parameters that can be measured include the size of the puff (puff
volume), the duration of the puff, and the interval between puffs. Inhalation
I parameters that can be measured include the amount of air that is mixed

] with the smoke as it is inhaled into the lungs (inhalation volume, also

I referred to as inhalation depth), the duration to peak inhalation, and any

151

Smoking and Tobacco Control Monograph No. 7

breath holding that occurs. Exhalation parameters include exhalation
volume and duration. These smoking parameters can now be measured
with technologies that have been developed over the past 20 years. Puffing
parameters can be measured with a plastic flowmeter, which is attached to
a pressure transducer; this system measures pressure differences between
two points in the flowmeter as the cigarette is puffed. Respiratory parameters
can be measured with noninvasive respiratory inductive plethysmography.
Essentially, the degree of movement of the chest and abdomen after
calibration procedures is directly proportional to volumes of smoky air
inhaled and exhaled. Thus, smoking is a complex behavior with a number
of discrete, measurable elements.

WHICH HUMAN It is important to identify which specific elements of

SMOKING BEHAVIORS smoking behavior influence smoke exposure to focus on

DETERMINE SMOKE relevant parameters of the FTC testing procedures vs.

EXPOSURE? human smoking comparison. Stitzer, Zacny, and other

colleagues over the past several years have conducted three studies (Zacny
et al., 1986 and 1987; Weinhold and Stitzer, 1989) that have examined
the relative importance of various smoking topography parameters in
determining smoke exposure. Smoke exposure is measured by determining
the amount of carbon monoxide (CO) and nicotine absorbed from smoking
a single cigarette — these parameters are called CO boost and nicotine boost,
respectively.

In these studies, smokers were trained to puff and inhale the cigarette in
a standardized fashion. The procedure of the standardization is simple: The
computer involved in the measurement of smoking topography parameters
can be programmed to beep when a specified level of a smoking parameter
has been reached. The investigator programs the computer to give the
smoker feedback as to when to (1) stop puffing (this controls puff volume),
(2) stop inhaling (this controls inhalation volume), and (3) start exhaling
(this controls breath-hold duration). After practice with this biofeedback
system, the smoker is able to reproduce a given smoking pattern that
includes a fixed puff volume, inhalation volume, and breath-hold duration.

In the first study (Zacny et al., 1986), an ultralow-yield cigarette was
smoked in this standardized fashion, and the number of vents that were
blocked was varied. In this way, the effect of vent blocking on smoke
exposure could be determined, as measured by CO boost. Either no vents
were blocked with tape, 50 percent of vents were blocked, or 100 percent of
the vents were blocked. Smokers took eight fixed-volume puffs (60 mL) from
the cigarette, inhaled to a certain volume (25 percent of vital capacity),
held the breath for a certain duration (10 seconds), and then exhaled. Any
differences in CO boost could be attributed to manipulation of vent blocking
because other smoking topography parameters were controlled. The authors
found a systematic increase in CO boost as a function of number of vents
blocked. In a second study, Weinhold and Stitzer (1989) varied the number
of puffs (from 8 to 16) taken from a cigarette. CO boost again increased in
a linear fashion as a function of number of puffs taken. In a third study

1.52

Chapter 1 1

(Zacny et al., 1987), three parameters were systematically manipulated:
puff volume (15, 30, 45, and 60 mL), inhalation volume (0, 20, 40, and
60 percent of vital capacity, respectively), and breath-hold duration (0, 4, 8,
and 16 seconds, respectively). As puff volume increased, the amount of
nicotine and CO absorbed from a cigarette increased in a systematic fashion.
However, varying the amount of air mixed with the smoke as it was inhaled
(inhalation volume) did not affect nicotine or CO boost; exposure was as
great with a shallow inhalation as with a deep inhalation. Breath-hold
duration increased CO boost but had no effect on nicotine boost. In
summary, the smoking topography parameters that appear to have the larger
effect on smoke exposure are vent blocking of low-yield cigarettes and the
number and size of puffs taken from any cigarette.

ARE HUMAN
SMOKING
PATTERNS
DYNAMIC OR
STATIC?

Much literature indicates that human smoking patterns are dynamic
and different from the static FTC smoking method. Puffing parameters
change during the course of smoking a single cigarette. Initially,
smokers take larger and longer puffs from the cigarette, but as they
smoke down the rod, the puffs get shorter and smaller. Interpuff
intervals are shortest at the beginning of the cigarette and longest near the
end of the cigarette. Smokers engage in activities that can have an influence
on smoking topography. Hatsukami and colleagues (1990) developed a
portable device that measures number of puffs, interpuff intervals, and puff
durations and assessed these parameters in a smoker's natural environment.
They found that variables, including mood of the smokers (relaxed vs.
stressed) and activities of the smoker (working vs. socializing), influenced
smoking topographies. Psychoactive drugs other than tobacco (e.g.,
stimulants, alcohol, opioids) also can influence smoking topographies.

Several investigators have noted changes in smoking topography as a
function of alcohol. Keenan and associates (1990) studied smoking
topography in alcoholic and nonalcoholic smokers: Alcoholic smokers
took more puffs from their cigarettes than did the nonalcoholic smokers,
indicating more intensive smoking and suggesting higher exposure levels
per cigarette.

Two other examples demonstrate that smoking is a dynamic process.

In the first example, Fant and associates (1995) studied smoking deprivation.
The number of cigarettes that subjects were permitted to smoke varied from
0 to 11 during a 6-hour period. The number of puffs taken was directly
related to the interval between cigarettes and inversely related to the number
of cigarettes smoked. In the second example, the authors reviewed studies
over the past 15 years that examined smoking topography as a function of
cigarette yield. We included only those studies that assessed the smoking of
commercially available, as opposed to research, cigarettes. We also arbitrarily
defined high-yield cigarettes as having nonventilated filters and an FTC
nicotine yield of 0.8 mg or more and low-yield cigarettes as having ventilated
filters and an FTC nicotine )deld of 0.6 mg or less. Table 1 summarizes the
seven studies that fit these criteria. A consistent finding in these studies is
that puff volume and puff number are both larger when low-yield compared
with high-yield cigarettes were smoked. Overall, it is clear that smoking

153

Smoking and Tobacco Control Monograph No. 7

Table 1

Studies that assessed smoking topography across different cigarette yields, using commercially
available cigarettes

Low-High

Nicotine

Yield

(mg)

Puff Volume

Puff Number

Reference

Number

Low

Yield

High

Yield

p Value

Low

Yield

High

Yield

p Value

Bridges et al., 1986

5 vs. 65^

0.3-1. 1

85.4

52.2

0.05

13.2

10.6

ns

Woodman et al., 1987

10

0.6-1. 4

59.5

43.6

0.05

14.0

12.1

ns

Zacny and Stitzer, 1988

10

0.1-1. 1

64.7

52.4

0.05

11.3

12.9

ns

Nil and Battig, 1989

15

0.5-0.8

25.7

26.6

ns

17.5

13.7

0.05

Hofer et al., 1991

36^

0.1-1.2

44.5

36.8

0.05

15.6

11.1

0.05

Kolonen et al., 1991

10

0.4-0.9

76.9

64.6

0.05

18.7

14.4

ns

Kolonen et al., 1992

8

0.3-1. 0

35.6

29.5

0.05

18.5

12.9

0.05

Mean

Range

56.0

25.7-85.4

43.7

26.6-64.6

15.5

11.3-18.7

12.5

10.6-14.4

® Cross-sectional study; the sample size in these studies represents each group of smokers studied within a yield
category.

Note: All studies were conducted with filtered, commercial brand cigarettes: low-yield brands were all ventilated and
ranged in nicotine yield from 0. 1 to 0.6 mg, and high-yield brands were all unventilated and ranged in nicotine
yield from 0.8 to 1.4 mg.

Key: ns = not significant.

topography is dynamic and changes in response to several factors, including
yield characteristics of the cigarette.

DOES THE FTC

METHOD

ACCURATELY

REFLECT

HUMAN

SMOKING

PATTERNS?

The FTC machine takes 2-second, 35-mL puffs every minute until
a certain point has been reached along the length of the cigarette (i.e.,
filter overwrap plus 3 mm). The length of the cigarette plays a large
role in how many puffs are taken by the smoking machine, although
porosity of the cigarette paper and tobacco burn rate also play roles.
How does the FFC method of smoking compare with how humans
smoke cigarettes? A table in the 1988 Surgeon General's Report on
smoking (U.S. Department of Health and Human Services, 1988) summarizes
results from 32 studies that assessed ad libitum human smoking topography.
Fable 2 lists the average values, along with the range of puffing parameters
observed in each study. Average puff duration across the 32 studies was
1.8 seconds, which is fairly close to the smoking machine value. Human
puff volumes tend to be larger than the 35 ml. used in standard FTC smoking
machine assays. The biggest difference between human and FTC machine
smoking parameters was in the rate of puffing. The average interpuff interval
in the human studies was 34 seconds, whereas FTC testing used a 60-second
interval. Fhus, humans took [)uffs at nearly twice the rate of smoking

154

'.i

I

I ' Chapter 1 1

Table 2

Published values of common measures of smoking

Number of

Reference Subjects

Puffs/

Cigarette

Interpuff

Interval

(seconds)

Puff

Duration

(seconds)

Puff

Volume

(mL)

Rawbone et al., 1978

12

10

41

1.8

Rawbone et al., 1978

9

10

35

2.1

43

Woodman et al., 1986

9

13

18

1.9

49

Nemeth-Coslett et al., 1986a

8

8

64

1.8

Nemeth-Coslett et al., 1986b

8

8

47

1.4

Nil et al., 1986a

132

13

28

2.2

30

Jarvik et al., 1978

9

10

Russell et al., 1980

10

11

35

Ashton et al., 1978

14

24

1.5

Schulz and Seehofer, 1 978

100

11

50

1.4

Schulz and Seehofer, 1 978

218

12

42

1.3

Henningfield and Griffiths, 1981

8

10

39

1.0

Stepney, 1981

19

13

38

Battig et al., 1982

110

13

26

2.1

40

Epstein et al., 1982

63

13

2.4

21

Russell et al., 1982

12

15

26

2.3

40

Gritz et al., 1983

8

9

47

2.2

66

Ossip-Klein et al., 1983

9

8

1.4

Ossip-Klein et al., 1983

9

12

1.9

Guillerm and Radziszewski, 1978

8

12

41

1.9

39

Gust et al., 1983

8

9

48

1.6

44

Adams et al., 1983

10

26

1.9

44

Moody, 1980

517

9

26

2.1

44

Nil et al., 1984

20

15

26

1.6

40

McBride et al., 1984

9

16

25

2.1

42

Medici et al., 1985

17

14

19

2.2

43

Burling et al., 1985

24

12

28

1.7

Nil et al., 1986b

117

13

22

2.1

42

Hughes et al., 1986

46

11

1.6

Bridges et al., 1986

108

11

56

Puustinen et al., 1986

11

13

22

2.3

44

Hilding, 1956

27

10

Mean

11

34

1.8

43

Range

8-16

18-64

1. 0-2.4

21-66

I

|i

'■i

Note: Data were taken from the baseline phase (or placebo treatment) of studies involving an experimental

manipulation with at least eight subjects. Values are rounded off to the nearest unit and, in some cases, were
calculated from other variables or estimated from data presented in figures; missing values indicate that the
variable was not measured or was not presented in the published study.

ii Source: U.S. Department of Health and Human Services, 1988.

I i

155

Smoking and Tobacco Control Monograph No. 7

machine rates used in standardized testing. Across the 32 studies, there
appears to be a large degree of variability in the values (as shown by the
range of values listed at the bottom of the table) that is not reflected in the
FTC method. The average number of puffs taken per cigarette by human
smokers was 11; FTC does not publish the number of puffs taken from a
cigarette by the machine. Differences in puffing rates suggest that the
FTC method probably underestimates the number of puffs taken from
a cigarette by humans.

It is possible to estimate the number of puffs used to determine FTC
cigarette yield by having cigarettes machine-smoked in a research laboratory.
The authors had a low-yield cigarette brand, Now, smoked according to the
FTC method at the Tobacco and Health Research Institute in Lexington,
Kentucky. Two hundred cigarettes were smoked; the average number of puffs
taken per cigarette was 6.8. This same procedure was repeated with a high-
yield cigarette. Camel, and an average of 8.3 puffs was taken. Thus, the
machine took more puffs from the high-yield than from the low-yield
cigarette, which is at odds with the human data presented in Table 1 in
which the opposite occurs. Therefore, there appears to be a discrepancy
between the FTC method of smoking and the way humans smoke different-
yield cigarettes: Machines tend to puff less smoke from low-yield than from
high-yield cigarettes, and humans tend to compensate for air dilution by
puffing more smoke from low-yield than from high-yield cigarettes. Thus,
humans smoke low-yield cigarettes in a manner that attenuates machine-
determined yield differences.

SUMMARY In conclusion, we have shown that the number and size of puffs are key
factors that determine per-cigarette smoke exposure. Vent blocking is
another important smoking behavior that can occur with low-yield
cigarettes. Human smoking behavior is dynamic, not static. There is
between-smoker variability in smoking topography, and there are dynamic
changes in response to smoking deprivation, cigarette characteristics, other
drugs, and situational determinants. The evidence suggests that the FTC
method does not accurately reflect human smoking patterns. The FTC
method takes smaller, fewer, and more widely spaced puffs than do humans,
on average. The underestimation of puff volume is exaggerated with low-
yield cigarettes because people tend to increase both the size and number of
puffs drawn from lower, as compared with higher, yield cigarettes, whereas
smoking machines decrease the number of puffs drawn while holding puff
size constant. In addition, the IH C method does not take into account the
important behavior of vent blocking of low-yield cigarettes. Fhus, there are
important differences between FTC and human smoking that result in the
machines underestimating the amount of smoke drawn by humans from
low-yield as compared with high-yield cigarettes.

156

Chapter 1 1

QUESTION- AND-ANSWER SESSION

DR. HOFFMANN: How did you cover half of the ventilation holes?

DR. ZACNY: We did not tape half of the cigarette. We put little pieces of
square tape all the way around it, so that approximately half the holes were
unblocked.

DR. HARRIS: Why is it that it is not how big a puff they took once it was in
their mouths, but how deeply they drag that puff into the lungs?

DR. STITZER: I think the explanation is that the dose of nicotine that is
drawn in with the puff is the critical determinant. The amount of air that is
breathed in along with it, which is what determines the depth, is how much
additional volume of air was breathed in with the smoke. That does not
seem to be relevant with nicotine.

DR. ZACNY: Even with a shallow inhalation, the surface volume of the lungs
is pretty huge.

DR. HARRIS: I understand. It is all in that 1.8-second drag on the cigarette.
DR. STITZER: Yes.

DR. HARRIS: Once the smoke is in your mouth, then you can jump up and
down; it does not matter.

DR. STITZER: No, once it is in your lungs. Once you have made that
inhalation maneuver, because if you just hold it in your mouth, that was
the zero inhalation condition.

DR. HARRIS: So, as long as you inhale it, it does not matter how much air
goes in with it.

DR. RICKERT: One important point is that you haven't looked at tar. Tar
may react differently. Depth of inhalation and volume of inhalation might
be more important. Deposition of tar is not nearly as efficient as for water-
soluble vapors and gases. In one of the documents that we received from
the tobacco industry, there is a study that has been cited by Stitzer, which
says that, on 1,631 cigarette butts, only .1 percent were completely blocked.
.The information that you have provided today suggests that it is somewhere
between 53 and 58 percent. What is the reason for the discrepancy?

DR. ZACNY: The reason for the discrepancy is that those 1,600 butts are
from only 10 subjects. We had them smoke the ultralow cigarettes for a
week and save the butts. We then analyzed the stain patterns.

We were looking at the acute effects of smoking these ultralow cigarettes
in the field, and there may be a lower incidence of blocking than what you
see when Lynn Kozlowski does his cross-sectional studies — when people have
been normally smoking these cigarettes for a long time. Plus, our data were
from only 10 subjects.

157

Smoking and Tobacco Control Monograph No. 7

DR. RICKERT: Do you feel that this blocking is something that we should be
concerned about?

DR. ZACNY: Yes.

DR. SHIFFMAN: One of the issues that you raised that we have not discussed
much is variability within a given smoker, due to brand switching, for
example. Can you give us some quantitative estimates of the degree of
variability?

DR. ZACNY: 1 believe Dr. Stitzer would be the best person to answer this
question.

DR. STITZER: In one example, it was shown for deprivation to be 10 to
15 puffs. And that makes quite a big difference when you multiply it by the
puff volumes, leading to a substantial difference in cumulative puff volume.

DR. HATSUKAMl: Also, one subject after meals typically took about eight
puffs per cigarette, whereas on the telephone, they would take about five
puffs from the cigarette.

DR. TOWNSEND: Dr. Zacny, 1 am a bit confused about the whole blockage
question. Is the measure that you used to determine hole blockage just the
staining at the mouth end of the filter?

DR. ZACNY: Yes. Different cigarettes have different types of what we call
tipping and different types of perforations. The perforations differ largely
in the number of holes and the size of those holes.

Those parameters of ventilation, in fact, determine to a large degree the
staining pattern in the first place. So, it is possible to make a highly air-
diluted cigarette with many ventilation holes that are very small and, in
fact, see relatively uniform staining patterns right at the mouth end of the
cigarette.

If you are interpreting that as vent blocking, then 1 think that is probably
an incorrect conclusion, because of the design of that specific filter. Filters
with large though very few holes will tend to force the smoke to the center
of the filter, and you will see that bullet shape right at the mouth end that
was shown in one of the slides.

Fhe concern is that not all cigarettes are built in the same way and so
that it is probably a bit premature to conclude that there is vent blocking
solely on the basis of filter observation.

DR. STITZER: The data that were presented in this talk showed what
happened to smoke exposure when the vents were experimentally blocked
with tape. Dr. 'I'ownsend is asking a different kind of question about
measurement of blocking in the natural environment.

DR. I'OWNSEND: So, these were not with actual subjects, then?

DR. S'n rZER: Fhey were with natural subjects, but we blocked the vents.

Chapter 1 1

DR. ZACNY: You were talking about the first study when we looked at 50
and 100 percent of hole blocking.

DR. TOWNSEND: 1 understand. 1 would like to talk with you some more
about this, because we have some data at R.J. Reynolds where we have gone
directly to an inhydrin staining test where the saliva on the filter, in fact,
stains with inhydrin; therefore, we can visually see how much saliva has
gotten up to the vents.

What we have seen in a study with a number of subjects is that the spent
butts show some blockage, but it is a very infrequent phenomenon. So, I
would like to talk to you further about that. Perhaps we can propose doing
some additional studies.

REFERENCES

Adams, L., Lee, C., Rawbone, R., Guz, A. Patterns of
smoking: Measurement and variability in
asymptomatic smokers. Clinical Science 65(4): 383-
392, 1983.

Ashton, H., Stepney, R., Thompson, J.W. Smoking
behaviour and nicotine intake in smokers presented
with a "two-thirds" cigarette. In: Smoking Behaviour.
Physiological and Psychological Influences, R.E.
Thornton (Editor). Edinburgh: Churchill
Livingstone, 1978, pp. 315-329.

Battig, K., Buzzi, R., Nil, R. Smoke yield of cigarettes
and puffing behavior in men and women.
Psychopharmacology 76: 139-148, 1982.

Bridges, R.B., Humble, J.W., Turbek, J.A., Rehm, S.R.
Smoking history, cigarette yield and smoking
behavior as determinants of smoke exposure.
European Journal of Respiratory Diseases 69[Suppl
146]: 129-137, 1986.

Burling, T.A., Stitzer, M.L., Bigelow, G.E., Mead, A.M.
Smoking topography and carbon monoxide levels
in smokers. Addictive Behaviors 10: 319-323, 1985.

Epstein, L.H., Dickson, B.E., Ossip, D.J., Stiller, R.,
Russell, P.O., Winter, K. Relationships among
measures of smoking topography. Addictive
Behaviors 7: 307-310, 1982.

Fant, R.V., Schuh, C.J., Stitzer, M.L. Response to
smoking as a function of prior smoking amounts.
Psychopharmacology 119: 385-390, 1995.

Glitz, E.R., Rose, J.E., Jarvik, M.E. Regulation of
tobacco smoke intake with paced cigarette
presentation. Pharmacology, Biochemistry and
Behavior 18(3): 457-462, 1983.

Guillerm, R., Radziszewski, E. Analysis of smoking
pattern including intake of carbon monoxide and
influences of changes in cigarette design. In:
Smoking Behaviour. Physiological and Psychological
Influences, R.E. Thornton (Editor). Edinburgh:
Churchill Livingstone, 1978, pp. 361-370.

Gust, S.W., Pickens, R.W., Pechacek, T.F. Relation of
puff volume to other topographical measures of
smoking. Addictive Behaviors 8(2): 115-119, 1983.

Hatsukami, D.K., Morgan, S.F., Pickens, R.W.,
Champagne, S.E. Situational factors in cigarette
smoking. Addictive Behaviors 15(1): 1-12, 1990.

Henningfield, J.E., Griffiths, R.R. Cigarette smoking
and subjective response: Effects of d-amphetamine.
Clinical Pharmacology and Therapeutics 30(4): 497-
505, 1981.

Hilding, A.C. On cigarette smoking, bronchial
carcinoma and ciliary action. I. Smoking habits
and measurement of smoke intake. New England
Journal of Medicine 254(17): 775-781, 1956.

Hofer, L, Nil, R., Battig, K. Nicotine yield as
determinant of smoke exposure indicators and
puffing behavior. Pharmacology, Biochemistry and
Behavior 40: 139-149, 1991.

Hughes, J.R., Pickens, R.W., Gust, S.W., Hatsukami,
D.K., Svikis, D.S. Smoking behavior of Type A and
Type B smokers. Addictive Behaviors 11: 115-118,
1986.

Jarvik, M.E., Popek, P., Schneider, N.G., Baer-Weiss,
V., Gritz, E.R. Can cigarette size and nicotine
content influence smoking and puffing rates?
Psychopharmacology 58(3): 303-306, 1978.

Keenan, R.M., Hatsukami, D.K., Pickens, R.W., Gust,
S.W., Strelow, LJ. The relationship between
chronic ethanol exposure and cigarette smoking
in the laboratory and the natural environment.
Psychopharmacology 100: 77-83, 1990.

Kolonen, S., Tuomisto, J., Puustinen, P., Airaksinen,
M.M. Smoking behavior in low-yield cigarette
smokers and switchers in the natural environment.
Pharmacology, Biochemistry and Behavior 40: 177-
180, 1991.

Kolonen, S., Tuomisto, J., Puustinen, P., Airaksinen,
M.M. Effects of smoking abstinence and chain-
smoking on puffing topography and diurnal
nicotine exposure. Pharmacology, Biochemistry
and Behavior 42: 327-332, 1992.

Kozlowski, L.T., Pillitteri, J.L., Sweeney, C.T. Misuse
of "light" cigarettes by means of vent blocking.
Journal of Substance Abuse 6: 333-336, 1994.

159

Smoking and Tobacco Control Monograph No. 7

Kozlowski, L.T., Pope, M.A., Lux, J.E. Prevalence of
the misuse of ultra-low-tar cigarettes by blocking
filter vents. American Journal of Public Health 78:
694-695, 1988.

McBride, M.J., Guyatt, A.R., Kirkham, A.J.T.,
Gumming, G. Assessment of smoking behaviour
and ventilation with cigarettes of differing nicotine
yields. Clinical Science 67(6): 619-631, 1984.

Medici, T.C., Linger, S., Riiegger, M. Smoking pattern
of smokers with and without tobacco-smoke-
related lung diseases. American Review of Respiratory
Disease 131(3): 385-388, 1985.

Moody, P.M. The relationships of qualified human
smoking behavior and demographic variables.

Social Science and Medicine 14A(1): 49-54, 1980.

Nemeth-Coslett, R., Henningfield, J.E., O'Keefe, M.K.,
Griffiths, R.R. Effects of mecamylamine on human
cigarette smoking and subjective ratings.
Psychopharmacology 88{4): 420-425, 1986a.

Nemeth-Coslett, R., Henningfield, J.E., O'Keefe, M.K.,
Griffiths, R.R. Effects of marijuana smoking on
subjective ratings and tobacco smoking.
Pharmacology, Biochemistry and Behavior 25(3): 659-
665, 1986b.

Nil, R., Battig, K. Separate effects of cigarette smoke
yield and smoke taste on smoking behavior.
Psychopharmacology 99: 54-59, 1989.

Nil, R., Buzzi, R., Battig, K. Effects of single doses of
alcohol and caffeine on cigarette smoke puffing
behavior. Pharmacology, Biochemistry and Behavior
20(4): 583-590, 1984.

Nil, R., Buzzi, R., Battig, K. Effects of different cigarette
smoke yields on puffing and inhalation: Is the
measurement of inhalation volumes relevant for
smoke absorption? Pharmacology, Biochemistry and
Behavior 24(3): 587-595, 1986b.

Nil, R., Woodson, P.P., Battig, K. Smoking behavior
and personality patterns of smokers with low and
high CO absorption. Clinical Science 71(5): 595-603,
1986a.

Ossip-Klein, D.J., Martin, J.E., Lomax, B.D., Prue,

D.M., Davis, C.J. Assessment of smoking
topography generaliz.ation across laboratory,
clinical, and naturalistic settings. Addictive Behaviors
8(1;: 11-17, 1983.

Puustinen, P., Olkkonen, 11., Kolonen, S., Tuomisto, J.
Microcomputer-assisted measurement of inhalation
parameters during smoking. Archives of Toxicology
(Suppl9j: 111-114, 1986.

Rawbrme, R.G., Murphy, K., Tate, M.E., Kane, S.J. The
analysis of smoking parameters: Inhalation and
absorption of tobacco smoke in studies of human
smoking behaviour. In: Smoking Behaviour.
Physiological and Psychological Influences, R.E.
Thornton (Ivditor). fxlinburgh: Churchill
Livingstone, 1978, pp. 171-194.

Russell, M.A.H., Sutton, S.R., Feyerabend, C., Saloojee,
Y. Smoker's response to shortened cigarettes: Dose
reduction without dilution of tobacco smoke.
Clinical Pharmacology and Therapeutics 27(2): 210-
218, 1980.

Russell, M.A.H., Sutton, S.R., Iyer, R., Feyerabend, C.,
Vesey, C.J. Long-term switching to low-tar low-
nicotine cigarettes. British Journal of Addiction 77(2):
145-158, 1982.

Schulz, W., Seehofer, F. Smoking behaviour in
Germany — the analysis of cigarette butts (KIPA).

In: Smoking Behaviour. Physiological and Psychological
Influences, R.E. Thornton (Editor). Edinburgh:
Churchill Livingstone, 1978, pp. 259-276.

Stepney, R. Would a medium-nicotine, low-tar

cigarette be less hazardous to health? British Medical
Journal 283(6302): 1292-1303, 1981.

U.S. Department of Health and Human Services. The
Health Consequences of Smoking: Nicotine Addiction.

A Report of the Surgeon General, 1988. DHHS
Publication No. (CDC) 88-8406. Rockville, MD:

U.S. Department of Health and Human Services,
Public Health Service, Centers for Disease Control,
Center for Health Promotion and Education, Office
on Smoking and Health, 1988.

Weinhold, L.L., Stitzer, M.L. Effects of puff number
and puff spacing on carbon monoxide exposure
from commercial brand cigarettes. Pharmacology,
Biochemistry and Behavior 33: 853-858, 1989.

Woodman, G., Newman, S.P., Pavia, D., Clarke, S.W.
Inhaled smoke volume, puffing indices and carbon
monoxide uptake in asymptomatic cigarette
smokers. Clinical Science 71(4): 421-427, 1986.

Woodman, G., Newman, S.P., Pavia, D., Clarke, S.W.
Response and acclimatization of symptomless
smokers on changing to a low tar, low nicotine
cigarette. Thorax 42: 336-341, 1987.

Zacny, J.P., Stitzer, M.L. Cigarette brand-switching:
Effects on smoke exposure and smoking behavior.
Journal of Pharmacology and Experimental
Therapeutics 246: 619-627, 1988.

Zacny, J.P., Stitzer, M.L., Brown, F.J., Griffiths, R.R.
Human cigarette smoking: Effects of puff and
inhalational parameters on smoke exposure. Journal
of Pharmacology and Experimental Theraj)eutics 240:
554-564, 1987.

Zacny, J.P., Stitzer, M.L., Yingling, J.E. Cigarette filter
vent blocking: FTfccts on smoking topography and
carbon monoxide exposure. I'harmacology,
Biochemistry and Behavior 2S: 1245-1252, 1986.

Chapter 12

Compensation for Nicotine by Smokers
of Lower Yield Cigarettes

Lynn T. Kozlowski and Janine L. Pillitteri

BACKGROUND The question has been asked whether brand-switching smokers
oversmoke lower nicotine cigarettes. The Federal Trade Commission
(FTC) testing method is a per-cigarette test and should be judged as such.
(Forty truly low-calorie candy bars together could be high calorie and still,
individually, be low calorie.) The FTC test cannot be blamed because
smokers smoke more cigarettes when they switch to those having a lower
5deld. Therefore, for this review compensation data were adjusted to per-
cigarette values. However, such per-cigarette adjustments only approximate
what would happen if the number of cigarettes were fixed for smokers.

If smokers have already compensated by smoking many more cigarettes,
then presumably they would have less need to smoke more of each cigarette.
In the five studies included in the authors' main review, the compensatory
percentage change in cigarettes per day averaged 15 percent (±6, 95-percent
confidence interval). No studies showed a decreased number of cigarettes
smoked with a lower yield brand of cigarettes.

Experimental brand-switching studies offering measures of nicotine
and cotinine were reviewed. An index of compensation was calculated
using a sequence of formulas developed by Russell and colleagues (1982).
Calculation of these formulas first requires information on the machine-
smoked nicotine yields of cigarettes to calculate (a) the percentage change
in nicotine yields. Information on the measured level of nicotine (or
cotinine) in body fluids is then used to calculate (b) the percentage change
in nicotine (or cotinine) intake. Finally, three consecutive formulas are
used to calculate (c) the actual compensatory increase in smoke intake
[(b/a - 1) X 100]; (d) the increase in smoke intake necessary for complete
compensation [(1- a)/a x 100]; and (e) using the values obtained in (c) and
(d) above, the degree of compensation [(c/d) - 100].

CIGARETTE BRAND Research on brand switching makes use of repeated-measures
SWITCHING IN designs. With these designs, the same smokers get different

EXPERIMENTAL cigarettes. This controls for individual differences in drug

RESEARCH metabolism (Benowitz et al., 1982) and for important biases

in brand selection, which usually are not controlled for in cross-sectional
research. This issue has been discussed by others (e.g., Giovino et al. [this
volume]; Cohen [this volume]). Wynder and coworkers (1984) explored the
demographics of smokers of the low-yield cigarettes and showed that age,
sex, race, education, and religion were strongly related to the selection of
low-tar cigarettes. Wynder and colleagues (1984) reported that education
is negatively associated with tar for males, but not for females. (Tar and

161

Smoking and Tobacco Control Monograph No. 7

nicotine are highly correlated across the full range of tar and nicotine yields.)
People who smoke low- and ultralow-yield cigarettes may be more health
conscious, have better diets, and be interested in smoking less. A random
sample of persons does not select ultralow-yield cigarettes.

Despite their advantages, experimental brand-switching studies have
important limitations. Outside of laboratories, smokers select their own
brands. There is a free market for most purchases of cigarettes. An
unsatisfying brand is likely to be rejected for a satisfying brand. Persons
trying an ultralow-yield cigarette may feel that they are puffing on air, so
they decide not to smoke these cigarettes and probably will not buy more
than one pack. Some compensatory smoking techniques (e.g., vent blocking
[Kozlowski et al., 1980 and 1989]) may take time to be learned by trial and
error. Short-term studies (i.e., less than 1 week of exposure on lower yield
brands) do not provide an adequate indication of the nature of compensatory
smoking in self-selected smokers. All reviewed studies involved brand
manipulations (change of "treatment" or brand in experimental study)
of more than 7 days.

Studies of brand switching also have biased samples. Who does and
does not volunteer for these studies? One of the five studies reviewed
(Guyatt et al., 1989) showed a dramatic number of dropouts following
informed consent. Of the people who went to at least one session in this
study, 81 percent dropped out. Another study on brand switching (Benowitz
et al., 1986a) required that participants be hospitalized for 14 days. Some
smokers, knowing that they were going to get ultralow-yield cigarettes,
either might not have wanted to smoke them or spend 14 days in the
hospital. One must wonder who would be available to participate in a
14-day study requiring confinement to a hospital room. Most studies of
brand switching also have small samples (mean = 22 subjects). As for
demographic differences, there is no way to represent the complexities of
age, sex, race, and education adequately in a sample of 22 participants.

According to the boundary model of dmg regulation, plasma nicotine
levels are not precisely regulated (Kozlowski and Herman, 1984); there are
aversive upper and lower limits or boundaries on intake for dependent
smokers. At the upper limit, when people are smoking a great deal, it is
difficult for them to smoke more due to overdose or toxic effects of nicotine.
When they are smoking a little, it is hard for them to smoke less than the
lower limit because of insufficient nicotine intake. However, within these
broad limits or boundaries, psychosocial factors primarily (i.e., the presence
of others smoking) determine nicotine ingestion, and dose manipulations
tend to have a smaller effect on smoking behavior (Kozlowski and Herman,
1984; Kozlowski, 1989) and how smokers feel (Benowitz et al., 1986b).

RESEARCH fable 1 shows the five studies reviewed and gives a summary of their

results. The following studies were not included in the review because they
were either too short term or used cigarette holders, which could interfere
with natural smoking behavior: Benowitz and colleagues (1986a), Kolonen

162

lary of five experimental brand-switching studies demonstrating changes in cigarette yields due to compensation

Chapter 12

OJ c

S E

■ m 3

I— t/5

C

o
'•*— *
ca
w
c
O)
Q.

E

o

O

CO

C/j

>

to

CD

E

CO

a

(D

CO

$

L_

nj

0-

0

05

_g

_l

b

0

0

T3

O)

(D

CD

c

CO

c

>-

-C

'c

1

0

0

CO

13

0^

0

CO

CO

E ^

in

- 5
0- o
c -I
(u .E

D) <l)
TO

-C o
O o

vP 2

0) c

O) •—

E "o
x: .03
O

W -d

o> E

o) -E
b -fo

"D ^

^ CO

CC >

II

CO ^

a> o)

F

a> ^
ra -E

05 -=

O 3

CD

>-

I

CO

13

(/)

Z)

O (/)

■5 w

>.

"O

■4-*

CO

CJ)

CM

CO

CO

>

CO

CO

CO

c

CD

E

o

C\J

iri

IT CO

CO

CD

c

C CD

O E

^ i

o >

CO

CO

in

CO

CO

>

o

o>

CM

CD

in

r--

r-. w

CO

00

m

CO

o

c\i

CO

CO

>

00

CM

CM

o

^ 8

I o

^ o
^ 8

CD

CO

CD

CD

CD

CD

00

(0

<0

CM

10

<D

Id

<D

lo

CD

CD

CO

05

CD

CO

lo

0

c

0

W

e

0

OD

q5

,

CO

in

~o

CO

CO

^ ^

CD

CO 05

C

CD

a

CO

CM

r^

^ 00

io

00

!q

00

00

CD

D CD

0

CD

0

(D

13

CD

ZZ-

0 :z-

cc

DC

ZZL-

CC

T~.

CD

CO

00

c\i

CM

CO

>

05

^

CD

CD

CD

^

, ,

CO

CD

05

CO

CO

■r^

T-^

CD

CD

T-‘

1-^

CO

r^

CO

r-.

CO

0

CO

CO

00 CO

>

CD

>

CD

>

>

If)

>

CD

1 —

■5t

00

1^

T~

CD

CD

CO

0

CD

CD

CD

CD

c

CD O

CO o
(D O

CM
CM ■>-

CO

CD ^

4-4 ^

CO oo
^ CD

■Q

§

O

cc

Q)

S

2

CO

.O)

CJ

.c

Q)

O)

S ,3>

•c: ?5

1 1
CQ CO
CO CO

S S

p c

^ Js

II

^ 3
to cp
O g
JO =r^

® 'co

CO O
CO CO

c ^
t CO

CO

c

o

.CO

w.

CO

%

o

o

*6

I

CO

t

CO

I

CO

.CO

s

CO

*o

g> 5

£5. CO

,Cn i;

sa

5 c

q> c]>

.§

CO
.3
CO

^ CO

c ts

CO 2>

o -2>
•2 o

c ^

q> ^
g CL

-Q 1£
CO

§

c ^
<0 00
Q. CJ>

E

8 'n
C '»
■is 3
^ 2>
5 ^
S4 O
O) O
CO "O

cb

*6 CO

I ^

Ll tc

Q>

163

Smoking and Tobacco Control Monograph No. 7

and colleagues (1991), Russell and colleagues (1975), and Zacny and Stitzer
(1988).

Figure 1 shows the pattern of results across the five studies. The solid
line summarizes results from the studies after adjusting for changes in the
number of cigarettes smoked. As nicotine yields go below the usual
"normal" levels (1.0 to 1.4 mg nicotine), more compensation takes place
until the lowest yield is reached. At this point, too much work may be
required of smokers to achieve substantial compensation. This kind of dose-
response pattern is consistent with that for other reinforcers. It may not be
important to compensate for a 0.9-mg nicotine cigarette; it easily provides
adequate levels of nicotine. The dashed line shows what happens when
there is no adjustment for changes in the number of cigarettes smoked.

This shows that compensation also is supported by an increase in cigarettes
per day in these brand-switching studies (the 0.4-mg nicotine cigarette now
shows close to 80 percent compensation).

Figure 1

Pattern of results illustrating percentage compensation across the five reviewed studies,
unadjusted (dashed line) and adjusted (solid line) for number of cigarettes

164

Chapter 12

The short-term study by Zacny and Stitzer (1988) (not included in our
review) examined smokers who had been given three different lower yield
brands (i.e., 0.1 mg, 0.4 mg, 0.7 mg nicotine). This study produced a pattern
of compensation similar to that in Figure 1.

For consumers, the average percentage compensation may be less
important than the likelihood of substantial compensation. If 1 in 2, 1 in 5,
1 in 10, 1 in 50, or even 1 in 100 smokers shows compensation of 25, 33, 55,
or 75 percent, then a problem exists. If automobile brakes failed at a rate of
even 1 in 1,000, this rate would be of great concern to manufacturers,
consumers, and regulatory agencies.

ONE REPEATED- Lynch and Benowitz (1987) conducted a self-selected brand-

MEASURES STUDY switching study of participants who spontaneously switched
OF SELF-SELECTED cigarette brands. The study included 62 people who had
BRAND SWITCHING lowered their standard yield. When they had been studied

earlier, they had had plasma measures taken, and they were recontacted 3 to
6 years later. In this group, the low-yield cigarette was 62 percent of the
former usual cigarette yield of nicotine (.68 mg versus 1.09 mg). Plasma
cotinine per cigarette was unchanged: 10.3 ng per mL for the low-yield
cigarette versus 10.2 ng per mL for the former usual cigarette. This represents
a compensation of 103 percent!

SMOKERS CAN GET
HIGH YIELDS FROM
THE LOWEST OF
THE LOW-YIELD
CIGARETTES:

MORE ON THE
ISSUE OF VENT
BLOCKING

Some points should be made about vent blocking and the
possibility of getting high yields from ultralow-yield brands.

In one study, 14 people were smoking ultralow-yield cigarettes
(Kozlowski et al., 1989), and half the smokers were vent
blockers. Two of the seven vent blockers smoked about
25 cigarettes per day and each blocker showed carbon
monoxide scores of 37 parts per million, which are very high.

Salivary cotinine levels of 303 and 385 ng per mL, from a
nominally .01-mg nicotine cigarette, are also very high. Therefore, there
were high exposures from a very-low-yield cigarette, clear evidence that some
smokers — if only two — were able to get substantial levels from the lowest of
the low-yield cigarettes.

Some submissions from the cigarette industry have indicated that vent
blocking is not a substantial problem. In contrast, four laboratories have
produced eight peer-reviewed studies that found evidence of vent blocking
(Hofer et al., 1991; Kozlowski et al., 1982a, 1988, 1989, and 1994; Lombardo
et al., 1983; Robinson et al., 1983; Zacny and Stitzer, 1988). In these studies,
the prevalence of "extreme" vent blocking ranged from 1 to 210 per 1,000
(median = 19 percent), and the prevalence of "at least some blocking" ranged
from 61 to 580 per 1,000 (median = 50 percent).

One submission from the cigarette industry notes that ventilation has
changed a great deal recently. However, invisible laser ventilation has been
available for at least a decade. From a consumer's point of view, it is unclear
why invisible ventilation techniques should be viewed as appropriate.

i

165

Smoking and Tobacco Control Monograph No. 7

Smokers can block the vents inadvertently if they do not know where the
vents are and what they do. If smokers know where the vents are located,
they can decide to avoid blocking the vents. There are real questions about
who is most advantaged by laser techniques and invisible perforations.

Marlboro Lights, Winston Lights, Camel Lights, and Newport Lights
("lights" in general) are ventilated-filter cigarettes. Much of the focus of
research has been on the ultralight cigarettes of 5 mg of tar or less. Unlike
the ultralights, these light cigarettes are best sellers, but like the ultralights,
they are ventilated-filter cigarettes. Therefore, the principle of informing the
consumer that these are ventilated cigarettes, discussing how the vents work,
and warning about blocking the vents with the fingers or lips is relevant to
lights as well as ultralights.

Anyone who is skeptical about vent blocking of ultralow-yield cigarettes
should take the lowest tar challenge: Light a 1-mg tar cigarette, placing your
lips on the filter as close to the smoker end as possible. Keep your fingers off
the filter so your fingers do not get in the way (i.e., do not block the vents
with your fingers) and take a puff. Consider its taste, temperature, and feel.
Now put your lips at least three-quarters of the way to the tobacco column
(i.e., block the vents with your lips) and take another puff of similar size.

(In our butt collection studies [Kozlowski et al., 1988 and 1994], we regularly
have found lipstick stains beyond the filter vents, on the filter end of the
cigarette, showing how far the cigarette had been put into the mouth.)
Compare the second puff to the first. See for yourself how easy it is to block
the vents and how much difference it makes to real tobacco pleasure by
doing this. Those onlookers who prefer not to take a puff of cigarette smoke
can usually see the difference in the smoke that is exhaled by someone else
because blocked vents produce a "juicy" mouthful of smoke that billows out
from a noninhaled puff of smoke. With unblocked vents, onlookers will see
only a little smoke exhaled.

GRAPHIC In 1982, a study was published on a color-matching technique to

rNFORMATION provide better information on tar and nicotine yields to smokers

ON TAR AND (Kozlowski et al., 1982b). The color-matching technique can be

NICOTINE YIELDS: used to estimate the number of puffs taken on a cigarette, and

THE COLOR- thus tar and nicotine yields, by comparing the color intensity of

MATCHING the end of a spent cigarette filter with a color scale. The study

TECHNIQUE demonstrated a strong relationship between the "darkness" of

color of the filter and the tar and nicotine yield of the cigarette. Figure 2
illustrates a modified version of the color-matching scale that the authors
incorporated on a cigarette package. Three different color papers (meant to
represent tar stains of low, standard, and high yields) developed by the
authors from the Fantone by Letraset Color-Matching System are used to
compare the filter stain colors from spent cigarettes. The low (Pantone
127U), standard (Pantone 117U), and high (Pantone 139U) colors are
mounted on the scale at points 2, 5, and 8, respectively. Smokers rated the
filter stain color on the O-to-lO scale, moving from the lower to the higher
intensity color blocks. Ihey decided "whether the filter looked lighter.

166

Chapter 12

Figure 2

The color-matching technique scale shown on a cigarette package. (Pantone-
colored papers representing low, standard, and high yields at scale locations 2, 5,
and 8, respectively.) The appearance of an unblocked vented filter is shown in the
bull's-eye stain; the uniform stain on the filter end indicates extreme vent blocking.

1 L

1

1 1

1

1 1

■

1

1^

1 1

0 1 2 3 4 5 6 7.

Low Medium i

B

Higt

9 10

1

To use scale, filter must look
like (^, not O

Not

Necessarily

Lights

darker, or about the same as each of the colored blocks, and then selected
the most appropriate scale number" (Kozlowski et al., 1982b).

Figure 2 also shows how stain patterns on spent filters can be used to
indicate whether vent blocking has taken place on a conventional ventilated
filter cigarette (Kozlowski et al., 1980). The bull's-eye tar stain on the left
indicates no vent blocking. (Diluting air rather than smoke has been drawn
through the periphery of the filter.) The uniform tar stain on the right
indicates extreme vent blocking. On ventilated-filter cigarettes, vent
blocking decreases filter efficiency so that the amount of stain left in the
filter underestimates the amount of smoke that has gone through the filter;

167

in other words, the color-matching technique requires the assumption that I i
vents remain unblocked. Because vent blocking alters yields dramatically, 1 1!
the graphic reminder not to block vents also may be useful in its own right. I !|

The color-matching technique is a tool that can be used in future studies
on compensation. Figure 2 also demonstrates how the color-matching lii

technique and the stain-pattern technique could be included on cigarette p;
packaging for consumer use. Color-matching information may better reflect I ji

the actual cigarette yields to smokers than the alternative FTC method. The | ji

FTC machine estimates of tar and nicotine yields can be unreliable given the || j

variability among smokers and the various methods of compensation.

Further developmental work is needed on this color-matching technique.

In the land of a largely blind FTC testing method, even a one-eyed color-
matching technique could be king (Kozlowski and Rickert, 1984). It is not
necessary to be perfect in providing the consumer with better information
about the tar and nicotine yields of cigarettes to improve on the current
standard method. A color scale attached to cigarettes can emphasize to the
consumer that the yields from a cigarette depend on how the cigarette is
smoked. Graphically, a color scale helps smokers see that yields are not
captured by any one tar or nicotine number, and thus smokers can get a
sense of where they stand in relation to the standard.

SUMMARY Our review of brand-switching studies indicated that smokers increase
nicotine intake from lower yield cigarettes by compensatory behavior,
including filter-vent blocking. This behavior is a neglected issue for smokers
of light and ultralight cigarettes. The current FTC testing method used to
estimate average tar and nicotine yields of cigarettes is compromised by
compensatory smoking behavior and individual variability among smokers. i
Graphic techniques (e.g., the color-matching technique and the stain-pattern
technique) also need to be explored as ways to provide estimates of tar and
nicotine yields to smokers of lower yield cigarettes. Simple graphic materials
may help these smokers realize that a low-yield cigarette can provide high
yields when smoked in certain ways. i

QUESTION-AND-ANSWER SESSION I

DR. BENOWITZ: Lynn, you said that in one study about half the people
were vent blockers. My work and the Gori study suggest that people are
taking in, on average, about .7 mg of nicotine per cigarette, which is 1

tremendously more than would be possible taking more puffs. So, I think '
virtually everyone who smokes ultralow-tar cigarettes must be blocking.

And how many of the holes do these ultralow-tar cigarette smokers block?

DR. KOZLOWSKI: 'Fhe story 1 like is the student of mine in class who said •'
his aunt, who smokes an ultralow-tar cigarette, keeps a roil of transparent fl

tape on her coffee table. When offering a cigarette to a friend she will say, ■

"Do you want that taped or untaped?" Bizarre as that might be; it happens. I

It illustrates that peo[)le do not understand what ventilation does to K

their cigarette. I had a call years ago from an angry executive as a result of l!

Chapter 12

some media exposure about the results of some of these studies. He said,

"I have a 1-mg tar cigarette, and yes, I block the vents on that cigarette, and
yes, it makes it taste better and it is easier to light, but 1 thought it was a
1-mg tar cigarette; it says so right on the pack."

Ventilation is not the only manufacturing technique that contributes to
an ultralow-yield cigarette. There can be other differences that mean that,
even with blocking, the smoker will not necessarily get the same really high
levels that you might with some other cigarettes. But it is clear that it is a
major factor; it is clear that smokers can subvert it completely or even
partially.

Lombardo did a study years ago with people staining their fingers with
printer's ink. And he found that, as the cigarette coal burns down, and your
fingers are getting away from it, they start to get in the way of the vent holes,
It is also interesting that those last few puffs are the richest, and if you were
to block those holes, that would be a particularly good time to do that to get
higher yields.

DR. TOWNSEND: Dr. Kozlowski, how did you measure the vent blockage?

DR. KOZLOWSKI: We have done it a few ways. Most of the time it is a stain
pattern method.

DR. TOWNSEND: On the mouth end of the filter?

DR. KOZLOWSKI: Yes.

DR. TOWNSEND: What I do not understand about something you just said
is that people will purposefully tape holes closed. I think my experience with
consumers is that they clearly know the tradeoffs between tar delivery of a
cigarette and taste characteristics.

It would really surprise me that consumers would make that purposeful
change to the design of a cigarette and not understand that they are
increasing tar. Besides, they have the choice to go out into the market and
buy a higher tar product if that is what they choose; so I do not understand
the rationale or the psychology here.

DR. KOZLOWSKI: I think it is something to be surprised about.

DR. TOWNSEND: About the compensation issue, there is another answer
that I do not completely understand.

Let's assume that compensation occurs to a very large degree, and people
get essentially the same deliveries from a low-tar cigarette that they get from
a higher tar cigarette. Then, why do consumers complain to us that the taste
of low-tar cigarettes is weaker, milder, less strong, and less acceptable?

Again, their perception is that tar and taste go together. As a smoker,

I can fairly accurately estimate the tar yield of a cigarette by smoking it, and
I can get within a couple of milligrams.

169

Smoking and Tobacco Control Monograph No. 7

I think many smokers, while they may not be as accurate in estimating
FTC tar yields, still can rank cigarettes by tar. Now, how could they possibly
do that if compensation were extensive?

DR. KOZLOWSKl: I think you do not want to think of compensation as
something that influences everybody's smoking behavior. What we found
in the Pharmacology, Biochemistry and Behavior study, half of the people block
vents quite a lot, and the other half did not block them at all.

And if you looked further at those who did not block, you found that
they did not smoke as many cigarettes per day. If you did taste ratings and
how they liked the taste of the cigarette, they seemed to be consumers who
were after a really low-yield smoke. They weren't blocking the holes. Not
everybody smoking a low-yield cigarette blocks the vents. But this gets back
to the issue of subject self-selection biases. We have to expect that there are
individual differences in how much nicotine a person might want and also
to the extent that a person is smoking for nicotine.

So, half of those subjects who were smoking ultralow-yield cigarettes
in the long term were not blocking vent holes; they did not smoke many
cigarettes per day; and they had low CO levels. The other half smoked
a lot more cigarettes a day, smoked earlier in the morning, and got higher
nicotine levels. You average them, and you get the kind of figures that are
commonly described as "intermediate." Some people were showing a lot
of compensation; some were showing very little; and that figure of mean
compensation can be misleading.

DR. DEBETHIZY: 1 think you have pointed out an important fact: No
machine-smoking method can predict individual behavior. This method
was never intended to predict individual behavior, and it does not. I think
people use different strategies when they smoke cigarettes, and it is rather
obvious in the data you presented today.

DR. COHEN: Is it your intuition that a great many people who compensate
are just following classic learning theory and do not even know they are
doing it?

DR. KOZLOWSKl: Some people are not aware they are doing it; that is clear.
They are not aware they are blocking the holes. I think that some people
find the cigarettes relatively difficult to light. You push it a bit further in
your mouth and it is a lot easier to light. Blocking could get started in a
number of ways.

DR. SHIFFMAN: We have all been struggling with the issue of variability
within a given product or products of equal FTC yield, I think, in talking
about compensation and in the difference between the machine yield and
the human biological exposure.

Now, with this issue of color matching, you are introducing something
that I think has to do with true exposure rather than FFC yield. 1 wonder
what you could tell us about the prospects of using a system like this to

170

Chapter 12

1

estimate exposure, the differences that might be due to things other than
number of puffs, and all the kinds of things we think a compensating smoker
might do.

DR. KOZLOWSKl: 1 think Dr. Rickert might be able to comment intelligently
on that.

DR. RICKERT: One of the things that we have done is to look at the yield of
a cigarette in relationship to the color of the filter itself. We have established
that there is an extremely good regression between the measured color
characteristic of the filter and the FTC yield. We have done that on the
smoking machine and for the actual filter on cigarettes. We have looked
at yields under 87 different conditions to cover a wide range of potential
behavioral conditions. And we have looked at the yields under those
conditions and have looked at the relationship between that and color.

And color of the filter is a very good predictor of yields under a wide variety
of conditions.

DR. HENNINGFIELD: 1 would just like to point out that the kinds of
compensation that you see are consistent with what Dr. Zacny was talking
about, the dynamic smoker, what 1 was talking about, the addicted and
behavior-modified smoker. But it is also very similar to what you see in the
animal laboratory, with addictive drugs like alcohol, sedatives, and opiates.

What you see is that as you push the dose up, you get some downward
compensation. As you decrease the dose, you get some upward. But it is
within a boundary. It is rarely perfect, because as you increase the dose, the
animals tend to get a little more drug. If you decrease the dose to a certain
point, the behavior can kind of just fall apart and get very erratic. It just
struck me how similar it was, what we see with animals and addictive drugs,
and what you are seeing.

DR. KOZLOWSKl: 1 agree.

DR. HENNINGFIELD: It looks like a basic biological phenomenon, in other
words.

REFERENCES

i Ashton, H., Stepney, R., Thompson, J.W. Self-titration
I by cigarette smokers. British Medical Journal 2: 357-
I 360, 1979.

I Benovvitz, N.L., Jacob, P., Jones, R.T., Rosenberg, J.

I Interindividual variability in the metabolism and
t cardiovascular effects of nicotine in man. Journal of
Pharmacology and Experimental Therapeutics 221:
368-372, 1982.

Benovvitz, N.L., Jacob, P., Kozlowski, L.T., Yu, L.
Influence of smoking fewer cigarettes on exposure
I to tar, nicotine, and carbon monoxide. New
I England Journal of Medicine 315: 1310-1313, 1986b.

j

Benovvitz, N.L., Jacob, P., Yu, L., Talcott, R., Hall, S.,
Jones, R.T. Reduced tar, nicotine, and carbon
monoxide exposure while smoking ultralow- but not
low-yield cigarettes. Journal of the American Medical
Association 256: 241-246, 1986a.

Guyatt, A.R., Kirkham, A.J.T., Mariner, D.C., Baldry,
A.G., Gumming, G. Long-term effects of switching
to cigarettes with lower tar and nicotine yields.
Psychopharmacology 99: 80-86, 1989.

Hofer, I., Nil, R., Battig, K. Ultralow-yield cigarettes and
type of ventilation: The role of ventilation blocking.
Pharmacology, Biochemistry and Behavior 40: 907-914,
1991.

i

171

Smoking and Tobacco Control Monograph No. 7

Kolonen, S., Tuomisto, J., Puustinen, P., Airaksinen,
M.M. Smoking behavior in low-yield cigarette
smokers and switchers in the natural environment.
Pharmacology, Biochemistry and Behavior 40: 177-180,
1991.

Kozlowski, L.T. Reduction of tobacco health hazards
in continuing users: Individual behavioral and
public health approaches. Journal of Substance Abuse
1: 345-357, 1989.

Kozlowski, L.T., Frecker, R.C., Khouw, V., Pope, M.

The misuse of "less-hazardous" cigarettes and its
detection: Hole-blocking of ventilated filters.
American Journal ofPiMic Health 70: 1202-1203,

1980.

Kozlowski, L.T., Heatherton, T.F., Frecker, R.C., Nolte,
H.E. Self-selected blocking of vents on low-yield
cigarettes. Pharmacology, Biochemistry and Behavior
33: 815-819, 1989.

Kozlowski, L.T., Herman, C.P. The interaction of
psychosocial and biological determinants of tobacco
use: More on the boundary model. Journal of Applied
Social Psychology 14: 244-256, 1984.

Kozlowski, L.T., Pillitteri, J.L., Sweeney, C.T. Misuse
of "light" cigarettes by means of vent blocking.
Journal of Substance Abuse 6: 333-336, 1994.

Kozlowski, L.T., Pope, M.A., Lux, J.E. Prevalence of the
misuse of ultra-low-tar cigarettes by blocking filter
vents. American Journal of Public Health 78: 694-695,
1988.

Kozlowski, L.T., Rickert, W.S. Kozlowski, Rickert reply.
American Journal of Public Health 74: 391, 1984.

Kozlowski, L.T., Rickert, W.S., Pope, M.A., Robinson,
J.C. A color-matching technique for monitoring tar/
nicotine yields to smokers. American Journal of Public
Health 72: 597-599, 1982b.

Kozlowski, L.T., Rickert, W.S., Pope, M.A., Robinson,
J.C., Frecker, R.C. Estimating the yield to smokers
of tar, nicotine, and carbon monoxide from the
"lowest yield" ventilated filter-cigarettes. British
Journal of Addiction 77(2): 159-165, 1982a.

Lombardo, T., Davis, C.J., Prue, D.M. When low tar
cigarettes yield high tar: Cigarette filter ventilation
hole blocking and its detection. Addictive Behaviors
8: 67-69, 1983.

Lynch, C.J., Benowitz, N.L. Spontaneous cigarette
brand switching: Consequences for nicotine and
carbon monoxide exposure. American Journal of
Public Health 78: 1191-1194, 1987.

Robinson, J.C., Young, J.C., Rickert, W.S., Fey, G.,
Kozlowski, L.T. A comparative study of the amount
of smoke absorbed from low yield ("less hazardous")
cigarettes. Part 2: Invasive measures. British Journal of
Addiction 78: 79-87, 1983.

Russell, M.A.H., Sutton, S.R., Iyer, R., Feyerabend, C.,
Vesey, C.J. Long-term switching to low-tar low-
nicotine cigarettes. British Journal of Addiction 77:
145-158, 1982.

Russell, M.A.H., Wilson, C., Patel, U.A., Feyerabend, C.,
Cole, P.V. Plasma nicotine levels after smoking
cigarettes with high, medium, and low nicotine
yields. British Medical Journal 2: 414-416, 1975.

West, R.J., Russell, M.A.H., Jarvis, M.J., Feyerabend, C.
Does switching to an ultra-low nicotine cigarette
induce nicotine withdrawal effects? Psycho-
phannacology 84: 120-123, 1984.

Wynder, E.L., Goodman, M.T., Hoffmann, D.
Demographic aspects of the low-yield cigarette:
Considerations in the evaluation of health risk.
Journal of the National Cartcer Institute 72: 817-822,
1984.

Zacny, J.P., Stitzer, M.L. Cigarette brand-switching:
Effects on smoke exposure and smoking behavior.
Journal of Pharmacology and Experimental Therapeutics
246(2): 619-627, 1988.

172

Chapter 13

Cigarette Design Technologies Reduce Smoke
Yield and Expand Consumer Ch oices: The
Role and Utility of the FTC Test AFethod

David E. Townsend

BACKGROUND The Federal Trade Commission (FTC) test method for measuring tar
and nicotine yields of cigarettes provides accurate and reliable information.
Comparison of yields of various brands is a key factor consumers use to make
objective choices in the marketplace. Another key factor is the taste of the
cigarette, which in most cases is related to the tar and nicotine yield.

Calls for reduced tar yields from cigarettes came from the popular press,
the scientific literature, and the public health community beginning in the
late 1950's. Many of these included statements that tar reduction would
reduce the relative risks for certain diseases.

The implementation of FTC testing for tar and nicotine in 1967 was an
important step for cigarette manufacturers to communicate information on
lower tar products to consumers for them to use to make informed decisions
in the marketplace.

Even at that time, FTC understood the limitations of standardized
machine smoking and recognized that no standard method would be able
to take into account the wide range in human smoking behavior:

No two human smokers smoke in the same way. No individual
smoker smokes in the same fashion. The speed at which one
smokes varies both among smokers, and usually also varies with
the same individual under different circumstances even within the
same day. Some take long puffs (or draws); some take short puffs.

That variation affects the "tar" and nicotine quantity in the smoke
generated (Federal Trade Commission, 1967).

The FTC also recognized that the FTC method could not predict the
absolute smoke yield any individual smoker might receive from a particular
cigarette:

No test can precisely duplicate conditions of actual human
smoking and, within fairly wide limits, no one method can be
said to be either "right" or "wrong." The Commission considers
it most important that the test results be based on a reasonable
standardized method and that they be capable of being presented
to the public in a manner that is readily understandable (Federal
Trade Commission, 1967).

173

Smoking and Tobacco Control Monograph No. 7

Daniel Oliver, chairman of FTC, confirmed FTC's position on cigarette
testing in a statement before a congressional committee:

As a general matter, I believe that advertisements that accurately
convey information on "tar" and nicotine content can be a
valuable source of information to consumers. Advertising that
provides comparative information on different "tar" and nicotine
levels can be especially useful (Oliver, 1988).

CIGARETTE The cigarette industry response to the public demand for

DESIGN AND reduced tar and nicotine cigarettes is evident in the dramatic

CILANGES IN THE decline in sales-weighted average tar yields over the past 40 years

CIGARETTE (Figure 1). In the early 1950's the average tar yield of cigarettes

MARKET was around 38 mg per cigarette. Today that average is about

12 mg per cigarette. Nicotine yields also have been reduced in a similar
fashion, although to a slightly different degree because the available
techniques reduce tar and nicotine yields with slightly different efficiencies.

The techniques to reduce tar over the years include filtration, more
efficient filtration (through different filter materials, fiber type and density,
and filter length), filter ventilation, expanded tobacco, tobacco weight
reduction, increased paper porosity or permeability, reconstituted tobacco,
faster burning cigarette papers, and reduction of cigarette circumference.

Figure 1

Sales-weighted average tar and nicotine yields, 1954-1993

3.5

3.0

2.5

2.0

1.5
1.0
0.5
0

174

Nicotine Yield (mg/cigarette)

Chapter 13

The physics and chemistry of a burning cigarette are exceedingly
complex. For example, filter ventilation involves two mechanisms. In the
first, fresh air from the outside is admitted to the filter and mixes with the
smoke. As a result of this ventilation, a smaller effective puff is drawn on
the burning end of the cigarette and less tobacco is then consumed during
the puff. In addition, the smoke velocity in the cigarette is dramatically
reduced, and the filter efficiency upstream of the ventilation holes increases.
Similarly, in the second, a higher paper porosity also allows more outside air
to enter the smokestream, also reducing the effective puff volume at the fire
cone. The various cigarette design parameters result in many interactive
effects on the performance of the cigarette.

The changes in cigarette design to reduce tar and nicotine yields have
not been limited to low-tar and ultralow-tar products. Even today's
nonfiltered cigarettes, the so-called high-tar brands, have about half the
tar yield of their 1950's counterparts.

As a result, consumers today have a much wider range of choices in tar
and nicotine than they did previously, and all cigarettes are substantially
lower in tar yields than they were in past years (Figure 2). Cigarette design
changes have resulted in an overall major reduction in smoke yields.

Figure 2

R.J. Reynolds Tobacco Company offers smokers a range of tar levels (1955-1993,
in 5-year intervals)

1955 1960 1965 1970 1975 1980 1985 1990 1993

Year

175

Smoking and Tobacco Control Monograph No. 7

UTILITY OF Although it has been clear that humans do not smoke like machines,

THE FTC TEST it is also clear that changes in the FTC smoking conditions do not

METHOD alter the relative ranking of cigarettes. The FTC conditions include

a 35-cc puff of 2 seconds duration, taken once per minute. If the puffing
conditions are changed, the relative ranking or yields of the cigarettes are
preserved.

For example. Figure 3 shows tar yield as a function of puff volume.

In this chapter, puff volumes of 35 cc, 45 cc, and 55 cc were chosen for
comparison of four cigarette products, one each from the lowest tar,
ultralight, full-flavor light, full-flavor categories. The 35-cc puff is the
FTC condition and is not intended to represent the lowest smoker puff
volume; the other two conditions were arbitrarily chosen. As the puff
volume increases, tar yield of the product in each category increases.
However, the ranking of the categories is preserved. For example, the
tar yield of an ultralow-tar product at a 55-cc puff volume is lower than
the tar yield of a low-tar product at the same puff volume.

Changing the puff frequency from one puff per minute to one puff every
45 seconds or one puff every 30 seconds also increases the tar yield in each
category, yet the ranking of the cigarettes is intact (Figure 4). Puff duration
has little, if any, effect on the actual yields (Figure 5).

Figure 3

Effect of puff volume on observed tar yields

a>

V

a>

w

(0

O)

o

O)

E

(0

Brand Style

Note The puff duration for this experiment was 2 seconds; the puff frequency was one every 60 seconds.

176

Chapter 13

j Figure 4

I Effect of puff frequency on observed tar yields

35

30

25

20

15

10

5

0

Puff frequencies studied
(one puff taken every “x” seconds):
11 60 seconds

□ 45 seconds

□ 30 seconds

Lowest Tar Ultralight Full Flavor Light Full Flavor

Brand Style

Note: The puff volume for this experiment was 45 cc; the puff duration was 1. 7 seconds.

Figure 5

Effect of puff duration on observed tar yields

177

Smoking and Tobacco Control Monograph No. 7

Figures 6 through 8 show smoke nicotine yields for the same set of
cigarettes; however, the smoking machine puffing conditions are different
for each figure. Like tar yield, the nicotine yield goes up with increased puff
volume and increased puff frequency, and the relative yields among the
categories remain ordered. Puff duration also has little if any effect on
nicotine yields.

Standard methods are used to provide information to consumers for
products other than cigarettes. A classic example is the estimated
Environmental Protection Agency (EPA) gas mileage ratings for vehicles.
Depending on driving habits, conditions, maintenance, and fuel type, a
vehicle may get more or less mileage than indicated by the EPA estimate.
Although few drivers will achieve the actual mileage listed for a vehicle, the
mileage ratings do provide a means of relative comparison among vehicles.

A potential buyer can use the information to determine if a particular vehicle
would fit into his or her particular transportation and economic needs.

Similarly, smokers have two primary considerations in making their
choices in the cigarette marketplace. The FTC method provides comparative
smoke yield information that is an essential part of that process. The second
factor of taste is an individual preference that is made with the comparative
information in mind.

Figure 6

Effect of puff volume on observed nicotine yields

Lowest Tar Ultralight Full Flavor Light Full Flavor

Brand Style

Note: The puff duration for this experiment was 2 seconds; the puff frequency was one every 60 seconds.

178

j Chapter 13

Figure 7

Effect of puff frequency on observed nicotine yields

Puff frequencies studied
(one puff taken every V seconds):

0>

i-

(D

O)

15

TO

E

2.0

1.5

1.0

0.5

0

H 60 seconds

□ 45 seconds

□ 30 seconds

Lowest Tar Ultralight Full Flavor Light Full Flavor

Brand Style

Note: The puff volume for this experiment was 45 cc; the puff duration was 1.7 seconds.

I

! Figure 8

I Effect of puff duration on observed nicotine yields

2.5

!

o

i—

(0

TO

lo

TO

I E

2.0

1.5

1.0

0.5

0

Puff durations studied:
0 1 .4 seconds

□ 1 .7 seconds

□ 2.0 seconds

Lowest Tar Ultralight Full Flavor Light Full Flavor

Brand Style

I

Note: The puff volume for this experiment was 45 cc; the puff frequency was one every 30 seconds.

Smoking and Tobacco Control Monograph No. 7

QUESTION-AND-ANSWER SESSION

DR. BENOWITZ: How do you reconcile the differences between your data
and other data that show when you get down to very low yields of nicotine,
you are getting to a cotinine level of 225 vs. maybe 325 or so at the higher
levels, in over 2,000 people. Do you think that is less accurate than your
data in 33 subjects?

DR. DEBETHIZY: 1 think that the method of measuring plasma cotinine is
a less accurate measure of nicotine uptake.

DR. BENOWITZ: What bias do you think there is in cotinine that could
explain the tremendous difference in findings?

DR. DEBETHIZY: 1 do believe the data that we have. And 1 am surprised at
them, based on what 1 know about the cotinine data in the field studies.

Now, that does not mean that the data we have are wrong; it may be that
at lower yielding products, that what people do over the course of an entire
24-hour period is different from what would be measured, say, at either
9:00 o'clock in the morning or 3:00 o'clock in the afternoon, depending
on where people are measuring those plasma samples. So, 1 would say that,
yes, the number is much larger for those plasma cotinine studies. But the
point is that, even with those studies, compensation is incomplete. People
smoking lower yielding cigarettes absorb less nicotine. Now, 1 would
conclude the same thing from both data sets.

DR. BENOWITZ: 1 want to go back to your statement when you said the
FTC method is accurately reflecting intake, because 1 think that is patently
wrong. It is not whether there is some reduction; it is whether you can look
at those cotinine levels, which would indicate that, when you get below 1 mg
nicotine, that the FTC method is underestimating consistently based on
cotinine levels in a couple of thousand people.

What ! am arguing about is, is it reasonable to generalize from your
33 subjects and say that is more valid than the 2,000?

Cotinine levels, as you know, vary throughout the day, but not more
than 10 or 15 percent if you are smoking regularly. So, there is no way that
10 or 15 percent can explain the difference, even if there were the worst bias
that you can imagine.

DR. DEBE'I’HIZY: 1 think that we can generalize from the data. I do not
think we can give those data the weight that 10 years of analysis has
provided us with. But I think that if we look at the [)lasma cotinine data,
peo[)le smoking lower yielding cigarettes absorb less nicotine. I'hose data
clearly show that.

1 hey do not show that peo[)le get the same amount of material from all
the wide range of nicotine-yielding cigarettes; would you agree with that?

180

Chapter 13

DR. BENOWITZ: There is a slope such that there is not 100 percent
compensation, although it may be close in some studies. But 1 would agree
that there is a slope.

DR. DEBETHIZY: And 1 think that what we have done is taken things to the
next step, what Dr. Henningfield called for earlier in his talk, which is to
apply a technique that is used for other materials.

If you were asking what the amount absorbed of a pharmaceutical
product in a 24-hour period would be, you would measure the total amount
excreted and sum it up. That is all we have done, and I think that the data
deserve consideration. I think that additional work will determine whether
that slope will stay as steep as it is now.

I think you will notice that there were people above and below that, so
there is wide variation. But the point is that people smoking lower yielding
products are absorbing less. How much less, I do not know, and I suspect,
even with our own data, we see some evidence that at the lowest yields, they
are absorbing more than FTC would predict.

DR. RICKERT: First of all, in looking at your data on the FTC yield and the
nicotine, one is impressed by the fact that it looks like there are basically two
points on that regression. One is at the very low .1 mg and the other one is
up at 1.4.

It seems to me that, for the bulk of the data, there is no relationship, that
it is really a two-pronged display, with one at the bottom and one at the top
and in the middle.

DR. DEBETHIZY: And I think that what you are looking at is the fact that the
two center groupings are very close in nicotine and tar yield, but of course
there is a full range of products out there on the market, and that is what we
wanted to address with that.

DR. RICKERT: In reading the industry documents, it has been stated time
and time again that consumers understand the FTC tar numbers. And my
reading of the literature and what I have heard today suggest that is not so.
And I was wondering whether there is industry information that supports the
hypothesis that consumers, indeed, understand FTC numbers of tar and
nicotine?

DR. TOWNSEND: How could the industry have changed so dramatically
over the years and people traded taste, if tar levels were not a consideration
in their choice?

People tell us, in focus groups and in other ways, "Yes, I am concerned
about what I believe are health risks in smoking." They have been told that
for 40 years, and they respond by looking at the tar levels of the products
that they choose in the marketplace.

181

Smoking and Tobacco Control Monograph No. 7

DR. RICKERT: I guess the question I am asking is, How do you reconcile
that point of view with the information that has been provided here today,
which suggests that the majority of smokers do not understand FTC
numbers?

DR. TOWNSEND: I am convinced that they understand tar ratings as a
relative comparison, the same as 1 think we understand ERA gas mileage.
When you go buy a new car, ERA gas mileage estimates are determined
by a number of factors, including the type of engine, how you drive, how
properly you inflate the tires, how well you maintain it, how good your
mechanic is, and even maybe the region of the country in which you live,
because gasoline engines are more efficient in certain climates than others.

1 do not take ERA gas mileage ratings to mean that is what 1 am going
to get. The same as 1 was referring to earlier, 1 do not think I am necessarily
getting the same efficiency on my hot water heater as it is rated.

1 think these ratings are for comparison purposes. And I think that is
what most consumers look for in the marketplace, and 1 think that is the
way they interpret the FTC numbers.

DR. RICKERT: Let's assume for a moment that the FTC numbers are a perfect
predictor of the amount of nicotine that is obtained from a cigarette. How
do you think that relates to the other components that may be of concern?
There are a whole host of chemicals that are related to various disease
phenomena. How do you think the blood nicotine levels will act as a
predictor of the absorption of these other constituents?

DR. DEBETHIZY: 1 think that one of the things about nicotine as a marker
is that nicotine represents probably the upper limit for a particular phase
for constituents. But one thing that 1 want to remind everybody is that the
FTC method was never intended to measure or assist people with the actual
uptake. It is to provide a relative ranking of cigarettes. And 1 think it has
reliably done that for a long time.

It also gives us a way that we can standardize the analysis of cigarette
smoking, so that we can compare work done in our laboratory with work
done in your laboratory. And it has done that very well for a long time.

DR. HARRIS: 1 wondered if you could show, once again, one of your slides
showing the trends in the sales-weighted tar and nicotine averages over time.
And also, you might put up the one of unfiltered Camels also, if you have it.
It is not essential which one, and I think Dr. Hoffmann even had one.

Dr. Hoffmann showed a similar slide, although the axes were labeled
somewhat differently. And 1 have also looked at data on sales-weighted
average nicotine, using the ITC numbers, at least those that were provided
at various times to the Federal Frade Commission.

And I have been led to the general conclusion that, while the F'FC-based
numbers declined substantially during the I9.S()'s and continued to decline.

182

Chapter 13

to some degree, during the 1960's, in the last 10 years, there has been
basically no change in the sales-weighted average, whether you measure it
by the distribution of tars with a percentage of brands under 15 mg or by
the sales-weighted average nicotine.

And in fact, data for 1992, and what 1 can estimate myself from 1993
and 1994, are a slight upturn in the sales-weighted average nicotine.

If you show your slide on Camel nonfilters, the graph ends in 1982 at a
tar level of 20.6. It would be interesting to have the data for after 1982.

DR. TOWNSEND: The tar delivery for Camel nonfilter is virtually flat from
1982 to the present. This is a chart that I prepared for something else, not
this.

DR. HARRIS: And my question is really. What, if any, observations you
might want to make about what appears to be progress in the decline in
FTC yields during the last 30 years, a progress of which is really confined
to the preceding two decades, with no change in the last decade?

If this had been a meeting on mileage in the car industry, somebody
would be waving a finger and saying, "What have you guys done in the
last 10 years with your car model?"

DR. TOWNSEND: My response to your question, first, is that I believe this
clearly points out to me the need for the industry to respond with low-tar
products that have improved taste characteristics. Clearly that is what
consumers have told me: "I trade in tar for what I perceive as a possible
benefit, and what I get is less taste." The concept of low tar, great taste,
does not wash with consumers.

There are taste deficiencies in the lowest end. Some people choose to
make that trade because, again, they weigh both factors in the marketplace:
taste and tar levels.

But I think both factors are important in their choice, because there are
many, many smokers who do buy products in the lowest category; again,
back to the need for some useful and valid comparative information on tar
levels, and that is already in the market. The FTC test method provides
valid and reliable information.

First, beginning in 1981 or 1982, the price of cigarettes in the United
States began to rise much faster than the rate of inflation. And at the same
time, we saw a dramatic increase in generic and branded discount cigarettes.

To some degree, the apparent stagnation in tar and nicotine levels
may reflect smokers choosing to go from branded to discount and generic
cigarettes. In fact, from what data I have seen, the major source of brand
switching in the last 10 years has been to the discount and generic segment
of the market, which, as you know, is about 35 percent of the market in
1993.

183

Smoking and Tobacco Control Monograph No. 7

Another possibility that has been raised by some is whether some brands
have had their nicotine levels and tar levels actually reconstituted upward
slightly.

A third is the changing demographics of the market. As some people
quit and other people start, the average smoker is a different person who
would intend, on average, to smoke a higher tar or higher nicotine cigarette.

Finally, there is the question of whether or not, in fact, there is a limit
that smokers are willing to tolerate, given the current cigarette array of
choices.

How does your hypothesis relate to the question at hand: What is the
meaningfulness of the FTC test, and is it useM in the market?

DR. HARRIS: There are several points you make. One is whether the test is
useful, but also whether the industry has accomplished anything in reducing
tar and nicotine levels. And I think that, since that was a preamble of both
talks and it is certainly an issue that I have been puzzling over, I thought it
was important to wrestle with the question of why have tar and nicotine
levels not fallen in the last 10 years, and is there anything that can be done
about it.

DR. SHIFFMAN: You have very much emphasized the issue of consumer
choice and consumers making a choice based on accurate information.

I take it, then, that if consumers could be provided with better, more accurate
information about yields, that is something that you would favor.

DR. TOWNSEND: I believe that what the consumer needs is there. The FTC
method provides reliable comparative information.

DR. SHIFFMAN: You do not want them to have better information?

DR. TOWNSEND: Convince me that there is better information. I am not
convinced that there is.

DR. SHIFFMAN: I am asking it as a hypothetical question.

DR. TOWNSEND: If there is important information that the consumer needs
to make choices in the marketplace, then I want to know it.

DR. SHIFFMAN: We have seen individual variability around the trend line,
and if we were able to provide individuals with information about where
they stood on that, then would that be an improvement that you might be
able to support?

DR. DEBETHIZY: That is a tall order for any sort of standardized method.

DR. RE'n rri: Your talk. Dr. deBethizy, referenced the historical context of
the development of the F'FC measures. And the historical context was the
claim that these measures would significantly reduce the risk of disease.

Do you think that the data so far support a claim that these ITC measures
predict or are meaningfully related to disease risk?

184

Chapter 13

DR. DEBETHIZY: Personally, I do not know. But what 1 would say is that
the data that I saw Dr. Samet present today, which were largely taken from
the 1981 Surgeon General's report, their conclusion was that people smoking
lower yielding products have reduced relative risk for lung cancer.

DR. PETITTI: I think that you saw the quotes from Dr. Wynder and some
of the earlier commentators. They mentioned 40-percent reductions in tar
might lead to large reductions in lung cancer. I just want to get a sense of
the magnitude in the reduction of disease risk with differences in tar levels
over the range we are talking about.

I just want to make the point that adjustment is the problem; when you
adjust, you assume people smoke the same number of cigarettes, whether
they smoke high yield or low yield. And it is very difficult to handle
statistically and is, I think, one of the problems in the original data that
were published in the 1981 report.

DR. DEBETHIZY: But you know, on average, that people smoking lower
yielding products do not smoke more cigarettes.

DR. BOCK: I am having a little bit of a problem. You had mentioned some
observations with staining of saliva regarding ventilation and hole blockage.
It seems to be the opposite from what was reported. Can you give me some
details of how you know you got saliva on the area covered by the lips in
every case, or most cases?

DR. TOWNSEND: What I can do is give you detailed information on the
whole experiment. And I will have to do that privately, because I don't have
the information with me today.

What I said earlier is the case. We saw infrequent hole blockage, but
there was hole blockage in some cases, and we determined that by an
inhydrin staining process. And I cannot recall the details and the numbers,
because I really was not responsible for that experiment. I would be happy
to follow up with you on that, if you are interested.

DR. COHEN: Let me quickly state what I think the premises are of your
presentation. Dr. Townsend, and see if you disagree with where I disagree.

Suppose we accept the premise that the FTC system provides useful
ranking information, everyone understanding the difference between
ranking and other kinds of information. Let's say the system does that.
Suppose we also accept the premise that truly individual smoking
characteristics are beyond the scope of such a rating system.

Now, you have established that there are product design features — type
of paper, type of tobacco, etc., that lead to different yields because of
smoking parameters that vary with such product design elements, such as
puff rate, puff volume, etc.

185

Smoking and Tobacco Control Monograph No. 7

Doesn't this mean that a numerical scale — say, from 1 to 27 — is
necessarily misleading and that a categorical rating system would be a more
valid way to report such information? You had four categories, 1 believe, and
you showed variance. What I am saying is, I was following that, and it
looked to me like you were about to recommend a four-category system.

DR. TOWNSEND: The whole idea of a categorical system, or so-called banded
system, has been put on the table by a number of people in the past.

Conceivably, that accomplishes the same endpoint with one exception,

1 believe. The same endpoint, of course, is that it provides a comparative
ranking for consumers. The flaw in that approach, if that is the only
ranking, and discrete numbers are not also included, is that, of course,
you would expect products to come up against the ceiling of each band.

DR. COHEN: I thought you established, with your own analysis, that
individual numbers were inherently misleading?

DR. TOWNSEND: I did not say that individual numbers were misleading.

In fact, 1 believe individual numbers — a numerical rating system — is, in fact,
the best and that is what the FTC test method is.

With those data, then, manufacturers have advertised their products as
light or ultralight, to fit some range of tar numbers.

DR. COHEN: I think you missed my point.

DR. TOWNSEND: The FTC method is the method that provides the useful
comparative information for the smoker.

DR. COHEN: 1 thought you established with your charts that there was
variance due to what is technically an interaction between product design
characteristics and smokers' adaptations to them. Is that correct?

DR. TOWNSEND: Okay, you are confusing me. Let me put the chart back
up. Standard deviation for replicate measures of that particular product.
Different particular puff frequencies. Puff frequency of 60 seconds happens
to be blue bars, 45 seconds is the red bars, 30 seconds is the yellow bars.

So, the ultralight product that was smoked at 30-second frequency — in fact,
this is the variability we saw in 10 replicate measures of that one cigarette.

DR. COHEN: Okay, then 1 did misunderstand that, but a lot of the
presentations today have essentially suggested that smokers respond to
product design characteristics by modifying behavior. And 1 am not talking
about idiosyncratic behavior, but standard ways that you put up — puff
frequencies, puff duration, number of puffs.

Ihat creates variance around a point estimate. And would it not be
more valid to acknowledge that those variances exist when you provide this
information to consumers? Ranking is the least informative scale.

DR. rOWNSEND: You are talking about something like the EPA gas mileage
ranking, where you have highway and city.

186

Chapter 13

DR. COHEN: Ranking information has only an ordinal property. The
absolute numbers have no significance, nor do the units of measurements
between — like 16, 17, 18, 19. In a ranking system, you never assume those
units are equal.

In this system, presented to consumers, consumers have a right to
assume equal appearing units — 16, 17, 18, 19. It goes beyond a ranking
system. If you just want a ranking system, then there are ways to do that,
to build on this kind of variance.

So, if all you want is a ranking system, the one that is in place now
attempts to do more than that, and I thought your evidence indicated that
it did not do it with great validity.

DR. TOWNSEND: What variance are you speaking of?

DR. COHEN: The interaction of human smoking topography and cigarettes
having different design features.

DR. TOWNSEND: Have you quantified that?

DR. COHEN: 1 have not done it. 1 think we have discussed it today. I am
not a technical expert on it. 1 am asking from the standpoint of consumer
usefulness.

If a ranking system could be preserved to meet your objectives that
you set out and, at the same time, it would have more validity because it
wouldn't represent units that do not exist because there is too much variance
around them; it is only a ranking system.

DR. DEBETHIZY: It sounds like you have put a proposal on the table, and
my impression is that is what we will do tomorrow.

DR. COHEN: It sounded like you were saying, "Well, if it achieves its
purpose as a ranking system and leaves consumers to know which brands
are lower and which are higher, and that is fine."

DR. DEBETHIZY: And we think that the current method does that. So, I will
be looking forward to the discussion tomorrow about alternative methods.

DR. HENNINGFIELD: This is actually a nice introduction to my point.
Compensation is one of the reasons that you get a good correlation with
machines and a lousy correlation with humans. It is not validated that the
FTC method predicts what humans get.

But you have seemed concerned that compensation was not perfect,
as though the fact that it was not imperfect rejected the notion of
compensation.

DR. DEBETHIZY: 1 think what 1 was doing was challenging the notion that
people get the same amount from every cigarette on the market.

DR. HENNINGFIELD: Do you know of any drug study with humans with
addictive drugs where you do get perfect compensation?

187

Smoking and Tobacco Control Monograph No. 7

DR. DEBETHIZY: I just do not even see a reason to address that question
because smoking is a habit that people engage in, and people enjoy smoking.
1 think we have a situation where people enjoy smoking; they have a wide
range of products to choose from; they can choose lower yielding products.
The data — whether it is plasma cotinine or total urinary uptake — show
that they get, when they smoke lower yielding cigarettes, less material,
on average.

Now, the FTC method was never intended to address that question.

It was intended to address the relative ranking of cigarettes.

DR. SHIFFMAN: Actually, what your data show is that within the broad
range in which sales are actually concentrated — let's say from .4 to about 1.0,
1.2, FTC nicotine — people are getting, on average, the same, regardless of the
nominal yield. So, in fact, your own data, as Dr. Rickert pointed out earlier,
suggest that, in fact, the numbers do not track.

DR. SHIFFMAN: What your data show is that people who smoke cigarettes
of .1 are very much lower than people smoking 1.4. But if you look at the
middle range, we could probably compute the correlation by tomorrow, but
1 would warrant that it is close to zero.

DR. DEBETHIZY: And the middle range is a very narrow range of tar and
nicotine yield.

DR. HENNINGFIELD: Let me just bring closure to my point, because what
I see — and I think Dr. Shiftman is pointing out, too — is that what you see
with humans with other addictive drugs, and what you see in animals, is
compensation that is partial. And that seems to be what we are seeing here:
another case of an addictive drug that controls behavior, and you do get
compensation, although it is not perfect. That seems to be the biological
process going on.

DR. TOWNSEND: And I think you and I fundamentally disagree about
that, because it is like coffee drinking. I do not see myself as being, if I drink
1 cup of coffee 1 day and 10 cups of coffee the next day, I do not see that
as compensation. I just see that as responding to different situations and
choosing to drink coffee under those different environmental conditions.
And I see smoking as a very similar activity that people engage in.

DR. KOZLOWSKI: I his is a supplement to Fred Bock's question and point.
Years ago, in Foronto, we tried to do a saliva test, 1 think similar to the one
you did. We collaborated with the forensic laboratory for the Province of
Ontario, and a Ph.D. student in pharmacology worked on it.

And we never ()ublished it because we did not find it useful, in part
because the lii)s, in general, were too dry and were not depositing that
much saliva and were doing it in a very soggy manner.

We found that a much more straightforward technique would be to take
cha()stick or li[) gloss, and that would stain the filter overwrap. You could

188

Chapter 1 3

then dissect the cigarette and count the holes that have been blocked by
that. We tried that technique and abandoned it as unsuccessful.

Just months ago, we published in Pharmacology, Biochemistry and
Behavior a validation of the stain pattern technique using Marlboro Lights,
Winston Lights, Marlboro Ultralights — I am not sure about the four or five
other brands. They worked just fine.

Even with ventilated filter cigarettes, which may be ventilated in the
20- to 30-percent range rather than 80 or 90, you can get a team of raters
to do reliable judgment.

DR. DEBETHIZY: I think what is important is that some individuals may
block holes when they smoke cigarettes. What is important is what happens
over the course of their entire smoking day and what happens, on average,
with people. And I think there has been an incredible emphasis placed on
blocking holes here today, and I think that Dr. Townsend's talk clearly
showed that there are many, many other techniques used to lower tar and
nicotine, techniques that could not be overcome by the smoking behavior
of the individual.

DR. FREEMAN: What I am going to do is say to cut this at this point and
maybe we can continue it tomorrow. We were due to be finished about
20 minutes ago. It may be a little hard on these two gentleman, who have
been very gracious in answering these questions. We do not want to put
them under too much, but we would like to bring you back tomorrow, if
you do not mind, for further discussion. But let's have these two questions.

DR. BENOWITZ: There has been some suggestion about providing the
consumer with more information, such as making ventilation holes visible
so that people can see them and not block them, and give them information
about what intense smoking would do, which you could simulate by
machines. Are there any negative aspects about doing those things, from
your perspective, and why not provide more information like that?

DR. DEBETHIZY: I think Dr. Townsend addressed it earlier when he
responded that we would consider any reasonable proposal as long as there
were some data to support that proposal.

And I am assuming that is what you all will do, and at least start the
process. And you all may conclude that the FTC method is fine as it is, but
I certainly do not have any problem with looking at reasonable proposals.

We have a motto at our company that we work for smokers. And if we
can be convinced that it is meaningful for the consumer, that might warrant
consideration.

DR. TOWNSEND: But you understand that we believe, today, that the FTC
test method is useful for the consumer; that its presence has been beneficial
for consumers making choices and also for driving the industry to reduce tar
and nicotine levels to this great extent that we have.

189

Smoking and Tobacco Control Monograph No. 7

Then the other point, too, ISO — the International Standards
Organization — of course, adopted a standard test method that is used
throughout the world, that is essentially the same. And they, in the
investigations, particularly in Germany and the UK, had many of the
same questions here.

DR. BENOWITZ: But there is no drawback for providing more information.

DR. TOWNSEND: Oh, absolutely. I think we always have to look at new
information that is available, and look at it in depth, critique it, and see if
anything is there.

DR. HOFFMANN: I am somewhat puzzled. I have seen the tobacco field for
decades, and I always found the tobacco industry to be flexible. The pressure
from the consumer and from the scientific community led to the industry
changing the cigarette. And you have in Science the outstanding study by
Dr. Benowitz.

We all agree that there is some compensation, and I find that you are
inflexible. We have to work on it; there is a way. I am surprised. I have,
in all the X years I have been working in this field, I have never seen such
an inflexible thing as this, where you stick to the FTC method.

I think we can always improve, and U.S. scientists know this just as
well as I.

DR. DEBETHIZY: Let me ask this question. Have you seen a proposal put
on the table that seriously would improve on the current method?

DR. HOFFMANN: The advances in the research have been done by both
sides — by industry as well as the scientific community. Suddenly, I find that
you say, "No, the FTC method is the final word."

DR. DEBETHIZY: 1 think that clearly what we have said is that, with 30 years
in a standardized method that has been incredibly valuable to consumers,
the industry, and the scientists that we have not seen any reason to walk
away from that. I would just give it back to the group and say, let's see a
serious proposal.

DR. TOWNSEND: 1 am sorry that you are misreading this as being inflexible.
1 think the fact is that the F'l'C method has worked for a long time, and it
continues to work.

But demonstrate what the proposal is, and how it adds to what we have
now, to make things better, or to provide some more information. For
example, one of the proposals that was floated this morning included a
min and max level. Okay, let's smoke a cigarette at FTC conditions, and
then let's go to a more intensive smoking condition, to report min and max.

What is going to happen from that is that you get exactly the same
ranking; it is just more numbers, but it is the same ranking. And consumers
are making their choices based on ranking.

190

Chapter 13

DR. HOFFMANN: But you have shown here, in your paper, it is very detailed.
It is for the low-yielding figure and compensation.

DR. DEBETHIZY: Yes, and you are right about that.

DR. HOFFMANN: So, this is not a reason that we should work together?

DR. DEBETHIZY: And we are quite willing to work with anybody, and that is
why we are here today.

DR. FREEMAN: Dr. Townsend, I just want to ask one question, and it may be
a little naive. Several times you have mentioned the value of tar in cigarettes
because you say it is associated with taste. Even conceding that although it
seems to be a thing that is killing people.

But what about nicotine? What it is the value of nicotine and cigarettes,
and why could it not be dramatically reduced?

DR. TOWNSEND: Nicotine, of course, is part of the smoking sensation. It
does provide a sensation to the smoker. I think one of our competitors found
that tobacco that had been treated to remove all the nicotine was not
successful in the marketplace.

More than that, just as I cannot look into the components of tar and say,
this is a very important, tasteful, and flavorful compound. You know, I am
not equipped as a chemist to say nicotine is an important compound for this
aspect of taste characteristics.

DR. FREEMAN: We will stop at this point.

REFERENCES

Federal Trade Commission. "FTC To Begin Cigarette Oliver, D. Prepared statement by the Federal Trade

Testing." News release. August 1, 1967, p. 2. Commission chairman before the U.S. House

Subcommittee on Transportation, Tourism, and
Hazardous Materials, Committee on Energy and
Commerce. May 4, 1988, p. 11.

191

Section II

Transcript of Second- Day D

iscussion

DR. FREEMAN: Good morning, everyone. I would like to welcome all of the
speakers and panel members and the audience back to these deliberations.
We will continue to deliberate concerning the FTC test method, and we will
begin this morning with continuation of the dialog that we were having
with Dr. Townsend and Dr. deBethizy, who represent the tobacco industry.

DR. BENOWITZ: 1 would like to follow up with two questions on why the
results from the study of 33 subjects relating the total nicotine recovery vs.
the FTC yield from cotinine studies are so different. The first question is,
since these results were so different, it would be very interesting to have
measured cotinine levels in these smokers, as well as looking at urinary
metabolites, and I would like to know if that was done and if we could
see those data. The second thing 1 was wondering was, since this question
of yield versus intake has been so important for so many years and since
R.J. Reynolds has the capability of doing it, I wonder if they have ever done
a study like the ones that 1 showed where they looked at cotinine levels
vs. yields in a large population just to see if their own work would replicate
the work of other people, and it seems like a very straightforward study
that would be something they might have done.

DR. DEBETHIZY: The answer to your first question is no, we did not measure
plasma cotinine in those studies. We were studying nicotine metabolism
interindividual variation. That was how we got into that work, and we
extended it then to ask the question across the tar categories. In the study
that we are currently doing, we are actually measuring salivary cotinine.

We made a conscious decision not to measure plasma cotinine because we
did not want to interfere by taking a blood sample. So, we are doing salivary
cotinine in that study to answer the exact question that you have raised.

I think that is a good question to ask. In subjects where we see lower total
nicotine output, are the plasma concentrations higher? That is a good
question. And your second question was?

DR. BENOWITZ: There were data from Dr. Gori's work that I presented that
were supported by Brown and Williamson, I believe. We basically looked at
cotinine levels vs. yields for a large population, and I think those data were
very important. I was wondering if R.J. Reynolds has ever done such
a study, and if any data are available addressing that question?

DR. DEBETHIZY: We have not done a field study. There were so many field
studies in the literature already, we just have never done a study like that.

DR. SHIFFMAN: Just to follow up on that study, a couple of us were pointing
out that the relationship seemed to be very much driven by the extremes,
and I took the liberty of computing what the correlation would be in those
same data if one excluded the very extremes. I had to impute the data from

193

Smoking and Tobacco Control Monograph No. 7

the graph, but if you look at the data above .13 and at or below 1.02 FTC
yield, the correlation is .16. In other words, except for the extremes in your
own data, there is no relationship.

DR. DEBETHIZY: Yes, it is interesting to me. 1 think what we have done
is taken the best available technology, done what Jack Henningfield asked
people to do, which is to take a look at things using modern techniques, and
we have done a study that 1 think causes us to stop and think about what
the previous data have shown. We can manipulate those data to get them
to look like what the other data look like, or we can take them on their own
merits, and 1 think that what we need to do is follow this study up with
further work, and that is why we are doing that, and 1 have encouraged
Dr. Benowitz to do the same thing. 1 have encouraged the Swedish Tobacco
Company, which has the capability to do the same thing.

1 think rather than doing some data selection on this particular study
we should take it on its own merit. It is a 33-person study. It suggests that
when smokers can freely do their activities, people consuming lower yielding
cigarettes absorb less nicotine. Now, it also suggests that there is large
interindividual variation, and I think lots of people have pointed that out.

When you do a study like this, you are going to get extremes because
people smoke cigarettes across a wide range, and I think you have to include
those people, and I think as we and others fill the data in over time we will
find out whether this correlation or the slope of this line is as steep as it is
now or whether it is shallower, and I just think we need to continue to do
that work. We have worked hard to develop a state-of-the-art technique,
and I think it has merit.

DR. SHIFFMAN: I think it has merit, too, and I applaud you for doing it.

At the same time we ought to be clear on what the data show, and the data
show that people smoking brands above 1.03 are getting more than people
smoking brands at about .13 and that in the middle range there is no
relationship to the FTC yield.

DR. FREEMAN: Dr. Benowitz, does that answer both of your questions?

DR. BENOWITZ: Yes.

DR. FREEMAN: Dr. Rickert?

DR. RICKERT: In the media recently, the cigarette Eclipse was described, and
it is obvious from the media description that the cigarette is going to pose
some challenges for the FTC methodology, in particular because it does not
burn down to a fixed butt length. J'here are some other challenges that may
be posed by that cigarette to the ITC methodology, and in particular I am
wondering about the distribution of nicotine between the gas phase and the
particulate phase. That is, in the current I'l'C methodology when testing the
Eclipse cigarette, will the amount of nicotine that is being delivered by that
cigarette be trapped using the traditional Cambridge Filter method?

194

Section II

DR. TOWNSEND: I do not think we really know the answer right now to
the question of nicotine distribution between the gas and particulate phase.

1 do think we have confidence that the FTC method can provide useful data
for Eclipse. Certainly the FTC method will have to be accommodated for
that product in much the same way that the FTC considered accommodating
the method for Barclay and the use of different holders. That proposal
was certainly up for discussion. The FTC method as it stands with some
modification, particularly for the fact that Eclipse does not burn down,
can provide useful data for that.

DR. FREEMAN: Dr. Henningfield?

DR. HENNINGFIELD: In dealing with the smokeless tobacco issue, which
has no labeling, and 1 think that is a problem, it struck me yesterday that I
am not sure what is worse, having no labeling or having labeling that might
be misleading to consumers about relative risks. In trying to deal with the
relative risk issue yesterday, you spent a lot of time talking about your
technologies that address health concerns and implying that there was some
health benefit, and I would like to know what your estimate is as to the
number of cancer deaths, for example, caused by standard cigarettes and
how many lives, if any, would be saved if people were using cigarettes with
these advanced technologies of filtration and so forth that you were talking
about yesterday? In other words, how many people die of cancer in your
estimation from the higher yield cigarettes, and how many fewer, if any,
would die from the lower yield cigarettes?

DR. TOWNSEND: I think what you did was completely mischaracterize what
I said yesterday. What 1 said was that the Surgeon General and the public
health community called for the reduction in tar and nicotine yields from
cigarettes, and I said that the industry and R.J. Reynolds responded to that
consumer demand through major design changes to the product, and we
successfully reduced the tar and nicotine yields from the very high 30's,
as a sales-weighted average, down to currently about 12 mg.

DR. FREEMAN: To follow up on the question, do you think that fewer
people die based on the changes that you have made?

DR. TOWNSEND: 1 am not an epidemiologist. The Surgeon General in 1981,
in his report, did say that reduced-tar products pose a reduced health risk.

DR. HENNINGFIELD: I would like to follow this though because the words
"light" and things like that are only used with foods when there is a health
benefit. Your industry is using those terms relating them to FTC yields, and
I would like to know what your estimate of the health benefit is. To know
that, we have to know what your estimate of the death rates are with the
different products.

DR. TOWNSEND: I just said that I was not an epidemiologist. I happen to
be a chemist. I do know what the Surgeon General and epidemiologists have
said. Many smokers have heard the same thing.

I

195

Smoking and Tobacco Control Monograph No. 7

DR. FREEMAN: Dr. Hughes?

DR. HUGHES: I would like to ask your opinion about ordinal vs. cardinal
scales because I think that was not made clear yesterday. You compared
the FTC method to the ERA gas mileage. If I buy a car that has 38 miles per
gallon and my sister buys a car with 19, 1 get twice what she gets. Now,
even your own data show that is not true with tar and nicotine yields.

When you have a tar yield that is twice another cigarette, you do not get
twice the tar. So, I find those numbers misleading. I think the normal
consumer when they see a cigarette that says, "1 milligram tar," and they
see another cigarette that says, "8," they think they are getting one-eighth
the tar, and that is not true. It seems to me if that is the case, and all you
want is rankings, that we should do away with the numbers because they
are misleading, and I would like to hear your thoughts about that.

DR. TOWNSEND: A relative ranking of cigarette yields is what is essential
in the marketplace. To date we believe that the FTC method provides useful
information for the consumer. Do you really believe that your car gets
19 miles per gallon when you drive it?

DR. HUGHES: I believe that my car that gets 19 gets half the mileage of
somebody else's car that gets 38, and I think most consumers would believe
that if they saw the numbers 19 and 38.

DR. DEBETHIZY: But what else is on that label? The other part that is on
there is, "Your actual mileage may vary," and that is important because again
this particular method was not set up to predict what an individual will get.

It was set up for relative ranking, and I think it is really important to stick
with that.

DR. HUGHES: I agree, and with relative ranking, when you have rankings
and ordinal categories, you do not have numerals attached.

DR. DEBETHIZY: You just gave a good example of that when you gave ERA
gas mileage.

DR. HUGHES: ERA gas mileage is a cardinal system. It is not a relative
ranking. The ERA gas mileage, 38, cars that have 38 miles per gallon do,
in fact, get twice the mileage as cars of 19.

DR. DEBE3 HIZY: Only if driven under standard conditions.

DR. HUGHES: No. You are confusing variability around the mean with
ordinal vs. ranking.

DR. DEBE'FHIZY: I think I understand the difference, and I think that we
could argue about this all day, but I think that the I'FC method was intended
as a machine-based standardized method to provide relative ranking, no
more than that.

DR. FREEMAN: Dr. Henningfield?

Section II

DR. HENNINGFIELD: Then if we play by your rules, what is wrong with
putting on the cigarettes, "Your intake may vary on a cigarette that is so-
called an 'ultralow'" and put right on the cigarette, "You may get up to
3 mg of nicotine and 80 mg of tar from this, depending on how you smoke
it"? What would be wrong with that? Wouldn't that just provide honest
information to consumers so that they would know? Maybe even giving
them a little bit of information that you folks know and we know about
what pushes it up there, such as smoking harder and things like that; what
is wrong with that?

DR. DEBETHIZY: As I said yesterday, we are quite willing to consider any
reasonable proposal, and 1 suspect, Mr. Chairman, we are going to move into
that mode eventually where we will discuss those proposals, and that is a
proposal to put on the table and discuss.

DR. HENNINGFIELD: I am not sure that is an issue right now, but so, you
would not object to that concept?

DR. DEBETHIZY: I would not object to putting that proposal on the table
because my understanding of what this panel is supposed to do is to make
recommendations like that for serious study and consideration.

DR. FREEMAN: Dr. Cohen?

DR. COHEN: Let me just say that a lot of the discussion proceeds from
what experts know about the FTC method and what it was designed to
do. I think the real question is what consumers think the numbers mean.

So, the important issue is how do consumers understand these numbers,
and I think it would come as a shock to them that these are only to be taken
as rankings. That would come as a great shock, and I think we should keep
that in mind.

DR. TOWNSEND: That is a point. Dr. Cohen, where we clearly disagree.

DR. COHEN: Do you have any data that show that consumers only think
about these numbers as rankings?

DR. TOWNSEND: It is clear to us that consumers look at tar information;
they also look at the category of cigarettes they smoke, whether it is a light
or an ultralight or regular, and they make decisions in the marketplace. The
actual fact is that in the market, sales-weighted tar and nicotine yields have
declined dramatically over the years, and people have traded taste to do that.

DR. COHEN: Let me say, in response, that that is perfectly consistent with
consumers believing that these are real numbers, not rankings. Your scenario
fits a situation in which consumers think that by going down to a very-low-
yield cigarette that these are cardinal numbers and real numbers. I am asking
you whether your company or any cigarette company has data that indicate
that consumers only think about these as rankings. The answer is either yes
or no.

I

I

I

I

197

Smoking and Tobacco Control Monograph No. 7

DR. TOWNSEND: It is clear to us from talking with consumers that they
understand the notion of tar and nicotine yield, that they make choices in
the marketplace. I am not about to talk about our consumer information at
this point. I am not a marketing expert, but it is clear to us that consumers
use the information from the FTC test method in one form or another, and
even in your words, some use the numbers, and yes, those people who
actually use the quantitative numbers may be more skewed to the ultralight
category, but consumers do use the numbers or they use the category
rankings of cigarettes, whether it be ultralight or regular.

DR. FREEMAN: Dr. Petitti?

DR. PETITTI: I am intrigued by the EPA mileage analogy and also by the
water heater analogy, but I think the difference is one about consumer
information. When you drive your 38-miles-per-gallon EPA-rated car, you
know as you drive, based on measurements that you can make, whether or
not you are, indeed, getting 38 miles per gallon or 19 miles per gallon based
on your mileage, the way you actually use that car. Similarly for your water
heater, you get your bill every month, and you can tell whether or not you
are exceeding or not exceeding the conditions that are printed on your label.
How does the cigarette consumer know whether or not they are or are not
getting what is on the package?

DR. TOWNSEND: I think the point of the analogy is not exactly where you
are coming from. The point of the analogy is that consumers do not expect
to get exactly that EPA gas mileage. I know I do not because yes, you are
right, I can measure it, and I do not, but I do use the EPA gas mileage
numbers to a degree in making choices in the marketplace in helping guide
my purchases. I said yesterday that I recently bought a new hot water heater.

I used the energy efficiency rating in helping me make that choice, and I
actually paid more for a more efficient hot water heater, but that tag also
said that the average price or the price you would expect to pay for running
this hot water heater is $358 per year. Do I believe that is what it is costing
me? No. So, it is a matter of providing me guidance for making choices
and in no way do I believe that represents an absolute number that predicts
my power bilk

DR. PETITTI: I think you already answered this question, but I do think that
both for the EPA mileage example and your water heater, and perhaps also
for cigarettes, that it would be useful to the consumer to know the specific
range that they might expect under certain specified driving conditions and
perhaps we are going to get to that in terms of the proposals that we
consider. |

DR. TOWNSEND: And I believe that is a question on the table because the }
ITC test method was not intended to do exactly that.

DR. PETHTI: Can you explain to me just once again your view of what the
FI’C method was meant to do in the context of health? You keep saying that
it was not meant to do that. 1 am having a hard time understanding your ■
view of what it was meant to do. f l

i'

198

DR. TOWNSEND: Consumers responded to the calls for reduced-health-
risk products for reduced-tar products that were made by the Surgeon
General, Wynder and Hoffmann, and other members of the public health
community. Consumers responded to that information, and they demanded
of the industry a reduction in tar and nicotine yields from cigarettes. A
standardized comparative, accurate, and reliable test method was required
to accomplish that, and that was the purpose of the FTC method, to provide
those comparative data.

DR. PETITTI: So they could make decisions about health?

DR. TOWNSEND: So they could make decisions about tar yields in the
marketplace, which they were told by the public health community were
related to health.

DR. KOZLOWSKl: 1 think there are great concerns about how many
consumers are getting the information about tar and nicotine yields as they
exist now. 1 think at the last testing, the Federal Trade Commission reported
tar and nicotine yields on upward of 900 cigarettes. According to the rules,
the tobacco industry is not required to print tar and nicotine yields on
cigarette packs. They are only in ads. What, in fact, is the percentage of
cigarettes that are not advertised at all so that there is, in fact, no way for
the consumer to know? 1 think that may, indeed, vary from manufacturer
to manufacturer, but also in a related point some data have shown that
it is on the ultralow tars that people are most likely to know the yields.

It is also the fact that it is on the ultralow tars that the yields themselves
are likely to be printed on the packs. Does the FTC know what percentage
of brands are unadvertised and therefore consumers have no access to
information on yields?

MR. PEELER: We do not have those data, but we can get them for the panel
if they like.

DR. FREEMAN: Dr. Shiftman?

DR. SHIFFMAN: I take it in a sense that we have a significant degree of
consensus. I have heard the gentlemen from R.J. Reynolds say that they
would be sympathetic to proposals that would provide more information for
the consumer to make informed choices, and so, accordingly, I suggest that
we shift from a mode of asking questions of them to a mode of considering
proposals that would accomplish that goal on which we seem to have some
consensus.

DR. DEBETHIZY: Mr. Chairman, it might help if Mr. Peeler would clarify
what the purpose of the FTC method is. I think there has been some
confusion here. I know I have been asked 10 times what the purpose is, and
if you do not mind doing that, I think if you could do that concisely, that
would help.

MR. PEELER: I would go back to the statement that I started with yesterday,
and that is to say that the purpose of the FTC rankings when they were put

199

Smoking and Tobacco Control Monograph No. 7

together was to establish a comparative basis for consumers choosing among
cigarettes. The reason the Commission asked the National Cancer Institute
to commence this review of the cigarette testing methodology was to review
whether that approach is still the correct approach, and obviously we are
very interested as an agency in the types of questions that Dr. Cohen's
research raises, which is how consumers actually use and view these data.

DR. FREEMAN: Mr. Peeler, is it implied in what you said that the FTC was
ultimately interested in what was happening to the American public in terms
of health? Is that implied in what you say, or was it separate from that or
were you as an agency concerned about what is happening to the American
people?

MR. PEELER: If you look at the history of the establishment of the current
tar and nicotine testing system, it was clearly driven by concerns about
health. It was clearly driven by the Surgeon General's findings that were
valid at that time, that lower tar, lower nicotine cigarettes had a health
benefit for consumers. So, clearly one of the issues that the Commission
asked this panel to address is whether those health considerations are still
valid in light of research that has occurred since the 1981 Surgeon General's
report reviewed those issues and reported them.

DR. FREEMAN: So, then we would conclude that there is a clear connection
in the work of this committee, not only to measure appropriately what
cigarettes contain, no matter what method is used, but the end result that
we are looking for is how can we help people in America with respect to
avoiding disease and death, which means we would have to communicate
appropriately to them in order to accomplish that. Is that a fair statement?

MR. PEELER: Most of the data that I have looked at indicate that there is a
large group of consumers who are concerned about tar and nicotine ratings
because of health reasons. So, clearly if the tar and nicotine ratings are
communicating that to consumers, the FTC would want to make sure that
these numbers are accurately delivering that benefit to consumers.

DR. TOWNSEND: Mr. Chairman, 1 think I am confused at this point because
what 1 just heard is different from what 1 heard in your opening statement,
and so 1 pulled out the copy of your opening statement, which says, "The
primary purpose of this meeting is not to redesign the FTC testing protocol
but rather to examine the protocol and make suggestions for improvements,
if warranted." Your opening statement does not really go to actual changes.

DR. FREF^MAN: You failed to go far enough in the opening statement. We
posed three questions, the third of which dealt with what 1 was just speaking
of, in other words, how does this translate to the American public in terms of
their perceptions in the opening statement.

DR. TOWNSEND: Ihank you.

DR. DEBE'FHIZY: I have a question for Mr. Peeler on the relationship
between the FTC and the Surgeon General's warnings. Could you clarify

200

Section II

that relationship for me? I was just curious about that because the questions
that have been asked of me and that we have been getting into are the health
implications of FTC numbers, and I was just curious about whether the FTC
method is there to clarify the Surgeon General's warning, or is it related at
all, or are they just two completely separate issues?

MR. PEELER: The FTC method as it was conceived and implemented was
designed to provide consumers with comparative information about the
relative tar and nicotine content of cigarettes. We know from the studies
that we have seen that some groups of consumers look at those numbers
as indicating a health benefit, which is why the Commission has asked
the panel to look at the question of whether there is, for example, a dose-
response relationship between the FTC tar and nicotine ratings and specific
smoking-related diseases.

DR. SHIFFMAN: Let me come back to what I consider to be items of
substantial consensus and maybe that will help us move on. I have not
heard anyone speak against providing the consumer with more information,
and so it seems to me that the appropriate education of the American
smoking consumer is something we can all agree on, and that it is part of the
intention of the FTC system, to give the consumer appropriate information.

It seems to me that an important aspect is providing appropriate education
to the consumer about the meaning of whatever information is conveyed in
this labeling.

The second item on which I think we have considerable consensus is
that in human smoking of particular cigarettes there is a considerable range
or variability in what the consumer will actually extract from the cigarette.
That was seen not only in some of the talks from past studies but also in
the R.J. Reynolds study. So it seems to me the second item of substantial
consensus is that no single number can completely represent the true
human yield from a cigarette. Therefore, it seems to me that the direction
in which we should be trying to move is to represent to the consumer the
sense of that range in variability and to accompany that with appropriate
educational measures so that we are providing the consumer with the kind
of information on which to make informed choices, and I think that is the
basis on which we ought to go forward.

DR. FREEMAN: Dr. Cohen?

DR. COHEN: Dr. Shiffman, could I ask if you would be willing to modify
your view just a bit? I think that to start exactly at that point is not the right
place to start because I think consumers want to know two things. They
want to know if 1 smoke at all, how risky is it, and does that level of risk vary
with the kind of cigarette I smoke. They want to know that. Now, that may
be impossible to provide. That is not my field, but they do want it. We
cannot finesse that issue.

DR. SHIFFMAN: I quite agree with you, and I think that part of what might
go into an educational campaign would be about the meaning of these
numbers or ranges in relation to health outcomes.

201

Smoking and Tobacco Control Monograph No. 7

DR. COHEN: But it begs the question: Is there a way to provide two types
of information, either in advertising or on the cigarette packages or both?

The first type is the level of risk for that particular kind of cigarette or the
level of harmfulness, and the second one is, can we give them a better sense
of the relative magnitudes? I do not know the answer to either question,
but I think that if we are going to wonder what we can do to help consumers,
I think we should think long and hard about the need for those two different
pieces of information.

DR. FREEMAN: At this point we are going to shift gears a little bit and get
into the essence of the deliberations, and we thank the members of the
tobacco industry for receiving those questions. We will go to the next phase
of this discussion, which is the main phase and that is, as you remember
from yesterday, we posed three questions that we were supposed to answer
during these deliberations, and we are going to look at each of those three
questions and get your comments on each one.

Question 1. Does the evidence presented clearly demonstrate that changes
are needed in the current FTC protocol for measuring nicotine,
tar, and carbon monoxide, and if so, what changes are required?

DR. GIOVINO: A lot has been made in this conference of the trends over
time in the FTC yield in terms of tar and nicotine with a very large decline
between the 1950's and 1980, roughly, and then a leveling off, and from the
data that have been presented at this conference, I have to wonder, especially
given Dr. Guerin's comments, what would that curve look like if consumer
changes in puff frequency, puff volume, hole blocking, and vent blocking
were incorporated? Dr. Guerin and Dr. Zacny have shown that the yields can
be changed, given various factors, and I see that trend as a measure of yes,
a standardized measure, but one that may not be as relevant now as it was
40 years ago.

So, I have to ask the panel to consider in its deliberations the issue of the
usefulness of those trend data, given as was demonstrated yesterday the wide
range of products now available and the different degrees of compensation
that can happen with those products.

DR. TOWNSEND: May I respond to that? I believe that the trends that
you saw yesterday in the chart are useful today as they always have been.

One thing that I think there is consensus on within this panel is that if
you change puffing conditions, what you do is shift the tar and nicotine
yields up or down depending on to what level you change those puffing
conditions. Even if you block the vents, you shift the tar and nicotine yields
up, but in general the relative ranking does not change. If the relative
ranking does not change, you are only changing the absolute values. Then
that is going to have no substantive effect on the trend charts that showed
nicotine and tar yield decreases over the years.

DR. GIOVINO: 1 have to wonder, given the situation 40 years ago when tar
and nicotine levels were so high, if those behaviors would have been so

202

Section II

,} common. The relative rankings may be accurate for any given year, but the

I range was so different 40 years ago than it is now that the actual amount of

I some of the compensatory behaviors may have been much less frequent

j 40 years ago than now.

The range in tar values and nicotine values was so much higher in the
I 1950's and the 1960's given the FTC yield that people may not have had to

perform the compensatory behaviors and wouldn't have even had the ability
I to hole block because it is my understanding that there were no holes then.

; So, my concern is that those trend data, while representing what the FTC

I has presented, are not representing even what the consumer is taking out

' of the cigarette, let alone getting into their lungs.

I

* DR. GUERIN: 1 am not sure that was ever the case anyway. In trend data,

what you are looking at are the characteristics of the average cigarette, not
how the cigarette was used, and that is all those data mean.

I DR. GIOVINO: Exactly. 1 think the panel understands this, that the trend

I data represent what the FTC method gives. The reality is that trends over

! time in terms of what the consumer is taking out are quite different.

' DR. BOCK: At the very earliest time that yield data were collected, the

standard deviations were given, which had big meaning for the analysts
but obviously did not have much meaning for people out in the street.

But the variability of smoking, which is part of the fact that people in the
street really need to know, the range of values for each cigarette, has not
been provided by the data. The labeling might have incorporated that
type of information, which would in large part, 1 think, answer some of
the criticisms.

DR. FREEMAN: Before you go on, let me follow up on that. Are you
suggesting then that might be a change in this Question 1 concern?

I DR. BOCK: It would indicate that maybe there should be a change in the

j protocol and the way the data are collected, and there should be provision

j made for a range.

I DR. FREEMAN: Measuring the same elements but giving the range.

i DR. BOCK: With different smoking parameters.

i

j DR. FREEMAN: That is a point of discussion. Dr. Woosley?

j DR. WOOSLEY: 1 think in answering that first one I have to agree with

Dr. Giovino that things have changed over the years, and I think that is
what the FTC is actually asking us. There was this huge range of difference
30 years ago or so, and the ability for this method to predict something was
great then, but now that most of the tobacco products have come down to
I some very homogeneous group, the variance is quite tight, and the ability

I of this numerical ranking to have any meaningful information or carry any

I meaningful information to the public is gone, in my estimation. Data

i yesterday were very convincing for me that numerical ranking does not

i

203

Smoking and Tobacco Control Monograph No. 7

really convey the exposure that occurs because of compensation. So, to me
the answer to the first question is pretty clear. The current system must be
changed in some way.

DR. TOWNSEND: 1 do not understand your comment that cigarettes today
are a more homogeneous group. From the data 1 showed yesterday the
spread in tar deliveries in 1954 was really quite narrow. Cigarettes were
really quite similar then. Today there is a huge range of products available
to the consumer. 1 see that as less homogeneous.

DR. WOOSLEY: 1 was referring to the potential range of intake, not the range
that the tobacco industry provided us.

DR. RICKERT: One of the things that people are concerned about is the fact
that consumers tend to misinterpret the information. One of the ways of
coming to grips with this problem is to deal with a range of potential values
rather than specific numbers. This problem was first noted, I think, back
in the 1981 Surgeon General's report when at that time there was a call for
publishing maximal values in addition to the values that are obtained under
FTC methodology. A more recent paper in 1994 has called for the same
approach, and 1 think serious consideration should be given to this question
of range, how one might express these upper limits, and if maximum were
to be used, how that maximum would be determined.

DR. GUERIN: If one examines Question 1 that we are addressing, the
question says, "Is there any evidence that changes are needed in the current
ITC protocol for measuring tar and nicotine and CO?" I have not necessarily
seen much evidence for changes in measuring it, but a lot of reasons for
changes in how we communicate it. Do we have to change the testing
protocol to achieve this, or do we have to have a better way of
communicating?

DR. WOOSLEY: It says, "Constituent yields," and 1 think the yield from that
method is probably inadequate. We need data on the yield to the smoker.

DR. ZACNY: 1 just want to go back to something that Dr. Townsend said
about 5 minutes ago and that we spent some time on yesterday when he
showed charts where you increase puff volume from, 1 guess 35 to 55 mL,
and the relative rankings would not change if puff volume were increased
across the different yields. 1 think things change when you talk about filter
vent blocking and maybe altering parameters for extensive filter vent
blocking because there is a fundamental difference between lower yield
cigarettes and high-yield cigarettes.

I he high-yield cigarettes do not have filter vents, and so you could,
by manipulating this parameter, turn a low-yield cigarette into a high-yield
cigarette; the relative rankings then would not be preserved.

DR. FREEMAN: Dr. Rickert?

204

Section II

DR. RICKERT: I think there is one issue that we have not really looked at,
and that is, there is something else that happens when you move from
standardized FTC testing conditions to other testing conditions. We always
consider what happens to the quantity of particulates, like tar, for example;
what should be also considered is what happens to the quality of that tar.

For example, in the Brown and Williamson documents that I received, it
seemed that moving from standard conditions to behaviorally defined
conditions resulted in an increase in mutagenicity of the tar fraction on a
gram-per-gram basis using the salmonella assay, and so I think focusing
totally on the changes in the relative ranking misses the point that the
biological activity on a per-gram basis may be changing as well.

DR. FREEMAN: Dr. Henningfield?

DR. HENNINGFIELD: I would like to follow up on a point that Dr. Woosley
made. I agree. My impression is that what people want to know is what gets
into people, not into machines. There is only one way to do that, and that is
to do what you do with any other drug: Test what gets in people, and that is
the only way that you can validate the upper range. There is no way you can
do that with a machine. You have to put the system in people to see what
they actually get.

Also, 1 think you need to do that because as we have seen, with almost
any system you come up with, the industry might come up with a creative
way to beat that machine. Testing in humans is essential. The question for
Dr. deBethizy is, when we were discussing providing a wide range of values,
and there seems to be some leaning that that would be a useful thing to do,
you seemed to agree that was worth considering, but that the FTC method
was not up to that task. 1 think that is what you said. If that is what you
said, why is the FTC method not up to the task?

DR. DEBETHIZY: I did not say that. What I said was that it seemed like a
reasonable proposal to put on the table. 1 did not say anything about not
providing the FTC number. 1 think the FTC number is a good number. It
has been a standardized number we have used for a long time to provide
relative ranking. If somebody is proposing that a range also be determined,
with a low and a high end, then let us discuss that.

DR. SHIFFMAN: Let us, indeed, discuss that. It seems to me that the FTC
assay method may be adaptable to this goal in the sense that if one looked
at a different set of parameters, if one included the potential vent blocking
as part of a protocol for maximum extraction or maximum yield, perhaps it
would be possible to use the machine testing method to adequately describe
or estimate the range of human exposures from a particular cigarette, and
that could well be more informative to smokers.

DR. FREEMAN: Before we get other comments, that seems to be a recurrent
point of discussion about the range being an important point to consider as
a possible suggestion, and 1 would like to zero in on that particular point and
discuss it. Is there anyone who wants to discuss that point?

I

205

Smoking and Tobacco Control Monograph No. 7

DR. KOZLOWSKI: A number of the studies in the literature argue that
the rankings would be preserved if you had a heavy smoke setting on the
machine. 1 think if you consider Zacny's data and other data, the idea of
tuning every cigarette up to the same maximum puff volume or maximum
puff rate is probably not a good model of human smoking behavior; the
higher yield cigarettes may, in fact, be undersmoked relative to the lower
yield cigarettes. Zacny's data on the puff volume show clearly that the puff
volumes are bigger on the ultralights than on the higher yield cigarettes, so
that when you have studies that just tune everything up, and you see the
ranking preserved, the fact of the matter is that if the human behavior is
more appropriately modeled you may well see that some of the higher yield
brands go down, the lower yield brands go up, and it would get a lot flatter
than simply jacking up all the settings of the machine regardless of what the
strength of that cigarette is to begin with.

DR. FREEMAN: 1 think perhaps the question and my own opinion as 1 am
posing it, given a particular cigarette, would be what is the range of possible
exposure of that cigarette compared to any other cigarette? The question
I would like some consideration of is, is that a reasonable thing to measure;
is that a reasonable approach to take as to what we should measure?

Dr. Benowitz?

DR. BENOWITZ: 1 think that is a reasonable approach. I think
Dr. Kozlowski's point of view is very well taken, but 1 think in practical
terms it really is impossible because you would have to study large numbers
of people smoking every single brand of cigarettes to be able to get individual
parameters. We know what he says occurs. It seems to me that the idea of
having the standard condition and an intensive smoking condition with and
without hole blocking would be very useful, but what it has to be coupled
with is information for consumers about how their smoking of the cigarette
will influence the yields. For example, if they block, this is what is going to
happen, and if they puff intensively or take a lot of puffs, then this is what
is going to happen, and I think that would be the best we could do to say,

"If you smoke in this way, you are going to get the maximum yields." And
that way we could pick what we think would be an intensive condition and
say, "If you smoke in this way, this is what your yield is going to be."

DR. FREEMAN: With a given cigarette?

DR. BENOWITZ: With a given cigarette, because I do not think it is feasible,
although I would like to, either to measure puffing parameters for every
brand of cigarette or even, as Dr. Henningfield suggested, to do human
exposure studies if you have 900 brands of cigarettes. I do not think that
would be practical. I think we should test maybe some brands of cigarettes
to see how well we are doing, but I do not think it is going to be feasible for
all these brands to do anything other than a standardized testing.

DR. FREEMAN: Putting that forward as a point of discussion, does anyone
disagree with what Dr. Benowitz has said, that we should perhaps

206

Section II

recommend that a given cigarette should be tested to see what its range of
possible exposure would be using whatever techniques make sense; is there
any disagreement with putting that forward? Dr. Hoffmann?

DR. HOFFMANN: 1 think that Dr. Benowitz' old studies concentrated on
nicotine, but 1 think when we test on humans and try to see how they
smoke, we should not limit it to nicotine. There are other carcinogenic
toxic agents, and 1 think that the work done by Dr. Benowitz on nicotine
is outstanding, but it is nicotine, and when we deal with cancer, at least,
that is my area of expertise, there are agents that are just as important.

DR. FREEMAN: 1 did not understand that you were speaking only of
nicotine. Were you not speaking of tar and the three things that are
mentioned?

DR. BENOWITZ: Yes, in terms of the machine testing I was certainly
speaking of all. In terms of human bioavailability testing, 1 think as
Dr. Hoffmann says, if we have tools to measure tar exposure, we definitely
should do that. Right now the only practical tools for large-scale studies
are nicotine and CO, but when we get tar measurement tools where we can
do it on hundreds of people, that should be included.

DR. TOWNSEND: If 1 could respond, my reaction to your proposal is that
if we provide to the consumer an FTC number and a maximum deliverable
number by hole blocking and a more intense puffing regimen, which 1
believe is your proposal, then from what I know about cigarette design those
two are going to very closely parallel each other, and the ranking of cigarettes
will be largely preserved. My question then is, does that provide additional
and useful information to the consumer?

DR. BENOWITZ: I would like to respond to that. 1 do not believe that
the ranking will be preserved. If you have the old-style cigarettes that are
nonfilter cigarettes, no matter what you do, the ranking is going to be
preserved, but if you are comparing a nonfilter cigarette and then a cigarette
that has extensive ventilation and you block holes, you might see one
surpass the other. I just do not believe that when you are dealing with a
cigarette that has 90 percent ventilation in a standard test and people have
the possibility of reducing that to zero percent ventilation, and you are
comparing that to cigarettes with no ventilating filters, the ranking will
be preserved.

DR. TOWNSEND: As a cigarette designer, 1 believe that it will be largely
preserved, and I guess what 1 am hearing you say is that this is your
suspicion, but 1 guess my question, is do you have data that support that,
and 1 guess the obvious direction 1 am going in with this question is, should
we collect data to see whether the ranking is largely preserved to convince
you?

DR. FREEMAN: I think we have a question over here.

207

Smoking and Tobacco Control Monograph No. 7

DR. HATSUKAMI: You mentioned changing the intensity of smoking.

I would like to know how you determine those parameters; how do you
determine the number of puffs that should be taken, the range of puffs that
should be taken, or the volume that should be taken; what should that be
based on?

DR. BENOWITZ: 1 have not looked at the current studies to see what is
available, but 1 think you could do that on the basis of looking at
observations, say in people who are smoking low-yield cigarettes and seeing
what they do.

DR. HATSUKAMI: With the minimum and the maximum ranges in terms of
number of puffs?

DR. BENOWITZ: 1 do not think a minimum is really necessary. I actually
believe that we should continue to report the standard FTC method mostly
because 1 would like to know how current cigarettes compare to cigarettes
marketed 20 years ago so we could have that as sort of a minimum because
in fact, you know, it is my belief, based on the evidence, that for the vast
majority of cigarettes the FTC method underestimates exposure. So, we
could still have that as a minimum exposure, and then we could have the
test method to show what a smoker might get if they smoke in an intense
way, which could then be specified.

DR. FREEMAN: Dr. Bock?

DR. BOCK: It does not make a lot of difference to me whether the
ranking is changed or not. When you put down an average and a standard
deviation, sophisticated people can understand whether the differences in
the average are important, and a range will give that kind of information
to unsophisticated people, and that is where I think the big advantage of
a range is. Whether it changes the ranking, it will say that if the cigarettes
are very closely ranked one above the other, it really does not matter very
much which is the reality of the situation.

DR. FREEMAN: Dr. Hughes?

DR. HUGHES: You asked if there is any disagreement. I have a little bit of
a disagreement, in that 1 am still worried about reporting numbers. 1 could
go along with numbers if there were a disclaimer that says that 10 mg of
tar does not cut your risk in half compared to 20 mg of tar. 1 am still very
concerned that even if we give ranges that people are going to look at the
averages and think that 10 gives you half the health risk of a 20, but it does
not.

DR. FREEMAN: Dr. Hughes, what would you recommend in that case?

DR. HUGH US: You could go with either the nonnumerical system where
you certainly had a band and put them in or numbers as long as there is a
disclaimer that these are not cardinal numbers. That can be communicated
fairly easily by doing that.

208

Section II

DR. FREEMAN: So, you favor going with numbers or not going with
numbers?

DR. HUGHES: We have a long morning ahead of us; I can be persuaded either
way. My only point is there must be some information to the consumer that
10 mg is not half the risk of 20. As long as that is in there, 1 am agreeable.

DR. FREEMAN: 1 think it is conceivable that whatever we decide here could
also be accompanied by something in writing to explain and educate the
public. 1 think we should assume that could be done.

DR. HUGHES: I would like to see if maybe Dr. Shiffman or Dr. Benowitz
could give us a more concrete proposal here to make sure that we know what
is going on because what 1 hear people saying, and I just want to make sure
we are all saying the same thing, is that having a testing method that has a
range of values, 1 do not know whether you want to call it the 95 percent vs.
the 5 percent or something like that, some range of values based on doing
different things, blocking holes and that sort of thing, but we are also talWng
about one thing (I was unclear), is still reporting a mean or not reporting a
mean? That is what I am confused about.

DR. SHIFFMAN: You are suggesting that the current FTC system would
represent a band?

DR. HUGHES: So, not a range. So, is it from the 50th to the 95th percentile?

DR. BENOWITZ: I do not think the current FTC is the 50th.

DR. DEBETHIZY: Yes, I do not think we want the current FTC method to be
the bottom. I mean if you are talking about a range, the range has a low and
a high, and the FTC number is in the middle. So, I think that is important.

If you are going to talk about a range, you have to talk about the whole range.

DR. HUGHES: I think you have to have a range, but whether the FTC ends
up in the middle I do not know. Let me suggest that I do not think that just
getting an upper and a middle is fair to the consumer because there are
consumers at the lower end who are getting more health benefit, if there is
any, from the low-yield cigarettes than the average smoker, and I do not think
it is fair to not portray that to them. So, I would like to see the full range.
Everybody thinks they are the average. I would like to see it not
have a mean.

DR. FREEMAN: Dr. Cohen?

DR. COHEN: I would guess that I am thinking ahead to what might happen
in the marketplace, both competitively and with respect to smokers who are
also consumers. I think that we have to understand that whatever analysis
is done for internal purposes among specialists is one thing, but when
information is presented to consumers in a form that they cannot handle,
we cannot underrate the difficulty of educating them about that. It is not
going to be easy to explain the idea of a range to consumers. I would ask
the panel to consider a slightly different alternative, and that would be to

209

Smoking and Tobacco Control Monograph No. 7

vary the test parameters and produce a range of reasonable smoking
responses on say, tar, maybe using the machine and then to pick some
number like the mean plus a standard deviation, let us say, just to throw
something out for discussion, because this would represent a number that a
reasonable number of smokers might really encounter. In other words, that
would be the tar level that a substantial number of smokers would actually
encounter in smoking a cigarette, and if that were presented to consumers,
yes, it would err a little bit on the high side for some consumers, but 1 think
our duty may be to give consumers information that serves to protect a
reasonable number of those who are ingesting more.

1 think that if that number were provided, 1 do not want to call it a
maximum, you would find that firms would have an incentive to modify
cigarettes. They have a lot of design features they can use to modify low-
yield cigarettes to be sure that the mean plus one standard deviation would
be as low as possible, and 1 do not think we should underestimate the
importance of what is done here on the design of cigarettes in the future.

1 do not think we should underestimate that, and I think if we give them
something along the lines of what we are talking about, they have the ability
to see that their cigarettes come in at as low a number as possible.

DR. FREEMAN: May 1 ask you. Dr. Cohen, how would you reach the mean in
such a method?

DR. COHEN: 1 am certainly not technically competent, but in listening to
the discussions and reading the papers, if the FTC testing method were
adjusted to deal with such things as puff number, puff interval, and puff
volume and this were done based on an observation of how smokers smoke,
just as it was done when the original Cambridge Filter method was set up
in the first place, then you would be able to know what the magic number
would be for two-thirds of the sample or some arbitrary number, and it
would be greater than the mean. I think that number would probably be
a lot easier to communicate than a range, and it would have the side benefit
of better informing smokers as to what their potential risk might be, and it
would also provide great incentives to the industry to make cigarettes that
came in at as low a number as possible.

If one of the major problems with the low-tar cigarettes is where the filter
holes are and how they work and the fact that they can be covered, then if
this testing protocol were followed and the mean plus one standard deviation
for that cigarette the way it was smoked were a fairly large range and if the
company making that cigarette did not like that large a number, it has the
capability of reducing that number by putting the filter holes in such a way
that they are not going to be blocked.

I would say that it is very important to consider the impact of what is
done here on what they do.

DR. FREEMAN: Dr. Shiftman?

210

Section II

DR. SHIFFMAN: I very much share your concern that we come up with a
system that is communicative and that is grasped by the smokers whom we
are trying to reach. Part of what is attractive about a range is that it also
communicates, to borrow a phrase, that their "mileage" will vary. My
concern about any one number, no matter where you put it on the spectrum,
is that it does not communicate that and implies that this is exactly what
this cigarette will deliver. So, 1 think it is a significant challenge to health
education, public education, and advertising people to design a system that
communicates this idea of range and the idea that range is to some degree
under the control of the smoker and his or her behavior. Part of what is
attractive to me about that range is communicating exactly that, that the
human yield is variable and that it is variable to some degree according
to the behavior of the smoker. That is something 1 would like to see
communicated.

DR. FREEMAN: Dr. Stitzer?

DR. STITZER: 1 just want to support that point. 1 think that we are dealing
with a situation where the public is very lacking in knowledge, and the one
particular thing that is not understood by smokers, 1 believe, is that the
way they smoke their cigarette determines the yield that they get from it.

1 think the basis of the system we design should be to convey a very basic
piece of information, and some of these ideas about ranges and so forth
are important. The fact of the matter is that, with low-yield cigarettes,
these ranges are going to be very wide. They are going to be completely
overlapping with the higher yield brands, but that is exactly the information
that we want the consumers to know.

Now, the unfortunate part is that consumers do not have any good
way of knowing where they in particular fit along any range that we might
present, and that is a different problem.

DR. FREEMAN: Dr. Benowitz?

DR. BENOWITZ: 1 would also like to support Dr. Shiftman's comments and
just say that we could be specific about this, and 1 think we should be. For
example, we can say that if you block these ventilation holes, this is what
your exposure will be, and if you do not, this is what your exposure will be,
and we can also request that ventilation holes be marked to make them
obvious to the smoker. Make them bright red or orange or something to
minimize your exposure; 1 am all for ultralow-yield cigarettes if people will
smoke them that way. 1 think that is great. You have to make it possible
for them to do that, and we could with labeling.

DR. FREEMAN: Dr. Henningfield?

DR. HENNINGFIELD: I also agree that the range is basic, honest, accurate
information, but it is clear that it has to be coupled with education on what
factors may affect your intake: how consumers can change their behavior
in ways that might be helpful. But a really important point of Dr. Cohen's
I think should be considered, and that is the importance of providing an

211

Smoking and Tobacco Control Monograph No. 7

incentive to the industry that may serve people. Drs. Benowitz and
Kozlowski and I had this embedded in our Journal of the American Medical
Association proposal: the notion that right now, in our estimation, virtually
all cigarettes you throw into the regular category, but by providing the
incentive to get that label of low, which could be a really nice selling point
and may be of health benefit, you would have to work to redesign cigarettes
in such a way that 1 think would be useful, and what you would have to
do is redesign them in a way that would make sure that the upper level was
lower. And that brings me again to the reason that I think we need some
bioavailability testing. 1 agree with Dr. Benowitz, not necessarily on 900
brands, but you need to anchor it at some point to what people get, and you
need an agency that can oversee that properly and also require it on demand;
that is the only way you are going to prevent another Barclay cigarette type
of scam, the notion that somebody comes up with a design that seems to
meet the low category, and they have just done it by beating the machine.

The only way you are going to check that is by seeing what people get.

DR. GUERIN: Dr. Henningfield, as a good example, the FTC test is what
discovered the Barclay scheme. The FTC test has been successful in
identifying those kinds of problems.

DR. SHIFFMAN: Dr. Henningfield, I take that perhaps as an additional
proposal that you address a different issue than we have been talking about,
which is the use of words like "light, low, ultralight" in advertising and the
importance of making those accurate and not deceptive or confusing to the i
consumer, and that I think is something we ought to address. That the |

information that is presented to the public in advertising goes beyond the
small numbers printed in the corner to the large "light," "ultralight," "low"
printed in bold print, and I think that is something we ought to look at. I

DR. HENNINGFIELD: Yes, I think those words should be banned. j

DR. SHIFFMAN: I would disagree that they should be banned. They should
be regulated so that they are accurate.

DR. FREEMAN: Dr. Rickert? '

DR. RICKERT: I think it is obvious that many consumers choose their brands
on the basis of some perceived risk to health, and it is also obvious that the
ITC numbers do not and never were designed for that particular purpose. ^

1 share Dr. Hoffmann's concern in that our measure of dose is often based on j
nicotine, which may or may not tell us about other constituents in tobacco ;
smoke. Specifically, at the level of molecular epidemiology, there are certain i
constituents that now can be tracked, and Dr. Hoffmann has mentioned i

NNK in urine. There is the constituent 4-amino-biphenyl, which is present [
in tobacco smoke and which in smokers ends up as a hemoglobin adduct. |

I here is, also, bcnzo(fl)pyrene, which in smokers ends up being bound to
albumin. So, there are a number of traceal)lc constituents in tobacco smoke ■
that have known toxic or carcinogenic properties, which also then can be i
related to uptake in smokers and nonsmokers alike. |

212

Section II

DR. FREEMAN; Dr. Cohen?

DR. COHEN: I would like to return to the point that Dr. Henningfield was
just talking about and the point that 1 raised a few minutes ago. 1 think it
is very important not to put the burden in the wrong place.

If we are going to put the burden on consumers to respond to whatever
design changes industry makes and then educate them each time industry
makes a clever change, as to where they put the holes or what kind of paper
they use or whatever, we are fighting a losing battle. 1 think that the best
approach here is really to allow the industry, which is able to modify its
product, to modify it in order to obtain the maximum benefit to their sales
from a low rating. They have an incentive to do that, and so, if the panel
deems it appropriate, 1 would suggest that coming up with a rating system
that reports to consumers a number within the range, which is not at the
midpoint of the range but is tilted toward those who do compensate, is the
smartest thing that can be done because that, in fact, will offer guidance
to consumers who after all should not have the primary responsibility for
outsmarting the designers of cigarettes, and 1 think that it would also provide
the cigarette industry an opportunity to modify their design in order to
achieve the numbers that are most beneficial for them.

DR. FREEMAN: Dr. Hoffmann?

DR. HOFFMANN: Before we got this upper limit and extreme, we should ask
Dr. Guerin how far we can go. We have heard yesterday that the low-tar
cigarette smoker may take up to 60 to 65 mL per puff and up to 6 or 7 puffs
per minute; is that possible with our current equipment?

DR. GUERIN: Current instrumentation would have to be modified
somewhat to reach some of the extremes in terms of volumes.

DR. HOFFMANN: You can do more than 50 mL?

DR. GUERIN: Right.

DR. HOFFMANN: With the machine?

DR. GUERIN: No, 1 said that you can, but it would require some
modification.

DR. HOFFMANN: New machinery?

DR. GUERIN: To reach some of the extremes in terms of puff volumes,
frequencies for a 20-port system would be too high. You can purchase
systems of smaller capacity that have that flexibility.

DR. HOFFMANN: But the standard machine we have now cannot go
through these extremes?

DR. GUERIN: It would not be able to go through all the extremes without
some modifications.

213

Smoking and Tobacco Control Monograph No. 7

DR. BOCK: The cost of the machine really is not something to use as a basis
for this discussion. It is so small compared to the cost of labor that it is
meaningless.

DR. FREEMAN: Dr. Hughes?

DR. HUGHES: We have been talking about these numbers and as scientists
we like to talk about numbers, and maybe we will get to this with the third
question. My major concern is conveying how much health benefit people
get by these lower nicotine, lower tar cigarettes because what 1 saw in the
1981 Surgeon General's report and what 1 saw Dr. Samet present yesterday
suggest to me that it is not great, and it is not very large, and 1 think when
the normal consumer switches to a 1-mg cigarette, they think they are doing
themselves a great benefit, and my concern is that the magnitude of that
effect be conveyed to the consumer.

DR. FREEMAN: Dr. Rickert?

DR. RICKERT: At the present time, FTC methodology is providing us with
information on tar and nicotine and CO; when we are talking about lower
yield cigarettes, we tend to link tar and nicotine explicitly together, and
while there is an obvious relationship between these variables, there is also
extreme variation.

All one has to do is take the 933 brands that were just published recently
in the FTC report and look at various plots of CO vs. nicotine or tar vs.
nicotine, and you cannot help but be struck by the fact that there is a wide
range of variation as far as specific nicotine level.

For example, if you look at that report and brands delivering .9 mg
of nicotine, there were 54 brands with varying tar yields. So, 1 think,
in addition to the issue of tar and nicotine, that the issue of how one
communicates simultaneously changes in all three variables because you
can have the situation where it could be high tar-low nicotine or low
tar-high nicotine, and by constantly linking the two together, 1 think
one is missing the point about the other two variables.

DR. FREEMAN: Yes, Dr. Stitzer?

DR. STITZER: I was going to pick up on Dr. Hughes' point because it is
an important one, but I think it is very much intertwined with the whole
discussion about reporting of the machine testing yields. If the smoker
can visualize the fact that the actual yield from this low-yield cigarette is
completely overlapping with the yield from this other high-yield cigarette,
then 1 think that can more easily bring home the other health message,
which is that switching to these cigarettes may not have any benefit
whatsoever. I think that there is a dose-response problem there.

DR. I REEMAN: Dr. Benowitz?

DR. BENOWITZ: I think that it would be great if we could put something
in about health risks. 1 think the data seem very clear that smoking any

214

Section II

cigarette is so much greater risk than smoking none that it will be impossible
to quantitate it, and 1 think that should be communicated. But at the same
time, even if there is a small difference in exposure from high- to low-yield
cigarettes, if you are talking about a huge population of smokers, it is
worthwhile to encourage as many possible to get as low a yield as possible,
even though it is not going to have nearly the effect of stopping smoking.

It still is of some benefit. So, 1 think we should warn people that switching
to low-yield cigarettes is not going to remove the risk of smoking, but still
try to encourage that somehow people do that.

DR. DEBETHIZY: Dr. Benowitz, you are raising an important issue. It is that
whatever change that gets recommended here today to the FTC, it is going to
require some research and some study to make sure that some unintended
things do not occur. For instance, if ranges were recommended, and put
on in advertising, would that have the effect of discouraging people from
switching down? As a scientist, 1 think it is important for us to understand
the ramifications, and 1 am assuming that this is just the start of a process,
that recommendations will be made, and that the FTC will consider those
using research techniques.

DR. FREEMAN: Yes, Dr. Shiftman?

DR. SHIFFMAN: Just to proceed on the point that Dr. Hughes and
Dr. Benowitz made, we have said a couple of times that the idea of educating
smokers is very important. 1 think educating them not only about these
numbers or ranges but also about the comparative benefit of not smoking
at all vs. lowering the received yield is an important part, and 1 think it
deserves some discussion, though perhaps not here, about the degree to
which that can be done in this sort of labeling rating system or whether,
in fact, we need other media as well. There is limited information we are
going to get on a pack or in an ad, but 1 think there is a responsibility to
educate smokers so that they do make those informed choices.

DR. FREEMAN: Dr. Woosley?

DR. WOOSLEY: 1 agree slightly with the representative from the tobacco
industry that we have to make sure there are no adverse consequences from
anything that we try to do in a meaningful way, and one of the most serious
concerns 1 have is that we do not want to give a false impression about
health risks. 1 think one of the most disturbing pieces of data that I saw
yesterday was the indication that people who were on the ultralows had
a lower cessation rate, and 1 am concerned that potential means that the
recommendations that come out of this panel may encourage people to go
to low yield instead of stopping smoking, and 1 think that overall will be
a terribly adverse health risk or adverse effect on the overall health of the
Nation.

DR. FREEMAN: Yes, Mr. Peeler?

MR. PEELER: In line with that discussion 1 just wanted to throw out two
pieces of information for the panel's consideration. The first is from the time

215

Smoking and Tobacco Control Monograph No. 7

the ban on tar and nicotine claims in advertising was lifted, the FTC basically
prohibited any health claims in the advertising. So, what you are seeing is
really what consumers are inferring from the low-tar and -nicotine systems.
The other thing that we have focused on and thought about in this area is
the fact that the Government's position for many, many years has been that
people who are concerned about their health should stop smoking. There
is actually a warning on the packages right now that says that. Stopping
smoking now increases your health, but it is a very difficult communications
conundrum, as Dr. Shiftman has indicated, about whether you can talk about
relative risk from tar and nicotine and not send an unintended message that
Dr. Woosley is talking about, that this is the better way to go.

DR. FREEMAN: Yes, Dr. Kozlowski?

DR. KOZLOWSKI: I think it is pretty clear that smokers currently are turning
to so-called "light" and "ultralight" brands, believing that they are doing
themselves a favor with respect to their health. 1 think there is a real concern
about health-conscious smokers who want to try to do something. They now
have the impression: How could a light cigarette kill anybody; how could
an ultralight cigarette kill anybody, and there is no deadly connotation to
the terms "light" or "ultralight." By providing better information the hope
is that some of the people who out of health concerns are turning to lower
yield cigarettes will see something of the risks of that, and they may be in a
fool's paradise, and they maybe should stop altogether. The pamphlet that
was passed out in an earlier version about 10 years ago was titled "Tar and
Nicotine Ratings May Be Hazardous to Your Health: Information for Smokers
Who Aren't Ready to Quit Yet." A smoker who is smoking a 1-mg tar
cigarette and enjoying it may think, "My God, 1 am smoking the lowest
yield cigarette on the market; how could that do me any harm? 1 have really
done something." If that smoker then sees, "1 blocked the vent holes," and
so on and so on, the hope is that becomes an inducement to stop. 1 think
you expressed some reservations about use of the term "consumers," but 1
think it is important that continuing smokers be treated in part as consumers
and be given information similar to what consumers have expected about
automobiles and things like that.

DR. COHEN: Could 1 follow up with some numbers? 1 have some evidence
exactly on the point that you just made, and it may be useful to the panel to
hear the evidence. 1 apologize, again, because you couldn't read the numbers
so clearly yesterday. In my survey, 83 percent of those smoking 1- to 5-mg tar
cigarettes thought that switching from a 20-mg tar cigarette to a 5-mg tar
cigarette would significantly lower that person's health risks due to smoking
for someone who smokes a pack a day. More than 25 percent of those
smoking cigarettes with 6 or more mg of tar thought that switching from a
20-mg tar cigarette to a 16-mg tar cigarette would significantly lower that
person's health risk due to smoking for someone who smokes a pack a day.

DR. DEBFTHIZY: I would not be surprised at that because the Surgeon
General said in 1981 that if you reduced your tar intake, you reduce risk.

216

Section II

It has been communicated pretty clearly that if you cannot stop or are
unwilling to stop, then reducing your tar intake is a good idea.

DR. GIOVINO: I would like the Surgeon General's comments to be put on
the record so that they could be stated exactly and hopefully they will be
used exactly as stated. The Surgeon General said:

The Public Health Service policy on lower tar and nicotine
cigarettes must remain unchanged. The health risks of cigarette
smoking can only be eliminated by quitting. For those who
continue to smoke, some risk reduction may result from a
switch to a lower tar and nicotine cigarette provided that
no compensatory changes in style of smoking occur.

I would ask that caveat be used when these types of statements are made.

I would also remind us that while the relative risk studies on lung cancer
may have controlled for number of cigarettes a day, and I am not sure of the
methodology on those, they certainly have not controlled for changes in puff
frequency or puff volume. So, one point I want to make is, let us make sure
that we provide in any statements we make about the Surgeon General's
statements the caveats that the Surgeon General's report provides, and the
second point I would like to make is that the categorization of light and
ultralight cigarettes in advertising and promotion is not always consistent.
There are many exceptions to those rules that Ron Davis pointed out in his
article in the American Journal of Public Health, and the current system of
light and ultralight seems not totally consistent at times with the tar and
nicotine ratings.

DR. FREEMAN: Dr. Headen?

DR. HEADEN: I want to go on record in support of the color-coded
representation of the FTC information for the consumer and to go on record
in support of a range rather than a single number. I would ask us to consider
the point that Dr. Cohen made. It is important to design this information in
a way that would encourage the tobacco industry to redesign cigarettes to
conform to whatever standard we adopt, but I do believe that if there is a
range that there will be an incentive on the part of the industry to lessen the
width of that range. A cigarette brand that has a very broad range gives a
very clear message to the consumer that the yield is variable, particularly
when they consider the upper limit, and that there would be a high incentive
for the industry to narrow the range of whatever yield there is for each of the
cigarettes.

DR. FREEMAN: Thank you very much. Because we have two other major
elements to consider today, and it does not mean we cannot discuss more
of this, we want to go to the second question, and you can continue to raise
ideas on the first question as we go along because they do overlap a bit.

217

Smoking and Tobacco Control Monograph No. 7

Question 2. Should constituents other than tar, nicotine, and carbon
monoxide be added to the protocol?

Dr. Henningfield?

DR. HENNINGFIELD: I do not think we have the information to decide the
entire list, but we probably have some ideas of things that should be added.

I would suggest that the procedure be used as the FDA uses for food labeling,
which is that substances that an organization or committee with specialty in
toxicology agrees are of toxicological significance be added. And with foods,
under the category of other flavorings and ingredients, industries are not free
of listing things that are of toxicological significance just because they call it
a flavoring, as occurs with cigarettes. Rather an outside body decides what is
of toxicological significance. I do not know if they are of toxicological levels,
but that should not be decided by the tobacco industry, in my opinion. That
should be decided by a regulatory agency with toxicology experts.

DR. FREEMAN: Dr. Benowitz?

DR. BENOWITZ: I think Dr. Henningfield's comments are well taken, but
I would just like to go on record in support of the sort of labeling that
Dr. Harris showed us yesterday, which I thought is very informative to
consumers when you see all the cyanide and arsenic and all those things in
cigarettes. I think it just helps to provide more information to a consumer
about the mix of what is in their tobacco smoke. I am sure they are not
going to read every bit of it, but anytime they are interested in looking and
they see a list of 30 cancer-causing compounds, I think it is useful for them
to know that. So, I am in favor of having that sort of listing available.

DR. FREEMAN: May I try to understand what you said, that you would not
present measurements in the way that we are measuring the three elements,
but you would simply list them as carcinogenic or harmful to health? How
would you do this?

DR. BENOWITZ: I think you could do it either way. It would be relatively
straightforward to measure those and just list how many micrograms or
whatever was there, or you could just list them. I do not have a very strong
feeling about it just as long as it is made clear to people what the types and
the mix of toxic chemicals they are taking in their smoke.

DR. FREEMAN: What I mean is, you would treat them differently from the
way we are treating tar, nicotine, and carbon monoxide?

DR. BENOWITZ: Yes, I do not think we need to provide information about
standard and maximum exposures, for example. I just think if we had one
number or one list it would be adequate.

DR. FREEMAN: I see. Thank you. Dr. Rickert?

DR. RICKER'F: I think there should be additional communication of
information. I question whether or not it should be in the form of an
absolute number. For example, it could be categorized in terms of

218

Section II

ciliatoxic agents and hydrogen cyanide, for example. It could be categorized
as carcinogens and then certain carcinogens. I think, given all the
discussion that we have had both yesterday and today about the potential
misinterpretation of numbers, it would seem to me that to add to that
confusion by a whole new set of numbers would not be serving the interests
of the consumer.

DR. FREEMAN: Thank you. Dr. Rickert. Dr. Stitzer?

DR. STITZER: I am totally in favor of more information being given to the
consumer, but I want to bring up priorities and to point out that cigarette
packs are not very big, and personally I think it is more important to convey
the information about the variability of yield in some prominent way on a
cigarette package rather than using that space to list hundreds of chemicals.

Now, what I would be interested in seeing is a kind of package insert
disclosure information that would be put in cigarette cartons. That might
be a better format for delivering the information.

DR. FREEMAN: Dr. Petitti?

DR. PETITTI: I would like to provide support for both of the prior
statements, particularly that perhaps cigarettes need to have prominently
displayed the fact that they contain some list of selected carcinogens but
not to portray that information in a quantitative fashion. I do not think we
have the ability to decide which of the numerous carcinogens in cigarette
smoke is the one or the ones that quantitatively are related to the various
forms of cancer caused by cigarettes and that we would be further misleading
the consumer by making them believe that, for example, a cigarette with low
levels of chemical X is better for them than a cigarette with a higher level of
chemical X. I do not think we will ever have the epidemiological database
that will allow us to link these various carcinogens quantitatively with risk
of any of the human cancers.

DR. FREEMAN: Yes, Dr. Guerin?

DR. GUERIN: If we had a list of common cigarette smoke toxins on
packages, would that not discourage the industry from producing products
that have undetectable quantities? Wouldn't we need some kind of a level
of detection to determine when we are at basically nothing because there
are advanced products that are being marketed or may be being marketed?

DR. FREEMAN: Let us go to Dr. Hughes first.

DR. HUGHES: People are interested in function. They do not care about the
names. They care about what these things do. So, my proposal would be to
get the word "cancer-forming" agents on there. That is the important thing,
not whether it is this long name or that long name or whatever because that
is the important thing that consumers need to know.

DR. FREEMAN: Yes, Dr. Townsend?

219

Smoking and Tobacco Control Monograph No. 7

DR. TOWNSEND: Thank you, Mr. Chairman. In the U.S. market, cigarettes
with comparable blends, in fact, show similar ratios of most smoke
constituents per milligram of tar. What that means is, as you bring the tar
level down in a cigarette, these constituents also come down more or less
proportionately.

I guess the question then that leads me to is, are we providing really
useful information for the consumer to use, or are we just giving them a lot
of information that they are going to ignore, like I think Dr. Harris indicated
that he did not read all the information on the food labeling. There is a lot
of information there. Is it useful information? That is just a question I am
proposing.

DR. FREEMAN: I understand. Dr. Rickert?

DR. RICKERT: I agree with what Dr. Guerin has said, in the sense that it
would seem that if there are new products being developed that would have
a zero yield of various carcinogens or ciliatoxic agents that must be allowed
for, and I think for that, for the agent to appear on the package, there would
have to be some consideration given to analytical detection limits and things
of this nature. So, I do not think it should necessarily be a blanket piece of
information that comes with every brand of cigarette, but there should be
some differentiation based on the individual products of the cigarettes.

DR. FREEMAN: Yes, Dr. Henningfield?

DR. HENNINGFIELD: There always has to be a threshold for whether you
list something, whether you are listing the lead in flour or rat droppings or
whatever it is. So that is a basic concept, but I think the important thing is
that an agency or panel with expertise, not the industry being regulated,
make the decision. So if BHT is listed on flour or cookies, you know, we do
not have to worry about how much of it is in there; I agree, that would be
confusing, but another group decides if the cyanide should be listed as
one of the "also contains" ingredients, for example, "also contains cyanide,
formaldehyde, lead." You have somebody else decide what is of potential
significance and therefore should be listed, and that is what is done with
food labeling.

DR. FREEMAN: So, in principle what are you saying that we should do?

DR. HENNINGFIELD: Of course there is a threshold, but that threshold does
not necessarily affect the labeling on the pack in the sense of the number
because you are not putting any numbers in the same way that for potato
chips you are not putting how much BHT is in there. You are listing the
milligrams of things that groups decide are very important to list by
milligrams like cholesterol and sodium, and then you have the list of other
ingredients as Dr. Harris showed on his label, and what the thresholds are;
to merit listing, another group decides that — a group with toxicology
background.

DR. FREEMAN: Dr. Rickert?

22tJ

Section II

DR. RICKERT: I am not really quite sure whether it is a toxicological
consideration or whether it is a chemistry consideration. If you take
cholesterol, for example, in order to earn a label of cholesterol-free, there
has to be a certain level. I mean in the instmment you can measure this
level, but it is cholesterol-free if it is less than that level. So, I think the
issue, from my point of view anyway is more of a chemical issue, rather
than a toxicological issue.

DR. FREEMAN: Dr. Shiftman?

DR. SHIFFMAN: I would like to come back to an issue that I think we
discussed but perhaps had not gotten closure on, which is the consumer
information or misinformation that is conveyed outside the formal label in
the form of brand names like "light" and "ultralight," and I guess what I was
hearing from several people was a proposal that the use of those terms be
regulated in a manner parallel to the FDA's recent regulations of such labels
on foods. That is my own view — that those labels ought to be allowed. They
have the potential to provide a smoker with meaningful information, but
they should be regulated so that they represent a particular number or range
in the ratings and so that they have a common meaning across brands and
across manufacturers. I would like to hear other people address that issue.

DR. FREEMAN: Dr. Hughes?

DR. HUGHES: I think that is a good comment. Dr. Shiftman. Where I keep
seeing the break is at .5. The ones less than .5 are different from those
above .5. So I would like to see the categories only be on those that are
less than .5 mg of nicotine with the comparable tar.

DR. FREEMAN: I think the general question here is, should constituents
other than tar, nicotine, and carbon monoxide be added to the protocol for
testing? It sounds to me as though you have said that they should not be
added for testing or, if testing is done, it should be done by either chemists or
toxicologists but not in the same way as for the other three major substances.
There is a question of whether certain substances should be listed in some
way on the tobacco pack or in some other way, so the American public
would know that there are harmful ingredients in tobacco other than the
three elements.

Let us try to get closure on that particular point before we go on.

DR. GUERIN: I think that tar and nicotine and CO in terms of quantitative
measurements are adequate. I have one question for Dr. Hoffmann. Of
all of the constituents that might not necessarily correlate very well with
tar, the N-nitrosamines stand out. Should we consider an N-nitrosamine
measurement?

DR. HOFFMANN: When you find another HCN, benzo(fl)pyrene, you will
always get from our friends of the tobacco industry, "Yes, but this can come
from air pollution." Carcinogens that derive from nicotine can come only
from tobacco, and I think that should be cited. Benzo(a)pyrene comes from

221

Smoking and Tobacco Control Monograph No. 7

every combustion. Hydrogen cyanide comes from every combustion of a
protein-containing component. Phenol comes from any combustion. The
tobacco-specific nicotine-derived nitrosamines, which are strong carcinogens,
come only from tobacco. I think the consumer should know that.

DR. GUERIN: To my knowledge that is the one class of chemicals that might
be considered in addition, although it would be very difficult to do, relative
to tar and nicotine and CO measurements.

DR. HOFFMANN: I personally do not think we confuse the smoker. 1 think
we can let them know on the package that cigarette smoke contains toxic
agents: hydrogen cyanide and known carcinogens, chloraminobiphenyls,
benzo(fl)pyrene, and the nicotine-derived nitrosamines. We should say that,
though not quantitatively. Giving numbers is only confusing here.

DR. FREEMAN: Before we entertain other questions, I would like to have the
committee's sense of whether you support what Dr. Hoffmann has just said,
which seems to be a good summary of what we have said so far? Is there any
disagreement with what Dr. Hoffmann has just said?

[NO RESPONSE]

Then we will take that as a consensus, and we will go on to consider it
further in the latter part of the day.

You had another comment. Dr. Petitti?

DR. PETITTI: Just to emphasize that this information should not be
quantitatively presented to the consumer because of the potential for
misinformation.

DR. FREEMAN: My understanding relative to Question 2 is that there are
certain elements that are proven to be harmful to human beings that are
within tobacco that should be listed, though this panel is not recommending
which specific compounds; that they should not be listed according to
quantity; and that they should not undergo the same testing that we are
recommending for tar, nicotine, and CO. I will take that as a consensus at
this point.

If that is sufficient for that question, I would like to go on now to the
third question that we have been asked to consider.

Question 3. Docs the FTC protocol provide information useful to consumers
in making decisions about their health?

Yes, Dr. Petitti?

DR. Idn rrri: I think that in the context of the purpose of the FTC protocol,
which is to provide information that allows consumers to choose cigarettes
that reduce their risk of disease, the current M C protocol is misleading in at
least two important ways. I’irst, it presents the consumer with a single
number, thus implying that the consumer will receive exactly that exposure.
.Second, I think that the numbers when [)resented as numbers implicitly

222

Section II

suggest that there is a ratio-scaled relationship between machine-measured
yield and disease risk. 1 think that these numbers could be made more useful
by remedying these misleading aspects of the current FTC protocol in ways
that we have been discussing throughout the morning.

DR. FREEMAN: Specifically what do you recommend?

DR. PETITTI: First, present a range of numbers, thereby correcting the
problem of a single number implying that the consumer will receive
exactly that amount of exposure; and, second, provide some kind of graphic
presentation of this information to the consumer in a way that takes away
the numerical aspect of "9 is 9 times higher than 1 and 20 is 20 times higher
than 1."

DR. FREEMAN: Dr. Rickert and then Dr. Ffughes.

DR. RICKERT: 1 agree. 1 think the issue is, can the FTC protocol provide
useful information rather than does it? In other words, are there ways that
we can take this kind of information and convey it? One of the options
that has been discussed and also appears in the literature is this idea of using
the color of tar to communicate the range of variation that one can get in
tar yields and, also, to allow individual smokers to gauge what they are
receiving from the cigarette. 1 think a lot of us feel that if it were possible
for low-yield smokers, that is those who are smoking the less-than-5-mg
cigarettes, to achieve a low-yield smoke from that cigarette, there may be
an accompanying health benefit to that achievement. At the present time,
however, there is no means for the smoker to ascertain what the yield is.

If there were some sort of graphic technique for them to visualize this
process, then that may confer a health advantage to them.

DR. FREEMAN: Dr. Hughes?

DR. HUGHES: I have a question for Mr. Peeler. I was struck by the remark
that talking about health benefits implies a health claim, which I understand
the tobacco industry has not made with this product. Therefore my question
to you is, can the FTC require such health education to come from the
tobacco manufacturers when they have not made that health claim?

MR. PEELER: The numbers clearly communicate some health benefit to
some portion of consumers. If we were able to determine that a significant
number of consumers were being misled by that, we could require some
corrective information or provide some corrective information. I think the
concern has always been the one that we started out with in 1960, that as
a result of that position, there basically were not any tar and nicotine
numbers being made available to consumers.

In terms of our present status right now, our jurisdiction is simply to
require that claims made in advertising be substantiated. Currently, there
is not an FTC rule, an FTC case, or any legislation requiring the disclosure
of tar and nicotine data either on labels or in advertising.

223

Smoking and Tobacco Control Monograph No. 7

DR. HUGHES: If some surveys or several surveys showed that most
consumers inferred a claim of health benefits from these cigarettes, would
that be important information for the FTC to know when considering
whether to force the tobacco companies to put in health information?

MR. PEELER: Consumer survey information is absolutely vital to everything
the FTC does in the regulation of advertising and labeling. Therefore, the
answer to the first question is yes. The next question is this point that we
raised earlier: whether there would be the ability to compel disclosure of
tar and nicotine information absent a health claim. You do get back to the
tension between having no information out there at all and having just the
accurate tar and nicotine.

DR. HUGHES: I think you are misunderstanding, because I am not talking
about whether they report the numbers or not. Suppose there was a proposal
that the FTC would require all cigarette advertisements that state anything
about tar to have a statement that reads something like, "Switching to a low-
tar cigarette is a very small health improvement compared to stopping
smoking."

MR. PEELER: There is a legal analysis under the Commission's unfairness
authority that could be used to require disclosure of that information,
assuming the correct factual predicate could be established.

DR. HATSUKAMI: Regarding health benefits, it seems that based on
Dr. Cohen's presentation there are significant numbers of people who
believe that there are benefits to switching to a lower tar and nicotine
cigarette, and yet some of the data do not show this relationship between
tar yield and health benefits, with the exception possibly of lung cancer.
Therefore, I agree with Dr. Hughes in requiring some kind of label on the
package explaining that switching to a low-tar and -nicotine cigarette
may or may not provide health benefits, thereby hopefully correcting
an apparent misconception.

DR. FREEMAN: Dr. Benowitz?

DR. BENOWITZ: I would just like to emphasize that such language has to
be stated very carefully; I believe for an individual there is very little benefit
to switching. For the society of all smokers there may be benefit, and I
would not want to lose that benefit. We have to walk that line of warning
individuals that this is not going to help you very much, but still encourage
the whole society of smokers to reduce their tar and nicotine intake. We
need to find language that will serve both purposes.

DR. FREEMAN: Dr. Kozlowski?

DR. KOZLOWSKI: I would like to encourage the panel to note in the report
that there is a fundamental deficiency in that the current procedures are
linked to cigarette advertising. A number of presenters mentioned that
currently one-third of cigarette brands are generics. There is no requirement
that a cigarette brand be advertised. I here is no law that says that you must

224

Section II

advertise a cigarette, and, if it is not advertised, there is no option for any
kind of brand-yield information or any kind of FTC method. I think one
might see the linking of consumer information solely to advertising as a
loophole in the system. 1 would encourage the panel to ask the FTC to try to
provide some estimate of what percentage of brands are not advertised at all.

DR. FREEMAN: Mr. Peeler, can you address that question?

MR. PEELER: As 1 said earlier, we do not have that information today.

We could certainly get that for the panel, if the panel would like it.

DR. FREEMAN: Are there a significant number of brands that are out in the
market but are not being advertised?

MR. PEELER: There has been a very large increase in the number of brands
that we have been reporting because of the increased number of generic
brands, which are frequently not advertised.

DR. FREEMAN: And is it true that the FTC does not require that those brands
undergo the same analysis?

MR. PEELER: Again, we have to go back to the beginning. The disclosure of
tar and nicotine is provided for under the FTC's general authority to regulate
advertising and to require substantiation of claims in advertising. At one
point there was an FTC proposal to require the disclosure of tar and nicotine
content in all cigarette advertising. Cigarette labeling is largely regulated
by separate Federal statute called the Federal Cigarette Labeling Act. The
rulemaking was suspended when the cigarette industry voluntarily agreed to
put this information in all their cigarette advertising. So that is the current
status of disclosure. Many companies do put that information on their
packs, particularly with respect to their lower yield cigarettes. 1 do not
have an estimate of how many packs that is.

DR. FREEMAN: Then it is conceivable that there could be cigarettes sold that
do not have this labeling?

MR. PEELER: 1 believe it is likely that there are many cigarettes, particularly
generics, that do not have these labels.

DR. FREEMAN: 1 think that is a very significant point. If the FTC does not
regulate that, who does?

MR. PEELER: The content of the cigarette label is largely regulated by the
Federal Cigarette Labeling Act, which is a Federal statute that requires a
certain number of disclosures, for example, the Surgeon General's disclosure,
and then basically says that, for other statements relating to smoking and
health, only Congress can impose those additional statements.

DR. FREEMAN: Thank you. Dr. Hughes?

DR. HUGHES: 1 think we have a model with the recent food pyramid. That
labeling change was accompanied by a massive educational campaign, and
1 do not know who did that, but it seems to me that we need a similar effort

225

Smoking and Tobacco Control Monograph No. 7

with tar and nicotine. I want my patients to know about nicotine yield
much more than I want them to know about riboflavin and cholesterol and
that sort of thing. My question is, why have we not had the same public
education campaign around nicotine yield, spending at least as much
money as we did on the food pyramid?

DR. KOZLOWSKl: To respond to that, there is a question of in whose
jurisdiction does that campaign fall. 1 think it is not within the FTC brief
in any explicit sense to do an extensive education campaign on it, and you
have just heard that cigarette labeling falls under quite a different procedure.

DR. FREEMAN: Dr. Townsend?

DR. TOWNSEND: In response to Dr. Kozlowski's concern and your obvious
concern about generics or unadvertised cigarettes being out there in the
marketplace without any information, that really is not true. While the
specific numbers are not advertised, the generic products are broken into
categories of tar deliveries, the same as other brands. For example, you can
find a generic sold as regular, lights or ultralights. There is information out
there, even if there is no advertising that carries with it specific absolute
FTC tar numbers.

DR. FREEMAN: Would you clarify this point because 1 am a little confused.
What would be the difference on the labeling of the generic product vs. the
one that is advertised?

DR. TOWNSEND: Let me take one example with the generic cigarette brand
Doral. The packages are in different colors, the same as other brands, with
dark green for the regular, light green for the lights, and a real light green
or a white for the ultralights; and their tar category is stated on the package.
That is a comparative measure of the FTC tar yield for those cigarettes even
though there is no advertising that carries the FTC number with it.

DR. FREEMAN: That is how they are similar. How do they differ?

DR. TOWNSEND: What 1 am saying is that there is a distinction in the
marketplace by virtue of what is written on the pack, that it is a light or
an ultralight or a regular.

DR. FREEMAN: 1 understand.

DR. TOWNSEND: And by the color of the pack.

DR. FREEMAN: But other than that, the numbers are not there?

DR. TOWNSEND: Because many of these products are not advertised, the
numbers in some cases are not available to consumers.

DR. FREEMAN: You have clarified my point. Dr. Benowitz?

DR. BENOWrrZ: One of the central recommendations of this panel should
be that cigarette labeling include these warnings. The other issue that
concerns all of us is consumer information, and even though it is not a

226

Section II

direct charge, I think we should make a strong recommendation that there be
labeling about yields and the other issues we have been talking about.

DR. FREEMAN: Are you speaking also on the generic cigarettes?

DR. BENOWITZ: Yes, on all cigarettes.

DR. FREEMAN: That is your recommendation?

DR. BENOWITZ: Yes, because the FTC can regulate advertising, but what
we are really talking about is labeling on the cigarette packs. If there is no
vehicle for doing that now, 1 think we should recommend there be one.

DR. FREEMAN: Yes, Dr. Rickert?

DR. RICKERT: 1 would certainly support what Dr. Benowitz is saying,
particularly given the amount of confusion that sometimes arises over the
use of the terms "light," "ultralight," and so on. Unless these terms have
been defined with some specific tar range associated with them, the use of
the terms without that tar information is certainly open to the potential for
misleading consumers.

DR. FREEMAN: Dr. Woosley?

DR. WOOSLEY: 1 agree with both of those statements wholeheartedly, but
1 am concerned by the reality of labeling being proscribed by legislation so
that it is not to be touched by anyone but Congress. If 1 am interpreting
that correctly, that is a terrible situation to be in. Let me just go on to say
that 1 believe the use of the terms "light" and "ultralight" in advertising are
perceived as a claim. 1 think there needs to be a very strong message from
this committee that those are perceived claims, and they carry with it the
impression of improved health, and 1 think that is a form of advertising.

DR. FREEMAN: Dr. Townsend, can you give the committee any sense of
what percentage of cigarettes sold in America are in the generic category as
opposed to the advertised category?

DR. TOWNSEND: No, 1 really cannot. There are some generic products or
low-cost products that are advertised; most, however, are not. 1 cannot give
you an exact percentage right now.

DR. FREEMAN: Thank you. Dr. Cohen?

DR. COHEN: Dr. Townsend, are you asserting that for cigarettes the color of
the package is intended to convey the tar level of the cigarette?

DR. TOWNSEND: In practice, if you look at products that are in the market
currently, in addition to having the category defined as regular, lights, or
ultralights on the pack and in the advertising, in many cases the packs are
different colors. If you look within one brand family, particular brands
within that brand family that are in the different tar categories do have
different pack colors.

227

Smoking and Tobacco Control Monograph No. 7

DR. COHEN: Is there an intention on the part of cigarette manufacturers to
convey information about tar yields by using color on packages, as well as
terms such as "light" and other descriptive adjectives?

DR. TOWNSEND: There is an intention, in my opinion, by the cigarette
industry in general to convey tar information to the consumer so that they
can make choices. In some brands the different colors are intended to
convey the different tar category in which they fall, and that category is
stated explicitly in the advertising.

DR. FREEMAN: I have been informed by staff that approximately 40 percent
of cigarettes sold in America are of the generic category. If that is true, then
I think this is a major issue to be considered here.

DR. KOZLOWSKI: It appears that Dr. Townsend is indicating that the
designations light and ultralight are in a sense being used as surrogates in
a broad sense for tar and nicotine ratings and that they are carrying tar and
nicotine rating information. My question is, are there industry standards or
R.J. Reynolds standards for what numbers are required before a cigarette is
called light or ultralight or is there variance across the industry and in what
products get the label "light"?

DR. TOWNSEND: It is my understanding that the definition, of course, has
changed a little bit over the years, but today the definition is really quite
consistent. Cigarettes under 6 mg constitute ultralights, those from 6 to 15
are lights, and above 15 are regulars.

DR. FREEMAN: Dr. Henningfield?

DR. HENNINGFIELD: I have a recommendation and a question for
Mr. Peeler. The recommendation is that, on discovering that FTC does not
have the means to put on the labeling this information that we are saying
is so important, I recommend that the FTC use all means at its disposal
to get this information and make it readily available to consumers. My
question for Mr. Peeler is, what kinds of things can you do? For example,
could you put your tar report information at all points of sale, or do people
just have to write for the catalog?

MR. PEELER: Let me clarify about the labeling. The place where we are
preempted on cigarette labeling and where everyone is preempted on
cigarette labeling is on statements relating to smoking and health, and there
is a question about exactly where that is. I would think if there were a
misrepresentation, for example, on a label of the tar content, and that is
all that there was, that would be something on which the PTC could take
action. The question of exactly how much information the ITC can
affirmatively require to be disclosed absent a representation by the company
is a whole other issue that we would have to look at in light of the panel's
recommendations. For example, in the 1970's when the PTC required
that health warnings start appearing in advertising, these requirements
were based on allegations that the advertising at that point was making

228

Section II

representations about the cigarette's healthfulness. The FTC's actions
resulted in settlements between the FTC and a number of companies that
provided for the health warnings in advertising, which ultimately became
required by statute in 1984.

The FTC does not have the FDA-type of regulatory power over the
cigarette industry. We have the power to prevent deceptive statements, and
we have the power to require the disclosure of certain types of information
when a failure to disclose that information would be unfair.

DR. HENNINGFIELD: Can you only regulate what appears in a magazine ad?
How do we get it to the consumers who do not read the magazine ads or for
the cigarettes that are not advertised?

MR. PEELER: If you are talking about information about the relationship of
cigarettes to health, then the advertising and the labeling right now contain
warnings, and our authority to require additional warnings or descriptions
would be triggered by what representations are made.

DR. HENNINGFIELD: So you could not require that a label like the one
Dr. Harris showed be put on cigarette packages?

MR. PEELER: Again, what we are here for is to hear the committee's
recommendations and take those back to the five commissioners who
run the agency. I think what you ought to be doing is making those
recommendations that you think are right, and then it will be up to the
five commissioners to sort through them in terms of what is within the
FTC's authority and what is not within the FTC's authority. Clearly the
focus of our concern and the reason that we are here is there has been a
lot of concern that the current tar and nicotine labeling system is not
serving its intended purpose, and because we are putting those numbers
out every year, that is going to be the first thing that we focus on.

DR. FREEMAN: Yes, you have a comment?

MS. WILKENFELD: In the seventies and early eighties when the Commission
published its number, the Office on Smoking and Health made a large chart
that was available at point of purchase in pharmacies and other places where
cigarettes were sold so that there was educational information at the point
of sale. Whether you call that labeling or the Commission would call it
advertising, I do not know. But the money ran out and that stopped.

DR. FREEMAN: I would like to just express a personal concern here. I think
all of us here should be concerned with the effect of a lethal product on the
American public with respect to morbidity and mortality. On the other
hand, this meeting has been called by the FTC along with the Congress. My
concern is that our human concerns do not become engulfed in bureaucratic
problems. The 40 percent of people in America who are smoking cigarettes
that you do not oversee still are smoking cigarettes and still have the same
lethality for those 40 percent. I hope that although we are governed by the
bureaucracy in a certain way, and you have limits, and we certainly respect

229

Smoking and Tobacco Control Monograph No. 7

that — we do have a Congress that can rule one way or another. I think we
should speak to the general problem while we are giving you direction.

MR. PEELER: 1 would certainly agree with that.

DR. FREEMAN: Thank you. Dr. Cohen?

DR. COHEN: Consumers want information that they can use to make
meaningful decisions. Assume the consumer wants to make the meaningful
decision as to whether switching to a particular kind of cigarette, say a 1- to
5-mg tar cigarette, would lead to a significant reduction in health risks. Can
this panel, given the state of the art, attempt to provide information that
would be helpful to the consumer as to the relative risk of smoking different
kinds of cigarettes? If the answer is no, then it is no, but I think that is more
important information than the information currently available through tar
numbers because tar numbers do not tell consumers information that is
meaningful.

DR. FREEMAN: Dr. Benowitz?

DR. BENOWITZ: If you tell people that if they take in less tar, there is a small
benefit, it is not one that should be denied, but I think it is misleading. I do
not think we can tell people that you are going to reduce your hazard; you are
going to live longer if you shift to low-yield cigarettes. So, we are caught in
a bind. We want to encourage people to minimize the risk, but we cannot
really tell them it is going to make a huge difference.

DR. COHEN: I think frankly that this is a subject that ought to command
the attention of the panel because consumers, like it or not, are using these
numbers as if they had absolute significance as numbers. The numbers mean
almost nothing. The panel has to address the question of is there a way of
informing consumers as to the relative risk, and perhaps the answer is to let
us inform them that there is not a lot of gain; doing that, frankly, would be
very useful and maybe more useful than telling them exactly what the tar
differences are.

DR. FREEMAN: Dr. Rickert?

DR. RICKERT: I think the committee has thought about that to a large
extent, and I think there is a consensus that what is really needed is a rather
large public health education campaign to try to communicate that very
information to consumers.

You are also asking for information about risk, and generally this comes
from epidemiological studies, which by their very nature are extremely
long and do not provide information that we can use immediately. The
information that we have today has been gleaned over a number of years
with respect to relative risk, and as we have seen, there is not much change.

Now, if one is going to ask what is the effect with today's cigarettes, then
we are talking about a period of time that will be measured in tens of years.

2'M)

Section II

Unfortunately, we have to come to grips with the problem today, and we
cannot wait for tens of years to pass to obtain accurate information about
today's cigarettes.

DR. FREEMAN: Dr. Hughes?

DR. HUGHES: 1 would like to respond very concretely to what you said.

Dr. Cohen, and it is my opinion that if there is not information that you
suggest about the health benefits of switching, the whole system should be
junked. My concern is that we are going to say, "Yes, there should be this
big public health campaign," and nothing will happen. NCI says, "We don't
have money for it." FTC says, "We don't have the bureaucracy to do it."

The FDA says, "We have other more important things to do," and this whole
education campaign does not get done. We come up with a range of values
that still has numbers on them, and people still think that they are doing
themselves a big benefit, and 1 would rather junk the entire system than
to have that happen.

The only way around it 1 can see is that the FTC decides that all
claims of light and ultralight imply a health claim and therefore require
a disclaimer. That is the only way out of it 1 can see, to make sure that
that happens.

DR. FREEMAN: What form of disclaimer?

DR. HUGHES: I do not want to micromanage with the wording. Whether
we say, "You may get a small benefit" or "You will get very little benefit
compared to stopping smoking," is a tough question, and 1 do not think
we need to decide on that wording. My point, again, is to reiterate what
Dr. Cohen said, which is that the system is bankrupt unless there is some
statement about the magnitude of health benefit that you will receive by
switching to a low-tar and -nicotine cigarette.

DR. FREEMAN: Dr. Stitzer?

DR. STITZER: I want to return to a point I made earlier. It seems to me
that we would be doing a great service if we could implement a new testing
technique that involved ranges that could display and convince smokers that
light cigarettes are the same or can be exactly the same as a regular cigarette.

I think the data show us that all the cigarettes from .4 mg up can look exactly
the same. They basically are occupying the same place in space. The range
of variability is the same, and that there is no health benefit for switching to
light cigarettes because of this dose variation, but there are also some data
suggesting that the ultralight cigarettes, those .1 mg and below, do produce
a different level of exposure.

Now, those cigarettes are not popular. They capture a very, very tiny
segment of the market, but they may make a difference. We do not have
the health data, but if there is a dose effect for health, those are the only
cigarettes that are going to make a difference, and it seems to me that a
new labeling system could potentially convey that kind of information.

231

Smoking and Tobacco Control Monograph No. 7

DR. FREEMAN: To play devil's advocate, there seems to be a point on the
other side that, if the American public becomes confused about the value of
low dose vs. high dose, would it defeat the purpose of encouraging people
to switch, assuming there is a benefit to low-dose cigarettes? 1 think I am
hearing those two arguments. Yes, Dr. Giovino?

DR. GIOVINO: 1 think part of any disclaimer that would be given if you
decided to do that would include the statement that I read earlier from
the Surgeon General's report that any cigarette smoking is dangerous, that
quitting is absolutely the best thing a person could do to protect his or her
health, and that reducing to these brands "may." And that is exactly what
the Surgeon General's report says, "May pose reduced risk, provided that
no compensation occurs." 1 think those two caveats are absolutely essential,
that quitting is better than switching and that provided no compensation
there may be reduced risk.

DR. FREEMAN: Dr. Benowitz?

DR. BENOWITZ: I think there is a second function that the FTC testing does
perform, and that is to mold what the tobacco companies provide. I do
think that there has been a reduction in lung cancer if you compare 1950's
cigarettes to modern cigarettes, and I would not want to lose that pressure
to keep yields as low as possible. I think whatever we do, we do not want
to lose that by saying that it does not matter at all. The other argument
is if there is a 10-percent reduction of health hazard — which would be very
difficult to measure by epidemiological means — if you are applying it to
about 40 million smokers, that can be substantial. And I would not want
to lose that for the population either.

1 do not want to be misleading. I certainly appreciate Dr. Hughes' point
of view, but 1 think we somehow should not let things slide the other way.

DR. FREEMAN: Is there any way that you can bring the two points of view
together? I think this is a very important point. We need to settle it here.

DR. STITZER: There is just one other thing. The only way that this
information can be relevant to the individual consumer is if there is a way
for that individual to judge where he or she falls along the dose continuum.
Now, that could be accomplished with a lovely sophisticated method like
Dr. Rickert has described with the color coding. I do not know whether that
technology is sufficiently available to incorporate into our recommendations,
but it could be part of our recommendation that we try to develop a system
that allows the individual smoker to know where they fall on the exposure
and dose continuum.

DR. FREEMAN: Yes, Dr. Hughes?

DR. HUGHES: I disagree a little bit with you. Dr. Stitzer. F,ven if people
could tell exactly where they were on the dose continuum, that does not
solve the problem of the dose response-health benefits curve being so
shallow. Alsr), 1 want to respond to your comment. 1 am not saying that

2.f2

Section II

there is no benefit, and I do not think we should say that there is no benefit
because I agree the public health argument is there. But the physician in
me says, "Always oversell your case," because people do not change very
much. Again, I do not care what kind of disclaimer or how it is worded,
but there has to be, again, something about health benefits to the consumer
in all of this.

DR. FREEMAN; Dr. Shiftman?

DR. SHIEFMAN; I think the two issues we have struggled with most recently
are related. While the system as now constituted has its focus on advertising
and on showing a single number for an FTC measure of tar and nicotine
values, in fact there are implicit claims being made both in advertising
proper and in brand names, which are a form of advertising that imply
a health claim. Therefore, it seems to me we ought very strongly to
recommend both that the use of those terms, like "light" and "ultralight,"
be regulated and that when they are used, they be accompanied in fair
balance by a disclosure of the sort that Dr. Hughes has suggested.

I think there is a middle ground that allows us to proceed based on what
we know and based, I think, on regulatory authority that already exists.

DR. FREEMAN: Yes, Dr. Zacny?

DR. ZACNY: Based on Dr. Rickert's point about state-of-the-art
epidemiological studies not being done at this time where you can say with
any degree of certainty what the relationship is between nicotine dose and
risk of disease with the brand of cigarettes we are dealing with now — the low-
yield cigarettes and the ultralow-yield cigarettes — I disagree slightly with
Dr. Hughes when he says that the relationship may be very shallow. As
scientists, in any claims we make we can just say that we do not know at
this time what the relationship is because it takes 10 or 20 years, but based
on what we know, it would be best not to block vent holes, to take smaller
puffs, etc.

DR. EREEMAN: That would be an educational campaign?

DR. ZACNY: Yes. Maybe I am wrong, but it seems that there may be a dose-
response relationship between risk of lung cancer and how much smoke
people take into their system, and they may realize a substantial benefit with
the ultralow-yield cigarette. I do not think we know with certainty, and in
the absence of that, I think the formulations that we are putting forth with
bands and ranges are a good idea.

DR. EREEMAN: Dr. Kozlowski?

DR. KOZLOWSKI: I would like to draw an analogy to the FDA nutritional
labeling. For a lot of the items on those labels, there is not persuasive
epidemiological research to show the dose-response curves for a lot of the
things that are listed as of interest. I think we do not want to be held to a
higher standard. Epidemiology takes time. It has its limitations, and the

233

Smoking and Tobacco Control Monograph No. 7

basic point is there is a lot of labeling that pertains to risks that are only
approximately known.

DR. FREEMAN: Dr. Benowitz?

DR. BENOWITZ: To follow up on Dr. Zacny's comments, what we really
would like to know is not brands vs. risk; we would like to know actual
exposure level. If we were able to measure cotinine or adducts of different
compounds or whatever in smokers vs. their yields, then we would have the
basis for recommending that individuals should reduce their exposure. Since
there is such an overlap with the yields as marketed now, 1 do not think we
are ever going to see a difference by yield — that does not mean that the
rationale for an individual reducing their intake is not valid, and so at this
point in time we may have to go forward based on scientific rationale and
plausibility for reducing exposure to toxic materials.

DR. FREEMAN: Dr. Hoffmann?

DR. HOFFMANN: 1 think there is a misunderstanding. It has been shown
in dozens of studies that there is a dose response with respect to cancer of
the lungs and the upper respiratory tract by number of cigarettes smoked
per day, length of time smoking, and groups of cigarettes. This continues
for ultralow, low, and average cigarettes. There is a dose response with
respect to cancer but not with respect to coronary artery disease. Any
cigarette is harmful. But ultralow cigarettes have a lower risk than
nonfiltered regular cigarettes when you smoke them for 10 or more years.

We should not say that there is no dose response.

DR. FREEMAN: Dr. Hoffmann, just to follow up on what you said, as far as
the science is concerned, you indicated that in there is no apparent dose
response for coronary heart disease?

DR. HOFFMANN: We do not know from the literature any benefit with
respect to coronary heart disease.

DR. STITZER: Dr. Benowitz can speak to this. Is there a dose effect based
on light vs. heavy smoking for coronary disease?

DR. BENOWITZ: No, there is not. I would like to add that 1 agree with
Dr. Hoffmann that this dose response would provide a rationale for what
we are doing here today. There are data in pregnancy showing the dose
response between cotinine level and the weight reduction of the newborn,
and therefore that is another rationale for another disease that there is a
dose-response relationship, and therefore, even though we cannot say that a
particular brand is going to be less hazardous than some other brand, we can
say that lowering your exposure in general will be beneficial, and then we
just have to help people to do that.

DR. FREEMAN: Didn't Dr. Samet state yesterday that there is no evidence
that coronary heart disease is reduced by lowering the nicotine or tar content
in cigarettes?

2.14

Section II

DR. BENOWITZ: He said that there was no evidence.

DR. HOFFMANN: That may be the case, but that we have no evidence.

DR. FREEMAN: I think we should respect those data because I think more
than 100,000 people die each year from cardiac death due to smoking. 1 do
not know the exact numbers, and that is a very significant point to argue
that even smoking low-nicotine, low-tar cigarettes will not protect you as far
as we know from dying from coronary heart disease. Is that a fair statement.
Dr. Hoffmann?

DR. HOFFMANN: No, I think we do not know.

DR. FREEMAN: Let me narrow down the point. Are you saying that there is
no distinction between the low-tar and -nicotine smokers and the high with
respect to death from coronary disease?

DR. HOFFMANN: There is no evidence.

DR. FREEMAN: No evidence that there is a difference?

DR. HOFFMANN: No evidence that 1 am aware of. We discussed it last
night, and nobody came up with any, but there may be. 1 am not aware of it.

DR. PETITTI: Since we are relying so heavily on the 1981 Surgeon General's
report, and very few data have come out since then, 1 would like to read the
statement specifically about cardiovascular disease:

The overall changes in the composition of cigarettes that
occurred during the last 10 or 15 years have not produced
a clearly demonstrated effect on cardiovascular disease, and
some studies suggest that a decreased risk of CHD may not have
occurred. Evidence on the association between CHD and filter
cigarettes is somewhat conflicting. One major study showed a
reduction of 10 to 20 percent of coronary deaths among persons
smoking lower tar and nicotine cigarettes as compared with
those smoking higher yield cigarettes.

That was the CPS-1 study, but other surveys have shown a slightly
increased risk of coronary mortality in people who smoke filter cigarettes
or those who smoke nonfilter cigarettes. Recent unpublished data from the
Framingham study that were ultimately published do not show a lower CHD
risk among smokers of filter cigarettes. It is not that there are no data. The
data that exist show no association of smoking lower yield cigarettes with
reduction in coronary heart disease morbidity and mortality.

DR. FREEMAN: 1 think that is a very, very important point that ought to be
factored into this discussion. Dr. Cohen?

DR. COHEN: 1 was very pleased. Dr. Hoffmann, to hear you clarify an issue
for me. 1 heard you say that we can inform people about lung cancer, at
least. Now, lung cancer is a major issue for people, and what 1 am wondering
is, since people use tar as a surrogate for relative harmfulness, might it make

235

Smoking and Tobacco Control Monograph No. 7

some sense to restrict their usage of tar by maybe getting rid of tar and
talking about cancer-causing compounds or some other component and let
these numbers communicate meaningful information about something we
do know something about, rather than where we do not know anything
about it that is conclusive? Would it be possible, in other words, to indicate
people's relative risk with respect to lung cancer for smoking these different
yield cigarettes?

DR. HOFFMANN: As 1 see it, the public associates tar with cancer.

DR. COHEN: 1 think they do it more broadly and associate it with overall
safety.

DR. HOFFMANN: They are not well informed. The American Cancer Society
did a fantastic job, publishing this over and over again, and the public is well
informed with respect to smoking and lung cancer. There has not been the
same level of information communicated about coronary risk.

DR. COHEN: In your view though, is it possible to relate the differences in
tar yield in cigarettes to a reduction in cancer risk? If the answer is yes, it
seems to me that one of the things the panel might consider doing is trying
to convey that or recommending that be conveyed.

DR. FREEMAN: Dr. Rickert?

DR. RICKERT: 1 would be somewhat reluctant to do that for several reasons.
First of all, the information with respect to lung cancer and risk reduction is
from the 1981 Surgeon General's report that relates to cigarettes that were
consumed 10 to 20 years prior to that date. My other concern is in terms of
"tar is tar is tar," that is, is the quality of the tar from today's ultralow-yield
cigarette the same as the quality or the carcinogenic potential of tar from a
cigarette from many years ago? 1 do not think we know that the tar of
today's cigarette is the same as the tar of the cigarette many years ago that
was being related to lung cancer.

DR. HOFFMANN: With respect to your first point. Dr. Rickert, there was
an International Agency for Research on Cancer monograph in 1986 that
presented 15 studies that all show between a 13- and 30-percent reduction.

I here are also reports from the World Health Organization, and there is the
Fluropean study by the National Cancer Institute. It did not end with 1981.

DR. RICKER F: i am not uncomfortable with the idea that there is a risk
reduction. What 1 am suggesting is that we really do not have enough
information about the tar characteristics of today's cigarettes to directly
compare them.

DR. HOFFMANN: You can do only risk biomarkers. Otherwise you cannot
do it.

DR. RICKERT: 1 agree.

DR. FREEMAN: Yes, Dr. Hoffmann.

236

Section II

DR. HOFFMANN: I believe that we should do better education. 1 do not
think that including whether it is plus or minus 3 mg or 5 mg helps the
public. The public has to be much better informed.

DR. FREEMAN: It appears from what I have heard here on the question of
does the FTC protocol provide information useful to consumers in making
decisions about their health, that the answer seems to be not sufficiently so,
and we have had an interesting discussion and debate here this morning
concerning various aspects of that decision.

Things that stand out in my mind are what do the numbers mean to
the public? Dr. Cohen has been eloquent in raising that issue. What is the
value of projecting a range per type of cigarette to the public? What is the
value of any kind of color coding to the public? What would be the value
of presenting a graph to the public so a person could see by graph form what
the differences in cigarettes are? We also have heard this morning, unknown
to me before, that approximately 40 percent of cigarettes are not looked at
by the Federal Trade Commission because they are not advertised.

MR. PEELER: They are looked at, and they are tested. The results are
reported, but unless the manufacturer either voluntarily puts that
information on the label or unless they advertise, those numbers aren't
necessarily communicated directly to consumers.

DR. EREEMAN: Thank you for that correction. The bottom line is the public
does not know those numbers.

237

Section III

Recommendations and Findings

DR. FREEMAN; Good afternoon, I am Dr. Harold Freeman. I am the
chairman of the President's Cancer Panel. At the request of the Congress
and the Federal Trade Commission, an ad hoc committee of the President's
Cancer Panel has met over the last 2 days to consider the Federal Trade
Commission test method for determining tar, nicotine, and carbon
monoxide levels in cigarettes. Before 1 get into our statement, 1 would like
to put the problem of tobacco into perspective. Tobacco use is the number
one cause of preventable death in America. Cigarette smoking is responsible
for more than 400,000 premature deaths every year in this country and
causes one-third of cancer deaths and one-third of heart disease deaths.

Although smoking is declining among adults in the United States, it is
discouraging that smoking is not declining among children, and in fact,
smoking prevalence among adolescents has changed little for more than
a decade.

1 have with me Dr. Saul Shiftman of the Department of Psychology,
University of Pittsburgh; Dr. Diana Petitti, director of the Division of
Research and Evaluation, Kaiser Permanente; and Dr. William Rickert of
Labstat, Inc.

This committee reviewed articles, studies, and other documents and
heard presentations from a variety of experts, including tobacco industry
scientists, on the subject of the FTC test method for determining tar,
nicotine, and carbon monoxide levels in U.S. cigarettes. We have deliberated
with the goals of answering questions and making recommendations. Our
deliberations centered around the following three summary questions:

1. Does the evidence presented clearly demonstrate that
changes are needed in the current FTC protocol for
measuring tar, nicotine, and carhon monoxide? If yes,
what changes are required?

2. Should constituents other than tar, nicotine, and carhon
monoxide he added to the protocol?

3. Does the FTC protocol provide information useful to
smokers in making decisions ahout their health?

1. The committee reached the following conclusions with respect to the

first question.

A. The smoking of cigarettes with lower machine-measured yields has
a small effect in reducing the risk of cancer caused by smoking, no
effect on the risk of cardiovascular diseases, and an uncertain effect
on the risk of pulmonary disease. A reduction in machine-measured

239

Smoking and Tobacco Control Monograph No. 7

tar yield from 15 mg tar to 1 mg tar does not reduce relative risk from
15 to 1.

B. The FTC test protocol was based on cursory observations of human
smoking behavior. Actual human smoking behavior is characterized
by wide variations in smoking patterns, which result in wide
variations in tar and nicotine exposure. Smokers who switch to lower
tar and nicotine cigarettes frequently change their smoking behavior,
which may negate potential health benefits.

C. Accordingly, the committee recommends the following changes to
the FTC protocol:

1. This system should also measure and publish information on
the range of tar, nicotine, and carbon monoxide yields that most
smokers should expect from each cigarette sold in the United
States.

2. This information should be clearly communicated to smokers.

3. A simple graphic representation should be provided with
each pack of cigarettes sold in the United States and in all
advertisements. The representation should not imply a one-to-
one relationship between measurements and disease risk.

4. The system must be accompanied by public education to make
smokers aware that individual exposure depends on how the
cigarette is smoked and that the benefits of switching to lower
yield cigarettes are small compared with quitting.

D. There should be Federal oversight of cigarette testing, but such testing
should continue to be performed by the tobacco industry and at
industry expense.

E. The questions involved in the purpose, methodology, and utility
of the FTC protocol are complex medical and scientific issues that
require the ongoing involvement of Federal health agencies,
including the National Institutes of Health, Food and Drug
Administration, and Centers for Disease Control and Prevention.

I’. The system should be reexamined at least every 5 years to evaluate
whether the protocol is maintaining its utility to the smoker.

G. When a cigarette manufacturer makes significant changes in cigarette
design that affect yields, it should notify the appropriate Federal
agency.

11. With regard to the second question, the committee recommends that
to avoid confusing smokers, no smoke constituents other than tar,
nicotine, and carbon monoxide be measured and published at the
j)resent time. Smokers should be informed of the presence of other
liazardous smoke constituents with each package and with all

240

Section III

advertisements. These constituents should be classified by toxic
effects.

111. In considering the third question, the committee reached the following
conclusions:

A. Information from the testing system is useless to smokers unless
they have ready access to it. The information from the testing system
should be made available to all smokers, including those who smoke
generic brands and other brands not widely advertised.

B. Brand names and brand classifications such as "light" and
"ultralight" represent health claims and should be regulated and
accompanied, in fair balance, with an appropriate disclaimer.

C. The available data suggest that smokers misunderstand the FTC test
data. This underscores the need for an extensive public education
effort.

1 would like to underscore two major points: First, the health benefits
of switching to low-tar and -nicotine cigarettes are minimal compared to
quitting entirely, and finally, in effect, how you smoke is much more
important than what you smoke.

We have deliberated for 2 days. We believe these findings are very
important to the health of the American public. We are dealing with a
product that is lethal, that needs to be controlled, and we believe that
these recommendations will lead to some control. 1 would open it up
for questions to my colleagues or to myself.

PARTICIPANT: Dr. Freeman, what do you expect to be the next step in the
educational process for consumers?

DR. FREEMAN: The findings from the deliberations of this committee will
be reported to the Director of the National Cancer Institute, who will then
formulate a report that will be passed on with the help of the President's
Cancer Panel to the appropriate agencies and the Congress.

PARTICIPANT: That is a lot of reporting. Can you predict what might
happen next?

DR. FREEMAN: I do not think we can predict what is going to happen in the
future, but our hope is that since the FTC methodology has been in effect
from 1967 and was based on findings that relate to 1936, and since in the
last 25 years there has been a considerable change in our knowledge through
research, as well as in the type of cigarettes that are being smoked, we now
believe that these changes are very essential and should be put into effect
very soon.

No one can predict because we are dealing with the FTC, possibly
other agencies of the Government, and the Congress, and no one on this
committee can predict how rapidly these changes may take place, but we
believe they are very important.

241

Smoking and Tobacco Control Monograph No. 7

PARTICIPANT: Are you saying that you are recommending keeping the
current FTC testing method and expanding it in some way or are you talking
about a whole new testing system?

DR. FREEMAN: Let me reemphasize that we are recommending the keeping
of the basic parts of the FTC testing methodology with the exception that
we want to expand testing to show the ranges of possible effects of the three
substances that are being measured. The reason that we believe this is
important is that the research has shown that people who smoke cigarettes
that, for example, are labeled as having low tar can get a much higher dose
from that cigarette than the label may indicate. For example, if you have a
low-yield cigarette, the way you smoke it, the rapidity of the puffing, the
depth of the puffing, whether you block the ventilation holes, etc., can have
an extraordinary effect on the real dose to the patient. Disease, we believe,
is related to the dose of carcinogens and other toxins.

PARTICIPANT: The impetus for this effort came from Congressman
Waxman, a Democrat in the Congress. Now the Congress is primarily
Republican. What effect do you think this is going to have on your
recommendations?

DR. FREEMAN: It is conceivable that people in power who have
philosophies that are different from Congressman Waxman's could present
barriers to our recommendations. We are hopeful though that even with
these changes that the logic of what we are saying will make sense even to
people who may disagree with what we are recommending in principle.
There are people, for example, who may wish to diminish the fight against
tobacco, and I am sure you are referring to them. I am hopeful that even
such people will listen to the logic of reporting to the American public the
truth of a finding that is responsible for 400,000 deaths a year and give the
public the chance of making an intelligent decision. We are not saying,
"Eliminate cigarettes." We are not saying, "Stop using the methodology that
has been present for 25 years." We are saying, "Give an honest report to the
American public and show them the range of the risk that they are subjected
to." 1 hope that everybody. Democrat, Republican, conservative, or liberal,
will follow that logic.

PAR'nCIFAN'F: You are suggesting, in addition, putting the CO on cigarettes
and also putting other ingredients?

DR. FREEMAN: One of the recommendations that I read to you indicates
that we believe that in addition to putting the ranges of the tar, nicotine,
and carbon monoxide that are now being measured with one number, we
want to change that to a range because that is a more truthful statement.

Fhis committee is also recommending that certain key harmful substances
known to be in cigarettes (we are not saying which ones should be listed)
should be given as information to everyone who buys a package of cigarettes.
We believe that if this is done in food, which does not apparently have the
toxic effect of tobacco, then we believe it should be done in this lethal
[)r(iduct.

242

Section III

PARTICIPANT: Could you explain the graphics that you would put on the
package?

DR. FREEMAN: I am going to refer this question to Dr. Rickert.

DR. RICKERT: There are a number of different ways of looking at that
particular problem. The graphics could involve a number of different issues;
for example, it could involve a color representation of the cigarette filter.

It could represent some icon that illustrates putting all of this information
together. A number of different possibilities were discussed, and I do not
think that the committee recommended any specific procedure. I think
the feeling was that there should be some way of communicating the
information to smokers without total reliance on numbers themselves.

PARTICIPANT: What would be the purpose? I do not understand the
purpose of the graphics overall.

DR. RICKERT: The graphics would make several points. First of all, the
point that yields to smokers depend on how the cigarette is used; that is,
if you have a graphic, it gets away from the idea that there is a fixed amount
of whatever the constituent happens to be. The purpose of the graphic is
to illustrate the variable nature of the smoking characteristics.

DR. SHIFFMAN: If I may add, we thought it was very important to
communicate to American smokers that what you get depends on how
you smoke and that any system that simply gives one number is, therefore,
inherently misleading. So, we envisioned a graph that would show you
a band within which your particular exposure might lie and that will
give smokers information on which they can make more accurate, more
reasonable comparisons among brands. We think they will find that there
is a good deal of overlap among brands that they now consider to be
different.

PARTICIPANT: You said "light" and "ultralight." Some people say that
those words represent health claims. Could you explain a little bit more
about that? How does that represent a health claim, and what kind of
disclaimer would be used?

DR. FREEMAN: It is the committee's belief that the public infers health
claim meanings from these labels, whether they be light, ultralight or
whether they be the numbers in tar and nicotine. It is anecdotal, and also
studied, that people look at these numbers and these claims and translate
them into what it means for their own destiny. The information gathered at
this meeting indicates that smokers should not be making these predictions,
first of all, and second, if the labeling by the cigarette industry of ultralight
implies that you are better off according to health, if that is so, and we
believe that this is so, then that represents a health claim on the part of the
advertiser. If it is a health claim, it should be followed by a disclaimer saying
that it is not a health claim, if it is inferred to be a health claim.

243

Smoking and Tobacco Control Monograph No. 7

DR. SHIFFMAN: I think, again, specifically we want to be sure that the
smoker understands that smoking a cigarette that is labeled as light or
ultralight does not necessarily protect them from the health risks of smoking
and that, in that sense, cutting down in this way may not keep people from
being cut down eventually by their smoking habit. We do think that the
public perceives those labels as implicit health claims.

DR. FREEMAN: It is even conceivable that a low-tar cigarette smoked in a
certain way may have the same health risk as a regular cigarette, and we have
pointed out in what I have already said that there is no scientific evidence
that any level of tar in cigarettes protects one against death due to coronary
heart disease.

PARTICIPANT: The other substances that you referred to, are you going to
talk about numbers?

DR. RICKERT: I think that what the committee felt in that area was that at
the present time, since there is evidence that consumers tend to misinterpret
the existing numbers, that to add additional numbers may add to that
confusion. At the same time there was the concern that there are additional
agents, other than tar, nicotine, and carbon monoxide, that have definite
implications for health. It was anticipated that these compounds would be
classified in various ways, for example, "carcinogens," and then there may
be a list of several carcinogens. There would be a list based on toxicological
effects but not including any numerical measurement.

PARTICIPANT: Can the machinery that is currently used to test cigarettes be
used?

DR. FREEMAN: We were told by an expert today that there may be some
fine tuning that will be necessary to use the current equipment to do this
kind of testing.

PARTICIPAN'F: Who would determine what that range was and how many
times the machine smoked or how long the puffs?

DR. FREEMAN: This committee did not go into that kind of detail. We are
talking about the principle, and the principle is that we know that human
smokers smoke in different patterns. Some smokers puff many times in a
minute, and some smokers may puff once a minute. Some smokers puff
deeply, and there are other factors that 1 could mention. While we are not
trying to micromanage how this should be done, the principle is that we
would like the machine measurement to more closely mimic the variation
that humans evidence in their patterns of smoking to give a more honest
range of what a given milligram of tar really represents in range. We do not
believe it is accurate at all; in fact, it is misleading to give one number when
the pattern of smoking can change that number radically with respect to
dose.

DR. SHIFFMAN: What the [)anel intended was that the range represent the
range of human smoking of |)articular brands so that the machine would

244

Section III

model that under different parameters, which might include things that are
now not dealt with in the FTC protocol, such as the blocking of ventilation
holes that are used to dilute the smoke in some brands that now list as being
low yield, but in fact can become high yield when a human finger or a
human lip blocks those vents.

PARTICIPANT: Can you tell us what the role of other Federal agencies is
going to be?

DR. FREEMAN: 1 am not an expert on the bureaucracy of America. However,
we did get somewhat of a description of the FTC role in our meeting here
today, which is a role that I understand deals with truth in advertising as
one of its major roles. And to make a personal statement here, I think that
is a limited role with respect to what we are trying to accomplish for the
American public.

We found out today that 40 percent of cigarettes smoked in America are
generics, and these for the most part are not advertised. However, the FTC
in most of its role is limited to making statements about cigarettes that are
advertised. So that if nearly half the cigarettes smoked in America are not
advertised, it diminishes the FTC's role. Yet, the American public needs to
know about the lethal nature of all cigarettes.

Now, as far as the FDA is concerned, again, I am not an expert on what
they do, but I think their role is different from the FTC and may get more
into the range of health concerns, hopefully. So, I cannot give you a finite
answer. Perhaps my colleagues can help me out.

DR. SHIFFMAN: I would just add that the current FTC system operates
under a voluntary agreement with the tobacco industry and cigarette
manufacturers, and the representatives of that industry who addressed us
during this meeting expressed an interest on the part of the industry of
keeping consumers and smokers informed. We expect that they would
follow through on that then in taking this step to make sure that accurate,
useful information is available to smokers.

PARTICIPANT: Is it your understanding that if the regulatory agencies
wanted to do this, that legislation would be necessary?

DR. FREEMAN: To do exactly what?

PARTICIPANT: To carry out your recommendations?

DR. FREEMAN: It is our belief that most of what we have recommended
could be carried out by the FTC without congressional change. Our worry
is that 40 percent of cigarettes are not regulated in a similar manner. Our
concern is about the health of the American public and that the bureaucracy
that we must go through to accomplish some of these things sometimes
is a barrier to that. The FTC has regulations; the FDA has regulations, but
sometimes what must be done or what should be done to save lives is
beyond the confines of a certain agency, and this is somewhat of a problem.

245

Smoking and Tobacco Control Monograph No. 7

PARTICIPANT: You said that this information would be useless unless
smokers had ready access to information, including smokers of generic
brands. In view of what you said about the FTC's jurisdiction, how do
you anticipate getting that?

DR. FREEMAN: This came up very honestly today. We have not had
time to think in depth about it. This was probably a surprise, even to this
committee, that that problem is so large, that 40 percent of cigarettes are of
the generic type, and honestly I do not have a good answer to that question.
It may be that the FDA and other agencies could help in some respects, but
I will refer this question to my colleagues to see if there is an answer to that.

DR. PETITTI: I could only repeat, I think, what we heard this morning,
which is that some of these changes might, in fact, require congressional
action, particularly if they resulted in changes in the labeling law. We are
saying that there may be the need to put things on cigarette packs in order
to adequately inform the American public about the FTC protocol.

PARTICIPANT: Looking ahead to the next 5-year review, do you see any gaps
in research areas that need to be addressed?

DR. FREEMAN: Yes, we do. First of all, we have this paper. Where we are
now with respect to our current knowledge, of course, is based on research,
not perfect research, but we know a lot more now than we knew in 1950,
when Ernst Wynder and others showed that tobacco is associated with
death. So, research is a critical element and at any time I think we must act
on what we know, but we must always move forward to finer knowledge.

For example, further research is needed to determine the extent to which
smokers of lower tar and nicotine cigarettes are less likely to attempt to quit
smoking. There is some preliminary evidence, for example, that low-tar
smokers may have less tendency to quit smoking. This would be very bad
if it turns out to be true. It needs further research.

Next, to adequately understand and evaluate the impact of what
is called compensation, research is required to assess the extent to which
other biomarkers are correlated with machine-measured yields of the same
substance. By compensation we refer to the point that low-tar smokers
frequently smoke more cigarettes apparently to get the physiological dose
of nicotine, which of course is an addictive substance. Compensation needs
to be studied further to see what effect it may have, and certain biomarkers
may come in handy to help us. Third, the differences in smoking patterns
in different ethnic groups should be studied for the implications for health
education and consumer information. We know, for example, that African-
Americans tend to smoke cigarettes that are higher in tar and tend to smoke
mentholated cigarettes. Other examples could be given. Poor Americans
tend to smoke higher tar cigarettes. Educated Americans tend to smoke
lower tar cigarettes. These are all very important questions that we only
have preliminary information on, and these things need to be studied much
more deeply. Finally, a system should be developed to help smokers gauge

246

Section III

where their individual smoking behavior places them on a dose continuum.
What diseases you develop, whether it be cancer or anything else, is often
associated with the dose that you receive, and individuals need to know
what dose they are receiving. There may be other research questions, and
I will open it up for Dr. Rickert and Dr. Shiffman to comment.

DR. SHIFFMAN: I would add only that in addition to refining our knowledge
in these areas that there may be some very different products for smokers on
the horizon. We heard some indication of those in the press, and the system
would have to be very carefully considered in order to properly evaluate new
kinds of products aimed at smokers.

DR. FREEMAN: I think using research in a different way, we need to better
understand the way the people in power deal with tobacco in America. It is
a substance that is high in the economy.

If cigarettes were invented today, they probably would be outlawed since
they kill 400,000 people a year. However, it is deeply integrated into our
economy. It affects policymaking. Sometimes there is a conflict, in my
opinion, between making regulations and trying to balance the budget.

America in one of its Government roles is saying that tobacco kills
400,000 American people. Other parts of Government are selling it overseas
and growing it in America. These are deep problems. They require further
research and knowledge and action.

Are there other questions?

If not, I would like to conclude by expressing my privilege of chairing
this committee. We brought together the best experts in America on the
subject. Dr. Dietrich Hoffmann, for example, is one of the pioneers in the
study of tobacco, and there were others, and it is a privilege to chair this
committee. It is our hope that these deliberations will have an effect on the
American public with respect to saving lives and preventing disease.

Thank you very much.

247

I

I

Section IV

Overview of 1980 to 1994 Research

Related to the Standard Federal Xrade
Commission Xest IMethod for Cigarettes

Michael D. Mueller

INTRODUCTION This chapter provides an overview of the major studies related
to the Federal Trade Commission (FTC) test method for determining tar,
nicotine, and carbon monoxide (CO) yields of cigarettes compared with
yields experienced by smokers, with special reference to low-tar and low-
nicotine cigarettes. Most of the studies reviewed here were published since
1980; studies published prior to 1980 were extensively reviewed in the 1981
Surgeon General's report (U.S. Department of Health and Human Services,
1981).

The apparent differences between stated yields, as measured by the
FTC test method, found in cigarette advertising and on some cigarette packs
and actual amounts received by smokers appear to be largely attributable to
compensation behaviors related to nicotine and possibly other substances
in cigarette smoke. For example, when smokers switch to low-tar and low-
nicotine cigarettes, they tend to increase the volume of inhaled smoke per
cigarette or increase the number of cigarettes smoked so as to maintain a
steady-state level of nicotine in their blood. They may also increase the
volume by changing their puffing behavior and increase yield by blocking
ventilation holes in filters.

Changes in puffing patterns can substantially alter tar and nicotine
yields, as reported by Rickert and colleagues (1983), who investigated the
impact of varying levels of butt length, puff duration, puff interval, puff
volume, and blocking of ventilation holes.

The differences in advertised tar and nicotine yields of cigarettes
compared with the amounts received by smokers result largely from
differences between the smoking parameters of the FTC test method and
actual smoking behaviors. These differences can substantially alter the
amounts of tobacco smoke constituents that smokers inhale. The FTC
method was devised in 1967, and it is not clear whether these parameters
were based on actual human smoking patterns and behavior. Furthermore,
cigarettes have undergone substantial changes in design and content over
the past 40 years. Also, much more is currently known about smoking
behavior; pharmacodynamics and pharmacokinetics; and the measurement
of tar, nicotine, CO, and other substances in cigarette smoke as well as in
blood, plasma, urine, and expired air in smokers.

249

Smoking and Tobacco Control Monograph No. 7

Rickert and Robinson (1981, p. 401) emphasize that

even if compensation [changes in smoking patterns to increase
smoke intake per cigarette] did not occur, it is likely that
smoking machine parameters fixed about 20 [years] ago no
longer represent the average smoker, who probably takes
puffs of more than 45 mL every 40 s instead of a 35-mL
puff every 58 s.

There are harmful substances in tobacco and tobacco smoke other than
tar, nicotine, and CO. These include hydrogen cyanide (HCN), acrolein, total
aldehyde, and tobacco-specific nitrosamines (TSNAs). Levels of some of these
harmful substances in low-tar and low-nicotine cigarettes probably differ
among brands and may also differ within brands when cigarettes are smoked
differently.

Smoking patterns may be influenced by factors other than nicotine
dependence. Pomerleau and Pomerleau (1984) pointed out that there is
substantial evidence that many cigarettes are smoked for reasons other than
to receive nicotine. They cite research indicating that smoking patterns are
influenced in part by environmental situations, emotions, personality, and
motivation.

Robinson and coworkers (1983) found that smoking compensation
behaviors may lead to disproportionate increases in CO and HCN when
smokers switch to low-nicotine cigarettes.

Thus, research over the past 15 years has created multiple arenas within
which scientists and policymakers may reexamine the accuracy and relevance
of the FTC testing method and, if necessary, redesign it.

PARAMETERS OF
THE FTC TEST
METHOD AND
CURRENT SMOKING
PATTERNS

The current FTC test method is based on four parameters:
puff frequency (every 60 seconds), puff volume (35 mL),
puff duration (2 seconds), and a butt length that varies with
cigarette type. Darrall (1988) noted that these parameters
were set as long ago as 1936 and were not based on observed
smoking patterns. For individual smokers, puff volume has been reported to
range from 23 mL to 60 mL; puff duration is known to vary from 0.8 seconds
to 3.0 seconds. Typically, butt length is set at 23 mm, or filter and overwrap
plus 3 mm, whichever is longer; however, the FTC reported that, for 135 of
176 brands tested, butt length was more than 30 mm (Kozlowski, 1981).

Cigarette design has undergone significant change over the past several
decades. Cigarette manufacturers can influence yields of tar, nicotine, and
other substances through changes in wrapping paper porosity; tobacco
packing density; and filter-related factors such as ventilation, particulate
matter retention, and pressure drop. Benowitz and colleagues (1983) noted
that delivery of tobacco substances also may be influenced by how fast the
paper burns because this may determine how long a cigarette is smoked.
Study results indicate substantial differences in yields when I'l'C test method
[)arameters are varied ( Fable 1 ).

250

Table 1

Variation of puffs and tar, nicotine, and CO yields with puff frequency

Section IV

00

03

in

O

00

CM

00

in

o

o

CD

CO

CO

in

CD

00

CD

o

CD

m

in

I”-

CM

CM

CO

'O’

CD

O'-

CM

o

o

t—

N-

r^

CD

o

CD

CD

in

20

CM

O

h-

o

t^

o

in

CD

CD

CM

CO

CD

o

o

CM

h-

T

O

C33

-1 —

CD

o

'O’

CO

CD

CM

CD

CM

CO

T—

03

CO

o

o

00

00

'O’

CO

o

Q_

03

03

-T

h-

CO

o

T

1 —

CD

CD

o

O

O

CO

CM

CM

co

r^

CD

CO

CD

CM

CD

O

CO

CO

CM

co

CD

h-

'O’

CO

CO

'O’

C33

r>-

CO

O

00

00

CM

00

r^

in

CO

in

'O’

CO

'O’

CO

CD

CO

in

30

O

O

</)

CX3

O

03

CM

'O’

'O’

"O’

'O’

CM

O

CM

■a

V-

CO

CO

CO

in

in

in

CD

00

in

CD

CD

CD

c

T

o

o

<D

r\

03

CM

CM

o

CM

CD

CD

O

o

CO

CO

in

in

in

CD

in

CO

in

CD

h-

>.

o

c

(U

13

CT

o

o

CD

CO

CO

CD

CO

CM

'O’

CO

CD

CD

03

o

CM

CM

CM

T

'O’

CO

CM

CM

CO

1 —

CO

Ll-

4=

3

C33

03

CM

CO

in

'O’

CM

CD

CD

in

CD

CM

0.

o

CO

CM

CO

CM

CO

CM

CO

CO

'O’

o

1-

O

CO

C33

in

*T“

T“

CD

CD

CD

in

CO

CO

CO

CM

CM

in

CO

CM

'O’

CM

'O’

n

1^

00

•M-

CD

00

h-

N.

O

00

~l —

CD

00

CM

CM

CM

CM

CM

't

CM

CO

CO

CO

CM

CO

o

CO

CM

CM

00

in

in

00

in

'O’

o

o

1

CM

O

CM

CO

CvJ

o

'O’

CD

CD

00

'O’

1T“

1 —

1 —

T—

1 —

1 —

T

CM

o

o

in

•M-

CO

O

in

C'-

00

CD

in

CO

CM

CM

CO

CM

T—

CM

n

O

CM

CM

CO

CM

(D

00

'O’

CO

CO

CM

CM

■o

c:

03

m

X

X

X

f*“

"k

k

k

_j

03

CO

' —

' —

<D

— '

h-

CO

CO

CO

X

CD

CD

03

(led)

(led)

(Id)

(Id)

> > ^

o

-E

CD

O)

CO

c

c

Ct:

Z3

X

>.

t-

CD

C

TD

CD

Id

C

03

W.

CJ3

03

C33

03

CD

03

‘c

3

CO

LU

CO

UJ

CO

CO

CO

CO

O

CD

o

CO

CD

^ OC

X

X

X ^

S

:3

CJ*

3

Cl

•o

c

o

o

CD

to

§

(TJ

T3

0)

.C

5

D)

C

q>

y

(b

CL

&

o

c:

*§

Q>

E

:3

c.

c:

e

:3

o

O

o

251

Key; P = puffs count; T = far yield; N = nicotine yield; CO = carbon monoxide yield; Mini = minicigarette; Reg = nonfilter; KS = king size; UM = ultramild;
EM = extra mild; IS = international size; PI = plain; V = ventilated; NV = nonventilated; M = middle tar; M-H = middle to high tar; L = low tar;

L-M = low to middle tar.

Source: Darrall, 1988.

Smoking and Tobacco Control Monograph No. 7

Gori (1990a) noted that machines, unlike humans, smoke each cigarette
in exactly the same way. Smokers usually inhale after taking a puff, and
inhalation seems to be largely under the influence of nicotine demand.

When smoking machines were invented, little was known about inhalation
patterns. Today, inhalation can be measured with various biological markers,
such as CO and cotinine (an indicator of nicotine intake).

In a study of eight smokers. Gust and colleagues (1983) observed that the
number of puffs and the duration, volume, and time between puffs varied
with each smoker. All these factors affect the amount of smoke constituents
to which the smoker is exposed. Gust and colleagues also noted that smoking
patterns can vary as a smoker smokes a single cigarette.

Observations of smoking behavior reveal that smoking patterns are
influenced by a wide range of factors, including degree of nicotine
dependence, environmental cues, stress levels, and personality variables.

A survey of 1,200 randomly selected smokers and ex-smokers in the
United States and Europe showed that consumers believe that the tar yields
stated on cigarette packages accurately represent what is received by the
smoker (Gori, 1990b). The majority of respondents indicated a belief
that the published yield is equal to the amount consumed per cigarette.
However, tar intake is related to nicotine intake, and individual intake
of tar varies according to the nicotine levels of cigarettes and the level of
nicotine dependence of smokers.

Guyatt and coworkers (1989a, p. 192) studied the changes in puffing
behavior during the smoking of a cigarette. The researchers reported

The most important change in puffing behavior during a single
cigarette is the reduction in puff volume since this directly
affects smoke uptake. Most subjects showed this effect, but the
proportional change was independent of the tar level of the
cigarette smoked or the sex of the subject and was consistent
between sessions. However, there were significant between-
subject differences indicating that each individual had [an]
idiosyncratic pattern. Most subjects control puff volume by
varying the duration, mostly by truncating the latter part of
the puff.

IMPACT OF CHANGING

paramf:ters of the ftc

I ES I ME I HOD ON ABSOELITE
YIELDS OF A CIGAREITE
liKAND AND RELATIVE
YIELDS OF DIFFERENT BRANDS

Schlotzhauer and Chortyk (1983) examined the
influence of varying smoking machine parameters
on yields of tar, nicotine, and other selected
smoke constituents from an ultralow-tar cigarette.
The smoking machine parameters were changed
to reflect the deeper inhalation, more frequent
puffs, and vent l)locking evident among smokers of lower yield cigarettes.
•Specifically, volume was varied from the standard 3vS ml. to 45 ml. and
55 ml.; frequency of {)uffs was doubled; and puff duration was increased
from 2.0 to 3.0 seconds. Only one parameter was varied at a time; yields
were measured with vent holes both unl)locked and completely blocked.

252

Section IV

As shown in Tables 2 through 4, changing one parameter at a time
produces substantial increases in yields, and when cigarettes were machine
smoked at the average of the parameters used in Tables 2 through 4, as
shown in Table 5, total particulate matter (TPM) yields were approximately
doubled, and increases of 96 to 271 percent in the individual components
were observed.

TAR AND NICOTINE
YIELD BY THE FTC
TEST METHOD AND
AMOUNTS DELIVERED
TO SMOKER

The issue of compensation has become a central concern in
assessing intake of tar, nicotine, CO, and other constituents
constituents of tobacco smoke, particularly with regard
to cigarettes described as low tar and low nicotine. Various
researchers have reported no correlation between cigarette
brand yield and actual exposure and substantially higher relative exposures
from low-delivery cigarettes than indicated by quantitative differences in
stated yields (Rickert and Robinson, 1981).

The current primary measurement of the carcinogenic potential of a
cigarette is its tar yield. Kozlowski and colleagues (1980a) noted that tar
yield depends in part on the number of puffs per cigarette and that a major
factor in tar reduction has been reduced cigarette length, which results in
fewer puffs per cigarette during standard FTC testing. Increasing the number
of puffs can lead to substantial increases in tar yields.

Table 2

Effect of increased puff volumes on cigarette mainstream smoke under FTC conditions of puff
frequency (60 seconds) and puff duration (2 seconds)

Results

35-mL

Volume

45-mL

Volume

Change (±%)
From FTC
Values

55-mL

Volume

Change (±%)
From FTC
Values

Cigarettes Smoked

20

20

—

20

—

Total Puffs

152

150

-1

150

-1

Puffs/Cigarette (average)

7.6

7.5

-1

7.5

-1

Total Volume Inhaled (mL)

5,320

6,750

+27

8,250

+55

TPM (mg)

86

95

+10

135

+57

TPM/Cigarette (mg)

4.3

4.7

+10

6.7

+56

TPM/Puff (pg)

566

633

+11

900

+59

Phenol/Cigarette (pg)

12

17

+41

23

+92

Glycerol/Cigarette (pg)

327

624

+91

1,000

+206

Catechol/Cigarette (pg)

28

28

0

43

+54

Hydroquinone/Cigarette (pg)

23

27

+17

41

+61

Nicotine/Cigarette (pg)

378

502

+33

713

+88

Neophytadiene/Cigarette (pg)

15

32

+113

39

+160

Palmitic Acid/Cigarette (pg)

35

63

+80

64

+83

C^g Acids/Cigarette (pg)

33

61

+85

55

+67

Key: TPM = total particulate matter.

Source: Schlotzhauer and Chortyk, 1983.

253

Smoking and Tobacco Control Monograph No. 7

Table 3

Effect of increased puff frequency and increased puff duration on cigarette mainstream
smoke composition

Results

60-Second

Frequency

30-Second

Frequency

Change

(+%)

3-Second Puff
Duration

Change

(±%)

Cigarettes Smoked

20

20

0

20

0

Total Puffs

152

281

+85

150

-5

Puffs/Cigarette (average)

7.6

14.0

+85

7.5

-5

Total Volume Inhaled (mL)

5,320

9,835

+85

7,800

+47

TPM (mg)

86

205

+138

166

+93

TPM/Cigarette (mg)

4.3

10.2

+138

8.3

+93

TPM/Puff (pg)

566

728

+29

1,106

+93

Phenol/Cigarette (pg)

12

20

+67

13

+8

Glycerol/Cigarette (pg)

327

1,542

+371

795

+143

Catechol/Cigarette (pg)

28

66

+136

70

+150

Hydroquinone/Cigarette (pg)

23

50

+117

40

+74

Nicotine/Cigarette (pg)

378

961

+154

618

+63

Neophytadiene/Cigarette (pg)

15

29

+93

53

+253

Palmitic Acid/Cigarette (pg)

35

41

+ 17

39

+11

C,g Acids/Cigarette (pg)

33

34

+3

30

-10

Key: TPM = total particulate matter.
Source: Schlotzhauer and Chortyk, 1983.

Table 4

Effect of obstructing tipping paper ventilations on cigarette mainstream smoke composition

Results

FTC

Conditions^

FTC With Obstructed
Perforations

Change

(±%)

Cigarettes Smoked

20

20

0

Total Puffs

152

131

-14

Puffs/Cigarette (average)

7.6

6.5

-14

Total Volume Inhaled (mL)

5,320

4,584

-14

TPM (mg)

86

256

+ 198

TPM/Cigarette (mg)

4.3

12.8

+ 198

TPM/Puff (pg)

566

1,969

+248

Phenol/Cigarette (pg)

12

19

+58

Glycerol/Cigarette (pg)

327

1,001

+206

Catechol/Cigarette (pg)

28

58

+ 107

Hydroquinone/Cigarette (pg)

23

53

+ 130

Nicotine/Cigarette (pg)

378

839

+ 122

Neophytadiene/Cigarette (pg)

15

50

+233

Palmitic Acid/Cigarette (pg)

35

85

+ 143

C,0 Acids/Cigarette (pg)

33

76

+ 130

• 35-mL puff volume, 60-second puff frequency, 2-second puff duration.
Key: TPM » total particulate matter.

Source: Schlotzhauer and Chortyk, 1983.

254

Section IV

Table 5

Effect of combined compensatory parameters on

yields of mainstream smoke components

FTC

New

Change

Results

Conditions

Conditions®

(±%)

Cigarettes Smoked

20

20

0

Total Puffs

152

236

+55

Puffs/Cigarette (average)

7.6

11.8

+55

Total Volume Inhaled (mL)

5,320

1 1 ,564

+117

TPM (mg)

86

169

+97

TPM/Cigarette (mg)

4.3

8.4

+95

TPM/Puff (ng)

566

716

+27

Phenol/Cigarette (ng)

12

30

+150

Glycerol/Cigarette (pg)

327

1,212

+271

Catechol/Cigarette (pg)

28

55

+96

Hydroquinone/Cigarette (pg)

23

53

+130

Nicotine/Cigarette (pg)

378

850

+125

Neophytadiene/Cigarette (pg)

15

52

+247

Palmitic Acid/Cigarette (pg)

35

86

+142

C^g Acids/Cigarette (pg)

33

71

+115

® Averaged, reported compensatory smoking parameters (49-mL puff, 38-second frequency, 2.5-second puff duration)
set on smoking machine.

Key: TPM = total particulate matter.

Source: Schlotzhauer and Chortyk, 1983.

In a subsequent study of four popular king size cigarettes (see Table 6),
Kozlowski (1981, p. 159) found that

the same cigarette can easily rise from a low-tar to a high-tar
category [through an increase in] the number of puffs taken
from it, within the range of puffs per minute consistent with
human smoking behavior. Based on the standard assay, brand
B has 1 7 percent more tar than brand C; however, based on a
10-puff estimate, their tar deliveries are identical. Those
smokers who take 14 puffs per cigarette are getting 58 percent
more tar than would be expected from the standard yields.

Rawbone (1984), in a study of 400 middle-tar and low-tar smokers in
the United Kingdom, found that tar delivery varied significantly between
middle- and low-tar cigarettes but noticeably less than expected. That is,
where a 46-percent lower tar delivery was expected with the low-tar
cigarettes, a 32-percent reduction was observed. Furthermore, with regard
to tar delivery, 98 percent of the middle-tar cigarette smokers fell within the
established bounds of 16.50 to 22.49 mg delivery, whereas only 70 percent

255

Smoking and Tobacco Control Monograph No. 7

Table 6

Tar yields (mg) as a function of number of puffs taken by smoking machines

Brand®

Number of Puffs

6

8.7^^

10

14

A

13

18

21

30

B

13

21

22

31

C"

13

18

22

31

D"

12

17

19

27

® Four of the most popular brands of king-size filter cigarettes.

“Mean number of puffs for the standard assay for these cigarettes: A, 8.6 puffs; B, 9.3; C, 8. 1; D, 8.9.
“ These brands are mentholated.

Source: Kozlowski, 1981.

of the low-tar cigarette smokers were experiencing a delivery at or below
the upper limit of 10.49 mg set for low -tar cigarettes (with 30 percent
experiencing a higher-than-expected tar delivery).

Rickert and colleagues (1986) machine-analyzed the nicotine, tar, and
CO yields of 10 cigarette brands under 27 different conditions (the standard
condition and 26 variations). Tar, nicotine, and CO yields increased with
volume of smoke produced per cigarette, but yields per liter of smoke were
relatively constant across the 27 conditions.

Woodward and Tunstall-Pedoe (1992) investigated the smoking patterns
of 2,754 smokers (1,133 males and 1,621 females) to determine intake of
smoke components by smokers of low-tar cigarettes. This study, perhaps
the largest naturalistic investigation of smoking behavior ever undertaken,
included smokers of low-, middle-, and high-tar cigarettes. The researchers
concluded that tar yield does not accurately reflect the amount of smoke
components consumed by the smoker. Specifically, tar intake increased
with tar yield but much less than anticipated; expired-air CO and cotinine
seemed to peak among middle-tar smokers. For women, thiocyanate
increased from low- to middle-tar smokers, and for men, from middle- to
high-tar smokers. I he researchers found that smokers of middle-tar cigarettes
may consume more of some smoke components than smokers of high-tar
cigarettes. Middle-tar smokers were noted to have higher levels of
expired-air CO and cotinine.

Armitage and colleagues (1988) investigated the influence of changes
in tar yield when nicotine yield was maintained. I’wenty-one smokers of
middle-tar cigarettes were studied, with randomization to three categories:
low tar and low nicotine, low tar and medium nicotine, and medium yields
of tar and nicotine. With regard to nicotine uptake, there were no significant
differences noted between rniddle-tar and nicotine-maintained cigarettes, but
there were significant differences between low-tar and nicotine-maintained

256

Section IV

cigarettes. The mean total puff volume of the nicotine-maintained cigarette
was significantly greater than that recorded for middle-tar cigarettes. There
was no difference in mean total puff volume between low-tar cigarettes and
nicotine-maintained cigarettes.

RELATIVE YIELDS OF
DIFFERENT BRANDS BY
THE FTC TEST METHOD
AND AMOUNT OF
NICOTINE ABSORBED
BY SMOKERS

Ebert and colleagues (1983) undertook a study of
76 smokers to determine correlations between levels of
plasma nicotine and alveolar CO and the nicotine and
CO yields of cigarettes. The correlations were found to be
poor (Figures 1 and 2). For the 24 smokers of low-nicotine,
low-tar cigarettes, nicotine levels were statistically lower
for smokers of low-nicotine cigarettes, but the levels were only slightly lower
and there was great overlap in individual plasma nicotine values; there was
no difference in the mean alveolar CO levels between the low-nicotine
smokers and smokers of regular cigarettes.

Research by Benowitz and colleagues (1983) on 272 subjects about to
enter a smoking treatment program revealed that the correlation between
stated nicotine yield and actual blood cotinine levels was not significant.
Furthermore, it was determined that nicotine concentration in the unburned
tobacco and amount of nicotine in an unburned cigarette are not correlated
positively with FTC-determined yields and that tobacco in low-yield cigarettes
did not contain less nicotine than tobacco in higher yield cigarettes.

Figure 1

Relationship between plasma nicotine concentration in smokers and nicotine yield
of cigarettes smoked

80 I

70

£ 60
w

S 50
0)

I 40

u

m 30
E

\A

iS 20

10

8

Source: Ebert etal., 1983.

• •

• •
>•

• ^

•

••••
• •
• •

• •

J L

J I I L

0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1,

Nicotine Yield (mg)

257

Smoking and Tobacco Control Monograph No. 7

Figure 2

Relationship between carbon monoxide (CO) concentration of alveolar air
in smokers and CO yield of cigarettes smoked

90

80

E

70

Q.

Q.

60

O

u

w

50

<

nj

40

o

(V

>

30

<

20

10

5 10 15 20 25

CO Yield (mg)

Source: Ebert etai, 1983.

Ventilation and burning characteristics are the primary determinants of
machine-measured yields, and these characteristics can be controlled by
smokers. Benowitz and colleagues acknowledged that blood cotinine is not
a "perfect marker," but a full range of cigarettes was included in the study
and there is no reason to suspect that brand is related to nicotine and
cotinine metabolism.

Russell and colleagues (1986) examined blood nicotine, cotinine, and
carboxyhemoglobin (COHb) levels among 392 smokers whose regular brands
varied from low tar to middle tar. Tar levels were estimated from blood
nicotine levels and cigarette tar yields. The authors reported

Smokers of LT [low-tarj cigarettes had a lower intake of tar,
nicotine, and CO than the smokers of higher yielding brands. On
average, their estimated intake of tar was about 25 percent lower,
their intake of nicotine was about 15 percent lower (17 percent
and 12 percent, as measured by blood nicotine and cotinine,
respectively), and their intake of CO was about 10 percent lower.

I'hese differences are substantially less than the reductions in the
standard machine-smoked yields of their cigarettes (47 percent,

39 percent, and 34 percent for tar, nicotine, and CO yields,
respectively), and this indicates the extent to which the LT
smokers were smoking and inhaling more intensively, {)resumably
to corn[)ensate for the lower yields. However, it is clear that

258

Section IV

despite such compensatory changes in smoking behavior,
their intake of the three major smoke components was still lower
to a statistically and clinically significant degree (Russell et al.,

1986, p. 83).

Maron and Fortmann (1987) examined the relationship of FTC
machine-estimated nicotine yield by cigarette brand with the level of
cigarette consumption and two biochemical measures of smoke exposure
(expired-air CO and plasma thiocyanate) in a population of 713 smokers.
These investigators found that the lower the nicotine yield, the greater
the number of cigarettes smoked per day. Smokers of ultralow-nicotine
cigarettes experienced smoke exposures that were not significantly different
from those of smokers of higher yield brands. Only after adjustment for
number of cigarettes smoked daily did nicotine yield become significantly
related to expired-air CO and plasma thiocyanate. The number of cigarettes
smoked per day accounted for 28 and 22 percent of the variance in observed
expired-air CO and plasma thiocyanate levels, respectively, whereas nicotine
yield accounted for only 1 and 2 percent of the variance, respectively. The
authors concluded that machine estimates suggesting low nicotine yield
underrepresent actual human consumption of harmful cigarette constituents.

In a study of 289 smokers of cigarettes in the 1-mg FTC tar class, Gori
and Lynch (1983) observed that nicotine intake (measured by plasma
cotinine) varied widely, from undetectable to about 800 ng/mL. The
findings indicated that smokers of low-yield brands tend to take in more
nicotine than posted FTC values. This observation is illustrated in Figure 3.
Brand A was .9 tar and .18 nicotine, whereas brand B was .5 tar and
.10 nicotine.

Coultas and colleagues (1993), working with a population of 298 mostly
Hispanic smokers, studied the relationship between yields of cigarettes
currently smoked and levels of salivary cotinine and expired-air CO.
Spearman's correlation coefficients (Snedecor and Cochran, 1980) between
the current number of cigarettes smoked and cotinine or CO were higher
than correlations between the FTC nicotine data and these same markers.

In multiple linear regression models, the current number of cigarettes
smoked was the most important predictor of cotinine and CO levels
(p < 0.0001), and the addition of FTC tar, nicotine, and CO to the models
explained little about the variability in cotinine and CO levels.

In a large-scale study of 2,455 cigarette smokers who smoked their usual
brands, Wald and colleagues (1984) observed that nicotine and CO intake
was relatively constant across brands, regardless of stated yield, although tar
intake appeared related to tar yield.

YIELD BY THE FTC TEST As pointed out by many researchers, cigarette

METHOD AND ABSORPTION smoking has the hallmarks of drug-dependent
OF NICOTINE IN SWITCHERS behavior, with strong evidence that nicotine is

the dependence-producing component (Benowitz et al., 1989). Nicotine is
rapidly absorbed into the blood and quickly delivered to the brain, where

259

Smoking and Tobacco Control Monograph No. 7

Figure 3

Observed and expected baseline plasma cotinine values as a function of FTC
nicotine delivery of brands A and B

Key: SEM = standard error of the mean.
Source: Gori and Lynch, 1983.

it produces a range of mental effects on the smoker. This quick absorption
and effect permit the smoker to control the nicotine level carefully; however,
nicotine is rapidly eliminated from the body, which means the smoker has
to deliver regular doses to the blood.

260

Section IV

Robinson and colleagues (1982 and 1983) studied the smoking patterns
of 22 cigarette smokers divided into treatment and control groups, with the
treatment group switching twice to cigarettes of successively lower nicotine
yields. Compensation behavior was measured noninvasively (average
number of daily cigarettes, daily mouth-level nicotine exposure, butt length,
expired-air CO, and saliva thiocyanate) and invasively (COHb, serum
cotinine, and plasma thiocyanate). As shown in Figure 4, there were no
major differences between smokers in treatment and control groups. The
near-complete compensation was attributed to upward changes in smoking
intensity, depth of inhalation, and cigarette consumption. In addition, there
was an observed tendency of smokers of lower delivery cigarettes to smoke
cigarettes down closer to the overwrap and to block ventilation holes.

In a different approach, Gritz and colleagues (1983) looked at the puffing
behavior of eight smokers presented with cigarettes at two and four times

Figure 4

Average daily exposure and standardized exposure measures by period for treatment and
control groups

Treatment groups Control groups

Note: Average nominal nicotine deliveries are shown as horizontal lines in Panel H. Abbreviated variable names have
been used. The increase of “standardized" exposure measure for the treatment group (Panels F and G) during
period three (P^) does not represent an increase in exposure. The exposure remains fairly constant during the
entire study, as Panels C and D indicate. Panels F and G illustrate the extent of compensation necessary to
maintain this constant exposure. See text for details.

Key: MLE = mouth-level exposure; CO = carbon monoxide.

Source: Robinson et al., 1982.

261

Smoking and Tobacco Control Monograph No. 7

their normal smoking rates. All eight smokers compensated to some degree.
Despite being presented with twice the usual number of cigarettes, the
smokers titrated their nicotine intake down, largely by changing their number
of puffs, puff volume, and puff duration per cigarette. Gritz and coworkers
disputed the view that some smokers may be compensators and others may
be noncompensators, arguing that these two groups of smokers represent the
opposite ends of a continuum.

Henningfield and Griffiths (1980) studied the effect of tobacco product
concentration on puffing rate and total number of puffs. Tobacco
concentration levels were set at 100, 50, 25, and 10 percent by means of
ventilated holders (identified in Figures 5 and 6 as holders 0, 1, 2, and 4).

As shown, puffs at holder 4 were about double those of holder 0. In addition,
there were substantial increases in puff rate.

Compensation via alterations in puffing patterns does not explain all
observed changes, however. In their investigation of puffing and inhalation
patterns and yields. Nil and colleagues (1986) found that changes in puff
volume account for only about one-fifth of the difference in smoke yields;
no significant changes were found in inhalation patterns. On the other hand,
with lower yield cigarettes, there was nearly complete compensation based
on alveolar CO uptake, and the degree of increased heart rate was viewed
as a nearly complete compensation for nicotine intake.

McBride and colleagues (1984) measured changes in smoking behavior
and ventilation when subjects smoked cigarettes of varying nicotine yields.
Nine smokers were studied, and the test order was randomized. Puff volume
was noted to increase significantly during the smoking of low-nicotine
cigarettes. In a study of 170 male smokers and 170 age-matched male
nonsmokers. Bridges and colleagues (1986) observed that total puff volume
was significantly greater for smokers of cigarettes lower in nicotine yields.

As shown in Figures 7 and 8, total puff volume was significantly correlated
with nicotine yield and plasma cotinine.

Researchers have observed that smokers can substantially alter tar,
nicotine, and CO delivery of cigarettes by blocking the ventilation holes in
the filters. In a two-part study of smokers of low-yield cigarettes, Kozlowski
et al. (1982a) observed hole-blocking behavior and measured tar, nicotine,
and CO levels. The investigators reported that 44 percent of 39 smokers of
low-yield cigarettes blocked the ventilation holes to various degrees with
their fingers or lips; 5 of 33 females left hole-blocking lipstick on the filters.

In the second part of their study, Kozlowski and colleagues (1982a)
evaluated the effect of hole blocking on the tar, nicotine, and CO yields of
American, British, and Canadian cigarettes of lowest or near-lowest yields.
After videotaping 48 smokers, the researchers defined actual smoking
behaviors and reset smoking machine parameters to reflect these real-life
patterns for puff interval (44 seconds) and puff duration (2.4 seconds).
Machine [)uff volume was set at 47 ml. (2 to 13 ml, below the smokers'
estimated average) because this is the maximum obtainable from most

262

Section IV

Figure 5

Mean total puffs per session (N = 4) and standard error values for each subject as
a function of cigarette holder number

I Ventilated Holder

Note: The approximate concentrations of delivered tobacco product are indicated by the holder
number in which 0= 100 percent, 1 = 75 percent, 2 = 50 percent, and 4 = 10 percent. The
’ abbreviations ST, ED, and GR represent three paid female volunteers who participated in the

1 study.

Source: Henningfield and Griffiths, 1980.

i machines. Ventilation holes were blocked with tape. The researchers

! compared standard yields of cigarettes to yields resulting from the study-

determined parameters and blocked ventilation holes; they observed that
"tar increases from 15- to 39-fold, nicotine from 8- to 19-fold and CO from
10- to 43-fold" (Kozlowski et al., 1982a, p. 159). Five cigarette brands similar
I in tar yield were found to differ substantially when parameters were changed

' and holes blocked.

I In a later study of 14 subjects, Kozlowski (1989) detected hole blocking

' by half the sample. Subjects blocking the ventilation holes of ultralow-yield

i

263

Smoking and Tobacco Control Monograph No. 7

Figure 6

Mean number (N = 4) of cigarettes smoked by each subject during 3-hour sessions
as a function of holder number (upper frame) and mean rate of puffing (puffs/
minute) per cigarette (lower frame)

Ventilated Holder

Note: Standard error values for each subject, in both frames, are indicated by the brackets. The

approximate concentrations of delivered tobacco product are indicated by the holder number in
which 0 = too percent, 1 = 75 percent, 2 = 50 percent, and 4=10 percent. The abbreviations
ST, ED, and GR represent three paid female volunteers who participated in the study.

Source: Henningfield and Griffiths, 1980.

264

Section IV

Figure 7

Relationship of total puff volume per cigarette with the nicotine yield of the
cigarette smoked

Nicotine Yield (mg/cigarette)

Source: Bridges et a!., 1986.

Figure 8

Relationship between plasma nicotine concentration and total volume puffed
per cigarette in a population smoking a single brand of cigarette (nicotine yield =
1 .05 mg/cigarette)

Total Puff Volume/Cigarette (mL)

Source: Bridges et a!., 1986.

265

Smoking and Tobacco Control Monograph No. 7

cigarettes were found to have higher CO and salivary cotinine levels.

Rickert and colleagues (1983) found that blocking half the ventilation holes
increased the delivery of TPM by 60 percent, and full occlusion increased
TPM delivery by 150 percent.

The effect of blocking on perforation ventilation (ventilation holes in the
filter) and channel ventilation (longitudinal air channels around the filter)
was studied by Hdfer and colleagues (1991). The researchers compared
results of lip smoking and holder smoking of cigarettes among 72 smokers,
divided equally by ventilation type of cigarette smoked. Hdfer and
colleagues (1991, p. 910) found that

under normal lip contact conditions, the CO and nicotine
deliveries of the channel-ventilated cigarettes were higher than
those of the perforation-ventilated cigarettes and higher than
with holder smoking. With holder smoking, both types of
cigarettes delivered comparable amounts of CO and nicotine
(t-tests, n.s.).

It appeared that the nicotine boost from channel-ventilated cigarettes was
twice that of perforation-ventilated cigarettes; differences in CO exposure
were less well defined. The researchers judged that there was evidence
of blocking in 86 percent of the channel filter cigarette smokers and in
33 percent of the perforated filter cigarette smokers.

In a novel approach to the study of hole blocking among smokers
of ultralow-tar cigarettes, Kozlowski and colleagues (1988) collected
135 discarded filters from ashtrays in shopping malls. It was found that
58 percent of the filters showed some evidence of hole blocking (as measured
by tar stain patterns); 19 percent showed evidence of extreme hole blocking;
and 42 percent showed no signs of hole blocking. Kozlowski and colleagues
(1994) extended this research to "light" cigarettes (about 9 to 12 mg tar,
about 15 to 30 percent vented): Twenty-seven percent of collected filters
indicated extreme blockage; 26 percent showed some blocking; and
47 percent showed no vent blocking. Although defeat of the air vents
will have a relatively small effect on light rather than ultralight cigarettes,
the greater sales of light cigarettes contribute to its significance for public
health. In an earlier report, Kozlowski and colleagues (1980b) examined
the effect of hole blocking on nicotine, tar, CO, and puffs (Table 7), noting
that ventilated filters have been developed primarily as a way to make less
toxic cigarettes but that smoking behavior can sabotage the benefits of
these filters.

Kozlowski and colleagues (1989) demonstrated that some smokers of
vented filter cigarettes are lighter smokers who appear to be seeking lower
smoke doses and do not block vents, whereas others are generally heavier
smokers who block vents and derive high daily doses of nicotine. Two
smokers, who were vent blockers, of a 1-mg tar, 0.1 -rng nicotine cigarette
achieved salivary cotinine levels (303 and 385 ng/mL) consistent with
smoking a high-yield cigarette.

266

Section IV

Table 7

Effects of blocking the ventilation holes on the yields of a popular, low-yield cigarette^

Characteristics

Unblocked

Holes

Half-Blocked

Holes

Fully Blocked
Holes

Constituents

Nicotine (mg)

0.45

0.73 ± .06

0.98 ± .06®

Tar (mg)

4.40

7.03 ± .04

12.60 ±.20®

Carbon Monoxide (mg)

4.50

7.80 ± .24

17.70 ±.40®

Puffs

11.10

10.50 ± .20

9.20 ± .40®

® Half-blocked vs. fully blocked comparison (\.-test, 2-tailed) p < .01. Values are means ± standard deviations.
Government figures for the June-July 1979 assay were used as the unblocked control; variances were not reported,
but those found in similar analyses imply that all within-row comparisons would be statistically significant. All
analyses in the table were performed by the same laboratory employing the same techniques.

Source: Kozlowski et al., 1980b.

Bridges and colleagues (1990) studied 170 male smokers to determine
the influence on yield of smoking topography (i.e., total smoking time per
cigarette, number of puffs, interpuff interval, puff duration, volume per puff,
total duration per cigarette, total volume per cigarette, flow rate). The
smokers were divided into six groups according to stated nicotine yields
of their cigarettes. The first four groups were most similar in age, smoking
history, and alcohol and coffee consumption. There were significant
negative correlations between nicotine yield and mean puff volume, total
duration and volume, and flow rate. That is, as nicotine yield decreased,
mean puff volume, total duration and volume, and flow rate increased
significantly. These statistical relationships are shown in Figure 9. Multiple
regression analysis showed that nicotine yield, alone or in combination with
other factors, is a significant predictor of number of puffs or total puff
volume per cigarette.

Figure 9 is of special interest because it represents smoking topography
changes in a subpopulation for which nicotine yield was held constant
to control for the possible confounding effects of nicotine on smoking
behavior. Cumulative puff volume for a cigarette is significantly correlated
with plasma nicotine, an indication that increased inhalation results in
increased absorption. For the same group, the interpuff interval was
negatively correlated with plasma nicotine levels (i.e., when time between
puffs went down, plasma nicotine level went up).

According to Bridges and colleagues (1990, p. 31)

Smokers smoking the lowest yield cigarettes (Group 1) had
significantly higher total puff volume per cigarette than did the
other groups, and significantly higher mean puff volume and
flow rate . . . than Groups 3 and 4. Smokers of lower yield

267

Smoking and Tobacco Control Monograph No. 7

Figure 9

Linear relationships between nicotine yield and puffing topography measures:
(A) number of puffs per cigarette, (B) total puff duration per cigarette, (C) total
puff volume per cigarette

y

Nicotine Yieid (mg/cigarette)

y

Nicotine Yieid (mg/cigarette)

y

Nicotine Yieid (mg/cigarette)

Note: The graphical representation for each of these relationships includes the equation for the
inserted least-squares best fit line, the correlation coefficient (r), and the level of significance
for the correlation. The data are for groups 1-4, n = 108.

Source: Bridges et al., 1990.

268

Section IV

cigarettes also tended to have higher numbers of puffs per
cigarette, decreased interpuff interval, increased duration per
puff, and increased duration per cigarette, but these differences
did not reach statistical significance. These results are consistent
with changes in puffing topography to compensate for lower
}deld cigarettes.

In addition, there were significant negative correlations between
nicotine yield and mean puff volume, total duration and volume, and flow
rate. That is, as nicotine yield decreased, mean puff volume, total duration
and volume, and flow rate increased significantly. In addition, multiple
regression analysis showed that nicotine yield, alone or in combination
with other factors, is a significant predictor of number of puffs or total puff
volume per cigarette.

Creighton and Lewis (1978) examined changes in smoking patterns
when cigarettes were varied according to nicotine delivery. Specifically,

16 smokers were monitored for 3 months. The first month, they all smoked
medium-delivery cigarettes of about 1.4 mg nicotine; then the group was
split for 1 month, with half smoking lower delivery cigarettes (about
1.0 mg nicotine) and half smoking higher delivery cigarettes (about 1.8 mg
nicotine). During the third month, the panel of 16 smokers returned to
the 1.4 mg nicotine cigarettes. Significant changes were found in smoking
patterns among the 16 smokers: either the increased smoking intensity
when smoking lower delivery cigarettes or decreased intensity when smoking
higher delivery cigarettes. However, the researchers reported that the
smokers did not equalize nicotine and TPM delivery when they switched
to lower delivery cigarettes, as was the case when they switched to higher
delivery cigarettes. The number of cigarettes smoked per day remained
about the same throughout the study.

Russell and colleagues (1982) looked at changes in nicotine, cotinine,
COHb, thiocyanate, and tar when 12 smokers switched to low-tar, low-
nicotine cigarettes for 12 weeks. Plasma nicotine and cotinine were both
reduced by about 30 percent and tar by 15 percent; plasma thiocyanate
and COHb did not change significantly. Although mouth level of nicotine
intake from low-tar, low-nicotine cigarettes was similar to the standard
machine yield, the blood levels of 30 percent were substantially less than
the anticipated level of 46 percent based on machine yields. There was
no compensatory increase in smoke intake at the mouth level, but blood
measures showed the increase in inhalation between 32.1 and 40.8 percent.

Similarly, Ashton and coworkers (1979) found that, when switched from
medium- to high- or low-nicotine brands, smokers compensated for about
two-thirds of the difference in standard yields. Specifically, when nicotine
yield was reduced by 50 percent, nicotine intake was about 15 percent lower.
Furthermore, based on machine yields, it was anticipated that the nicotine
yield of low-nicotine cigarettes would be 32.6 percent that of high-nicotine
cigarettes; however, in the laboratory the observed yield was 59 percent that
of high-nicotine cigarettes.

269

Smoking and Tobacco Control Monograph No. 7

Benowitz and colleagues (1986) looked at differences in tar, nicotine,
and CO exposure when smokers switched from their regular brand to high-,
low-, and ultralow-yield cigarettes. The researchers detected no differences
in exposure among the high- and low-yield smokers. However, for smokers
of ultralow-yield cigarettes, there were substantial reductions in exposure to
tar (49 percent), nicotine (56 percent), and CO (36 percent). Despite these
reductions, the investigators reported that the relative exposure to tar and
nicotine from ultralow-yield compared with higher yield cigarettes was much
greater than predicted by FTC machine-determined yields.

Kolonen and colleagues (1991) examined puffing patterns of 36 smoking
students, with different smoking histories, in a natural environment. The
subjects included 18 smokers of low-yield cigarettes, 10 smokers of medium-
yield cigarettes, and 8 smokers who had switched from medium- to low-yield
cigarettes. Subjects smoked their regular brand for the first week, a low-yield
brand for the second week, and a medium-yield brand for the third week.

All three groups had the highest daily puff volumes when smoking low-yield
cigarettes, and the correlations between urine cotinine concentration and
daily puffing in the three groups were poor. However, the urinary cotinine
concentration was significantly lower for low-yield smokers compared with
the switchers. The investigators concluded that cotinine excretion results in
the switchers' group were in line with earlier reports showing that long-term
switchers have no significant decreases in plasma and urine cotinine.

In a longer study of switching effects, Guyatt and coworkers (1989b)
monitored 28 smokers who switched to cigarettes with lower tar and nicotine
yields. The researchers concluded, after monitoring subjects for about
1 year, that most effects of the switch to lower yield cigarettes did not persist
beyond 36 weeks. The drop in cotinine levels was only 40 percent of what
was expected from stated nicotine yields; mean puff volume increased by
16 percent; and smokers seemed to achieve about 60 percent compensation
when smoking lower tar cigarettes.

YIELDS BY THE FTC Carbon monoxide yields follow somewhat surprising

TEST METHOD AND OTHER dynamics. For example, as Rickert and colleagues
CONSTITUENTS USING (1980) reported, efficient filters may substantially

FTC PUFF PROFILE reduce tar yields of cigarettes but lead to increased

delivery of CO.

In a study of reduced-draw-resistance cigarettes, Dunn (1978) found that
smokers can substantially vary their inhalation patterns, leading to marked
changes in the amount of smoke that reaches the lungs as measured by
alveolar CO levels. Although increased levels of alveolar CO were expected
with reduced draw resistance, CO levels decreased, possibly because of
increased delivery of nicotine. Dunn proposed that the level of CO in
exhaled air may be a good measure of depth of inhalation.

There appears to be substantial natural variation in the amount of
CO inhaled by smokers, even when numbers of cigarettes smoked are
approximately equal. Burling and colleagues (1985) studied 12 matched

270

Section IV

pairs of smokers, each pair smoking a similar number of cigarettes but with
different levels of CO (one high-CO-level subject and one low-CO-level
subject). The CO boost per cigarette was found to be significantly different
for the matched pairs of smokers. The CO boost for the high-CO group was
6.9 ppm per cigarette and for the low CO group 4.4 ppm.

The study found no differences between the high-CO and low-CO groups
in terms of number and duration of puffs. Given the significant differences
in CO levels, the researchers speculate that the difference may reside in puff
intensity, puff volume, or inhalation characteristics. These influences on CO
levels are relevant to low-nicotine yields and changes in smoking behavior;
Herning and colleagues (1983) reported that CO boost appears correlated to
blood nicotine levels.

In an earlier study. Burling and coworkers (1983) found that a smoker's
CO level is influenced by factors other than the FTC-determined CO yield
of cigarettes. The researchers reported that the CO level is significantly
related to interpuff interval, cigarette duration, time since last cigarette, and
self-rated estimate of depth of inhalation. This research underscores the
likelihood that CO levels may be determined by multiple factors, not just
stated yield. However, the finding suggests that, when numbers of cigarettes
are held equal, a person smoking cigarettes with a higher CO yield will likely
have higher CO levels than a person smoking cigarettes of lower CO yield.
Furthermore, Wald and colleagues (1984) reported that smokers of filter
cigarettes have a 60-percent higher intake of CO than do those who smoke
nonfilter cigarettes.

Russell and colleagues (1982), in a study of long-term switching to low-
tar, low-nicotine cigarettes, observed complete compensation as measured by
CO uptake, and Robinson and coworkers (1983) reported that COHb levels
did not change significantly after smokers switched to cigarettes with 15- and
72-percent lower CO deliveries.

Robinson and colleagues (1984) examined exposure among 22 smokers
of high-nicotine cigarettes who switched to cigarettes of similar nicotine
yield but with reduced 5delds for tar, CO, and hydrogen cyanide. Cotinine
levels remained about the same; however, although reductions of 40 to
50 percent in CO and HCN were expected, the measured reductions were
5.3 percent for expired-air CO, 12.2 percent for COHb, 2 percent for saliva
thiocyanate, and 1 percent for plasma thiocyanate.

Darrall (1988) found that a 50-percent blockage of ventilation holes
produced small changes in tar and nicotine yields but greater changes in CO.
Nil and coworkers (1986), in a study of 117 regular smokers, reported that
the CO boost of cigarettes appeared to remain steady among smokers despite
controlled switching to cigarettes of higher or lower yields.

Fischer and colleagues (1989), in an investigation of six different
cigarette brands (filter and nonfilter and very low to medium tar yields),
found that puff volume and puff frequency, the key determinants of total

271

Smoking and Tobacco Control Monograph No. 7

volume inhaled, significantly affect the smoker's exposure to tobacco-specific
nitrosamines. According to the investigators,

The medium-tar cigarette using standard smoking conditions
delivered TSNA values that were close to the calculated average
intake by smokers. The calculated average TSNA intake for the
low-tar cigarette, however, was about double the value determined
under standard smoking conditions (Fischer et al., 1989, p. 1065).

The researchers concluded that

since the standard smoking conditions cannot reflect the real
behavior for low- and very-low-tar cigarettes, especially with
respect to the total inhalation volume, risk evaluation has to
consider the increase in TSNA intake with increasing total
volume (Fischer et al., 1989, p. 1065).

In a subsequent study, Fischer and colleagues (1991) investigated
170 types of American, European, and Russian cigarettes. The findings
revealed that the amounts of two TSNAs — NNN (N-nitrosonornicotine)
and NNK (4-methylnitrosamino-l-[3-pyridinyl]-l-butanone) — in cigarette
smoke are not correlated with tar or nicotine delivery and the amounts of
TSNAs in mainstream smoke are related to the amount of preformed
nitrosamine in the tobacco.

In an investigation of compensation behaviors among smokers switching
to lower delivery cigarettes, Robinson and coworkers (1983) noted
disproportionate increases in HCN levels. The researchers concluded that
machine-determined "standardized" deliveries do not reflect potential
exposure to HCN.

Rickert and Robinson (1981), in a study of delivery by low-hazard and
high-hazard brands and actual levels, found that differences in HCN and
CO yields of the two different delivery types varied much more widely than
actual levels of COHb and plasma thiocyanate obtained from smokers of
each. High-hazard cigarette smokers had nearly four times the HCN of
low-hazard cigarette smokers; however, the actual levels differed by only
20 percent. These differences were not statistically significant, possibly
due to small sample size (n = 31).

Rickert and colleagues (1983) looked at variations in smoking patterns
and reported that HCN delivery is influenced by blocking of ventilation
holes and, to a lesser degree, by puff duration, puff volume, and butt length.
Blocking half the ventilation holes increased HCN yield by 70 percent;
covering all the holes produced a 250-percent increase in yield. Fhese
investigators determined that HCN yield for the cigarette brand investigated
ranged from 5 to 241 pig, depending on variations in smoking parameters,
although the mean HCN yield was 39 ng. For 115 Canadian cigarettes, the
average HCN yield varied from 2 to 233 pig. Fhis impact of smoking pattern
on HCN yield was cited by Rickert and colleagues as a possible explanation

272

Section IV

for the poor correlation between HCN yield and levels of plasma thiocyanate
and saliva thiocyanate.

Rickert and colleagues (1980) indicated that aldehydes, gas-phase
constituents of tobacco smoke, are known to be ciliatoxic and may not be
removed to a substantial degree from cigarette smoke by filters. Acrolein, a
toxin restricted in occupational and industrial settings, also may contribute
to the chemical toxicity of tobacco smoke. In a study of 102 brands of
Canadian cigarettes, Rickert and colleagues found that tar level was a poor
predictor of total aldehydes and acrolein delivery. The effect of changes in
smoking patterns on phenol, glycerol, catechol, hydroquinone, palmitic acid,
and neophytadiene are shown in Tables 2 through 5 (Schlotzhauer and
Chortyk, 1983).

PROPOSALS TO At least three proposals have been published for changes in the
CHANGE THE FTC FTC cigarette test method. Kozlowski and colleagues (1982b)
TEST METHOD made a proposal addressing the issue of the variability in human
smoking behavior. These investigators suggested a three-level (i.e., light,
average, and heavy) machine regimen linked to a color-matching technique
to help smokers gauge the extent of puffing on a given cigarette — the darker
the stain, the greater the exposure, with the tar stains keyed to a range
of tar doses. Rickert and colleagues (1986) proposed an estimate based
on average yields of tar, nicotine, and carbon monoxide per liter of smoke.
Henningfield and coworkers (1994) proposed that multiple tests be used:
an average smoking test and a heavy smoking test. The heavy smoking test
would include vent-blocking conditions for those cigarettes incorporating
ventilation holes and if it is possible for those holes to be blocked by the
smoker's lips or fingers.

REFERENCES

Armitage, A.K., Alexander,!., Hopkins, R., Ward, C.
Evaluation of low to middle tar/medium nicotine
cigarette designed to maintain nicotine delivery to
the smoker. Psychopharmacology 96{4): 447-453,
1988.

Ashton, H., Stepney, R., Thompson, J.W. Self-titration
by cigarette smokers. British Medical Journal 2(6186):
357-360, 1979.

Benowitz, N.L., Hall, S.M., Herning, R.I., Jacob, P. Ill,
Jones, R.T., Osman, A.L. Smokers of low-yield
cigarettes do not consume less nicotine. New
England Journal of Medicine 309(3): 139-142, 1983.

Benowitz, N.L., Jacob, P. Ill, Yu, L., Talcott, R., Hall,

S., Jones, R.T. Reduced tar, nicotine, and carbon
monoxide exposure while smoking ultralow- but
not low-yield cigarettes. Journal of the American
Medical Association 256(2): 241-246, 1986.

Benowitz, N.L., Porchet, H., Jacob, P. 111. Nicotine
dependence and tolerance in man:

Pharmacokinetic and pharmacodynamic
investigations. Progress in Brain Research 79: 279-
287, 1989.

Bridges, R.B., Combs, J.G., Humble, J.W., Turbek, J.A.,
Rehm, S.R., Haley, N.J. Puffing topography as a
determinant of smoke exposure. Pharmacology,
Biochemistry and Behavior 37: 29-39, 1990.

Bridges, R.B., Humble, J.W., Turbek, J.A., Rehm, S.R.
Smoking history, cigarette yield and smoking
behavior as determinants of smoke exposure.
European Journal of Respiratory Disease 146(Suppl):
129-137, 1986.

Burling, T.A., Lovett, S.B., Richter, W.T., Frederiksen,
L.W. Alveolar carbon monoxide: The relative
contributions of daily cigarette rate, cigarette
brand, and smoking topography. Addictive
Behaviors 8(1): 23-26, 1983.

Burling, T.A., Stitzer, M.L., Bigelow, G.E., Mead, A.M.
Smoking topography and carbon monoxide levels
in smokers. Addictive Behaviors 10: 319-323, 1985.

Coultas, D.B., Stidley, C.A., Samet, J.M. Cigarette
yields of tar and nicotine and markers of exposure
to tobacco smoke. American Review of Respiratory
Disease 148: 435-440, 1993.

273

Smoking and Tobacco Control Monograph No. 7

Creighton, D.E., Lewis, P.H. The effect of different
cigarettes on human smoking patterns. In: Smoking
Behavior. Physiological and Psychological Influences,
R.E. Thornton (Editor). Edinburgh: Churchill
Livingstone, 1978. pp. 289-300.

Darrall, K.G. Smoking machine parameters and
cigarette smoke yields. Science of the Total
Environment 74: 263-278, 1988.

Dunn, P.J. The effects of a reduced draw resistance
cigarette on human smoking parameters and
alveolar carbon monoxide levels. In: Smoking
Behavior. Physiological and Psychological Influences,
R.E. Thornton (Editor). Edinburgh: Churchill
Livingstone, 1978. pp. 203-207.

Ebert, R.V., McNabb, M.E., McCusker, K.T., Snow, S.L.
Amount of nicotine and carbon monoxide inhaled
by smokers of low-tar, low-nicotine cigarettes.
Journal of the American Medical Association 250(20):
2840-2842, 1983.

Fischer, S., Spiegelhalder, B., Preussmann, R. Influence
of smoking parameters on the delivery of tobacco-
specific nitrosamines in cigarette smoke — A
contribution to relative risk evaluation.
Carcinogenesis 10(6): 1059-1066, 1989.

Fischer, S., Spiegelhalder, B., Preussmann, R. Tobacco-
specific nitrosamines in commercial cigarettes:
Possibilities for reducing exposure. lARC Scientific
Publications (105): 489-492, 1991.

Gori, G.B. Cigarette classification as a consumer
message. Regulatory Toxicology and Pharmacology
12(3 Pt 1): 253-262, 1990a.

Gori, G.B. Consumer perception of cigarette yields: Is
the message relevant? Regulatory Toxicology and
Pharmacology 12(1): 64-68, 1990b.

Gori, G.B., Lynch, C.J. Smoker intake from cigarettes
in the 1-mg Federal Trade Commission tar class.
Regulatory Toxicology and Pharmacology 3(2): 110-
120, 1983.

Gritz, E.R., Rose, J.E., Jarvik, M.E. Regulation of
tobacco smoke intake with paced cigarette
presentation. Pharmacology, Biochemistry and
Behavior 18: 457-462, 1983.

Gust, S.W., Pickens, R.W., Pechacek, T.F. Relation of
puff volume to other topographical measures of
smoking. Addictive Behaviors 8: 1 15-1 19, 1983.

Guyatt, A.R., Kirkham, A., Mariner, D.C., Baldry, A.G.,
Gumming, G. Ixmg-term effects of switching to
cigarettes with lower tar and nicotine yields.
Psychopharmacology 99: 80-86, 1989b.

Guyatt, A.R., Kirkham, A.J. I ., Baldry, A.G., Dixon, M.,
Gumming, G. How does puffing behavior alter
during the smoking of a single cigarette?
Pharmacology, Biochemistry and Behavior 33( 1 ): 1 89-
195, 1989a.

Menningfield, J.E., Griffiths, R.U. Effects of ventilated
cigarette holders on cigarette smoking by humans.
Psyihopharrnacology GH(2): 115-119, 1980.

Henningfield, J.E., Kozlowski, L.T., Benowitz, N.L.

A proposal to develop meaningful labeling for
cigarettes. Journal of the American Medical Association
272(4): 312-314, 1994.

Herning, R.L, Jones, R.T., Benowitz, N.L., Mines, A.H.
How a cigarette is smoked determines blood
nicotine level. Clinical Phannacology and Therapeutics
33: 84-90, 1983.

Hofer, L, Nil, R., Battig, K. Ultralow-yield cigarettes
and type of ventilation: The role of ventilation
blocking. Pharmacology, Biochemistry and Behavior
40(4): 907-914, 1991.

Kolonen, S., Tuomisto, J., Puustinen, P., Airaksinen,
M.M. Smoking behavior in low-yield cigarette
smokers and switchers in the natural environment.
Pharmacology, Biochemistry and Behavior ^0(1): 177-
180, 1991.

Kozlowski, L.T. Tar and nicotine delivery of cigarettes:
What a difference a puff makes. Journal of the
American Medical Association 254(2): 158-159, 1981.

Kozlowski, L.T. Evidence for limits on the acceptability
of lowest tar cigarettes. American Journal ofPiMic
Health 79(2): 198-199, 1989.

Kozlowski, L.T., Frecker, R.C., Khouw, V., Pope, M.A.
The misuse of "less-hazardous" cigarettes and its
detection: Hole-blocking of ventilated filters.
American Journal of Public Health 70(11): 1202-1203,
1980b.

Kozlowski, L.T., Heatherton, T.F., Frecker, R.C., Nolte,
H.E. Self-selected blocking of vents on low-yield
cigarettes. Pharmacology, Biochemistry and Behavior
33(4): 815-819, 1989.

Kozlowski, L.T., Pillitteri, J.L., Sweeney, C.T. Misuse of
"light" cigarettes by means of vent blocking. Journal
of Substance Abuse 6: 333-336, 1994.

Kozlowski, L.T., Pope, M.A., Lux, J.E. Prevalence of the
misuse of ultra-low-tar cigarettes by blocking filter
vents. American Journal of Public Health 78(6): 694-
695, 1988.

Kozlowski, L. r., Rickert, W.S., Pope, M.A., Robinson,
j.C. A color-matching technique for monitoring tar/
nicotine yields to smokers. American Journal of Public
Health 72(6): 597-599, 1982b.

Kozlowski, L.'F., Rickert, W.S., Pope, M.A., Robinson,
J.C., Frecker, R.C. Estimating the yield to smokers
of tar, nicotine, and carbon monoxide from the
"lowest yield" ventilated filter cigarettes. British
Journal of Addiction 77(2): 159-165, 1982a.

Kozlowski, L.'F., Rickert, W.S., Robinson, J.C.,
Grunberg, N.E. Have tar and nicotine yields of
cigarettes changed? Science 209(4464): 1550-1551,
1980a.

Maron, D.J., Fortmann, S.P. Nicotine yield and
measures of cigarette smoke ex[)osure in a large
[)opulation: Are lower-yield cigarettes safer?

American Journal of Public Health 77(5): 546-549,

1987.

274

Section IV

McBride, M.J., Guyatt, A.R., Kirkham, A.J., Gumming,
G. Assessment of smoking behavior and ventilation
with cigarettes of differing nicotine yields. Clinical
Science 67: 619-631, 1984.

Nil, R., Buzzi, R., Battig, K. Effects of different cigarette
smoke yields on puffing and inhalation: Is the
measurement of inhalation volumes relevant for
smoke absorption? Pharmacology, Biochemistry and
Behavior 24: 587-595, 1986.

Pomerleau, O.F., Pomerleau, C.S. Neuroregulators
and the reinforcement of smoking: Towards a
biobehavioral explanation. Neuroscience and
Biobehavioral Reviews 8(4): 503-513, 1984.

Rawbone, R.G. Switching to low-tar cigarettes: Are the
tar league tables relevant? Thorax 39(9): 657-662,
1984.

Rickert, W.S., Collishaw, N.E., Bray, D.F., Robinson,
J.C. Estimates of maximum or average cigarette tar,
nicotine, and carbon monoxide yields can be
obtained from yields under standard conditions.
Preventive Medicine 15(1): 82-91, 1986.

Rickert, W.S., Robinson, J.C. Estimating the hazards
of less hazardous cigarettes. II. Study of cigarette
yields of nicotine, carbon monoxide, and hydrogen
cyanide in relation to levels of cotinine,
carboxyhemoglobin, and thiocyanate in smokers.
Journal of Toxicology and Environmental Health 7:
391-403, 1981.

Rickert, W.S., Robinson, J.C., Collishaw, N.E., Bray,
D.F. Estimating the hazards of "less hazardous"
cigarettes. III. A study of the effect of various
smoking conditions on yields of hydrogen cyanide
and cigarette tar. Journal of Toxicology and
Environmental Health 12(1): 39-54, 1983.

Rickert, W.S., Robinson, J.C., Young, J.C. Estimating
the hazards of "less hazardous" cigarettes. I. Tar,
nicotine, carbon monoxide, acrolein, hydrogen
cyanide, and total aldehyde deliveries of Canadian
cigarettes. Journal of Toxicology and Environmental
Health 6(2): 351-365, 1980.

Robinson, J.C., Young, J.C., Rickert, W.S. A
comparative study of the amount of smoke
absorbed from low-yield ("less hazardous")
cigarettes. Part 1. Non-invasive measures. British
Journal of Addiction 77(4): 383-397, 1982.

Robinson, J.C., Young, J.C., Rickert, W.S. Maintain
levels of nicotine but reduce other smoke
constituents: A formula for "less-hazardous"
cigarettes? Preventive Medicine 13(5): 437-445, 1984.

Robinson, J.C., Young, J.C., Rickert, W.S., Fey, G.,
Kozlowski, L.T. A comparative study of the
amount of smoke absorbed from low-yield ("less
hazardous") cigarettes. Part 2. Invasive measures.
British Journal of Addiction 78(1): 79-87, 1983.

Russell, M.A.H., Jarvis, M.J., Feyerabend, C., Saloojee,
Y. Reduction of tar, nicotine, and carbon monoxide
intake in low-tar smokers. Journal of Epidemiology
and Community Health 40(1): 80-85, 1986.

Russell, M.A.H., Sutton, S.R., Iyer, R., Feyerabend, C.,
Vesey, C.J. Long-term switching to low-tar low-
nicotine cigarettes. British Journal of Addiction 77(2):
145-158, 1982.

Schlotzhauer, W.S., Chortyk, O.T. Effects of varied
smoking machine parameters on deliveries of total
particulate matter and selected smoke constituents
from an ultra-low-tar cigarette. Journal of Analytical
Toxicology 7(2): 92-95, 1983.

Snedecor, G.W., Cochran, W.G. Correlation. In:
Statistical Methods, G.W. Snedecor and W.G.
Cochran (Editors). Ames, lA: Iowa State University
Press, 1980.

U.S. Department of Health and Human Services. The
Health Consequences of Smoking: The Changing
Cigarette. A Report of the Surgeon General. DHHS
Publication No. (PHS)81-50156. Rockville, MD: U.S.
Department of Health and Human Services, Public
Health Service, Office on Smoking and Health,

1981.

Wald, N.J., Boreham, J., Bailey, A. Relative intakes of
tar, nicotine, and carbon monoxide from cigarettes
of different yields. Thorax 39(5): 361-364, 1984.

Woodward, M., Tunstall-Pedoe, H. Do smokers of
lower tar cigarettes consume lower amounts of
smoke components? Results from the Scottish
Health Study. British Journal of Addiction 87(6): 921-
928, 1992.

275

• LIBRARY

Amazing Help.
http://nihlibrary.nih.gov

1 0 Center Drive
Bethesda, MD 20892-1150
301-496-1080

AllONAL

ANCJK

SITUIK

Mil Piibikalioii No. »(>- 102S
August IfHMi

'K)(»7