ooLc
&EPA
Agency
OfficG of Research and
Development
Washington, DC 20460
Respiratory Health
Effects of Passive
Smoking:
Lung Cancer and
Other Disorders
EPA/600/6-90/006B
May t 992
SAB Review Draft
Review
Draft
(Do Not
Cite or
Quote)
Notice
This document is a preliminary draft. It has not been formally released
by EPA and should not at this stage be construed to represent Agency
policy. It is being circulated for comment on its technical accuracy and
policy implications.
-------
-------
EPA/600/6-90/006B
May 1992
SAB Review Draft
RESPIRATORY HEALTH EFFECTS
OF PASSIVE SMOKING:
LUNG CANCER AND OTHER DISORDERS
NOTICE
THIS DOCUMENT IS A PRELIMINARY DRAFT. It has not been formally released by the U.S.
Environmental Protection Agency and should not at this stage be construed to represent Agency
policy. It is being circulated for comment on its technical accuracy and policy implications.
This report has been supported by the Office of Health and Environmental Assessment, Office of
Research and Development, and the Indoor Air Division, Office of Atmospheric and Indoor Air
Programs.
Office of Health and Environmental Assessment
. Office of Research and Development
U.S. Environmental Protection Agency
Washington, D.C.
Printed on Recycled Paper
-------
DRAFT-DO NOT QUOTE OR CITE
DISCLAIMER
This document is a draft for review purposes only and does not constitute Agency policy.
Mention of trade names or commercial products does not constitute endorsement or recommendation
for use.
n
06/01/92
-------
JUN-17-1992 11:20 FROM EPfl OHEft TIS M-PFAFF
TO
B5135697566
P. 02
U.S. Environmental Protection Agency
Science Advisory Board
Indoor Air Quality and Total Human Exposure
Open Meeting
July 21-22, 1992
Pursuant to the Federal Advisory Committee Act,' P,L. 92-463, notice is hereby given
that the Science Advisory Board's (SAB) Indoor Air Quality and Total Human Exposure
, "'.Committee (IAQTHEC) (hereafter, the Committee) will meet on July 21-22, 1992 in the
Main Ballroom of the Holiday Inn, 15th Street and Jefferson Davis Highway, Arlington, VA
22202. The meeting will begin on both days at 9:00 a.m., and end no later than 5:00 p.m.
on July 22, The meeting is open to the public and seating is on a first-come basis.
BACKGROUND
The purpose of the meeting is for the Committee to review the U.S. Environmental
Protection Agency's (EPA) draft report Respiratory Health Affects of Passive Smoking: Lung
Cancer and Other Disorders (EPA/6OO/6-9Q/Q06B). This document was prepared by the
Agency's Human Health Assessment Group, Office of Research and Development (ORD), at
the request of the Agency's Indoor Air Division, Office of Air and Radiation (OAR), under
the authority of Title IV of Superfund (The Radon Gas .and Indoor Air Quality Research Act
of 1986) to provide information and guidance on the potential hazards of indoor air
pollutants. This report is a revision of an earlier report titled, Health Effects of Passive
Smoking: Assessment of Lung Cancer in Adults and Respiratory Disorders in Children
(EPA/600/6-90/006A), which the SAB reviewed in public session on December 4-5, 1990.
As a result of that review, the SAB suggested several areas in which the health risk
assessment could be improved, and offered to provide additional advice on a revised
document (See the SAB's report issued as a result of that review: An SAB Report: Review of
Draft Environmental Tobacco Smoke Health Affects Document, EPA-SAB-IAQC-91-Q07, April
IP&p.^The Agency has now completed its revision of the document and has requested that
the SAB review the revised draft
-------
-------
JUN-17-1992 11!21 FROM EPfl OHEft TIS M.PFPFF
TO
85135697566 P. 03
- CHARGE TO THE COMMITTEE
As pan of the tentative Charge to the Committee, the Agency has
answer the following questions (Chapter numbers refer to the rev:
mem, EPA/6QO/6-90/006B):
that the
EPA
I - ETS EXPOSURE (Chapter 3)
1) Do the conclusions on the chemical similarities of ETS and
mainstream smoke warrant the lexicological comparison between
passive and active smoking made as pan of the biological plausibility
arguments for lung cancer (Chapter 4) and non-cancer respiratory
disorders (Chapter 7)?
2) Is the extent of ETS exposure in various environments adequately
characterized?
3) Are the methods of assessing ETS exposure and the uncertainties
associated with each accurately described?
H - LUNG CANCER
A. HAZARD IDENTliFICATION (Chapters 4 and 5)
4) Is the evidence for the lung carcinogenicity of ETS presented
adequately?
5) Does any of the new information alter the SAB conclusion
regarding the categorization of ETS as an EPA Group A carcinogen?
B. POPULATION IMPACT (Chapter 6)
6) Is fee approach used to derive estimates of U.S. female never-
smoker lung cancer risk scientifically defensible?
7) Is the approach used to extrapolate lung cancer risk from female
never-smokers to male never-smokers and former smoker of both sexes
scientifically defensible?
8) Are the assumptions used to derive these lung cancer population
estimates and the uncertainties involved characterized adequately?
9) Is the degree of confidence in these estimates as stated appropriately
characterized?
IH - NONCANCER RESPIRATORY DISORDERS
A. HAZARD IDENTIFICATION (Chapter 7; Sections 8. land 8,2)
-------
-------
JUN-17-1992 1H21 FROM EPft DHEA TIS M.PFPFF
TO
B51356975SG P.04
Have the feiolofic^plfluaibiHtywgTunaitsbMn adequately
11) Have the me* in^ortam cofci^^
12) Has the weight of evidence been property characterized? Axe the
ooBdosions sdaitiflcsliy defeasible? ' ' U
• ' -"^silF^
13) Ik the evidence with respect to maternal sraoMnf and Hidden fafcnt
deeihiyndrome properly characteiiied? Should fills evidence be
included i& t**^° report?
B. POPUXAttON IMPACT (Chapter 8)
14) It lite prwe&ted population impact ofETS
infections and aathma in children adeatifically defensibie?
IS) Aze the assumptions, unoertaiaties, and degree of ooofid
ranges of pomilation impact ertfanatei
in the
This Charge is subject to change and the Committee nay elect to investigate other area* as
well, .
AVAILABILITY OF DOCUMENTS AND INFORMATION
1) The present EPA draft document (Respiratory Bto&iqfftestf fastfve Smoking;
Lung Cancer and O&tr Disorders (EEA/6QQ/6-9W006B)) win be made available to the
interested public and the Committee on or about June 22 , 1992. Copter ef^dt^ft
document are not avrihhk item tf^ ggfo"sfe AfriTMiy "fttrin? Single copies may be obtained
from the following souroa(s):
ft) Center for Environmental Research InfennatSon (CEKt-FRN), U.S.
Enviroameatal Protection Afency, 26 W. Marti& Luther King Drive,
Cincinnati, OH 45268i tnlaphoffie: <513) 569-7562; FAX: (513) 569-7566.
Please provide the document number (^A/6tXVi^O/006^f and your name
and mailing addie«. Availability may be United, however, frufiyjdialif who
a eoay of the earlier SPA fr«ft ^ffl fi^mttiC^Iv he M«t a eopy^of
b) National TechntcaJ Dofbrmatiott Service (NT1S), 5285 Port Royal Road,
Springfield, VA 22161; telephone: (703) 487-4650. Availability date nay
vaiy, plesue check with imS. The NHS ordering number if FB92*182344.
(oo* $59,00 paper; $19.00 microfiche).
-------
-------
JUN-17-1992 11:22 FROM EPS OHEfl TIS M.PFAFF
TO
S5135697566
P.05
c) The revised draft document will also be available for inspection at the ORD
*ubUc Information Shelf, U.S. EPA Headquarters Library, ^M Street,
S,W., Washington, DC 20*50; the EPA Regional Ubrarief|JSi Ihe Federal
Depository Libraries.
«
*-
2) The earlier EPA draft document (Health Effects of Passive Smoking: Assessment
Cancer in Adults and. Respiratory Disorders in Children (BPA/600/6-90/006A)) is
available only from the following source: National Technical Information Service (NTIS),
' 5285 Port Royal Road, Springfield, VA 22161; telephone: (703) 487-4650. The NTIS
ordering number is PB90-261-652/AS. (cost $35.00 paper; $12.50 microfiche). This
V document was reviewed earlier by the SAB and is flat a subject of the present review.
3) The Science Advisory Board report: Review of Draft Environmental Tobacco
Smoke tfealth Affects Document (EPA-SAB-MQC-92-007) April 1991, is available in single
copies only, from: U.S. EPA, Science Advisory Board (A-101), Office of the Staff Director,
ATTN: Ms. Lori Gross, 401 M Street, SW, Washington, DC 20460 (street and mailing
address are the same); telephone: (202) 260-4126 and FAX: (202) 260-9232. Please provide
the report title, SAB report number and your name and mailing address to obtain a copy.
4) For further information concerning the meeting including a draft agenda, or to
reserve speaking time on the agenda (see below), please contact Mr. Robert Flaak, Assistant
Staff Director, (mailing address: Science Advisory Board Staff Office (A-101F), U.S.
Environmental Protection Agency, 401 M Street, SW, Washington, DC 20460; street
address: Suite 508, 499 South Capitol Street, Washington, DC 20460), telephone: (202) 260-
6552 and FAX: (202) 260-7118- COPIES OF THE EPA DRAFT DOCUMENTS AND .
THE SAB REPORT ARE NOT AVAILABLE FROM THE SAB STAFF OFFICE.
PROCEDURES FOR PROVIDING COMMENTS
The Agency is not soliciting public comment on its dsaft document However, as a
Jural matter, the Science Advisory Board normally accepts either written or oral
; on issues that are under its review. To be most useful, the comments should be
• focused on the particular issues before the Committee, as summarized in the Charge to the
Committee above. Comments submitted to the SAB will be provided to the Committee for
-------
-------
JUN-17-1992 11:22 FROM EPft O-EPl TIS M.PFAFF
TO
B5135697566
P.06
consideration during the feview process. The SAB does not acknowledge receipt of nor does
it provide a response to any public comments received.
1) Oral Cofnmem; Oral comment is taken during a specified period during the
lie meeting (this will be announced in the agenda). Members of the^tifetic who wish to
> a brief oral presentation to the Committee must contact Mr. Flaak in writing (via letter
WFAX) no later than 4:00 p.m. (eastern time) on July 14, 1992 in order to reserve time on
the Agenda. The request must include the name of the person making the presentation,
organizational affiliation represented, a summary of the issue to be discussed (cf.t the Charge
to the Committee above), and identification of any audio-visual requirements. Phone calls
/.are welcome to clarify the process, however, a reservation to speak must still be made in
writing. The SAB expects that public statements presented at its meetings will not be
repetitive of previously submitted oral or written statements. In general, each individual or
group making an oral presentation will be limited to a total time of five minutes. A copy of
the text and copies of any visuals used musH be provided to Mr. Flaak at the time of the
presentation, and will be made part of the public record.
2) Written Cftmnn»nfr Written statements of any length may be provided to the
Committee up until the meeting. Copies of these statements received in the SAB Staff office
by noon (eastern time) on July 6, 1992 will be mailed to the Committee before the meeting;
copies received after that date will be provided to the Committee at the meeting. Members
of the public who submit written comments either before or at the meeting are requested to
provide at least 50 copies of any such documents to Mr. Flaak to allow for adequate
distribution of their position or information. Copies of all comments provided to the SAB as
a result of this review will be made part of the public record and will also be provided to the
Agency for their information.
Date
Dr. Donald Barnes
Staff Director
Science Advisory Board
TOTAL P.06
-------
_
-------
DRAFT-DO NOT QUOTE OR CITE
CONTENTS
Tables • • vm
Figures . . . • '• • • xi"
Preface xv
Authors, Contributors, and Reviewers . . xvi
1. SUMMARY AND CONCLUSIONS 1-1
1.1. BACKGROUND ; 1-1
1.2. PRIMARY FINDINGS 1-4
1.2.1. ETS and Lung Cancer • • 1-5
1.2.2. ETS and Noncancer Respiratory Disorders 1-11
2. INTRODUCTION 2-1
2.1. FINDINGS OF PREVIOUS REVIEWS 2-2
2.2. EPA'S 1992 DOCUMENT 2-4
2.2.1. Scope 2-4
' 2.2.2. Use of EPA's Guidelines 2-6
2.2.3. Contents of This Document 2-8
3. ESTIMATION OF ENVIRONMENTAL TOBACCO SMOKE EXPOSURE 3-1
3.1. INTRODUCTION 3-1
3.2. PHYSICAL AND CHEMICAL PROPERTIES 3-3
3.3. ASSESSING ETS EXPOSURE 3-4
3.3.1. Markers for Environmental Tobacco Smoke 3-4
3.3.2. Measured Exposures to ETS-Associated Nicotine and RSP 3-7
3.3.2.1. Personal Monitors 3-7
3.3.2.2. Measurements Using Stationary Monitors 3-9
3.3.3. Biomarkers of ETS Exposure 3-13
3.3.4. Questionnaires for Assessing ETS Exposures 3-16
3.4. MODELS FOR ASSESSING ETS EXPOSURE . . ; 3-19
3.5. SUMMARY 3-21
4. HAZARD IDENTIFICATION I: LUNG CANCER INACTIVE SMOKERS,
LONG-TERM ANIMAL BIOASSAYS, AND GENOTOXICITY STUDIES 4-1
4.1. INTRODUCTION 4-1
4.2. LUNG CANCER IN ACTIVE SMOKERS 4-2
4.2.1. Time Trends 4-2
in
06/01/92
-------
DRAFT-DO NOT QUOTE OR CITE
CONTENTS (continued)
4.2.2. Dose-Response Relationships 4.3
4.2.3. Histological Types of Lung Cancer and
Associations With Smoking 4.4
4.2.4. Proportion of Risk Attributable to Active Smoking 4-5
4.3. LIFETIME ANIMAL STUDIES 4-5
4.3.1. Inhalation Studies 4.5
4.3.2. Intrapulmonary Implantations of Cigarette Smoke
Condensates ; 4-7
4.3.3. Mouse Skin Painting of Cigarette Smoke Condensates ... 4-7
4.4. GENOTOXICITY 4.8
4.5. SUMMARY AND CONCLUSIONS '.'.'.'.'.'.'.'.'.'. 4-9
5. HAZARD IDENTIFICATION II: INTERPRETATION OF EPIDEMIOLOGIC
STUDIES ON ETS AND LUNG CANCER 5-1
5.1. INTRODUCTION 54
5.2. RELATIVE RISKS USED IN STATISTICAL INFERENCE .'.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'. 5-6
5.2.1. Selection of Relative Risks 5-6
5.2.2. Downward Adjustment to Relative Risk for
Smoker Misclassification Bias 5.5
5.3. STATISTICAL INFERENCE ' ... .1 ........... 5-8
5.3.1. Introduction 5.3
5.3.2. Outcomes by Study and Country 5.9
5.3.2.1. Tests for Association ,..'.. 5-9
5.3.2.2. Confidence Intervals 5.11
5.3.2.3. Tests for Trend 5-12
5.3.2.4. Statistical Conclusions 5-14
5.4. EXTENDED DATA INTERPRETATION '.'.'.'.'.'.'. 5-15
5.4.1. Introduction 5.45
5.4.2. Potential Confounders 548
5.4.2.1. History of Lung Disease 5-19
5.4.2.2. Family History of Lung Disease 5-19
5.4.2.3. Heat Sources for Cooking or Heating . 5.30
5.4.2.4. Cooking With Oil ..'.'.'.'.'. 5-21
5.4.2.5. Occupation . 5_2i
5.4.2.6. Dietary Factors 5_22
5.4.2.7. Summary on Potential Confounders . 5-24
5.4.3. Potential Sources of Bias and Other Uncertainty 5-24
5.4.4. Potential Effects on Individual Studies 5-27
5.4.5. Analysis by Tier and Country 5.39
5.5. CONCLUSIONS FOR HAZARD IDENTIFICATION .'.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'.'. 5-40
5.5.1. Criteria for Causality 5.49
5.5.2. Assessment of Causality 5.41
5.5.3. Conclusion for Hazard Identification 5.43
IV
06/01/92
-------
DRAFT-DO NOT QUOTE OR CITE
CONTENTS (continued)
6. POPULATION RISK OF LUNG CANCER FROM PASSIVE SMOKING . 6-1
6.1. INTRODUCTION 6-1
6.2. PREVIOUS APPROACHES TO ESTIMATION OF POPULATION RISK . 6-1
: 6.2.1. Examples Using Epidemiologic Data . 6-2
6.2.2. Examples Based on Cigarette-Equivalents 6-5
6.3. THIS REPORT'S ESTIMATE OF LUNG CANCER MORTALITY
ATTRIBUTABLE TO ETS IN THE UNITED STATES 6-8
6.3.1. Introduction and Background 6-8
6.3.2. Parameters and Formulae for Attributable Risk 6-10
6.3.3. U.S. Lung Cancer Mortality Estimates Based on Results of
Combined Estimates from 11 U.S. Studies ......... ^ . 6-15
6.3.3.1. U.S. Lung Cancer Mortality Estimates for Female
Never-Smokers 6-15
6.3.3.2. U.S. Lung Cancer Mortality Estimates for Male
Never-Smokers 6-16
6.3.3.3. U.S. Lung Cancer Mortality Estimates for Long-Term
(5+ Years) Former Smokers 6-17
6.3.4. U.S. Lung Cancer Mortality Estimates Based on Results of the
Fontham et al. 1991 Study (FONT) 6-19
6.3.5. Sensitivity to Parameter Values . . . 6-21
6.4 SUMMARY AND CONCLUSIONS ON POPULATION RISK 6-23
7. PASSIVE SMOKING AND RESPIRATORY DISORDERS
OTHER THAN CANCER 7-1
7.1. INTRODUCTION 7-1
7.2. BIOLOGICAL MECHANISMS 7-2
7.2.1. Plausibility ..' 7-2
7.2.2. Effects of Exposure In Utero and During the First
Months of Life , 7-3
7.2.3. Long-Term Significance of Early Effects on
Airway Function 7-6
7.2.4. Exposure to ETS and Bronchial Hyperresponsiveness 7-7
7.2.5. ETS Exposure and Atopy 7-8
7.3. EFFECT OF PASSIVE SMOKING ON ACUTE RESPIRATORY
ILLNESSES IN CHILDREN 7-9
7.3,1. Recent Studies on Acute Lower Respiratory Illnesses . . 7-10
7.3.2. Summary and Discussion of Acute Respiratory Illnesses 7-15
7.4. PASSIVE SMOKING AND ACUTE AND CHRONIC
MIDDLE EAR DISEASES 7-17
7.4.1. Recent Studies on Acute and Chronic Middle Ear Diseases . 7-17
7.4.2. Summary and Discussion of Middle Ear Diseases 7-20
06/01/92
-------
DRAFT-DO NOT QUOTE OR CITE
CONTENTS (continued)
7.5. EFFECT OF PASSIVE SMOKING ON COUGH, PHLEGM,
AND WHEEZING 7-21
7.5.1. Recent Studies on the Effect of Passive Smoking on Cough,
Phlegm, and Wheezing 7-22
7.5.2. Summary and Discussion on Cough, Phlegm, and
Wheezing . . 7-28
7.6. EFFECT OF PASSIVE SMOKING ON ASTHMA 7-30
7.6.1. Recent Studies on the Effect of Passive Smoking on
Asthma in Children 7-30
7.6.2. Summary and Discussion on Asthma 7-34
7.7. ETS EXPOSURE AND SUDDEN INFANT DEATH SYNDROME 7-35
7.8. PASSIVE SMOKING AND LUNG FUNCTION IN CHILDREN 7-39
7.8.1. Recent Studies on Passive Smoking and Lung Function
in Children 7.39
7.8.2. Summary and Discussion on Pulmonary Function
in Children 7-42
7.9. PASSIVE SMOKING AND RESPIRATORY SYMPTOMS AND
LUNG FUNCTION IN ADULTS 7-42
7.9.1. Recent Studies on Passive Smoking and Adult Respiratory
Symptoms and Lung Function 7-43
7.9.2. Summary and Discussion on Respiratory Symptoms and
Lung Function in Adults 7-45
8. ASSESSMENT OF INCREASED RISK FOR RESPIRATORY ILLNESSES IN
CHILDREN FROM ENVIRONMENTAL TOBACCO SMOKE 8-1
8.1. POSSIBLE ROLE OF CONFOUNDING 8-1
8.2. MISCLASSIFICATION OF EXPOSED AND UNEXPOSED SUBJECTS 8-2
8.2.1. Effect of Active Smoking in Children 8-2
8.2.2. Misreporting and Background Exposure 8-3
8.3. ADJUSTMENT FOR BACKGROUND EXPOSURE 8-3
8.4. ASSESSMENT OF RISK 8-8
8.4.1. Asthma 8-8
8.4.2. Lower Respiratory Illness 8-11
8.4.3. Sudden Infant Death Syndrome ; 8-12
8.5 CONCLUSIONS 8-12
REFERENCES R-l
APPENDIX A: REVEEWS OF EPIDEMIOLOGIC STUDIES ON ETS AND
LUNG CANCER A-l
vi 06/01/92
-------
DRAFT-DO NOT QUOTE OR CITE
CONTENTS (continued)
APPENDIX B: METHOD FOR CORRECTING RELATIVE RISK FOR
SMOKER MISCLASSIFICATION
APPENDIX C: REVIEW FORMAT FOR CASE-CONTROL STUDIES
APPENDIX D: LUNG CANCER MORTALITY RATES ATTRIBUTABLE TO
SPOUSAL ETS IN INDIVIDUAL EPIDEMIOLOGIC STUDIES
B-l
C-l
D-l
APPENDIX E: STATISTICAL FORMULAE E-l
Vll
06/01/92
-------
DRAFT-DO NOT QUOTE OR CITE
TABLES
3-1
3-2
3-3
3-4
3-5
4-1
4-2
4-3
4-4
4-5
4-6
4-7
4-8
4-9
Distribution of constituents in fresh, undiluted mainstream smoke- and
diluted sidestream smoke from nonfilter cigarettes .......... ............... 3-24
Studies measuring personal exposure to airborne nicotine associated
with ETS for nonsmokers .......................................... 3_2y
Studies measuring personal exposure to paniculate matter associated
with ETS for nonsmokers .... .................................. 3_2g
Weekly average concentrations of each measure of exposure by parental
smoking status in the cross-sectional study, Minnesota, 1989 ..... ............. 3-29
Approximate relations of nicotine as the parameter between
nonsmokers, passive smokers, and active smokers ................... . ..... 3.30
Main characteristics of major cohort studies on the
relationship between smoking and cancer ............................... 4_U
Lung cancer mortality ratios-prospective studies ................ . ...... ... 4.13
Lung cancer mortality ratios for men and women, by current
number of cigarettes smoked per day-prospective studies .................... 4-14
Relationship between risk of lung cancer and duration of smoking in
men, based on available information from cohort studies ..................... 4-15
Lung cancer mortality ratios for males, by age of
smoking initiation-prospective studies . . .................. . . ........... 4.15
Relationship between risk of lung cancer and number of years
since stopping smoking, in men, based on available information
from cohort studies ................... ....................... 4_17
Relative risks of lung cancer in some large cohort studies among
men smoking cigarettes and other types of tobacco ......................... ,4-18
Age-adjusted lung cancer mortality ratios for males and females,
by tar and nicotine (T/N) in cigarettes smoked ........ .................... 4_20
Relative risk for lung cancer by type of cigarette smoked (filter vs.
nonfilter), in men, based on cohort and case-control studies ................... 4-21
vm
06/01/92
-------
DRAFT-DO NOT QUOTE OR CITE
TABLES (continued)
4-10 Main results of studies dealing with the relationship between .
smoking and different histological types of lung cancer 4-22
4-11 Lung cancer death attributable to tobacco smoking in certain countries . 4-27
5-1 Epidemiologic studies on ETS and lung cancer in this report 5-44
5-2 Studies by location, time, size, and ETS exposure 5-46
5-3 Case-control studies of ETS: characteristics . 5-48
5-4 Estimated relative risk of lung cancer from spousal ETS as reported
by epidemiologic study • • 5-51
5-5 Effect of statistical adjustments for cofactors on risk estimates
for passive smoking • 5-55
5-6 Alternative estimates of lung cancer relative risks associated
with active and passive smoking . • • • 5-58
5-7 Estimated correction for smoker misclassification , • • • • 5-60
5-8 Statistical measures by individual study and pooled by country,
corrected for smoker misclassification . .,-. . . ... 5-63
5-9 Case-control and cohort studies: exposure response trends for females 5-66
5-10 Reported p values of trend tests for ETS exposure by study 5-70
5-11 P values of tests for effect and for trend by individual study ..;......... 5-71
5-12 Other risk-related factors for lung cancer evaluated in selected studies ..........: 5-73
5-13 Dietary effects in passive smoking studies of lung cancer in females ....... . . . . : 5-74
5-14A Study limitations and sources of uncertainty 5-77
5-14B Study limitations and sources of uncertainty • • 5-79
5-15 Diagnosis, confirmation, and exclusion of lung cancer cases 5-80
5-16 Classification of studies by tier • 5-82
IX
06/01/92
-------
DRAFT-DO NOT QUOTE OR CITE
TABLES (continued)
5-17 Summary data interpretation by country and by tier 5.04
6-1 Definition and estimates of relative risk of lung cancer for
11 U.S. studies combined for various exposure sources and
baselines , _'
o-2o
6-2 Estimated female lung cancer mortality by attributable sources
for United States, 1985, using the pooled relative risk estimate
from 11 U.S. studies ^ 0~
0-27
6-3 Female and male lung cancer mortality estimates by attributable
ETS sources for United States, 1985, using 11 U.S. studies
(never-smokers and former smokers who have quit 5+ years) 5.29
6-4 Female lung cancer mortality estimates by attributable sources
for United States, 1985, using both the relative risk estimates
and Z-values from the Fontham et al. study 6.31
6-5 Female and male lung cancer mortality estimates by attributable
ETS sources for United States, 1985, using the Fontham et al. study
(never-smokers and former smokers who have quit 5+ years) 6-33
6-6 Effect of single parameter changes on lung cancer mortality due
to ETS in never-smokers and former smokers who have quit 5+ years 6-35
7-1 Studies on respiratory illness referenced in the Surgeon
General's and National Research Council's reports of 1986 7.43
7-2 Recent epidemiologic studies of effects of passive smoking on
acute lower respiratory tract illnesses (LRIs) 7.49
7-3 Studies on middle ear diseases referenced in the Surgeon
General's report of 1986 7 54
7-4 Recent epidemiologic studies of effects of passive smoking on
acute and chronic middle ear diseases 7 55
7-5 Studies on chronic respiratory symptoms referenced in the Surgeon
General's and National Research Council's reports of 1986 •. 7.53
7-6 Recent epidemiologic studies of effects of passive smoking on
cough, phlegm, and wheezing 7_59
06/01/92
-------
DRAFT-DO NOT QUOTE OR CITE
TABLES (continued)
7-7
7-8
7-9
7-10
7-11
Recent epidemiologic studies of effects of passive smoking on
asthma in childhood
Epidemiologic studies of effects of passive smoking on
incidence of sudden infant death syndrome (SIDS)
Studies on pulmonary function referenced in the Surgeon
General's and National Research Council's reports of 1986
Recent epidemiologic studies on the effects of passive smoking
on lung function in children
8-3
B-l
B-2
B-3
B-4
B-5
B-6
B-7
Recent epidemiologic studies on the effects of passive smoking
on adult respiratory symptoms and lung function . . . .,
Adjusted relative risks for "exposed children" . •.
Behavior variations in adjusted relative risks from equation 8-1 when the
observed relative risks and Z ratios are close together
Range of estimates of adjusted relative risk and attributable
risk for asthma induction in children based on both threshold
and nonthreshold models
Observed ratios of occasional smokers to current smokers
(based on cotinine studies)
Differences in smoker misclassification bias between EPA estimates and
those of P.N. Lee regarding passive smoking relative risks for females .
Misclassification of female current smokers
Misclassification of female former smokers reported as never-smokers
based on discordant answers
Misclassification of female lung cancer cases
Notation for proportionate distribution of reported female lung cancer
cases and controls by husband's smoking status
Proportionate distribution notation for subjects by observed
and true smoking status
7-65
7-68
7-70
7-71
7-73
8-15
8-15
8-16
B-13
B-14
B-16
B-18
B-19
B-20
B-21
XI
06/01/92
-------
DRAFT-DO NOT QUOTE OR CITE
TABLES (continued)
B-8 Observed lung cancer relative risks for exposed and nonexposed wives
by the wife's smoking status using average never-smoking wives
as the reference category _
B-9 Prevalences and estimates of lung cancer risk associated with active
and passive smoking , -. .„
O-2.3
B-10 Observed smoking prevalence among the controls in Correa example B.27
B-ll Observed relative risks-Correa example B 27
B-12 Crude case table-prevalence of cases by smoking status in Correa example ....... B-28
B-13 Normalized case table-prevalence of cases by smoking status in
Correa example -
B-28
B-14 Proportionate distribution of observed and true smoking status for wives
in Correa example
B-15 Deletions from the never columns in Tables B-10 and B-13 B_30
B-16 Observed ratios of female former smokers to ever-smokers
in the USA, UK, and Sweden: populations or controls (numbers or %) B-31
B-17 Observed ratios of current smoker lung cancer risk to ever-smoker
risk for females
B-32
D-l Female lung cancer mortality from all causes in case-control studies D-4
D-2 Parameter values used to partition female lung cancer mortality
into component sources _ ,
£)_(j
D-3 Female lung cancer mortality rates by attributable source D_7
D-4 Lung cancer mortality rates of female ever-smokers (ES) and never-smokers (NS)
by exposure status : ~ 0
D-8
xn
06/01/92
-------
DRAFT-DO NOT QUOTE OR CITE
FIGURES
3-1 Cumulative frequency distribution of RSP mass concentrations from
central site ambient and personal monitoring of smoke-exposed and
nonsmoke-exposed individuals < . 3-31
3-2 Mean, standard deviation, maximum, and minimum nicotine values measured
in different indoor environments with smoking occupancy . 3-32 •
3-3 Weeklong RSP mass and nicotine measurements in 96 residences
with a mixture of sources. . 3-34
3-4 Mean, standard deviation, maximum, and minimum concentrations
of RSP mass measured in different indoor environments for smoking and
nonsmoking occupancy 3-35
3-5 Range of average nicotine concentrations and range of maximum
and minimum values measured by different indoor environments
for smoking occupancy from studies shown in Figure 3-2 3-36
3-6 Range of average RSP mass concentrations and range of maximum
and minimum values measured by different indoor environments
for smoking occupancy from studies shown in Figure 3-4 3-37
3-7 Cumulative frequency distribution and arithmetic means of RSP vapor-phase
nicotine levels, measured over a 1-week period in the main living area in
residences in Onondaga and Suffolk Counties in New York State between
January and April 1986 3-38
3-8 Cumulative frequency distribution and arithmetic means of RSP mass levels
by vapor-phase nicotine levels, measured over a 1-week period in the main
living area in residences in Onondaga and Suffolk Counties in New York State
between January and April 1986 3-39
3-9 Monthly mean RSP mass concentrations in six U.S. cities 3-40
3-10a Weeklong nicotine concentrations, measured in the main living area
of 96 residences versus the number of questionnaire-reported cigarettes
smoked during the air-sampling period 3-41
3-10b Weeklong RSP mass concentrations, measured in the main living area
of 96 residences versus the number of questionnaire-reported cigarettes
smoked during the air-sampling period 3-41
3-11 Average cotinine tw by age groups 3-42
xin
06/01/92
-------
DRAFT-DO NOT QUOTE OR CITE
FIGURES (continued)
3-12
3-13
3-14
3-15
3-16
4-1
4-2
4-3
5-1
5-2
5-3
5-4
5-5
5-6
Distribution of individual concentration of urinary cotinine
self-reported exposure to ETS 3_43
Urinary cotinine concentrations by number of reported exposures to
tobacco smoke in the past 4 days among 663 nonsmokers, Buffalo
New York, 1986 ; 3_44
Average cotinine/creatinine levels for subgroups of
nonsmoking women defined by sampling categories of exposure
or by self-reporting exposure to ETS from different sources
during the 4 days preceding collection of the urine sample 3.45
Diagram for calculating the RPS mass from ETS emitted into any
occupied space as a function of the smoking age and removal rate (N) 3.45
Diagram to calculate the ETS-associated RSP mass concentration in
a space as a function of total mass of ETS-generated RSP emitted (determined
from Figure 3-15) and the volume of a space (diagonal lines) 3.47
Age-adjusted cancer death rates for selected sites, males,
United States, 1930-1986 4.28
Age-adjusted cancer death rates for selected sites, females,
United States, 1930-1986 4_29
Relative risk of lung cancer in ex-smokers, by number of years
quit, women, Cancer Prevention Study II 4_30
Test statistics for hypothesis RR = 1, all studies 5.35
Test statistics for hypothesis RR = 1, USA only 5.35
Test statistics for hypothesis RR = 1, by country 5_86
Test statistics for hypothesis RR = 1, China without WUWI and LIU 5-86
90% Confidence Intervals, by country 5.07
90% Confidence Intervals, China without WUWI and LIU 5.37
xiv
06/01/92
-------
DRAFT-DO NOT QUOTE OR CITE
PREFACE
This assessment of the respiratory health effects associated with passive smoking has been
prepared by the Human Health Assessment Group, Office of Health and Environmental Assessment,
Office of Research and Development, which is responsible for its scientific accuracy and conclusions.
The assessment was prepared at the request of the Indoor Air Division, Office of Atmospheric and
Indoor Air Programs, Office of Air and Radiation, which defined its scope and provided funding.
The document has been developed under the authority of Title IV of Superfund (The Radon
Gas and Indoor Air Quality Research Act of 1986) to provide information and guidance on the
potential hazards of indoor air pollutants.
An earlier draft of this document was made available for public review and comments in June
1990, and was reviewed by the Agency's Science Advisory Board in December 1990. This revision
reflects the comments received from those reviews, plus additional comments from an internal review
conducted in February and March 1992.
A comprehensive search of the scientific literature for this revision is complete through
September 1991. In addition, a few studies published since then have been included in response to
recommendations made by reviewers.
Due to both resource and time constraints, the scope of this report has been limited to an
analysis of respiratory effects, primarily lung cancer in nonsmoking adults and noncaricer respiratory
illnesses in children, with emphasis on the epidemiologic data. Further, because two thorough reviews
on passive smoking were completed in 1986 (by the U.S. Surgeon General and the National Research
Council), this document provides a summary of those reports with a more comprehensive analysis of
the literature appearing subsequent to those reports and an integration of the results.
It is the Agency's intention with the release of this draft to seek additional advice from its
Science Advisory Board in preparation for release of a final report later this year.
xv
06/01/92
-------
DRAFT-DO NOT QUOTE OR CITE
AUTHORS, CONTRIBUTORS, AND REVIEWERS
This document was prepared by the Office of Health and Environmental Assessment (OHEA)
within the Office of Research and Development, with major contract funding provided by the Indoor
Air Division within the Office of Air and Radiation's Office of Atmospheric and Indoor Air Programs,
Steven P. Bayard served as OHEA project manager with overall responsibility for contents of this
report and its conclusions.
Other OHEA staff responsible for the scientific content of sections of this document are
Apama M. Koppikar1 and Jennifer Jinot1. Jennifer Jinot also served as a contributor and technical
editor for a major portion of this report.
AUTHORS
Major portions of this revised report were prepared by ICF Incorporated, Fairfax, Virginia,
under EPA Contract No. 68-00-0102. A list of authors follows:
Chapter 1: Steven P. Bayard1
Chapter 2: Jennifer Jinot1
Chapter 3: Brian P. Leaderer2
Chapter 4: Jennifer Jinot
Chapters 5/6: Kenneth G. Brown3
'Human Health Assessment Group, Office of Health and Environmental Assessment, U.S EPA
Washington, DC 20460.
2J. P. Pierce Foundation Laboratory, Department of Epidemiology and Public Health, Yale
University School of Medicine, New Haven, CT 06520. Subcontractor to ICF Inc.
3Kenneth G. Brown, Inc., P. O. Box 16608, Chapel Hill, NC 27516. Subcontractor to ICF Inc.
xvi
06/01/92
-------
DRAFT-DO NOT QUOTE OR CITE
Chapters 7: Fernando D. Martinez4
Chapters 8: Fernando D. Martinez and Steven P. Bayard
Appendix A: Kenneth G. Brown, Neal R. Simonsen3, and A. Judson Wells3
Appendix B: A. Judson Wells
Appendix C: Kenneth G. Brown
Appendix D: Kenneth G. Brown and Neal R. Simonsen
Appendix E: Kenneth G. Brown
CONTRIBUTORS
Numerous persons have provided helpful discussions or responded to requests for pre-prints,
data, and other material relevant to this report. The authors are grateful to W.J. Blot, N. Britten, R.C.
Brownson, P.A. Buffier, T.L. Butler, D.B. Coultas, K.M. Cummings, J. Fleiss, E.T.H. Fontham, Y.T.
Gao, L. Garfinkel, S. Glantz, N.J. Haley, T. Hirayama, D.J. Hole, C. Humble, G.C. Kabat, J.C.
Kleinman, L.C. Koo, M. Layard, P.N. Lee, M.D. Lebowitz, P. Macaskill, G.J. Knight, G.E. Palomaki,
J.P. Pierce, J. Repace, H. Shimizu, W.F. Stewart, D. Trichopoulos, A. Wu-Williams, and R.W. Wilson.
3Kenneth G. Brown, Inc., P. O. Box 16608, Chapel Hill, NC 27516. Subcontractor to ICF Inc.
4Division of Respiratory Sciences, University of Arizona Medical Center, Tucson, AZ 85724.
Subcontractor to ICF Inc. '
xvii
06/01/92
-------
DRAFT-DO NOT QUOTE OR CITE
REVIEWERS ;
This document is a revision of an earlier External Review praft ,(EPA/600/6-90/006A) that was
released for public review and comment on June 25, 1990. The draft document subsequently was
reviewed by the EPA Science Advisory Board (SAB) on December 4 and 5, 1990. Many of the
revisions follow closely the valuable advice presented in the SAB's April 19, 1991, report to the
Agency. Other revisions are based on comments received from peer reviewers and the public. In
addition, many reviewers both within and outside the 'Agency provided assistance at various internal
review stages.
The following members of the SAB's Indoor Air Quality and Total Human Exposure
Committee (IAQTHEC) participated in the review of the External Review Draft.
Chairman , .
'-'•i"'-:';'^;::;:; < ~ i.,: , ,,
Dr. Morton Lippmann, Professor, .Institute of Environmental Medicine^ New York Wversity Medical
Center, Tuxedo, NY 10987 '
Vice Chairman :
Dr. Jan AJ. Stolwijk, Professor, School of Medicine, Department of Epidemiology and Public Health
Yale University, 60 College Street, New Haven, CT 06510
Members of the IAQTHEC
Dr. Joan Daisey, Senioor Scientist, Indoor Environment Program, Lawrence Berkeley Laboratory One
Cyclotron Road, Berkeley, CA 94720
Dr. Victor G. Laties, Professor of Toxicology, Environmental Health Science Center-Box EHSC
University of Rochester School of Medicine, Rochester, NY 14642
Dr. Jonathan M. Samet, Professor of Medicine, Department of Medicine, The Univeristy of New
Mexico School of Medicine, and The New Mexico Tumor Registry, 900 Camino De Salud
ME, Albuquerque, NM 87131
Dr. Jerome J. Wesolowski, Chief, Air and Industrial Hygiene Laboratory, California Department of
Health, Berkeley, CA 94704
Dr. James E. Woods, Jr., Professor of Building Construction, College of Architecture and Urban
- -es, 117 Burress Hall, Virginia Polytechnic Institute and State University, Blacksburg, VA
XVlll
06/01/92
-------
DRAFT-DO NOT QUOTE OR CITE
Consultants to the IAQTHEC
Dr. Neal L. Benowitz, Professor of Medicine, Chief, Division of Clinical Pharmacology and
Experimental Therapeutics, University of California, San Francisco, Building 30, Fifth Floor,
San Francisco General Hospital, 1001 Potrero Avenue, San Francisco, CA 94110
Dr. William J. Blot, National Cancer Institute, 9000 Rockville Pike, Bethesda, MD 20892 (Federal
Liaison to the Committee)
Dr. David Burns, Associate Professor of Medicine, Department of Medicine, University of California,
San Diego Medical Center, 225 Dickenson Street, San Diego, CA 92103-1990
Dr. Delbert Eatough, Professor of Chemistry, Brigham Young University, Provo, UT 84602
Dr. S. Katharine Hammond, Associate Professor, Environmental Health Sciences Program,
Department of Family and Community Medicine, University of Massachusetts Medical School,
55 Lake Avenue, North, Worcester, MA 06155
Dr. Geoffrey Kabat, Senior Epidemiologist, American Health Foundation, 320 East 43rd Street, New
York, NY 10017
Dr. Michael D. Lebowitz, Professor of Internal Medicine, University of Arizona College of Medicine,
Division of Respiratory Sciences, Tucson, AZ 85724
Dr. Howard Rockette, Professor of Biostatistics, School of Public Health, 318 Parran Hall, University
of Pittsburgh, Pittsburgh, PA 15261
Dr. Scott T. Weiss, Charming Laboratory, Harvard University School of Medicine, Boston, MA 02115
xix
06/01/92
-------
-------
DRAFT—DO NOT QUOTE OR CITE
1. SUMMARY AND CONCLUSIONS
1.1. BACKGROUND
Tobacco smoking has long been recognized (e.g., U.S. DHEW, 1964) as a major cause of
mortality and morbidity, responsible for an estimated 434,000 deaths per year in the United States
(CDC, 199la). Tobacco use is known to cause cancer at various sites, in particular the lung (U.S.
DHHS, 1982; IARC, 1986). Smoking can also cause respiratory diseases (U.S. DHHS, 1984, 1989)
and is a major risk factor for heart disease (U.S. DHHS, 1983). In recent years there has been
concern that nonsmokers may also be at risk for some of these health effects as a result of their
exposure ("passive smoking") to the tobacco smoke that occurs in various environments occupied
by smokers. Although this environmental tobacco smoke (ETS) is dilute compared to the
mainstream smoke (MS) inhaled by active smokers, it is chemically similar, containing many of
the same carcinogenic and toxic agents.
In 1986, the National Research Council (NRC) and the Surgeon General of the U.S. Public
Health Service independently assessed the health effects of exposure to ETS (NRC, 1986; U.S.
DHHS, 1986). Both of the 1986 reports conclude that ETS can cause lung cancer in adult
nonsmokers and that children of parents who smoke have increased frequency of respiratory
symptoms and acute lower respiratory tract infections, as well as evidence of reduced lung
function.
More recent epidemiologic studies of the potential associations between ETS and lung
cancer in nonsmoking adults and between ETS and noncancer respiratory effects more than
double the size of the database available for analysis from that of the 1986 reports. This U.S. EPA
document critically reviews the current database on the respiratory health effects of passive
smoking, and these data are utilized to develop a hazard identification for ETS and to make
quantitative estimates of the public health impacts of ETS for lung cancer and various other
respiratory diseases.
The weight-of-evidence analysis for the lung cancer hazard identification is developed in
accordance with U.S. EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA, 1986a) and
established principles for evaluating epidemiologic studies. The analysis considers animal
bioassays and genotoxicity studies, as well as biological measurements of human uptake of tobacco
smoke components and epidemiologic data on active and passive smoking. The availability of
abundant and consistent human data, and especially human data at actual environmental levels of
exposure to the specific agent (mixture) of concern, allow a hazard identification to be made with
1-1
05/31/92
-------
DRAFT—DO NOT QUOTE OR CITE
a high degree of certainty. The conclusive evidence of the dose-related lung carcinogenicity of
MS in active smokers (Chapter 4), coupled with information on the chemical similarities of MS
and ETS and evidence of ETS uptake in nonsmokers (Chapter 3), is sufficient by itself to establish
ETS as a known human lung carcinogen, or "Group A" carcinogen under U.S. EPA's carcinogen
classification system. In addition, this document concludes that the overall results of 30
epidemiologic studies on lung cancer and passive smoking (Chapter 3), using spousal smoking as a
surrogate of ETS exposure for female never-smokers, similarly justify a Group A classification.
The weight-of-evidence analyses for the noncancer respiratory effects are based primarily
on a review of epidemiologic studies (Chapter 7). Most of the endpoints examined are respiratory
disorders in children, where parental smoking is used as a surrogate of ETS exposure. For the
noncancer respiratory effects in nonsmoking adults, most studies used spousal smoking as an
exposure surrogate. A causal association was concluded to exist for a number of respiratory
disorders where there was sufficient consistent evidence for a biologically-plausible association
with ETS that could not be explained by bias, confounding, or chance. The fact that the database
consists of human evidence from actual environmental exposure levels gives a high degree of
confidence in this conclusion. Where there was suggestive but inconclusive evidence of causality,
as was the case for asthma induction in children, ETS was concluded to be a risk factor for that
endpoint. Where data were inconsistent or inadequate for evaluation of an association, as for
acute upper respiratory tract infections and acute middle ear infections in children, no conclusions
were drawn.
This report has also attempted to provide estimates of the extent of the public health
impact, where appropriate, in terms of numbers of ETS-'attributable cases in nonsmoking
subpopulations. Unlike for qualitative hazard identification assessments where information from
many sources adds to the confidence in a weight-of-evidence conclusion, for quantitative risk
assessments the usefulness of studies usually depends on how closely the study population
resembles nonsmoking segments of the general population. For lung cancer estimates among U.S.
nonsmokers, the substantial epidemiology database of ETS and lung cancer among U.S. female
never-smokers was considered to provide the most appropriate information. From the large
number of similarly designed studies, pooled relative risk estimates were calculated and used in
the derivation of the population risk estimates. The large number of studies available, the
generally consistent results, and the condition of actual environmental levels of exposure increase
the confidence in these estimates. Even with these conditions, however, uncertainties remain,
such as in the use of questionnaires and current biomarker measurements to estimate past
1-2
05/31/92
-------
DRAFT—DO NOT QUOTE OR CITE
exposure, assumptions of exposure-response linearity, and extrapolation to male never-smokers
and to exsmokers. Still, given the strength of the evidence for the lung carcinogenicity of tobacco
smoke and the extensive human database from actual environmental exposure levels, fewer
assumptions are necessary than is usual in U.S. EPA quantitative risk assessments and confidence
in these estimates is rated medium to high.
Population estimates of ETS health impacts are also made for certain noncancer respiratory
endpoints in children, specifically lower respiratory tract infections (LRIs, i.e. pneumonia,
bronchitis, and bronchiolitis) and episodes and severity of attacks of asthma. Estimates of ETS-
attributable cases of LRI in infants and young children are thought to have a high degree of
confidence because of the consistent study findings and the appropriateness of parental smoking
as a surrogate measure of exposure in very young children. Estimates of the number of asthmatic
children whose condition is aggravated by exposure to ETS are less certain than those for LRIs
because of different measures of outcome in various studies and because of increased
extraparental exposure to ETS in older children. Estimates of the number of new cases of asthma
in previously asymptomatic children also have less confidence because at this time the weight-of-
evidence for asthma induction, while suggestive of a causal association, is not conclusive.
Most of the ETS population impact estimates are presented in terms of ranges, which are
thought to reflect reasonable assumptions about the estimates of parameters and variables required
for the extrapolation models. The validity of the ranges is also dependent on the appropriateness
of the extrapolation models themselves.
While this report focuses only on the respiratory health effects of passive smoking, there
may also be other health effects of concern. Recent analyses of more than a dozen epidemiology
and toxicology studies (Steenland, 1992; NIOSH, 1991) suggest that ETS exposure may be a risk
factor for cardiovascular disease. In addition, there were a few studies in the literature linking
ETS exposure to cancers of others sites; at this time, that database appears inadequate for any
conclusion. This report does not develop an analysis of either the nonrespiratory cancer or the
heart disease data and takes no position on whether ETS is a risk factor for these diseases. If it is,
the total public health impact from ETS will be greater than that discussed here.
1-3
05/31/92
-------
DRAFT—DO NOT QUOTE OR CITE
1.2. PRIMARY FINDINGS
A. Lung Cancer in Nonsmoking Adults
1. Passive smoking is causally associated with lung cancer in adults, and ETS, by
the total weight-of-evidence, belongs in the category of compounds classified
by EPA as Group A (known human) carcinogens.
2. An estimated range of 2,500 to 3,300 lung cancer deaths per year among
nonsmokers (never-smokers and former smokers) of both sexes are
attributable to ETS in the United States. The confidence in this range is
medium to high with approximately 3,000 annual lung cancer deaths
representing the best estimate.
B. Noncancer Respiratory Diseases and Disorders
1. Exposure of children to ETS from parental smoking is causally associated
with:
a. increased prevalence of respiratory symptoms of irritation
(cough, sputum, and wheeze),
increased prevalence of middle ear effusion (a sign of middle ear
disease), and
a small but statistically significant reduction in lung function as
tested by objective measures of lung capacity.
ETS exposure of young children and particularly infants from parental (and '
especially mother's) smoking is causally associated with an increased risk of
lower respiratory tract infections (pneumonia, bronchitis, and bronchiolitis).
This report estimates that exposure to ETS contributes 150,000 to 300,000
lower respiratory tract infections annually in infants and children less than
18 months of age, resulting in 7,500 to 15,000 hospitalizations. These higher
risks continue at a decreasing rate for children until about age 3, but no
estimates are derived for children over 18 months.
a. Exposure to ETS is causally associated with additional episodes and
increased severity of asthma in children who already have the disease.
This report estimates that ETS exposure exacerbates symptoms in
approximately 20% of this country's 2 million to 5 million asthmatic
children and is a major aggravating factor in approximately 10%.
b.
c.
2.
3.
1-4
05/31/92
-------
DRAFT—DO NOT QUOTE OR CITE
b. In addition, the epidemiologic evidence is suggestive but not conclusive
that ETS exposure increases the number of new cases of asthma in
children who have not previously exhibited symptoms. Based on this
evidence and the known ETS effects on both the immune system and
lungs (e.g. atopy and airway hyper responsiveness), this report concludes
that ETS is a risk factor for the induction of asthma in previously
asymptomatic children. Data suggest that relatively high levels of
exposure are required to induce new cases of asthma in children. This
report estimates that previously asymptomatic children exposed to ETS
from mothers who smoke at least 10 cigarettes per day will exhibit a
probable range of 8,000 to 26,000 new cases of asthma annually. The
confidence in this range is medium and is dependent on the conclusion
that ETS is a risk factor for asthma induction.
4. Passive smoking has subtle but significant effects on the respiratory health of
nonsmoking adults, including coughing, phlegm, chest discomfort, and
reduced lung function.
This report also has reviewed data on the relationship of maternal smoking and sudden
infant death syndrome (SIDS), which is thought to involve some unknown respiratory
pathogenesis. The report concludes that while there is strong evidence that infants whose mothers
smoke are at an increased risk of dying from SIDS, available studies do not allow us to
differentiate whether and to what extent this increase is related to in utero versus postnatal
exposure to tobacco smoke products. Consequently, at this time this report is unable to assert
whether or not ETS exposure by itself is a risk factor for SIDS independent of smoking during
pregnancy. Postnatal exposure may potentiate effects of in utero tobacco smoke exposure, or it
may not have any additional effect.
Regarding an association of parental smoking with either upper respiratory tract infections
(colds and sore throats) or acute middle ear infections in children, this report finds the evidence
inconclusive.
1.2.1. ETS and Lung Cancer
The Surgeon General (U.S. DHHS, 1989) estimated that smoking was responsible for more
than one of every six deaths in the United States and that it accounted for about 90% of the lung
cancer deaths in males and about 80% in females in 1985. Smokers, however, are not the only
1-5
05/31/92
-------
DRAFT—DO NOT QUOTE OR CITE
ones exposed to tobacco smoke. The sidestream smoke (SS) emitted from a smoldering cigarette
between puffs (the main component of ETS) has been documented to contain many of the same
carcinogenic compounds (known and suspected human and animal carcinogens) that have been
identified in the mainstream smoke (MS) inhaled by smokers. Exposure concentrations of these
carcinogens to passive smokers are variable but much lower than for active smokers. An excess
cancer risk from passive smoking, however, is biologically plausible.
Based on the firmly established causal association of lung cancer with active smoking with
a dose-response relationship down to low doses (Chapter 4), passive smoking is considered likely
to affect the lung similarly. The widespread presence of ETS in both home and workplace and its
absorption by nonsmokers in the general population have been well documented by air sampling
and by body measurement of biomarkers such as nicotine and cotinine (Chapter 3). This raises the
question of whether any direct evidence exists for the relationship between ETS exposure and
lung cancer in the general population and what its implications may be for public health. This
report addresses that question by reviewing and analyzing the evidence from 30 epidemiologic
studies of effects from normally occurring environmental levels of ETS (Chapter 5). Because
there is widespread exposure and it is difficult to construct a truly unexposed subgroup of the
general population, these studies compare individuals with higher ETS exposure to those with
lower exposures. Typically, female never-smokers who are married to a smoker are compared
with female never-smokers who are married to a nonsmoker. Some studies also consider ETS
exposure of other subjects (i.e., male never-smokers and long-term former smokers of either sex)
and from other sources (e.g., workplace and home exposure during childhood), but these studies
are fewer and represent fewer cases, and they are generally excluded from the analysis presented
here. Use of the female never-smoker studies provides the largest, most homogeneous database
for analysis to determine whether an ETS effect on lung cancer is present. This document
assumes that the results for female never-smokers are generalizable to all nonsmokers.
Given that ETS exposures are at actual environmental levels and that the comparison
groups are both exposed to appreciable background (i.e., nonspousal) ETS, any excess risk for lung
cancer from exposure to spousal smoke would be expected to be small. Furthermore, the risk of
lung cancer is relatively low in nonsmokers, and most studies have a small sample size, resulting in
a very low statistical power (probability of detecting a real effect if it exists). Besides small
sample size and low incremental exposures, other problems inherent in several of the studies may
also limit their ability to detect a possible effect. Therefore, this document examines the data in
several different ways. After downward adjustment of the relative risks for smoker
1-6
05/31/92
-------
DRAFT—DO NOT QUOTE OR CITE
misclassification bias, the studies are individually assessed for strength of association and
exposure-response trend. Then the study results are pooled by country using statistical techniques
for combining data, including both positive and nonpositive results, to increase the ability to
determine whether or not there is an association between ETS and lung cancer. .Finally, in
addition to the previous statistical analyses that weigh the studies only by size, regardless of design
and conduct, the studies are qualitatively evaluated for potential confounding, bias, and likely
utility to provide information about any lung carcinogenicity of ETS. Based on these qualitative
considerations, the studies are categorized into one of four tiers and then statistically analyzed
successively by tier.
Results from all of the analyses described above strongly support a causal association
between lung cancer and ETS exposure. The overall proportion of individual studies found to
show an association between lung cancer and ETS exposure is unlikely to occur by chance
(p < 0.005). Similarly, the proportion showing a statistically significant dose-response-trend
(p < 10'9) is highly supportive of a causal association. Combined results by country showed
statistically significant associations for Greece (2 studies), Hong Kong (4 studies), Japan (5
studies), and the United States (11 studies), and in that order of strength of relative risk. Pooled
results of the four Western European studies (three countries) actually showed a slightly stronger
association than that of the United States, but it was not statistically significant, probably due to
the smaller sample size. The combined results of the Chinese studies do not show an association
between ETS and lung cancer; however, two of the four Chinese studies were designed mainly to
determine the lung cancer effects of high levels of other indoor air pollutants indigenous to those
areas, which would obscure a smaller ETS effect. These two Chinese studies do, however, provide
very strong evidence on the lung carcinogenicity of these other indoor air pollutants, which
contain many of the same components as ETS. When results are combined only for the other two
Chinese studies, they demonstrate a statistically significant association for ETS and lung cancer.
The relative risks for Greece and Japan of 2.00 and 1.44, respectively, are probably the
best estimates, because both female smoking prevalence and nontobacco-related lung cancer risks,
which tend to dilute the estimates of ETS effects, are low in these two countries. Also, for the
time period for which ETS exposure was of interest, spousal smoking is considered to be a better
surrogate for ETS exposure in these societies than in Western countries, where other sources of
ETS exposure (work, public places, and other nonhome environments) are generally higher.
1-7
05/31/92
-------
DRAFT—DO NOT QUOTE OR CITE
Based on these analyses and following the U.S. EPA Guidelines for Carcinogen Risk
Assessment (U.S. EPA, 1986a), EPA concludes that environmental tobacco smoke is a Group A
(known human) carcinogen. This conclusion is based on a total weight-of-evidence, principally:
• Biological plausibility. ETS is taken up by the lungs, and components are
distributed throughout the body. The presence of the same carcinogens in ETS
and mainstream smoke, along with the established causal relationship between
lung cancer and active smoking with the dose-response relationships exhibited
down to low doses, make it reasonable to conclude that ETS is also a lung
carcinogen.
• Supporting evidence from animal bioassays and genotoxicity experiments. The
carcinogenicity of tobacco smoke has been established in lifetime inhalation
studies in the hamster, intrapulmonary implantations in the rat, and skin painting
in the mouse. There are no lifetime animal inhalation studies of ETS; however,
the carcinogenicity of ETS condensates has been demonstrated in intrapulmonary
implantations and skin painting experiments. Positive results of genotoxicity
testing for both MS and ETS provide corroborative evidence for their carcinogenic
potential.
• Consistency of response. All 4 of the cohort studies and 20 of the 26 case-control
studies observed a higher risk of lung cancer among the female never-smokers
classified as exposed to ETS. Of the 17 studies judged to be of higher utility
based on study design, execution, and analysis (Appendices A and C), 15 observed
higher risks, and 6 of these increases were statistically significant, despite most
having low statistical power. Evaluation of the total study evidence from several
perspectives leads to the conclusion that the observed association between ETS
exposure and increased lung cancer occurrence is not attributable to chance.
• Broad-based evidence. These 26 case-control and 4 prospective studies provide
data from 8 different countries, employ a wide variety of study designs and
protocols, and are conducted by many different research teams. Results from all
countries, with the possible exception of two areas of China where high levels of
other indoor air lung carcinogens were present, show small to modest increases in
lung cancer associated with spousal ETS exposure. No alternative explanatory
variables for the observed association between ETS and lung cancer have been
indicated that would be broadly applicable across studies.
1-8
05/31/92
-------
DRAFT—DO NOT QUOTE OR CITE
Upward trend in dose-response. Both the largest of the cohort studies, the
Japanese study of Hirayama—200 lung cancer cases, and the largest of the case-
control studies, the U.S. study by Fontham and associates (1991)--420 lung cancer
cases and two sets of controls, demonstrate a strong dose-related statistical
association between passive smoking and lung cancer. This upward trend is well
supported by the preponderance of epidemiology studies. Of the total of 17
studies in which data are classified by exposure level, 11 were statistically
significant for the trend despite most having low statistical power.
Detectable association at environmental exposure levels. Within the population of
married women who are lifelong nonsmokers, the excess lung cancer risk from
exposure to their smoking husbands' ETS is large enough to be observed.
Carcinogenic responses are usually detectable only in high-exposure
circumstances, such as occupational settings, or in experimental animals receiving
very high doses. In addition, effects are harder to observe when there is
substantial background exposure in the comparison groups, as is the case here.
Effects remain after adjustment for potential bias. Current and ex-smokers may
be misreported as never-smokers, thus inflating the apparent cancer risk for ETS
exposure. The evidence remains statistically significant and conclusive, however,
after adjustments for smoker misclassification. For the United States, the
Summary estimate of relative risk from nine case-control plus two cohort studies is
1.19 (90% confidence interval [C.I.] = 1.04-1.35) after adjustment for
misclassification (p < 0.05). For Greece, 2.00 (1.42, 2.83), Hong Kong, 1.61 (1.25,
2.06) and Japan, 1.44 (1.13, 1.85), the estimated relative risks are higher than those
of the United States and more highly significant after adjusting for the potential
bias.
Confounding cannot explain the association. The broad-based evidence for an
association found by independent investigators across several countries, as well as
the positive dose-response trends observed in most of the studies that analyzed for
them, make any single confounder highly unlikely, as an explanation for the
results. In addition, this report examined potential confounding factors (history of
lung disease, home heat sources, diet, occupation) and concluded that none of
these factors could account for the observed association between lung cancer and
ETS.
1-9
05/31/92
-------
DRAFT—DO NOT QUOTE OR CITE
The individual risk of lung cancer from exposure to ETS does not have to be very large to
translate into a significant health hazard to the U.S. population because of the large number of
smokers and the widespread presence of ETS. Current smokers comprise approximately 26% of
the U.S. adult population and consume more than one-half trillion cigarettes annually (1.5 packs
per day, on average), causing nearly universal exposure to at least some ETS. As a biomarker of
tobacco smoke uptake, cotinine, a metabolite of the tobacco-specific compound nicotine, is
detectable in the blood, saliva, and urine of persons recently exposed to tobacco smoke. Cotinine
has typically been detected in 50% to 75% of reported nonsmokers tested (50% equates to 63
million U.S. nonsmokers of age 18 or above).
The best estimate of approximately 3,000 lung cancer deaths per year in U.S. nonsmokers
age 35 and over attributable to ETS (Chapter 6) is based on data pooled from all 11 U.S.
epidemiologic studies of never-smoking women married to smoking spouses. Use of U.S. studies
should increase the confidence in these estimates. Some mathematical modeling is required to
adjust for expected bias from misclassification of smoking status and to account for ETS exposure
from sources other than spousal smoking. Assumptions are also needed to relate responses in
female never-smokers to those in male never-smokers and ex-smokers of both sexes, and to
estimate the proportion of the nonsmoking population exposed to various levels of ETS. Overall,
however, the assumptions necessary for estimating risk add far less uncertainty than other EPA
quantitative assessments. This is because for ETS the extrapolation is based on a large database of
human studies, all at levels actually expected to be encountered by much of the U.S. population.
The components of the 3,000 lung cancer deaths figure include approximately 1,500 female
never-smokers, 500'male never-smokers, and 1,000 former smokers of both sexes. More females
are estimated to be affected because there are more female than male nonsmokers. These
component estimates have varying degrees of confidence; the estimate of 1,500 deaths for female
never-smokers has the highest confidence because of the extensive database. The estimate of 500
for male never-smokers is less certain because it is based on the female never-smoker response
and is thought to be low because males are generally subject to higher background ETS exposures
than females. Adjustment for this higher background exposure would lead to higher risk
estimates. The estimate of 1,000 lung cancer deaths for former smokers of both sexes is
considered to have the lowest confidence, and the assumptions included are thought to make this
estimate low as well.
Workplace ETS levels are generally comparable to home ETS levels, and studies using body
cotinine measures as biomarkers demonstrate that nonhome exposures to ETS are often greater
1-10 05/31/92
-------
DRAFT—DO NOT QUOTE OR CITE
than exposure from spousal smoking. Thus, this report presents an alternative breakdown of the
estimated 3,000 ETS-attributable lung cancer deaths between spousal and nonhome exposures. By
extension of the results from spousal smoking studies, coupled with biological measurements of
exposure, more lung cancer deaths are estimated to be attributable to ETS from combined
nonhome exposures - 2,200 of both sexes - than from spousal exposure - 800 of both sexes. This
home-versus-other-sourtes partitioning depends on current exposure estimates that may or may
not be applicable to the exposure period of interest. Thus, this breakdown contains this element
of uncertainty in addition to those discussed above with respect to the previous breakdown.
Other estimates of annual U.S. nonsmoker lung cancer deaths attributable to ETS
developed in this document give a range of 2,500 to 3,300. These other estimates use both
mortality and cotinine exposure data from the largest and best-designed U.S. study (Fontham et
al., 1991). Relatively small differences in cotinine ratios, as measures of exposure from spousal
smoking, can result in substantial variability in population risk estimates. The range suggested
above provides an estimation of the uncertainty in these estimates. Overall, however, considering
the multitude, consistency, and quality of all these studies, the weight-of-evidence conclusion that
ETS is a known human lung carcinogen, and the limited amount of extrapolation necessary, the
confidence in the estimate of approximately 3,000 lung cancer deaths is medium to high.
1.2.2. ETS and Noncancer Respiratory Disorders
Exposure to ETS from parental smoking has been previously linked with increased
respiratory disorders in children, particularly in infants. Several studies have confirmed the
exposure and uptake of ETS in children by assaying saliva, serum, or urine for cotinine. These
cotinine concentrations were highly con-elated with smoking (especially by the mother) in the
Child's presence. Nine million to twelve-million American children under 5 years of age, or one-
half to two-thirds of all children in this age group, may be exposed to cigarette smoke-in the
home (American Academy of Pediatrics, 1986).
With regard to the noncancer respiratory effects of passive smoking, this report focuses on
epidemiologic evidence appearing since the two major reports of 1986 (NRG and U.S. DHHS) that
bears on the potential association of parental smoking with detrimental respiratory effects in their
children. These effects include symptoms of respiratory irritation (cough, sputum, or wheeze);
acute diseases of the lower respiratory tract (pneumonia, bronchitis, and bronchiolitis); acute
middle ear infections and indications of chronic middle ear infections (predominantly middle ear
effusion); reduced lung function (from forced expiratory volume and flow-rate measurements);
1-11
05/31/92
-------
DRAFT—DO NOT QUOTE OR CITE
incidence and prevalence of asthma and exacerbation of symptoms in asthmatics; and acute upper
respiratory tract infections (colds and sore throats). The more than 50 recently published studies
reviewed here essentially corroborate the previous conclusions of the NRC and Surgeon General
regarding respiratory symptoms, respiratory illnesses, and pulmonary function, and they
strengthen support for those conclusions by the additional weight-of-evidence (Chapter 7). For
example, new data on middle ear effusion strengthen previous evidence to warrant the stronger
conclusion in this report of a causal association with parental smoking. Furthermore, recent
studies establish associations between parental smoking and increased incidence of childhood
asthma. Additional research also supports the hypotheses that in utero exposure to mother's smoke
and postnatal exposure to ETS alter lung function and structure, increase bronchial
responsiveness, and enhance the process of allergic sensitization, changes that are known to
predispose children to early respiratory illness. Early respiratory illness can lead to long-term
.pulmonary effects (reduced lung function and increased risk of chronic obstructive lung disease).
This document also summarizes the evidence for an association between parental smoking
and SIDS, which was not addressed in the 1986 NRC or Surgeon General reports. SIDS is the most
common cause of death in infants ages 1 month to 1 year. The cause (or causes) of SIDS is
unknown; however, it is widely believed that some form of respiratory pathogenesis is generally
involved. The current evidence strongly suggests that infants whose mothers smoke are at an
increased risk of dying of SIDS, independent of other known risk factors for SIDS, including low
birthweight and low gestational age, which are specifically associated with active smoking during
pregnancy. However, available studies do not allow this report to conclude whether that increased
risk is related to in utero versus postnatal exposure to tobacco smoke products, or to both.
The 1986 NRC and Surgeon General reports conclude that both the prevalence of
respiratory symptoms of irritation and the incidence of lower respiratory tract infections are
higher in children of smoking parents. In the 18 studies of respiratory symptoms subsequent to
the 2 reports, increased symptoms (cough, phlegm, and wheezing) were observed in a range of
ages from birth to midteens, particularly in infants and preschool children. In addition to the
studies on symptoms of respiratory irritation, nine new studies have addressed the topic of
parental smoking and acute lower respiratory tract illness in children, and eight have reported
statistically significant associations. The cumulative evidence indicates strongly that parental
smoking, especially the mother's, causes an increased incidence of respiratory illnesses from birth
up to the first 18 months to 3 years of life, particularly for bronchitis, bronchiolitis, and
1-12
05/31/92
-------
DRAFT—DO NOT QUOTE OR CITE
pneumonia. Overall, the evidence confirms the previous conclusions of the NRC and Surgeon
General.
Recent studies also solidify the evidence for the conclusion of a causal association between
parental smoking and increased middle ear effusion in young children. Middle ear effusion is the
most common reason for hospitalization of young children for an operation.
At the time of the Surgeon General's report on passive smoking (U.S. DHHS, 1986), data
were sufficient only to conclude that maternal smoking may influence the severity of asthma in
children. The recent studies reviewed here strengthen and confirm these exacerbation effects. In
addition, the new evidence is conclusive that ETS exposure increases the number of episodes of
asthma in children who already have the disease. It is also suggestive that ETS exposure increases
the number of new cases of asthma in children who have not previously exhibited symptoms,
although the results are statistically significant only with children whose mothers smoke 10 or
more cigarettes per day. While the evidence for new cases of asthma itself is not conclusive of a
causal association, the consistent strong associations of ETS with both increased frequency and
severity of the asthmatic symptoms and the established ETS effects on both the immune system
and airway hyperresponsiveness lead to the conclusion that ETS is a risk factor for induction of
asthma in previously asymptomatic children.
Regarding the effects of passive smoking on lung function in children, the 1986 Surgeon
General and NRC reports both conclude that children of parents who smoke have small decreases
in tests of pulmonary output function of both the larger and smaller air passages when compared
with the children of nonsmokers. As noted in the NRC report, if ETS exposure is the cause of the
observed decrease in lung function, the effect could be due to the direct action of agents in ETS
or an indirect consequence of increased occurrence of acute respiratory illness related to ETS.
Results from eight studies on ETS and lung function in children that have appeared since
those reports add some additional confirmatory evidence suggesting a causal rather than an
indirect relationship. For the population as a whole, the reductions are small relative to the
interindividual variability of each lung function parameter. However, groups of particularly
susceptible or heavily exposed subjects have shown larger decrements. The studies reviewed
suggest that a continuum of exposures to tobacco products starting in fetal life may contribute to
the decrements in lung function found in older children. Exposure to tobacco smoke products
inhaled by the mother during pregnancy may contribute significantly to these changes, but there
is strong evidence indicating that postnatal exposure to ETS is an important part of the causal
pathway.
1-13
05/31/92
-------
DRAFT—DO NOT QUOTE OR CITE
With respect to lung function effects in adults exposed to ETS, the 1986 NRC and Surgeon
General reports found the data at that time inconclusive, due to high interindividual variability
and the existence of a large number of other risk factors, but compatible with subtle deficits in
lung function. Recent studies confirm the association of passive smoking with small reductions in
lung function. Furthermore, new evidence also has emerged suggesting a subtle association
between exposure to ETS and increased respiratory symptoms in adults.
There is some evidence suggesting that the incidence of acute upper respiratory tract
illnesses and acute middle ear infections may be more common in children exposed to ETS.
However, several studies failed to find any effect. In addition, the possible role of confounding
factors, the lack of studies showing clear dose-response relationships, and the absence of a
plausible biological mechanism preclude more definitive conclusions.
In reviewing the available evidence indicating an association (or lack thereof) between ETS
exposure and the different noncancer respiratory disorders analyzed in this report, the possible
role of several potential confounding factors was considered. These include other indoor air
pollutants; socioeconomic status; effect of parental symptoms; and characteristics of the exposed
child, such as low birthweight or active smoking. No single or combined confounding factors can
explain the observed respiratory effects of passive smoking in children.
For diseases for which ETS has been either causally associated (lower respiratory tract
infections) or indicated as a risk factor (asthma cases in previously asymptomatic children),
estimates of population attributable risk can be calculated. A population risk assessment
(Chapter 8) provides a probable range of estimates that 8,000 to 26,000 cases of childhood asthma
per year are attributable to ETS exposure from mothers who smoke 10 or more cigarettes per day.
The confidence in this range of estimates is medium and is dependent on the suggestive evidence
of the database. While the data show an effect only for children of these heavily smoking
mothers, additional cases due to lesser ETS exposure are also a possibility. If the effect of this
lesser exposure is considered, the range of estimates of new cases presented above increases to
13,000 to 60,000. Furthermore, this report estimates that the additional public health impact of
ETS on asthmatic children includes over 200,000 children whose symptoms are significantly
aggravated and as many as 1,000,000 children who are affected to some degree.
This report estimates that ETS exposure contributes 150,000 to 300,000 cases annually of
lower respiratory tract illness in infants and children younger than 18 months of age and that
7,500 to 15,000 of these will require hospitalization. The strong evidence linking ETS exposure to
increased incidence of bronchitis, bronchiolitis, and pneumonia in young children gives these
1-14
05/31/92
-------
DRAFT—DO NOT QUOTE OR CITE
estimates a high degree of confidence. There is also evidence suggesting a smaller ETS effect on
children between ages 18 months and 3 years, but no additional estimates have been computed for
this age group. Whether or not these illnesses result in death has not been addressed here.
In the United States, more than 5,000 infants die of SIDS annually. It is the major cause of
death in infants between the ages of 1 month and 1 year and the linkage with maternal smoking is
well established. The Surgeon General and World Health Organization estimate that more than
700 U.S. infant deaths per year from SIDS are attributable to maternal smoking (U.S. CDC,
199la). However, this report concludes that at present there is not enough direct evidence
supporting the contribution of ETS exposure to declare it a risk factor or to estimate its population
impact on SIDS.
1-15
05/31/92
-------
-------
DRAFT—DO NOT QUOTE OR CITE
2. INTRODUCTION
An estimated 434,000 deaths per year in the United States, or more than one of every six
deaths, are directly attributable to tobacco use, in particular cigarette smoking (CDC, 1991a;
figures for 1988). Approximately 112,000 of these smoking-related deaths are from lung cancer,
accounting for an estimated 87% of U.S. lung cancer mortality (U.S. DHHS, 1989; percentage for
1985). Cigarette smoking is also causally related to cancer at various other sites, such as the
bladder, renal pelvis, pancreas, and upper respiratory and digestive tracts (IARC, 1986). Roughly
30,000 deaths per year from cancers at these sites are attributable to smoking (CDC, 199la).
Furthermore, smoking is the major cause of chronic obstructive pulmonary disease (COPD), which
includes emphysema, and is thought to be responsible for approximately 61,000 COPD deaths
yearly, or about 82% of COPD deaths (U.S. DHHS, 1989). Tobacco use is also a major risk factor
for cardiovascular diseases, the leading cause of death in the United States. It is estimated that
each year 156,000 heart disease deaths and 26,000 deaths from stroke are attributable to smoking
(CDC, 199la). In addition to this substantial mortality, the association of smoking with these
conditions also involves significant morbidity.
Smoking is also a risk factor for various respiratory infections, such as influenza, &
bronchitis, and pneumonia. An estimated 20,000 influenza and pneumonia deaths per year are
attributable to smoking (CDC, 199la). Smokers also suffer from lung function impairment and
numerous other respiratory symptoms, such as cough, phlegm production, wheezing, and shortness
of breath. In addition, smokers are at increased risk for a variety of other conditions, including
pregnancy complications and ulcers.
-i
Although the exact mechanisms and tobacco smoke components associated with these
health effects are not known with certainty, more than 40 known or suspected human carcinogens
have been identified in tobacco smoke. These include, for example, benzene, nickel, polonium-
210, 2-napthylamine, 4-aminobiphenyl, formaldehyde, various N-nitrosamines,
benz[a]anthracene, and benzo[a]pyrene. Many other toxic agents, such as carbon monoxide,
nitrogen oxides, ammonia, and hydrogen cyanide, are also found in tobacco smoke.
Smokers, however, are not the only ones at risk from exposure to these tobacco smoke
toxicants. In utero exposure from maternal smoking is known to be associated with low
birthweight and increased risk of fetal and infant death (U.S. DHHS, 1989). Furthermore,
nonsmokers might be at risk for smoking-associated health effects from passive exposure to
environmental tobacco smoke (ETS). When a cigarette is smoked, approximately half of the
2-1
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
smoke generated is sidestream smoke (SS) emitted from the smoldering cigarette between puffs.
This SS constitutes roughly 85% of ETS (Fielding, 1985).
Twenty-eight percent of the U.S. adult population (CDC, 1991b), or about 50 million
Americans, are smokers, and so virtually all Americans are exposed to some amount of ETS in the
home, at work, or in public places. In view of the high levels of mortality and morbidity
associated with smoking and the considerable potential for exposure of nonsmokers to ETS,
passive smoking is potentially a substantial public health concern.
2.1. FINDINGS OF PREVIOUS REVIEWS
The first epidemiologic results associating passive smoking with lung cancer appeared in
the early 1980's. Since then, two major comprehensive reviews of the health effects of passive
smoking, and several less extensive ones have been published. One of the major reviews was
conducted by the National Research Council (NRC) in 1986. At the request of two Federal
agencies, the U.S. Environmental Protection Agency and the U.S. Department of Health and
Human Services, the NRC formed a committee on passive smoking to evaluate the methods for
assessing exposure to ETS and to review the literature on all of the potential health consequences
of exposure. The committee's report (NRC, 1986) addresses the issue of lung cancer risk in
considerable detail and includes summary analyses from 10 case-control studies and 3 cohort
(prospective) studies. The report concludes that "[considering the evidence as a whole, exposure
to ETS increases the incidence of lung cancer in nonsmokers." Combining the data from all the
studies, the committee calculated an overall observed relative risk estimate of 1.34 (95%
C.I. - 1.18-1.53).
The NRC committee was concerned about potential bias in the study results caused by
current and former smokers incorrectly self-reported as lifelong nonsmokers (never-smokers).
Using plausible assumptions for misreported smoking habits, the committee determined that
smoker misclassification cannot account for all of the increased risk observed in the epidemiologic
studies. Furthermore, the upward bias on the relative risk of lung cancer caused by smoker
misclassification is counterbalanced by the downward bias from background ETS exposure to the
supposedly unexposed group. Correcting for smoker misclassification and background ETS
exposure, the committee calculated an overall adjusted relative risk estimate of 1.42 (range of 1.24
to 1.61) for lung cancer in nonsmokers from exposure to ETS from spousal smoking plus
background sources.
2-2
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
The NRC committee also found evidence for noncancer respiratory effects in children
exposed to ETS. It recommended that "[i]n view of the weight of the scientific evidence that ETS
exposure in children increases the frequency of pulmonary symptoms and respiratory infections, it
is prudent to eliminate smoking and resultant ETS from the environments of small children."
Furthermore, the committee concluded that "[household exposure to ETS is linked with increased
rates of chronic ear infections and middle ear effusions in young children." The NRC report also
notes that M[e]vidence has accumulated indicating that nonsmoking pregnant women exposed to
ETS on a daily basis for several hours are at increased risk for producing low-birthweight babies,
through mechanisms which are, as yet, unknown."
The second major review, the Surgeon General's report on the health consequences of
passive smoking, also appeared in 1986 (U.S. DHHS, 1986). This review covers ETS chemistry,
exposure, and various health effects, primarily lung cancer and childhood respiratory disease. On
the subject of lung cancer, the report concludes:
The absence of a threshold for respiratory carcinogenesis in active smoking, the
presence of the same carcinogens in mainstream and sidestream smoke, the
demonstrated uptake of tobacco smoke constituents by involuntary smokers, and
the demonstration of an increased lung cancer risk in some populations with
exposures to ETS leads to the conclusion that involuntary smoking is a cause of
lung cancer.
With respect to respiratory disorders in children, the Surgeon General's report determined that
"[t]he children of parents who smoke, compared with the children of nonsmoking parents, have an
increased frequency of respiratory infections, increased respiratory symptoms, and slightly smaller
rates of increase in lung function as the lung matures."
In 1987, a committee of the International Agency for Research on Cancer (IARC) issued a
report on methods of analysis and exposure measurement related to passive smoking (IARC, 1987).
The committee reviewed the physicochemical properties of ETS, the toxicological basis for lung
cancer, and methods of assessing and monitoring exposure to ETS. The report borrows the
summary statement on passive smoking from a previous IARC document that dealt mainly with
tobacco smoking (IARC, 1986). The working group that produced the 1986 report had found that
the epidemiologic evidence then available on passive smoking was compatible with either the
presence or the absence of a lung cancer risk; however, based on other considerations related to
biological plausibility, it concluded that passive smoking gives rise to some risk of cancer.
Specifically, the 1986 IARC report states:
Knowledge of the nature of sidestream and mainstream smoke, of the materials
absorbed during "passive smoking," and of the quantitative relationships between
2-3
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
dose and effect that are commonly observed from exposure to carcinogens . ..
leads to the conclusion that passive smoking gives rise to some risk of lung cancer.
More recently, the Working Group on Passive Smoking, an independent international panel
of scientists supported in part by RJR Reynolds Nabisco, reported the findings of its
comprehensive "best evidence synthesis" of over 2,900 articles on the health effects of passive
smoking (Spitzer et al., 1990). The group concluded that "[t]he weight of evidence is compatible
with a positive association between residential exposure to environmental tobacco smoke
(primarily from spousal smoking) and the risk of lung cancer." It also found "strong evidence that
children exposed in the home to environmental tobacco smoke have higher rates of hospitalization
(50 to 100%) for severe respiratory illness" and that the "evidence strongly supports a relationship
between exposure to environmental tobacco smoke and asthma among children." In addition, the
working group reported that there is evidence for associations between home ETS exposure and
many chronic and acute respiratory illnesses, as well as small decreases in physiologic measures of
respiratory function, in both children and adults. Evidence demonstrating an increased prevalence
of otitis media (inflammation of the middle ear) in children exposed to ETS at home was also
noted. With respect to in utero exposure, the group concluded that active maternal smoking is
associated with reduced birthweight and with increased infant mortality.
A recent review of the health effects associated with adult workplace exposure to ETS
conducted by the National Institute for Occupational Safety and Health (NIOSH, 1991) determined
that "the collective weight of evidence (i.e., that from the Surgeon General's reports, the
similarities in composition of MS [mainstream smoke] and ETS, and the recent epidemiologic
studies) is sufficient to conclude that ETS poses an increased risk of lung cancer and possibly
heart disease to occupationally exposed workers." Furthermore:
Although these data were not gathered in an occupational setting, ETS meets the
criteria of the Occupational Safety and Health Administration (OSHA) for
classification as a potential occupational carcinogen [Title 29 of the Code of
Federal Regulations, Part 1990]. NIOSH therefore recommends that exposures be
reduced to the lowest feasible concentration.
The classification of "potential occupational carcinogen" is NIOSH's category of strongest evidence
for carcinogenicity.
2.2. EPA's 1992 DOCUMENT
2.2.1. Scope
Due to the serious health concerns that have arisen regarding ETS, the most ubiquitous
indoor air pollutant, and the wealth of new information that has become available since the
2-4
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
extensive 1986 reviews, the EPA has performed its own analytical hazard identification and risk
assessment for the respiratory health effects of passive smoking, based on a critical review of the
data currently available, with an emphasis on the abundant epidemiologic evidence. The number
of lung cancer studies analyzed in this document is more than double the number reviewed in
1986 (31 vs. 13), with a total of about 3,000 lung cancer cases in female nonsmokers now reported
in case-control studies and almost 300,000 female nonsmokers followed by cohort studies.
Furthermore, the database on passive smoking and respiratory disorders in children contains more
than 50 new studies, including 8 additional studies on acute lower respiratory tract illnesses, 9 on
acute and chronic middle ear diseases, 18 on respiratory symptoms, 9 on asthma, and 8 on lung
function. This report also discusses six recent studies of the effects of passive smoking on adult
respiratory symptoms and lung function. Finally, eight studies of maternal smoking and sudden
infant death syndrome (SIDS), which was not addressed in the NRC report or the Surgeon
General's report, are reviewed. (Although the cause of SIDS is unknown, the most widely
accepted hypotheses suggest that some form of respiratory pathogenesis is usually involved.)
First, this document reviews information on the nature of ETS and human exposures.
Then, in accordance with the Guidelines for Carcinogen Risk Assessment (U.S. EPA, 1986a), it
critically analyzes human, animal, and genotoxicity data to establish the weight-of-evidence for
the hazard identification of ETS as a human lung carcinogen and to characterize the U.S.
population risk. Similarly, it reviews studies of passive smoking and noncancer respiratory
disorders, particularly in children, and provides both hazard identification and population risk
estimates for some of these effects.
While this report restricts analysis to ETS-associated respiratory effects because of time
and resource considerations, several recent studies have also linked passive smoking with an
increased risk of heart disease or cancers at sites other than the lung. For cancers of other sites,
the available evidence is quite limited (e.g., Hirayama, 1984; Sandier et al., 1985), but three recent
analyses, examining over 15 epidemiologic studies and various supporting mechanistic studies,
suggest that ETS is an important risk factor for heart disease, accounting for as many as 35,000 to
40,000 deaths annually (Wells, 1988; Glantz and Parmley, 1991; Steenland, 1992). This report
takes no position on ETS and heart disease.
Other health effects of active smoking may also have passive smoking correlates of public
health concern. Maternal smoking during pregnancy, for example, is known to affect fetal
development. Studies on passive smoking during pregnancy are far fewer but have demonstrated
an apparent association with low birthweight (e.g., Martin and Bracken, 1986). Furthermore,
2-5
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
passive exposure to tobacco smoke products both in utero and postnatally may result in other
nonrespiratory developmental effects in children~for example, decrements in neurological
development (Makin et al., 1991). Again, this document takes no position on these potential
nonrespiratory effects.
2.2.2. Use of EPA's Guidelines
The lung cancer hazard identification and risk characterization for ETS are conducted in
accordance with the EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA, 1986a). In fact,
tobacco smoke is a mixture of over 4,000 compounds and could be evaluated according to the
Guidelines for the Health Risk Assessment of Chemical Mixtures (U.S. EPA, 1986b). Such a
highly complex mixture, however, is not easily characterized with respect to chemical
composition, levels of exposure, and toxicity of constituents. Furthermore, the effects and
mechanisms of interactions among chemicals are insufficiently understood.
The Guidelines for the Health Risk Assessment of Chemical Mixtures acknowledges these
inherent uncertainties and recommends various assessment approaches, depending on the nature
and quality of the data. When adequate data are available on health effects and exposure for the
actual mixture of concern, as is the case with both MS and ETS, the preferred approach,
according to the mixtures guidelines, is to adopt the procedures used for single compounds
described by the Guidelines for Carcinogen Risk Assessment, as is done here. The EPA has also
used this strategy for assessments of diesel exhausts, PCBs, and unleaded gasoline. The
compilation of health effects and exposure information for all the mixture components of interest
is considered optional. In the case of tobacco smoke, compiling this information would be highly
impractical due to the large number of components and the highly complex and changing nature
of this mixture. It is also considered unnecessary, given the abundant epidemiologic data on ETS
and lung cancer.
The Guidelines for Carcinogen Risk Assessment provide a general framework for the
analysis of carcinogenic risk, while permitting "sufficient flexibility to accommodate new
knowledge and new assessment methods as they emerge" (U.S. EPA, 1986a). According to the
guidelines, a qualitative risk assessment, or hazard identification, is performed by evaluating all of
the relevant data to determine if a compound has carcinogenic potential. Then, a dose-response
assessment is made by using mathematical models to extrapolate from high experimental or
occupational exposures, where risks are usually detected, to lower environmental exposure levels.
2-6
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Finally, the dose-response assessment and an exposure assessment are integrated into a risk
characterization, providing risk estimates for exposed populations.
The enormous database on active and passive smoking provides more than sufficient
human evidence on which to base a hazard identification of ETS. The use of human evidence
eliminates the uncertainty that normally arises when one has to base hazard identification on the
results of high-dose animal experiments. Furthermore, the epidemiologic data on passive smoking
provide direct evidence from environmental exposure levels, obviating the need for a dose-
response extrapolation from high to low doses. These low-level environmental exposures,
however, are associated with low relative risks that can only be detected in well-designed studies
of sufficiently large size. For this reason, new assessment methods are used to categorize studies
on the basis of quality criteria and to combine studies to increase the statistical power.
As an alternative to using actual epidemiologic data on ETS, an ETS risk assessment could
have used "cigarette equivalents" to correlate ETS exposure with lung cancer risk based on dose-
response models from active smoking. This would have involved using measures such as cotinine
or respirable suspended particles to compare smoke uptake between smokers and ETS-exposed
nonsmokers in order to equate passive smoking to the active smoking of some quantity of a
cigarette(s). Then the carcinogenic response associated with that exposure level would be
estimated from extrapolation models based on the dose-response relationships observed for active
smoking. This procedure was not used for several reasons. Although MS and ETS are
qualitatively similar with respect to chemical composition (i.e., they contain most, if not all, of the
same toxicants and carcinogens), the absolute and proportional quantities of the components, as
well as their physical state, can differ substantially. Many tobacco smoke compounds partition
preferentially into the MS component of smoke emissions; others, however, such as certain highly
carcinogenic N-nitrosamines, are preferentially produced at lower temperatures and appear in
much greater amounts in the ETS fraction. In addition, active and passive smokers have different
breathing patterns, and particles in ETS are smaller than those in MS. Therefore, the distribution
and deposition of smoke constituents in the respiratory tracts of active and passive smokers will
not be identical. Furthermore, it is not known which of the chemicals in tobacco smoke are
responsible for its carcinogenicity. Clearly the comparison of a small number of biomarker
measures cannot adequately quantify differential distributions of unknown carcinogenic
compounds.
Another area of uncertainty in the "cigarette equivalents" approach relates to potential
metabolic differences between active and passive smokers. Active smoking is known to induce
2-7
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
chemical- and drug-metabolizing enzymes in various tissues to levels that significantly exceed
those found in nonsmokers. Thus, the dose-response relationships for tobacco smoke-associated
health effects are likely to be nonlinear. In fact, evidence suggests that a linear dose-response
extrapolation might underestimate the risk of adverse health effects from low doses of tobacco
smoke (Remmer, 1987). Because of these uncertainties, the data from active smoking are more
appropriate for qualitative hazard identification than for quantitative dose-response assessment.
Furthermore, at least for lung cancer and other respiratory effects, we have substantial
epidemiologic data from actual exposure of nonsmokers to environmental levels of genuine ETS,
which constitute a superior database from which to derive quantitative risk estimates for passive
smoking, without the need for low-dose extrapolation.
2.2.3. Contents of This Document
ETS is chemically similar to MS, containing most, if not all, of the same toxicants and
known or suspected human carcinogens. A major difference, however, is that ETS is rapidly
diluted into the environment, and consequently, passive smokers are exposed to much lower
concentrations of these agents than are active smokers. Therefore, in assessing potential health
risks attributable to ETS, it is important to be able to measure ETS levels in the many
environments where it is found and to quantify actual human ETS exposure. The physical and
chemical nature of ETS and issues related to human exposure are discussed in Chapter 3. The use
of marker compounds and various methods for assessing ambient ETS concentrations, as well as
the use of biomarkers, questionnaires, and modeling techniques to determine human exposure, is
described. Furthermore, measurements of ET&eomponents in various indoor environments and of
ETS constituents and their metabolites in adult and child nonsmokers are presented, providing
evidence of actual nonsmoker exposure and uptake.
Chapter 4 reviews the major evidence that conclusively established that the tobacco smoke
inhaled from active smoking is a human lung carcinogen. Unequivocal dose-response
relationships exist between tobacco smoking and lung cancer, with no evidence of a threshold
level of exposure. Supporting evidence for the carcinogenicity of tobacco smoke from animal
bioassays and genotoxicity experiments is also summarized, including data from the limited animal
and mutagenicity studies pertaining specifically to ETS.
The chemical similarity between MS and ETS and the measurable uptake of ETS
constituents by nonsmokers (Chapter 3), as well as the causal dose-related association between
tobacco smoking and lung cancer in humans, extending to the lowest observed doses, and the
2-8
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
corroborative evidence for the carcinogenicity of both MS and ETS provided by animal bioassays
and genotoxicity studies (Chapter 4), clearly establish the biological plausibility that ETS is also a
human lung carcinogen. In fact, this evidence is sufficient in its own right to establish weight-of-
evidence for ETS as a Group A (known human) carcinogen under EPA guidelines.
In addition to the evidence of human carcinogenicity from high exposures to tobacco
smoke from active smoking, there are now more than 30 epidemiologic studies investigating lung
cancer in nonsmokers exposed to actual ambient levels of ETS. The majority of these studies
examine never-smoking women, with spousal smoking used as a surrogate for ETS exposure.
Female exposure from spousal smoking is considered to be the single surrogate measure that is the
most stable and best represents ETS exposure.
For the purposes of the hazard identification analysis in Chapter 5, which is based
primarily on the epidemiologic studies of ETS, this document extensively and critically evaluates
31 epidemiologic studies from 8 different countries, including 11 studies from the United States
(Appendix A). More than half of these studies have appeared since the NRC and Surgeon
General's reviews were issued in 1986. Two U.S. studies are of particular interest. The recently
published five-center study of Fontham et al. (1991) is a well-designed and conducted case-
control study with 429 never-smoking female lung cancer cases and two separate sets of controls.
This is the largest case-control study to date, and it has a high statistical power to detect the small
increases in lung cancer risk that might be expected from ambient exposures. Another large U.S.
case-control study was the recent study by Janerich et al. (1990), with 191 cases. Both of these
studies were supported by the National Cancer Institute.
In evaluating epidemiologic studies, potential sources of bias and confounding must also be
addressed. Smoker misclassification of current and former smokers as never-smokers is the one
identified source of systematic upward bias to the relative risk estimates. Therefore, prior to the
statistical analyses of the epidemiologic data that are conducted in Chapter 5, the relative risk
estimates from each study are adjusted for smoker misclassification using the methodology of
Appendix B. Other potential sources of bias and confounding are discussed extensively in the
course of Chapter 5.
Chapter 5 quantitatively and qualitatively analyzes the epidemiologic data to determine the
weight-of-evidence for the hazard identification of ETS. First the individual studies are
statistically assessed using tests for effect (i.e., association between lung cancer and ETS) and tests
for dose-response trend. Then various combining analyses are performed to examine and compare
the epidemiologic results for separate countries. The studies are also categorized into four tiers
2-9
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
according to the utility of the study in terms of its likely ability to detect a possible effect, based
on specific criteria for evaluating the design and conduct as displayed in Appendix C and the
critical reviews in Appendix A. These tiers are integrated one at a time into statistical analyses, as
an alternative method for evaluating the epidemiologic data that also takes into account qualitative
considerations. Chapter 5 concludes with an overall weight-of-evidence determination for lung
cancer based on the analyses in Chapters 3, 4, and 5.
In Chapter 6, the relative risk estimates from 11 U.S. studies of passive smoking and lung
cancer are adjusted upward to correct for the systematic downward bias caused by background
exposure to ETS from sources other than spousal smoke. Using additional assumptions to extend
the results from female never-smokers to male never-smokers and long-term former smokers of
both sexes, the population risk for U.S. nonsmokers is characterized by estimating the annual
number of lung cancer deaths that are attributable to exposure from all sources of ETS. Separate
estimates are calculated for background (workplace and other nonhome exposures) and spousal
(home) exposures, as well as for female and male never-smokers and former smokers. Chapter 6
also discusses the sources of uncertainty and sensitivity in the lung cancer estimates.
The final two chapters address passive smoking and noncancer respiratory disorders. Both
the NRC and Surgeon General's reports concluded that children exposed to ETS from parental
smoking are at greater risk for various respiratory illnesses and symptoms. This document
confirms and extends those conclusions with analyses of more recent studies. New evidence for
an association between ETS and middle ear effusion, and for a role of ETS in the cause as well as
in the prevalence and severity of childhood asthma, is reviewed. In addition, the evidence for an
association between maternal smoking and SIDS is examined.
Chapter 7 reviews and analyzes epidemiologic studies of passive smoking and noncancer
respiratory disorders, mainly in children. Possible biological mechanisms, additional risk factors,
and the potential long-term significance of early effects on lung function are discussed. Then, the
evidence indicating relationships between childhood exposure to ETS and acute respiratory
illnesses, middle ear disease, chronic respiratory symptoms, asthma, and lung function
impairment, and between maternal smoking and SIDS, is evaluated.
Passive smoking as a risk factor for noncancer respiratory health effects in adults is also
analyzed in Chapter 7. The NRC and Surgeon General's reports concluded that adults exposed to
ETS may exhibit small deficits in lung function but noted that it is difficult to determine the
extent to which ETS impairs respiration because so many other factors can similarly affect lung
2-10
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
function. More recent evidence and new statistical techniques allow the demonstration of subtle
effects of ETS on lung function and respiratory health in adults.
Chapter 8 discusses potential confounding factors and possible sources of bias in the ETS
studies that might affect the conclusions of Chapter 7. Chapter 8 also describes methodological
and data considerations that limit quantitative estimation of noncancer respiratory health effects
attributable to ETS exposure. Finally, the chapter develops population impact assessments for
ETS-attributable childhood asthma and for infant/toddler bronchitis and pneumonia. Acute
respiratory illnesses are one of the leading causes of morbidity and mortality during infancy and
early childhood, and an estimated 2 to 5 million children under age 18 are afflicted with asthma.
Therefore, even small increases in individual risk for these illnesses can result in a substantial
public health impact.
2-11
05/15/92
-------
-------
DRAFT—DO NOT QUOTE OR CITE
3. ESTIMATION OF ENVIRONMENTAL TOBACCO SMOKE EXPOSURE
3.1. INTRODUCTION
This chapter considers some of the major issues relevant to assessing human exposure to
environmental tobacco smoke (ETS). Current information on the nature of ETS, use of marker or
proxy compounds for ETS, measured personal exposures to ETS proxies, measured concentrations
of ETS proxy air contaminants in various indoor environments, use of ETS biomarkers, current
models for assessing ETS proxy concentrations, and use of questionnaires for assessing exposure to
ETS is summarized and discussed.
In the course of a typical day, an individual spends varying amounts of time in a variety
of microenvironments (e.g., residences, industrial and nonindustrial workplaces, automobiles,
public access buildings, outdoors). While in these microenvironments, individuals are exposed to a
broad and complex spectrum of organic and inorganic chemicals in gaseous and particle forms, as
well as a range of viable particles.
ETS is a major source of indoor air contaminants because of the large, though decreasing,
number of smokers in the population and the quantity and quality of the contaminants emitted
into the environment from tobacco combustion (NRC, 1981, 1986). Although no national surveys
have been published, the ubiquitous nature of ETS in indoor environments indicates that some
unintentional inhalation of ETS by nonsmokers is unavoidable. The combustion of tobacco results
in the emission of a particularly complex array of air contaminants into indoor
microenvironments. The nature of the resultant ETS contaminant mix and eventual human
exposure is the product of the interaction of several interrelated factors associated with the source,
transport, chemical transformation, dispersal, and removal, as well as human activities. Efforts to
determine adverse health and nuisance effects of ETS must address the issue of exposure to a
complex mixture that occurs in a number of microenvironments and must recognize that assessing
ETS exposures is inherently complicated. Fully assessing ETS exposures would involve
determining the time-weighted sum of exposures to each constituent in a multiplicity of
microenvironments. Because this cannot be done, a simplified approach using biological or
atmospheric markers, or questionnaires is generally used.
Accurate methods of assessing ETS exposures are needed for conducting epidemiologic
studies, for calculating risks, and for developing effective control measures to reduce or eliminate
risks. In epidemiologic studies of ETS, accurate exposure information is crucial to minimize the
effects of misclassification and the influence of confounders and to improve the probability of
3-1
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
revealing exposure-response associations. In risk assessment, exposure assessment provides basic
information on the exposure-distribution curve (populations at a range of exposures) for ETS and
essential information to calculate dose. ETS exposure assessment is essential in developing cost-
effective mitigation efforts to reduce or minimize ETS-associated risks and then to monitor
progress toward the targeted risk reduction.
Effective exposure assessment efforts require the identification of the health or nuisance
effect under study, the specification (when possible) of the biological response time of
the effect, and the ascertainment of the individual air contaminant, general group of air
contaminants, or contaminant source thought to be associated with that effect. It is difficult to
identify a single effect associated with a single air contaminant exposure and even more difficult
to determine a dose-response relationship. The outcome variable under study is generally part of
an effect complex related to other risk variables (e.g., health status, age, race, diet, personal
habits, occupation) and a variety of air contaminants emitted from a number of potential sources.
It is important to specify the duration, frequency, and magnitude of exposure to specific
contaminants or categories of contaminants on a time scale corresponding to the health or comfort
effect. Such a specification is necessary if an appropriate and adequate exposure assessment effort
is to be undertaken. For example, studies of ETS-associated chronic effects (e.g., cancer) would
ideally have ETS exposure measures integrated over periods of years, whereas studies of ETS-
associated acute effects (e.g., odor, eye irritation) require exposure measures over a period of a
few minutes. Specification of the biological response time under study is important in developing
an ETS exposure assessment strategy.
Exposure to individual air contaminants, categories of air contaminants, or sources of air
contaminants found outdoors and indoors can be assessed by direct and indirect methods. Direct
methods include personal monitoring and use of biological markers, measured in the subject
population. The indirect method employs models to estimate exposures. The modeling approach
can employ the use of stationary monitoring and questionnaires. Stationary monitoring, with
passive or active methods, is used to measure concentrations of air contaminants in different
environments. These measured concentrations are then combined with time activity patterns (time
budgets) to determine the average exposure of an individual as the sum of the concentrations in
each environment weighed by the time spent in that environment. Monitoring of contaminants
might also be supplemented with the monitoring of factors in the environment that impact the
contaminant levels measured (e.g., meteorological variables, primary compounds, ventilation).
Measurement of these factors in a carefully chosen set of conditions can lead to models that
3-2
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
predict concentrations in the absence of measured concentrations and provide a means of assessing
the impact of efforts to reduce or eliminate exposures. Questionnaires are used to determine time
activity patterns of individuals, to provide a simple categorization of potential exposure, and to
obtain information on the properties of the environment that have an impact on the measured
levels (e.g., presence of sources, source use).
3.2. PHYSICAL AND CHEMICAL PROPERTIES
ETS comprises aged exhaled mainstream smoke (MS) from the smoker, diluted sidestream
smoke (SS) emitted from the smoldering tobacco between puffs, contaminants emitted into the air
during the puff, and contaminants that diffuse through the cigarette paper (NRC, 1986; U.S.
DHHS, 1986; Guerin et al., 1992). SS is the principal contributor to ETS.
Chemical characterization of MS and SS air contaminant emissions from cigarettes, cigars,
or pipes is derived from laboratory-based studies that have typically used standardized testing
protocols (Brunnemann et al., 1976; Wynder and Hoffman, 1967; Dube et al., 1982). The data
available are primarily for tobacco combustion in cigarettes. These protocols employ smoking
machines, set puff volumes and frequencies, and standardize air contaminant collection protocols
(e.g., small chambers, Cambridge filters, chamber airflow rates). Existing protocols reflect
conditions representative of human smoking practices of more than 30 years ago and do not
reflect current human smoking parameters (NRC, 1986; U.S. DHHS, 1986). MS and SS air
contaminant emission rates determined in these studies can be affected by a number of factors,
such as puff volume, air dilution rate, paper porosity, and moisture content of the tobacco.
Variability in any of the factors can affect the nature and quantity of the emissions.
Results of laboratory evaluations have indicated substantial similarities and some
. differences between MS and SS emissions from cigarettes (NRC, 1986; U.S. DHHS, 1986). The air
contaminants emitted in MS and SS are very similar in their chemical composition. Differences in
SS and MS emissions are attributable to differences in the temperature of combustion of the
tobacco, pH, and degree of dilution with air, which is accompanied by a corresponding rapid
decrease in temperature. SS is generated at a lower temperature (600°C vs. 900°C) and at a higher
pH (6.7-7.5 vs. 6.0-6.7) than is MS. SS is diluted rapidly with air. The size of SS particles is
smaller than MS particles (SS particle size is 0.01-1.0 jim, whereas MS particle size is 0.1-1.0 /on).
At the higher pH of SS, the proportion of unprotonated nicotine in the smoke increases, with SS
nicotine predominantly in the vapor phase, while, in MS, nicotine is principally particle phase.
3-3
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
More than 4,000 compounds have been identified in laboratory-based studies of tobacco
smoke (Dube et al., 1982). Part of the data available from these studies is shown in Table 3-1. It
is immediately obvious from Table 3-1 that SS and MS contain many of the same notable air
contaminants, including several known or suspected human toxic and carcinogenic agents (e.g.,
carbon monoxide, ammonia, nicotine, tobacco-specific nitrosamines, benzo[a]pyrene, benzene,
cadmium, nickel, aromatic amines). More than 20 carcinogens have been identified in ETS.
Many of these toxic and carcinogenic compounds are emitted at levels higher in SS than in MS.
For example, #-Nitrosodimethylamine, a potent animal carcinogen, is emitted in quantities 20 to
100 times higher in SS than in MS. A number of recent studies indicate that although filtering MS
(filter cigarettes) will reduce the MS emissions for a number of compounds, it does not
substantially reduce the emission rates for most SS constituents, particularly for known toxic and
carcinogenic compounds (Adams et al., 1987; Guerin et al., 1987; Higgins, 1987).
The available data indicate that tobacco combustion will result in the emission of known
toxic and carcinogenic contaminants into the environment, resulting in exposure to these
contaminants by nonsmokers. It is important to note, however, that although the SS emissions are
higher than MS emissions for many compounds, the dilution rate into the environment of SS is
rapid, thus substantially lowering actual exposure concentrations of the contaminants. In cases
where the SS emissions or exhaled MS emissions are in direct proximity to a nonsmoker (e.g., an
infant held by a smoking mother or father), the nonsmoker's exposure to ETS contaminants will
be high.
Few emission data have been collected under conditions more typical of actual smoking
conditions (e.g., using smokers rather than smoking machines). It is not known how the MS and
SS air contaminant emission data for specific compounds generated by the standardized testing
protocols compare to data gathered under conditions more representative of actual smoking.
3.3. ASSESSING ETS EXPOSURE
3.3.1. Markers for Environmental Tobacco Smoke
Although ETS is a major source of indoor air contaminants, the actual contribution of ETS
to indoor air is difficult to assess in the background of many contaminants contributed from a
variety of other indoor and outdoor sources. Relatively few of the thousands of individual
constituents of the ETS mix have been identified and characterized. In addition, little is known
about the role of individual ETS constituents in eliciting the adverse health and nuisance effects
observed. However, the issue is not how to fully characterize the exposure to each ETS-related
3-4
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
contaminant, but rather how to obtain accurate quantitative measures of exposure to the entire
ETS mixture. The measurement of all components in ETS is not feasible, practical, or even
desirable because of limitations in knowledge of the mixture components related to the effects of
interest as well as the feasibility and cost of sampling. It is necessary then to identify a marker
(also referred to as a tracer, proxy, indicator, or surrogate) for ETS that, when measured, will
accurately represent the frequency, duration, and magnitude of exposure to ETS. These markers
can be chemicals measured in the air, biomarkers, models, or simple questionnaires.
There are important issues related to the measurement of a given marker compound to
represent exposure to ETS. Ideally, an air contaminant marker for ETS should (1) vary with
source strength, (2) be unique to the source, (3) be easily detected in air at low concentrations, (4)
be similar in emission rates for a variety of tobacco products, (5) occur in a consistent ratio in air
to other ETS components in the complex mix, and (6) be easily, accurately, and cost-effectively
measured (Leaderer, 1990). The marker can be a specific compound (e.g., nicotine) or much less
specific (e.g., respirable suspended particle mass). These criteria for selecting a suitable marker
compound are the ideal criteria. In practice, no single contaminant or class of contaminants has
been identified that would meet all the criteria. Selection of a suitable marker for ETS is reduced
to satisfying as many of the criteria for judging a marker as is practical. In using a marker, it is
important to state clearly the role of the marker and to note its limitations.
A number of marker or proxy compounds have been used to represent ETS concentrations
in both field and chamber studies. Nicotine, carbon monoxide, 3-ethenylpyridine, nitrogen
dioxide, pyridine, aldehydes, nitrous acid, acrolein, benzene, toluene, myosmine, and several
other compounds have been used or suggested for use as markers or proxies for the vapor phase
constituents of ETS (NRC, 1981, 1986; U.S. DHHS, 1986; Hammond et al., 1987; Eatough et al.,
1986; Lofroth et al., 1989; Leaderer and Hammond, 1991). Tobacco-specific nitrosamines,
particle phase nicotine and cotinine, solanesol, polonium-210, benzo[a]pyrene, potassium,
chromium, and respirable suspended particle (RSP) mass (RSP mass < 2.5 /an) are among the air
contaminants used or suggested for use as markers for particle phase constituents of ETS (NRC,
1981, 1986; U.S. DHHS, 1986; Leaderer and Hammond, 1991; Benner et al., 1989; Hammond et
al., 1988; Rickert, 1984). All the markers employed to date have some problems associated with
their use. For example, carbon monoxide, nitrogen oxides, benzene, and RSP have many indoor
and outdoor sources other than the combustion of tobacco, while other compounds, such as
nitrosamines and benzo[a]pyrene, are sufficiently difficult to measure (e.g., concentrations in
3-5
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
smoking environments are low and the cost of collection and analysis of samples is high) that their
use is very limited.
At the present time, vapor phase nicotine and respirable suspended particulate matter are
widely and most commonly used as markers of the presence and concentration of ETS for a
variety of reasons associated with their ease of measurement, existing knowledge on the their
emission rates from tobacco combustion, and their relationship to other ETS contaminants.
Vapor phase nicotine, the dominant form of nicotine in ETS (NRC, 1986; U.S. DHHS,
1986; Hammond et al., 1987; Eatough et al., 1986; Eudy et al., 1985), accounts for approximately
95% of the nicotine in ETS and is a good marker air contaminant for ETS. It is specific to tobacco
combustion and emitted in large quantities in ETS (NRC, 1981, 1986; U.S. DHHS, 1986; Rickert
et al., 1984; Eatough et al., 1990). Chamber measurements have shown that nicotine
concentrations vary with source strength (Hammond et al., 1987; Hammond and Leaderer, 1987)
and show little variability among brands of cigarettes despite variations in MS emissions (Leaderer
and Hammond, 1991; Rickert et al., 1984). Field studies have shown that weekly nicotine
concentrations are highly correlated with the number of cigarettes smoked (Leaderer and
Hammond, 1991; Hammond et al., 1987; Mumford et al., 1989; Hammond et al., 1989). One large
field study (Leaderer and Hammond, 1991) showed that weekly nicotine concentrations were
strongly correlated with measured RSP levels as well as with reported number of cigarettes
smoked. In this study, the slope of the regression line was 10.8, similar to the RSP/nicotine level
seen in chamber studies. The RSP intercept was equal to background levels in homes without
smoking. A comparable study by Miesner et al. (1989) of particulate matter and nicotine in
workplaces found a similar ratio between RSP and nicotine. The utility of nicotine as an ETS
marker is enhanced by the fact that recent advances in air sampling have resulted in the
development of a variety of validated and inexpensive passive and active monitoring methods for
measuring nicotine in indoor air environments and for personal monitoring (Hammond et al.,
1987; Hammond and Leaderer, 1987; Marbury et al., 1990; Eatough et al., 1989a; Koutrakis et al.,
1989; U.S. DHEW, 1977; Muramatsu et al., 1984; Oldaker and Conrad, 1987). In addition,
measurements of nicotine and cotinine, a metabolite of nicotine, in blood, urine, and saliva are
used extensively as biomarkers of exposure to ETS.
The combustion of tobacco results in substantial emissions of RSP. One small chamber
study using a smoking machine found the average particle emission rate for 15 Canadian cigarettes
to be 24.1 mg per cigarette with a range of 15.8 to 36.0 mg per cigarette (Rickert et al., 1984). A
large chamber study using smokers reported an average particle emission rate of 17.1 mg for 12
3-6
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
brands of American cigarettes (Leaderer and Hammond, 1991). This study noted that emission
rates among brands are similar. Included in the RSP are a number of compounds of direct health
concern—for example, many of the polycyclic aromatic hydrocarbons (NRC, 1986; U.S. DHHS,
1986; Table 3-1). There are a number of accepted methods to measure personal RSP exposures
and concentrations in indoor environments (Ogden et al., 1989).
Studies of personal exposures to RSP and of RSP levels in indoor environments have
shown elevated levels of RSP when any ETS exposure was reported (NRC, 1986; U.S. DHHS,
1986; Leaderer and Hammond, 1991; Repace and Lowrey, 1980; Ishizu, 1980). One study found a
strong correlation between weekly residential RSP levels and reported number of cigarettes
smoked (Leaderer and Hammond, 1991). At low smoking and high ventilation rates, however, it
may be difficult to separate the ETS-associated RSP in a background of RSP from other indoor
sources (e.g., kerosene heaters) or even from outdoor sources. Efforts to model ETS exposures for
the purpose of assessing risks and the impact of various mitigation measures have often focused on
predicting ETS-associated RSP concentrations (e.g., Leaderer, 1988; Repace and Lowrey, 1980).
3.3.2. Measured Exposures to ETS-Associated Nicotine and RSP
3.3.2.1. Personal Monitors
Personal monitoring allows for a direct integrated measure of an individual's exposure.
Personal air monitoring employs samplers (worn by individuals) that record the integrated
concentration of a contaminant that individuals are exposed to in the course of their normal
activity for time periods of several hours to several days. The monitors can be active (employing
pumps to collect and concentrate the air contaminant) or passive (working on the principle of
diffusion). As with biomarkers, personal monitoring provides an integrated measure of exposure
to air contaminants across a number of environments in which an individual spends time, but it
does not provide direct information on concentrations of the air contaminant of interest in
individual environments or on the level of exposure in each environment unless samples are taken
in only one environment or are changed with each change of environment. Supplemental
information (e.g., air monitoring of spaces, time-activity patterns) is needed to determine the
contribution of each microenvironment to total exposure.
There are relatively few studies reported that have measured personal exposures to ETS-
associated nicotine and RSP for nonsmoking individuals. The few reported studies for personal
exposure to nicotine are summarized in Table 3-2. Personal exposures associated with specific
indoor environments are presented. The indoor environments include the nonindustrial
3-7
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
workplace, homes, restaurants, public buildings, and transportation-related indoor spaces. Table
3-2 also highlights the wide range of indoor environments in which ETS exposures take place and
the wide range of personal exposures encountered in those environments. It is important to note,
however, that there are relatively few observations available and that the observations for non-
workplace nicotine exposures is dominated by the Japanese data (Muramatsu et al., 1984), which
may not be representative of personal exposures in the United States. Because the data are
limited, specific conclusions related to the contribution of different indoor environments to
personal nicotine exposures associated with passive smoke cannot be drawn. The data do indicate,
however, that a wide range of exposures to ETS occur in a variety of indoor environments where
smoking is permitted. The data also indicate that the occupational and residential environments
are important sources of exposure to ETS because of the levels encountered, which are
comparable, and the length of time individuals spend in them.
Those studies of personal exposure to RSP for nonsmoking individuals that have attempted
to stratify the collected data by ETS exposure are shown in Table 3-3. Three of the five studies
represent exposures integrated over several microenvironments (e.g., residential, public buildings,
occupational), while two studies report exposures for the workplace only. Individuals reporting
exposure to ETS have substantially higher integrated exposures to RSP than do those reporting no
exposure. Passive smoke exposure resulted in increases in personal RSP exposures beginning at 18
to 64 jtg/m3. It is difficult to assess the ETS contribution to personal RSP levels for each indoor
environment for the 24-hour RSP personal exposures. The contribution of each of these indoor
environments must be substantially higher than the 24-hour averages presented, because exposures
presumably did not take place during sleeping hours or in all microenvironments. Table 3-3
demonstrates that the contribution of ETS-related RSP in the work environment to personal
exposure is important and variable.
The most extensive study of personal exposures to RSP clearly demonstrates the impact on
RSP levels from exposure to ETS (Spengler et al., 1985). In this study, outdoor, indoor, and
personal 24-hour concentrations of RSP (particle diameter < 3.5 /on) were obtained for a
nonsmoking sample of 101 individuals. Of the 101, 28 persons reported some exposure to ETS in
either the home or workplace, while 73 reported no ETS exposure. The cumulative frequency
distributions of RSP for the ETS-exposed and non-ETS-exposed individuals and measured
outdoor levels are shown in Figure 3-1. Those reporting ETS exposure had mean personal RSP
levels 28 fig/m3 higher than those reporting no ETS exposure (Table 3-3). A larger variation in
RSP concentrations was also seen for those reporting ETS exposure.
3-8
05/15/92
-------
DRAFT-DO NOT QUOTE OR CITE
3.3.2.2. Measurements Using Stationary Monitors
Concentrations of nicotine, RSP, and other ETS constituents in an enclosed space can
exhibit a pronounced spatial and temporal distribution. The concentration is the result of a
complex interaction of several important variables, including (1) the generation rate of the
contaminants from the tobacco, (2) location in the space that smoking occurs, (3) the rate of
tobacco consumption, (4) the ventilation or infiltration rate, (5) the concentration of the
contaminants in the ventilation or infiltration air, (6) air mixing in the space, (7) removal of
contaminants by surfaces or chemical reactions, (8) reemission of contaminants by surfaces, and
(9) the effectiveness of any air cleaners that may be present. The choice of location for obtaining
an RSP or nicotine measurement, the timing of sample collection, and the duration of sampling
should take into consideration the aforementioned factors.
In the past several years, numerous studies have been conducted in a variety of indoor
environments to determine the impact of tobacco combustion on levels of nicotine and RSP.
These studies have employed a variety of protocols that used a diversity of air sampling
techniques (e.g., passive, active, continuous integrative), sampled over highly varying timeframes
(from minutes to several days), and collected highly variable information on factors affecting the
measured concentrations (e.g., number of cigarettes smoked, volume of building, ventilation
rates). In an attempt to present an overall view of the contribution of ETS to indoor air quality,
only the summary results of the measured concentrations of ETS-associated nicotine and RSP will
be discussed here. Several reviews of the studies evaluating the impact of ETS on indoor RSP
levels have been conducted over the past few years, and a number of recent reports have discussed
measured indoor levels of nicotine (e.g., NRC, 1986; U.S. DHHS, 1986; Leaderer and Hammond,
1991). More detailed information is provided in those reports and the individual study reports.
A summary of measured nicotine concentrations in various indoor environments where
smoking was noted is summarized in Figure 3-2. The mean concentration, standard deviation,
and maximum and minimum nicotine values recorded are presented. Also given in Figure 3-2 is
the number of locations in which the measurements were taken and the reference in which the
data were reported. Elevated nicotine levels were measured in all microenvironments in which
smoking was reported. Measured nicotine levels, as would be expected, were highly variable,
covering several orders of magnitude.
The home and workplace environments may represent the most important environments
for exposure to ETS because of the length of time individuals spend there. For the four studies
reported, nicotine levels in homes where smoking occurs ranged from less than 1 j^g/m3 (Leaderer
3-9
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
and Hammond, 1991) to more than 14 /tg/m3 (U.S. DHEW, 1977). For two of the studies
(Leaderer and Hammond, 1991; U.S. DHEW, 1977), nicotine concentrations represent weekly
averages. Actual concentrations in the homes during nonsleeping occupancy (i.e., while smoking
would be occurring) would be considerably higher than the levels presented in Figure 3-2 (a
factor of 3 or higher). Workplace nicotine also demonstrated a wide range of concentrations, from
near zero to more than 33 /tg/m3. In other environments, nicotine concentrations demonstrated
considerable variability. It is important to note that short-term concentrations (on the order of
minutes) are likely to show considerably more variability, resulting in considerably higher short-
term peak exposures.
In one large study of residential levels of ETS-associated nicotine and RSP (Leaderer and
Hammond, 1991), both were found to be highly correlated with reported number of cigarettes
smoked. This study found that, consistent with chamber data, measured nicotine concentrations
predicted the contribution to residential RSP levels from tobacco combustion (Figure 3-3). The
data in Figure'3-3 might be used to estimate the RSP levels associated with tobacco combustion
from the nicotine levels shown in Figure 3-2.
A substantial number of studies examining the impact of tobacco combustion on
concentrations of RSP in various indoor environments have been reported. Many of these studies
have reported outdoor RSP concentrations and indoor RSP levels without smoking as well as
concentrations when smoking occurs. These studies are summarized in Figure 3-4. The sampling
time for the presented data ranged from 1 minute to more than several days. A major portion of
the data is for the residential indoor environment. Where smoking is reported, RSP levels are
considerably higher than where it is not. RSP levels associated with smoking, like those for
nicotine, demonstrated considerable variability ranging from a few /tg/m3 to more than 1 mg/m3.
Workplace RSP levels associated with smoking occupancy are comparable to residential RSP levels.
Indoor levels of nicotine and RSP associated with the combustion of tobacco are a function
of several factors related to the generation, dispersal, and removal of ETS in enclosed
environments. Thus, measured levels of these air contaminants indicate a wide range of
concentrations. Figures 3-5 and 3-6 present a summary of the range of nicotine and ETS-
associated particle concentrations measured by type of environment. The figures present the
range of average values reported for each study and the minimum and maximum values reported.
Only studies reporting sampling times over 4 hours were included in the residential and office
summaries in Figures 3-5 and 3-6 because averaging time is more likely to represent the exposures
associated with occupancy time (this included most of the studies for residential spaces shown in
3-10
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Figures 3-2 and 3-4). Because occupancy time in other environments (e.g., restaurants) is likely
to be much shorter (for patrons, but not for service personnel), averaging times on the order of
minutes or greater were considered for the other indoor environments presented in the figures.
Indoor particulate levels associated with smoking occupancy (Figure 3-6) were calculated by
subtracting particle levels for nonsmoking occupancy (presented in the studies) from the smoking
occupancy levels. Thus, the increase in particle mass concentrations associated with ETS is
presented in Figure 3-6.
The summary nicotine data (Figure 3-5) suggest that average nicotine values in residences
with smoking occupancy will range from 2 jig/m3 to approximately 10 /ig/m3, with high values up
to 14 /ig/m3 and low values down to 0.1 /tg/m3. Average nicotine concentrations in offices with
smoking occupancy show a range of average concentrations similar to that of residences, but with
considerably higher maximum values. The data from other indoor spaces suggest considerable
variability, particularly in the range of maximum values. The cumulative distribution of weekly
nicotine measured in one study (Leaderer and Hammond, 1991) of a sample of 96 homes, with the
levels for smoking occupancy emphasized, is shown in Figure 3-7.
Residential particle mass concentrations will increase from 18 to 95 /ig/m3 with smoking
occupancy, while the recorded increases can be as high as 560 /tg/m3 or as low as 5 /ig/m3 (Figure
3-6). Figure 3-8 (Leaderer and Hammond, 1991) highlights the distribution of weekly RSP
concentrations for residences with smoking occupancy. In that study, smoking residences had RSP
concentrations approximately 29 /ig/m3 higher than nonsmoking homes. Average concentrations
in offices with smoking occupancy will be lower on average than in residences. Restaurants,
transportation, and other indoor spaces with smoking occupancy will result in a considerably
wider range of average, minimum, and maximum increases in particle concentrations than in the
residential or office environments.
As noted earlier, indoor air contaminant concentrations are the result of the interaction of
a number of factors related to the generation, dispersal, and elimination of the contaminants.
Source use is no doubt the most important factor. Few studies have measured contaminant
concentrations as a function of the smoking rate in residences or offices, but some data are
available. One study estimated an average weekly contribution to residential RSP of 2 to 5 /ig/m3
per cigarette (Leaderer et al., 1990), while another study estimated that a pack-a-day smoker
would add 20 /ig/m3 to residential levels (Coghlin et al., 1989). Variations in residential RSP
levels as a function of the number of smokers and over a period of several months are
demonstrated in Figure 3-9 (Spengler et al., 1981). An association between the reported number
3-11
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
of cigarettes and weekly residential nicotine and RSP levels for a sample of 96 homes (Leaderer
and Hammond, 1991) is shown in Figure 3-10. Smoking clearly increases indoor concentrations of
both nicotine and particle mass. Residential levels of both nicotine and particle mass increase
with increasing levels of smoking. Because nicotine and particle mass are proxies for the complex
ETS contaminant mix, it is expected that other ETS air contaminants, including the toxic and
carcinogenic contaminants, will be elevated with smoking occupancy.
Children have been identified as a particularly sensitive group at health risk from exposure
to ETS in the residential indoor environment (NRC, 1986; U.S. DHHS, 1986). One study has
measured smoking status of the parents and weekly nicotine concentrations in the activity room
and bedroom of 48 children under the age of 2 years. The results, shown in Table 3-4, indicate
that activity room and bedroom concentrations of nicotine in the children's homes increase with
the reported number of cigarettes smoked in the home by parents. Concentrations also increased
with the number of reported smokers in the household. Correlation coefficients of more than 0.7
were calculated between nicotine concentrations and number of cigarettes smoked.
It is important to note that while measurements of nicotine and ETS-associated RSP are
excellent indicators of the contribution of ETS to air contaminant levels in indoor environments,
their measurement does not directly constitute a measure of total exposure. The concentrations
measured in all indoor environments have to be combined with time-activity patterns in order to
determine average exposure of an individual as the sum of the concentrations in each environment
weighted by the time spent in that environment. Both the home and the work environment (those
without policies restricting smoking) have highly variable ETS concentrations, the ranges of which
are largely overlapping. Which environment is most important in determining total exposure will
vary with individual circumstances. For example, one who lives in a smoker-free home but works
in an office with smokers will receive most ETS exposure at work; however, for those exposed
both at home and at work, the home may be more important because, over the course of a week,
more time is spent at home (assuming equal exposure concentrations).
An additional issue to be considered is how well the general indoor concentrations
represent exposures of individuals who may be directly exposed to the SS plume of ETS. Small
children, particularly infants, being held by smoking parents may receive exposures considerably
higher than those predicted from concentrations reported for indoor spaces. Special consideration
must be given to these significant subpopulations.
3-12
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
3.3.3. Biomarkers of ETS Exposure
Biomarkers of exposure are actually a measure of dose or uptake and hence an indicator
that an exposure has taken place. Biomarkers, within the context of assessing exposure to air
contaminants, refer to cellular, biochemical, or molecular measures that are obtained from
biological media such as human tissues, cells, or fluids and are indicative of human exposure to air
contaminants (Collier et al., 1990; Goldstein et al., 1987). The relation between the biomarker
and exposure, however, is complex and varies as a function of several factors, including
environmental factors and the uptake, distribution, metabolism, and site and mode of action of the
compound or compounds of interest.
Ideally, a biomarker of exposure for a specific air contaminant should be chemically
specific, have a long half-life in the body, be detectable in trace quantities with high precision, be
measurable in samples easily collected by noninvasive techniques, be inexpensive to assay, be
either the agent that is associated with the effects or strongly associated with the agent of interest,
and be quantitatively relatable to a previous exposure regimen. Ideal biomarkers for air
contaminants, such as markers for complex mixtures, do not exist.
Numerous biomarkers have been proposed as indicators for ETS (e.g., thiocyanate,
carboxyhemoglobin, nicotine and cotinine, N-Nitrosoproline, aromatic amines, protein or DNA
adducts) (NRC, 1986; U.S. DHHS, 1986). Although these biomarkers demonstrate that an
exposure has taken place, they may not be directly related to potential for development of the
adverse effect under study (not the contaminant directly implicated in the effect of interest), they
can show considerable variability from individual to individual, and they represent only fairly
recent exposure (potentially inadequate for chronic outcomes). Furthermore, some of these
markers may not be specific to ETS exposure (e.g., carboxyhemoglobin), while others (e.g.,
thiocyanate) may not be sensitive enough for ETS exposures.
Nicotine and its metabolite, cotinine, in the saliva, blood, and urine are widely used as
biomarkers of active smoking and exposure to ETS and are valuable in determining total or
integrated short-term dose to ETS across all environments (NRC, 1986; U.S. DHHS, 1986).
Nicotine and cotinine are specific to tobacco and are accurately measured by gas chromatography,
radioimmunoassay, or high-pressure liquid chromatography in concentrations down to 1 ng/ml.
Nicotine has a half-life typically of about 2 hours in the blood and is metabolized to cotinine and
excreted in the urine. The short half-life of nicotine makes it a better indicator of very recent
exposures rather than a measure of integrated exposure.
3-13
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Cotinine in saliva, blood, and urine is the most widely accepted biomarker for integrated
exposure to MS or ETS (NRC, 1986; U.S. DHHS, 1986). Cotinine is the major metabolite of
nicotine, is specific to tobacco, and has a longer half-life for elimination from the body. The
elimination half-life in smokers is approximately 20 hours (range of 10 to 37 hours), while it is
typically longer in nonsmokers with ETS exposure, particularly in children (Figure 3-11) (Elliot
and Rowe, 1975; Collier et al., 1990; Goldstein et al., 1987; Etzel et al., 1985; Greenberg et al.,
1984). The longer half-life of cotinine makes it a good indicator of integrated ETS exposure over
the previous day or two. Laboratory studies of nonsmokers exposed to acute high levels of ETS
over varying times have shown significant uptake of nicotine by the nonsmokers and increases in
their cotinine levels (NRC, 1986; U.S. DHHS, 1986; Hoffmann et al., 1984; Russell and
Feyerabend, 1975).
Several studies have been conducted of cotinine levels in free-living populations of
smokers, nonsmokers reporting passive smoke exposure, and nonsmokers reporting no passive
smoke exposure (NRC, 1986; U.S. DHHS, 1986; Greenberg et al., 1984; Wald et al., 1984; Wald
and Ritchie, 1984; Jarvis, et al., 1985; Coultas et al., 1987; Riboli et al., 1990; Cummings et al.,
1990). These studies have found that exposure to ETS is highly prevalent even among those living
with a nonsmoker (e.g., Cummings et al., 1990). Saliva, serum, and urine cotinine levels in ETS-
exposed nonsmokers were generally found to be higher than those in nonsmokers reporting no
ETS exposure, and levels of cotinine in smokers are considerably higher than those levels in
nonsmokers passively exposed (Table 3-5). Cotinine levels in nonsmokers exposed to ETS are on
the order of approximately 1% of the levels in active smokers. Cotinine levels of nonsraokers have
been found to increase with self-reported ETS exposure (Figures 3-12 and 3-13).
In a 10-country study of ETS exposure of 1,369 nonsmoking women (Riboli et al., 1990),
average urinary levels of cotinine/creatinine by country ranged from approximately 2.5 ng/mg for
Shanghai to approximately 14 ng/mg for Trieste. Eighty percent of those women sampled had a
detectable level of cotinine. Statistically significant differences were observed between centers,
with the lowest values observed in Honolulu, Shanghai, and Chandigarh and the highest values in
Trieste, Los Angeles, and Athens. This study also found a linear increase in cotinine/creatinine
levels for the group of women reporting no ETS exposure either at home or work to the group
reporting ETS exposure both at home and at work (Figure 3-14). Urinary cotinine levels were
also found to increase with the number of questionnaire-reported passive smoke exposures in a
group of 663 persons who never smoked and ex-smokers (Cummings et al., 1990). In that study,
76% of the subjects reported passive smoke exposure, with 27% reporting exposure at home and
3-14
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
28% reporting exposure at work. Jarvis et al. (1983) studied the increase of cotinine in 7
nonsmokers after 2 hours' exposure to ETS in a "smoky public house." They found highly
statistically significant increases of cotinine in all body fluids: from 1.1 to 7.3 ng/mL in plasma,
from 1.5 to 8.0 ng/mL in saliva, and from 4.8 to 12.9 ng/mL in urine. Because the samples were
taken immediately post-exposure, they do not indicate peak cotinine concentrations, however.
Cotinine values in smokers and nonsmokers measured in either the laboratory or field
setting show considerable variability attributable to individual differences in the uptake
distribution, metabolism, and elimination of nicotine. An additional issue that has to be
considered in interpreting the field data is that exposure status is determined by respondent self-
reporting. This can lead to a misclassification error, which tends to reduce the differences in
cotinine levels measured in the ETS-exposed versus non-ETS-exposed groups and to increase the
variability in the levels within any exposure category. Within the exposed group, this
misclassification error could either increase or decrease the average cotinine levels measured.
It is important to recognize that nicotine and cotinine are actually proxy biomarkers. They
may not be the active agents in eliciting the adverse effect under study but merely indicative of
the level of passive smoke exposure. Using these measures to estimate cigarette equivalents or to
determine equivalent active smoking exposure could result in over- or underestimating exposure
to individual or classes of compounds that may be more directly related to the health or nuisance
effect of concern. The use of different biomarker proxies (e.g., protein adducts) could result in
estimates of much larger cigarette equivalent doses.
Nicotine and cotinine levels in ETS-exposed nonsmokers measured in laboratory and field
studies have been used to estimate cigarette equivalent exposures and to equate ETS exposures
with active smoker exposures (NRC, 1986; U.S. DHHS, 1986; Jarvis, 1989). On an equivalent
cigarette basis, an upper-bound estimate of nicotine dose of 2.5 mg per day for passive smoke
exposure has been proposed (Jarvis, 1989). This would translate into the equivalent of about one-
fifth of a cigarette per day, or about 0.7% of the average smoker's dose of nicotine (cigarette
equivalent dose of other toxins or carcinogens would be different, as described above).
Comparisons of cotinine values in ETS-exposed nonsmokers with those measured in smokers
ranged from 0.1% to 2%. One analysis proposed that, on average, nonsmokers' cotinine levels are
0.5% to 0.7% of those found in cigarette smokers (Jarvis, 1989). It should be noted that these
estimations are based on a number of assumptions that may not hold (e.g., the half-life of nicotine
and cotinine in smokers and nonsmokers is the same).
3-15
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
One of the protein adducts that has been used as a biomarker of active and passive
smoking is the 4-aminobiphenyl adduct of hemoglobin (4-ABP-Hb). One advantage of
hemoglobin adducts is that their half-life is quite long, and they will persist through the life of a
red blood cell, which is approximately 120 days. Therefore, levels of 4-ABP-Hb reflect exposures
over the past several weeks, rather than the day or two of exposure-integration reflected by
cotinine measurements.
Tobacco smoke is the primary environmental source of 4-aminobiphenyl (its use in the dye
industry was discontinued decades ago), and smokers have between five and eight times as much
4-ABP-Hb adducts as nonsmokers (Hammond et al., 1990; Perera et al., 1987; Maclure et al.,
1989). That nonsmokers appear to have approximately 10% to 20% the adduct level as smokers
may at first appear to be contradictory to the urinary cotinine ratios of about 1%, but in fact both
results are quite consistent with our knowledge of the emissions of various contaminants in MS
and SS. Approximately twice as much nicotine is emitted in SS as in MS, but about 31 times as
much 4-ABP is emitted in SS as in MS. Thus, compared to MS, SS is 15 times more enriched in 4-
ABP than in nicotine. The ratio of biomarkers in those exposed to ETS compared to smokers is 15
times greater for the biomarker 4-ABP-Hb than for the biomarker cotinine, a metabolite of
nicotine.
The above discussions indicate that the "cigarette equivalent" dose of those exposed to ETS
varies with the compound, so that a passive smoker may receive 1% as much nicotine as an active
smoker but 15% as much 4-ABP. These commentaries on the data are preliminary and warrant
further investigation, but they do suggest the importance of careful interpretation of biomarkers
in estimating dose.
3.3.4. Questionnaires for Assessing ETS Exposures
Questionnaires are the most commonly used method to assess exposure to ETS in both
retrospective studies of acute and chronic effects and in prospective studies. They are the least
expensive method of obtaining ETS exposure information for large populations. They can be used
to provide a simple categorization of ETS exposure, to determine time-activity patterns of
individuals (e.g., how much time is spent in environments where smoking occurs), and to acquire
information on the factors or properties of the environment affecting ETS concentrations (e.g.,
number of cigarettes smoked, size of indoor environments, subjective evaluation of level of
smokiness). The time-activity pattern information is combined with measured or estimated
concentrations of ETS in each environment to provide an estimate of total exposure. Information
3-16
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
on the factors affecting ETS concentrations is used to model or predict ETS levels in those
environments.
Questionnaires are used most extensively to provide a simple categorization of potential
ETS exposure (e.g., Do you live with a smoker? Are you exposed to ETS at your place of work?
How many hours a week are you exposed to ETS?) and to obtain information on possible
confounders (e.g., occupational history, socioeconomic status). When used simply to determine a
dichotomous exposure (ETS exposed vs. unexposed), any misclassification tends to bias measures
of association toward the null. Thus, any effect that may be present will be underestimated or
may even be undetectable. If there are more than two exposure categories (e.g., light, medium, or
heavy exposure), the intermediate categories of exposure may be biased either away from or
toward the null. Misclassification errors may arise from respondents' lack of knowledge, biased
recall, memory failure, or intentional alteration of information. In addition, there are
investigator-based sources of misclassification. Errors may arise if semiquantitative levels are
incorrectly imputed to answers; for example, even if house exposures are higher than occupational
exposures, on average, for any given individual, the ranking may well be reversed from that of
the average.
In using questionnaires to assess exposure categories to ETS to determine time-activity
patterns and to acquire information on the factors affecting concentrations, it is important to
minimize the uncertainty associated with the estimate and to characterize the direction and
magnitude of the error.
Unlike those for active smoking, standardized questionnaires for assessing ETS exposures
in prospective .or retrospective studies of acute or chronic health or nuisance effects do not exist.
Questionnaires used to assess ETS exposure have typically not been validated. There is no "gold
standard" with which to validate the questionnaire. Various strategies, however, have been used to
assess the validity of diverse types of questionnaires used to assess ETS exposure. Efforts to
validate questionnaires have used survey data, air monitoring of nicotine in various
microenvironments, and nicotine or cotinine in body fluid samples.
One report (NRC, 1986) estimated an error rate of 5% in using surrogate responses in the
simple classification of an individual as ever/never smoked. Such a classification scheme (e.g.,
married to a smoker) has been used to assess a nonsmoking spouse's exposure in the home for
ETS-associated cancer outcome. A recent study (Leaderer and Hammond, 1991) of 96 homes
using a questionnaire to assess residential smoking and a passive nicotine air monitor found that
13% of the residences reporting no smoking had measurable levels of nicotine, while 28% of the
3-17
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
residences reporting smoking had nondetectable levels of nicotine. A good level of agreement
between questionnaire-reported number of cigarettes smoked and residential levels of ETS-related
RSP and nicotine was observed in this study (Figure 3-10).
Studies (Hammond et al., 1989; Schenker et al., 1990; Coultas et al., 1987; Riboli et al.,
1990; Cummings et al. 1990; Coultas et al., 1990a) comparing various measures of ETS exposure
(e.g., location of exposure, intensity of exposure, duration of exposure, number of cigarettes
smoked) with cotinine levels measured in physiological fluids generally meet with only moderate
success (explained variations on the order of 40% or less). The largest such study (Riboli et al.,
1990) was a collaborative effort conducted in 10 countries; correlations in the range of 0.3 to 0.51
(p < 0.01) were found between urinary cotinine levels and various measures of exposure derived
from questionnaire data. Using cotinine as a biomarker of exposure, studies indicated that a
substantial percentage of those persons reporting no ETS exposure by questionnaire do have
measurable exposure. Differences in the uptake metabolism and excretion of nicotine among
individuals make it difficult to use this measure as a "gold standard" in validating questionnaires.
Also, the recent exposure lasting 1 to 2 days that is measured by cotinine may differ from usual
exposure.
In one effort to develop a validated questionnaire (Schenker et al., 1990), 53 subjects were
asked detailed questions about their exposures to ETS, including location of exposures, number of
smokers, ventilation characteristics, number of hours exposed, proximity of smokers, and intensity
of ETS. They then wore a passive sampler for nicotine for 7 days and recorded the same
information regarding each exposure episode in daily diaries. Formulae were developed to score
the exposures on both the questionnaire and the diary, and these scores were then correlated to the
average nicotine concentrations measured over the 7-day period. Excellent correlation was found
(r2 « 0.83 for the questionnaire and 0.90 for the diary). However, the simple questions that have
most frequently been used in epidemiologic studies (e.g., whether a subject lived with a smoker,
number of hours the subject was exposed) were not nearly as well correlated with the measured
exposures. These results indicate that reliable questionnaires can be developed but that those used
in most studies in the past will lead to some random misclassification of exposure and, hence,
underestimation of any effect that may be present.
ETS exposures take place across a number of environments, with an individual's total
exposure a function of the amount of time spent in each environment and the concentration in
that environment. Questionnaires need to assess exposures across indoor environments. Personal
air monitoring or the measurement of a biomarker provides a method to validate ETS exposure
3-18
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
assessment questionnaires and to assess the contribution of each environment to total current
exposure.
Personal air monitoring and cotinine measurements in combination with questionnaires
have highlighted the importance of obtaining information on spouses' smoking status, smoking at
home, smoking at work, smoking in various other indoor environments (e.g., social settings,
vehicles, public places) amount of time in environments where smoking occurs, and the intensity
of the exposure (Hammond et al., 1989; Schenker et al., 1990; Coultas et al., 1987; Riboli et al.,
1990; Cummings et al. 1990; Coultas et al., 1990a).
3.4. MODELS FOR ASSESSING ETS EXPOSURE
Epidemiologic studies of ETS ideally should have direct measurements of the ETS
exposures for the target individuals or populations. It is, however, neither practical nor possible
in most instances to obtain such measurements. For example, in retrospective studies of lung
cancer or respiratory illnesses, air samples of contaminant levels in various microenvironments or
personal air sampling cannot be obtained. Current measurements may not be directly relevant
because exposures have changed over the past 20 years. In such cases, past and present ETS
exposures can be modeled. Models that predict ETS concentrations in various microenvironments
can be used either to estimate total exposure in combination with time-activity patterns or to
estimate the impact of variations in factors (e.g., number of cigarettes smoked, changes in
ventilation rates) that have an impact on microenvironmental concentrations. Models used for
predicting ETS concentrations in indoor spaces will be discussed here.
Predictive or exploratory models for indoor concentrations of ETS-associated air
contaminants are generally either physical/chemical or statistical in nature. The physical/chemical
model usually follows some form of the general mass balance equation. This approach requires
detailed information on the input parameters (e.g., ETS source strengths, infiltration rates,
mixing, reaction rates) to predict the indoor concentrations. The input parameters are either
measured in chamber studies and in homes or are estimated. This approach has been extensively
utilized in chamber studies of ETS-associated air contaminants (Repace and Lowrey, 1980; Hoegg,
1972; Leaderer et al., 1984). In one report, the mass balance equation was used to estimate the
range of indoor concentrations of RSP associated with ETS over a range of assumptions related to
the input parameters (NRC, 1986). The results of that effort are shown in Figures 3-15 and 3-16.
(These figures were taken directly from NRC, 1986.) Figures 3-15 and 3-16 allow for the easy
calculation of RSP mass from ETS in indoor environments for a range of conditions. These
3-19
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
figures highlight the large impact of tobacco combustion on indoor RSP levels and indicate that
variations in input parameters (e.g., smoking occupancy) can have substantial impacts on predicted
RSP levels.
The mass balance model in its original form requires the measurement or estimation of a
number of parameters and, hence, is not easily applied to field studies. A condensed version of
the mass balance equation for estimating ETS-generated RSP levels in a variety of indoor
microenvironments has been developed by using known emission rates of RSP for tobacco
combustion in combination with data from several sources, including both measured and estimated
parameters (e.g., RSP emission rates, smoking densities, infiltration or ventilation rates, deposition
rates)(Leaderer, 1988). The condensed model is given by:
Ceq=K(Dhs/Nv)
where: C,^ is the equilibrium RSP concentration in a space due to smoking in /*g/m3; Dhs = the
number of active smokers (burning cigarettes) per 100 m3; Nv is the infiltration/ventilation rate
for the space in air changes per hour; and K is calculated from standard conditions (smoking rates,
RSP emission rates, mixing rates, ventilation rates, and particle loss rates to surfaces) and is equal
to 217 for spaces with three or more smokers, 145 for two smokers, and 72 for one smoker
(Repace, 1987). The authors of this approach are currently modifying the model to incorporate
nicotine measurements (Repace and Lowrey, in preparation). This simplified model offers an
easy method to estimate exposures to RSP-associated ETS. While the model has not been fully
validated, it does offer an easy method by which RSP-associated ETS in various indoor spaces can
be easily estimated.
The second modeling approach is statistical in nature and based on empirical
measurements. These models make simple assumptions with little or no transformations of the
independent input variables to the model. The statistical models use, as input parameters, data
obtained in large field studies through both measurement and estimation (questionnaires). The
statistical models are typically simple linear models where the independent variables are used as
they are recorded from the questionnaires to explain variations in the concentrations of the air
contaminants measured.
The statistical approach has not been widely used. In a study of 96 homes in New York
State (Leaderer and Hammond, 1991), measured weekly levels of RSP and vapor phase nicotine
were compared to the number of cigarettes reported smoked (obtained by questionnaire), house
3-20
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
volume, and measured infiltration rates. Respondent-reported number of cigarettes predicted
both residential RSP and vapor phase nicotine-associated ETS levels very well (Figure 3-10). The
inclusion of house volume and infiltration rate as independent variables in the models occasionally
proved to be significant at the 0.05 level, but it explained only a small amount of the variation in
the model between measured RSP and nicotine levels.
3.5. SUMMARY
Environmental tobacco smoke is a major source of indoor air contaminants. The
ubiquitous nature of ETS in indoor environments indicates that some unintentional inhalation of
ETS by nonsmokers is unavoidable. Environmental tobacco smoke is a dynamic complex mixture
of more than 4,000 chemicals found in both vapor and particle phases. Many of these chemicals
are known toxins and carcinogenic agents. Nonsmoker exposure to ETS-related toxic and
carcinogenic substances will occur in indoor spaces where there is smoking occupancy. Many of
the ETS compounds are emitted in higher concentrations in sidestream smoke than mainstream
smoke. Sidestream emissions, however, are quickly diluted into the environment where ETS
exposures take place. Individuals close to smokers (e.g., an infant in a smoking parent's arms)
may be directly exposed to the plume of sidestream smoke or exhaled mainstream smoke and thus
be more heavily exposed.
Given the complex nature of ETS, it is necessary to identify marker or proxy compounds
that, when measured, will allow for the quantification of exposure to ETS. Vapor phase nicotine
and respirable suspended particle mass are two such markers that are suitable indicators of
exposure to ETS. Nicotine and RSP have been measured in personal monitoring studies and in
studies of a variety of indoor environments. The results of these studies clearly demonstrate that
reported exposure to ETS, even under the conditions of low frequency, duration, and magnitude,
will result in RSP and nicotine values above background levels. These studies indicate that ETS
exposures take place in a wide range of microenvironments (e.g., residences, workplaces,
restaurants, airplanes) where smoking occurs. Indoor levels of RSP and vapor phase nicotine have
been shown to vary in a linear fashion with reported tobacco consumption. Nicotine levels
measured indoors have ranged from less than 1 /tg/m3 to more than 500 /tg/m3, while RSP-
associated ETS levels have ranged from less than 5 /ig/m3 to more than 1 mg/m3. Nicotine
exposures greater than 100 ng/m3 are exceedingly rare; most environments measured have ranged
from less than 0.3 pg/m3 (smoke free) to 30 /tg/m3; bars and smoking sections of planes may reach
3-21
05/15./92
-------
DRAFT—DO NOT QUOTE OR CITE
50 to 75 pg/m3. Thus, the normal range of ETS exposures is approximately hundredfold: 0.3 to 30
/zg/m3 for nicotine and from 5 to 500 /ig/m3 for RSP.
In residences with smoking occupancy, average daily or weekly nicotine values might
typically range from less than 1 to 10 /tg/m3, varying principally as a function of number of
smokers or number of cigarettes smoked. Average daily or weekly residential concentrations of
ETS-associated RSP could be expected to increase from 18 to 95 /*g/m3 (added to background
levels) in homes where smoking occurs. Like nicotine, ETS-associated RSP increases with
increased smoking. Average levels of nicotine and RSP in offices with smoking occupancy are
roughly comparable to those in homes.
Cotinine in saliva, blood, and urine is the most widely accepted biomarker of ETS
exposure. It is not directly related to the air exposure to nicotine because of substantial
differences in the time course of exposure uptake, metabolism, and elimination of nicotine in
exposed individuals. In addition, the ratio of cotinine in smokers versus nonsmokers may not be
the same as the ratio for the active agents in ETS responsible for the adverse effects. Cotinine,
however, is an excellent indicator that ETS exposure has taken place and may be a good indicator
of dose. The available data indicate that as many as 80% of nonsmokers are exposed to ETS, that
there is variability in average exposure levels among different cities, and that cotinine levels vary
as a function of passive smoke exposure. Comparisons of cotinine levels in smokers and ETS-
exposed nonsmokers have led to estimates that nonsmokers receive from 0.1% to 7% of the dose of
nicotine of an average smoker. The dose of active agents may be quite different (e.g., nonsmokers
may receive 10% to 20% of the dose of 4-ABP that smokers inhale). These estimates, however,
are based on a number of assumptions that may not hold.
Questionnaires are the most commonly used method to assess exposure to ETS in both
retrospective studies of acute and chronic effects and in prospective studies. They have been used
not only to establish simple categories of ETS exposure, but also to obtain information on activity
patterns of exposed individuals and to obtain information on environmental factors affecting
concentrations in different indoor environments. No standardized or validated questionnaires
have yet been developed for assessing ETS exposure. A number of studies have compared
questionnaire responses to measured air concentrations of nicotine and RSP and cotinine levels.
These efforts have indicated that a significant percent of individuals reporting no exposure had
actually been exposed. In general, questionnaires had moderate success in assessing exposure
status and level of exposure. Misclassification errors must be addressed in using questionnaires to
assess ETS exposure.
3-22
05/15/92
-------
DRAFT-DO NOT QUOTE OR CITE
Physical/chemical and statistical models provide a viable means for predicting
concentrations of ETS-related contaminants in situations when it is impractical to obtain direct
measurements. The utility of these models will be enhanced when they are better validated.
Environmental tobacco smoke represents an important source of indoor air contaminants.
The available data suggest that exposure to ETS is widespread with a wide range of exposure
levels.
3-23
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 3-1. Distribution of constituents in fresh, undiluted mainstream smoke and diluted
sidestream smoke from nonfilter cigarettes1
Constituent
Vapor phase:2
Carbon monoxide
Carbon dioxide
Carbonyl sulfide
Benzene3
Toluene
Formaldehyde4
Acrolein
Acetone
Pyridine
3-Methylpyridine
3-Vinylpyridine
Hydrogen cyanide
Hydrazine4
Ammonia
Methylamine
Dimethylamine
Nitrogen oxides
Af-Nitrosodimethylamine4
/V-Nitrosodiethylamine4
#-Nitrosopyrrolidine4
Formic acid
Acetic acid
Methyl chloride
Amount in MS
10-23 mg
20-40 mg
12-42 jtg
12-48 /rg
100-200 ng
70-100 ng
60-100 /*g
100-250 fig
16-40 ng
12-36 ng
11-30/ig
400-500 ng
32 ng
50-130 ng
11. 5-28.7 fig
7.8-10 Atg
100-600 fig
10-40 ng
ND-25 ng
6-30 ng
210-490 fig
330-8 10 ng
150-600 /xg
' " Range in SS/MS
2.5-4.7
8-11
0.03-0.13
5-10
5.6-8.3
0.1--50
8-15
2-5
6.5-20
3-13
20-40
0.1-0.25
3
3.7-5.1
4.2-6.4
3.7-5.1
4-10
20-100
<40
6-30
1.4-1.6
1.9-3.6
1.7-3.3
(continued on the following page)
3-24
05/15/92
-------
Table 3-1. (continued)
DRAFT-DO NOT QUOTE OR CITE
Constituent
Particulate phase:2
Particulate matter8
Nicotine
Anatabine
Phenol
Catechol
Hydroquinone
Aniline4
2-Toluidine
2-Naphthylamine3
4-Aminobiphenyl3
Benz[a]anthracene5
Benzo[a]pyrene4
Cholesterol
7-Butyrolactone5
Quinoline
Harman6
N '-Nitrosonornicotine5
NNK7
AT-Nitrosodiethanolamine4
Cadmium4
Nickel3
Zinc
Polonium-2103
Benzoic acid
Lactic acid
Glycolic acid
Succinic acid
Amount in MS
15-40 mg
1-2.5 mg
2-20 /*g
60-140 #tg
100-360 ng
1 10-300 /ig
360 ng
160 ng
1.7 ng
4.6 ng
20-70 ng
20-40 ng
22 Mg
10-22 ng
0.5-2 pg
1.7-3.1 /ig
200-3,000 ng
100-1, 000 ng
20-70 ng
HOng
20-80 ng
60 ng
0.04-0.1 pCi
14-28 /ig
63-174 /ig
37-126jtg
110-140 jig
Range in SS/MS
1.3-1.9
2.6-3.3
<0.1-0.5
1.6-3.0
0.6-0.9
0.7-0.9
30
19
30
31
2-4
2.5-3.5
0.9
3.6-5.0
3-11
0.7-1.7
0.5-3
1-4
1.2
7.2-
13-30
6.7
1.0-4.0
0.67-0.95
0.5-0.7
0.6-0.95
0.43-0.62
3-25
(continued on the following page)
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 3-1. (continued)
Data from Elliot and Rowe (1975); Schmeltz et al. (1979); Hoffman et al. (1983); Klus and
Kuhn (1982); Sakuma et al. (1983, 1984a,b); Hiller et al. (1982). Diluted SS is collected with
airflow of 25 ml/s, which is passed over the burning cone; as presented in the NRC report on
passive smoking (1986).
Separation into vapor and particulate phases reflects conditions prevailing in MS and does not
necessarily imply same separation in SS.
Known human carcinogen, according to U.S. EPA or IARC.
Probable human carcinogen, according to U.S. EPA or IARC.
Animal carcinogen (Vaino et al., 1985).
l-methyl-9#-pyrido[3,4-6]-indole.
NNK = 4-(^-methyl-AT-nitrosamino)-1 -(3-pyridyl)-1 -butanone.
Contains di- and polycyclic aromatic hydrocarbons, some of which are known animal
carcinogens.
3-26
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
it
•SI
0-3
•o
CD en
« £
O cS
"3 ts
u-S
vq
o
01
-H
o c/•>
— HTT>ntS
-* cs
cl
'is
en
T)
sS
£
o
>
c
I
§
g
a
oo
1
ea
I
w
—
2
et
H
DO
.s
1
a.
• i-<
<
1
i
a
1
§
O CO
ffitf
d
o
o.
en
§
"3 ,_f
CD
l-c
CD
CD O
-
CD
3 °
§2 aS
3-27
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
o
S
1
E
•2J
ctt
c3
S
CD
•^f
"I
.«
•4-*
S
O.
O
CO
O
O.
S
•a
I
CD
Q.
00
a
'C
I
§
CO
0>
1
CO
v
I
w
i
5
a
oo
oo
on
-,
05
a!
g 2 2
to
S
a
X
a
x
jo
S
I
1
-H o
s
vo
oo
VO
oo
IN
00
rr
O\
1
a IH
CS fi)
>-) o
O o
_g
e
•^ s^
g aj ca
o>
oa
t
I
o
3
ca
8
S
00
t>0 «
o -i
cu ca
O. 4-.
00 co
00
oo «
a — •
a) ca
a ^
00 D
a>
X 00
£2
C8
O\
^«
fi
o o
x o\
0 ON
&0 -H
CD
« . S
d> 'O co
f"| fl> O
CO *^
S 2
a.S
S a s
•o "O -a
CD CD OJ
•M *! **
oJ eg d
3-28
o
a
t-t
o
i H
BJ 43 O Qi
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 3-4. Weekly average concentrations of each measure of exposure by parental smoking
status in the cross-sectional study, Minnesota, 1989
Smoking Status
Non- Light Father / Mother Both
smokers smokers only only parents
Number of subjects
Total cigarettes (no./week)
Activity room nicotine (jig/m3)
Bedroom nicotine (jug/m3)
23 4 86 7
0.9 28.8 68.6 58.8 227.6
0.15 0.32 2.45 5.50 12.11
0.30 1.21 2.66 5.32
3-29
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
1
s
C/3
53
0)
J4
I
a
o
-
JS
2
Ri
O.
•
o
tn
n
•1
Rt
a
o.
in
CO
to
SJ
^rtj
O ****»•
S ^*"
W**« c^
R
1 ^
0
j
Exposure
__
S... J
52 ""
0
i
I
s
J3
en
O
£X
£
S
.— ^
7 *c^
o ^
ta
ui
,v^
1
cn
C
&
>
rt
i
^-
w> 73
"r> '^
<*jj «^
In
«M 0>
O tia*
* I
CO
E*
73
1
.«
a
0>
a
1
oo
^3" rn ^D *O ^D ^3
—I t~- W-> • t-. -H O\
*o t^ n oo r-; r-; oo vo
c> c>
* * { J
00 10 -H O V> V-
O >X CN rN cN r~
o vq oj rn r>j -^
o' o o o o d
O oq o\ oq t-; vo
»-< en en o o ^H
^S. J
DO 00
K rt C 0i
g | | i ¥ 1 1 .1
'§ "s. "i i 's a "i i
.y a a d '£ a d d
z ~ .« --, y .« .« .-
O
O
o
V
(U
o
D.
I
?
g
1
o
g
o
u
O
Q.
X
0)
a
§
g
a)
u
§
r-
00
.a
O
C/3
3-30
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
IU
o
>-
z
111
o
01
a.
in
P
CUMU
80 -
60 -
40 -
20 -
f ..•''"*
) •** j
_ 1 .• •
; /
! • •
i : •'
- i t f
* ./
- / / /"
/./^ i
-x
• **
AntbMfll
a i
NQft-&nak»-£*p««d
««.*_« Personal,
I 1 ! ._! 1
40 80 120 160
RESPIRABLE PARTICULATE CONCEHTHATION
200
240
Figure 3-1. Cumulative frequency distribution of respirable suspended particle (RSP) mass
concentrations from central site ambient and personal monitoring of smoke-exposed and
nonsmoke-exposed individuals. Reprinted from Spengler et al., 1985.
3-31
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
•
m
to
•
§___ to
^^ tn
•
s •
10 •
s-
• •
'
N^^_ '
I • 1
«- — __
c •
i i i i • i i i
69 4 10 69 T 10 1121 4
l"H5161"t "t 16 if 9 12 17 "> 12 9 18 8
,g_ —-Airplanes Train Car- -Bus- -Sub.- Sub
e -Smoking- Nonsmoking
"o c
CM
— H
O ts O **
10 d-
a " \
Md>
«* P* J3 tJ
M U, £C°
1 .
~" •"• as
:-i
-S s
; i
CM 01 V>
n a, S
CGI C*.
OB -*
• * S
2 * °
o
! « S
1 *^ *«i
$ H £
1
W
3-2. Mean, standard deviation, maximum, and minimum nicotine values measured in different indoor environments with
g occupancy. References from which observations are reported and the number of environments monitored are also given.
3-32
05/15/92
-------
DRAFT--DO NOT QUOTE OR CITE
REFERENCES FOR FIGURES 3-2 AND 3-4
Figure 3-2
1 NRC, 1986
2 U.S. DHHS, 1986
3 Brunnemann et al., 1976
4 Wynder and Hoffmann, 1967
5 Dube and Green, 1982
6 Adams, 1987
7 Guerin, 1987
8 Higgins, 1987
9 Leaderer, 1990
10 NRC, 1981
11 Hammond, 1987
12 Eatough et al., 1986
13 Lofroth et al., 1989
14 Eatough et al., 1989
15 Leaderer and Hammond, 1991
16 Benner et al., 1989
17 Hammond et al., 1988
18 Rickert et al., 1984
Figure 3-4
19 Eudy et al., 1985
20 Eatough et al., 1990
21 Hammond, 1987
22 Hammond, 1987
23 Mumford et al., 1989
24 Hammond et al., 1989
25 Marbury, 1990
26 Eatough, 1989
27 Koutrakis et al., 1989
28 DHEW, 1977
29 Muramatsu et al., 1984
30 Oldaker and Conrad, 1987
31 Ogden et al., 1989
32 Leaderer, 1988
33 Repace and Lowrey, 1980
34 Ishizu, 1980
35 Schenker et al., 1990
36 Coghlin et al., 1989
37 Spengler et al., 1981
38 Spengler et al., 1985
39 Sexton et al., 1984
40 Weber, 1980
3-33
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
200
^ 150
r>
E
« 100
VI
o
O_
-------
a- -
to O
s
DRAFT—DO NOT QUOTE OR CITE
a
§H
jgl 1
8
1 JT '»
** 2 «
S "5
<3 S •
*»i
S"S e
ift
o
a
a
e
KM
1 • 1
KM
e
. a
»••!••« >» W
O
Sirt o « ® iG
^ £1 2 ^* ***
3-35
1 CM
r^.
ol
31 LD
& °]
31 "5
"
-------
DRAFT—DO NOT QUOTE OR CITE
24
20
16
I
£ 12
o
o
8
\ Max. Value
Range of
Average
Values
'Mln. Value
70
83
25
Residential Offices Restaurants Transportation
Occupied Spaces with Smoking
Others
Figure 3-5. Range of average nicotine concentrations and range of maximum and minimum
values measured by different indoor environments for smoking occupancy from studies shown in
Figure 3-2. Only those studies with sampling times of 4 hours or greater are included in the
residential and office indoor environment summaries.
3-36
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
t
13
I
120
100
o
I 80
60
40
20
560
0 Max. value
Range of
Average
Values
' Mm. value
1370
986
1140
850
r
I
Residential Office Restaurants Transportation
Occupied Spaces with Smoking
Other
Figure 3-6. Range of average respirable suspended particle (RSP) mass concentrations and range
of maximum and minimum values measured by different indoor environments for smoking
occupancy from studies shown in Figure 3-4. RSP values represent the contribution to
background levels without smoking. Background levels were determined by subtracting reported
indoor concentrations without smoking. Only those studies with sampling times of 4 hours or
greater are included in the residential and office indoor environment summaries.
3-37
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
o
100
80
60
I «
x
• All data (N«96) I.I (2.0)
a Nicotine «OQ (N«49) 0
* N!cotlnt«CO (N«47) JUT (2.43)
234567
Vapor Phase Nicotine
8
10
Figure 3-7. Cumulative frequency distribution and arithmetic means of vapor-phase nicotine
levels, measured over a 1-week period in the main living area in residences in Onondaga and
Suffolk Counties in New York State between January and April 1986. Reprinted from Leaderer
and Hammond, 1991.
3-38
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
lOOr
• All date (N«96) 29.4 (25.9)
a NIcotine*0.0 (N«49) 15.2(7.4)
a. N!cotlne»0.0(N«47) 44.1 (29.9)
-1
20 40 60 80 100 120
Respirable Particle Mass
140 160
Figure 3-8. Range of average nicotine concentrations and range of maximum and minimum
values measured by different indoor environments from studies shown in Figure 3-2. Only those
studies with sampling times of 4 hours or greater are included in the residential and office indoor
environment summaries.
3-39
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
tu
^
H" —
o: ^
Ul ...
SI
c: Z
_ m
5 o
ISO
no
ICO
so
so
7O
6O
SO
•SO
2O
2O
10 -
* « tndoor, no smokers,
i 3 Indocr, I smoker .-.
• ' Indeor, >l srncker / V
No*. OiJcn.fi
1976
1977
i'ul. J«g. .Sep. C
-------
DRAFT—DO NOT QUOTE OR CITE
10
o
u
Nicotint= 0.065+0.028 T -
N = 96
rz=0.67 -'-
a,.
0 50 IOO ISO 200 230 300
Total numbtr of cigarettes
zoo
IOO
a
E
a.
50
RSP-I7.7+OJ22T
N-96
rZ-0.55
I ...
SO IOO ISO 200 Z5O
Total number of cigarettes (T)
3OO
Figure 3-10 a. and b. Weeklong nicotine and respirable suspended particle (RSP) mass
concentrations, measured in the main living area of 96 residences versus the number of
questionnaire-reported cigarettes smoked during the air-sampling period. Numbers 1-9 refer to
the number of observations at the same concentrations. Closed circles indicate that cigar or pipe
smoking was reported in the houses, with each cigar or pipe smoked set equal to a cigarette. Data
from residences in Onondaga and Suffolk counties in New York State between January and April
1986.
3-41
05/15/92
-------
1
IfiOy
140- -
120--
100- •
ao--
60--
40--
20- .
• Q--
DRAFT--DO NOT QUOTE OR CITE
Neonate
Under 18 mo.
Over 18 mo.
Adult
Age Group
Figure 3-11. Average continue t1/2 by age groups. Reprinted from Collier et al., 1990.
3-42 05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
40
35
30
— 25
111
§
O
=3
10
A Y~ •
£T "* - I r-
NONE A UTTLE SOUE A LOT
AT ALL.
Figure 3-12. Distribution of individual concentrations of urinary cotinine by degree of self-
reported exposure to ETS. Horizontal bars indicate median values. Reprinted from Jarvis and
Russellls 1985.
3-43
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
03
16
15
14
13
12
li
10
9
8
^
6
5
4
3
2
1
n
—
-
-
-
-
-
-
-
-
1-6.22
1-7.75 .
x»9.75
x-12.3
J-
none 1 or 2 3 to 5 6 or more
n=162 n=2O8 n=152 n=14f
Number of Exposures in the Past 4 Days
Figure 3-13. Urinary cotinine concentrations by number of reported exposures to tobacco smoke
in the past 4 days among 663 nonsmokers, Buffalo, New York, 1986. Reprinted from Cummings
et al., 1990.
3-44
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
a) Sampling categories of b) Self-reported exposure
exposure
I all
I cot > 0
exposing
AT won*:
AT n»e:
-|20
15
10
c) Self-reported
extreme patterns
of exposure
-i 20 .T20
16
10
J. o
15
o
2
10
to ewosuie
EXPOSURE CWM AU.
SOWTBO SOURCES
ALL
0 N".
i-CAN
271
5.0
26S
t.l
J<«5
10.6
269
11.0
<*29
5.4
172
7.1
335
1C.I
117
12.i
209
3t
II.S
Figure 3-14. Average cotinine/creatinine levels for subgroups of nonsmoking women defined by
sampling categories of exposure or by self-reporting exposure to ETS from different sources
during the 4 days preceding collection of the urine sample. Reprinted from Riboli et al., 1990.
3-45
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Total RSP Emitted (mg)
Ol -J O
o o o
o o o
r\i in tn -«j O
o o o o o
O O O O Q
o o o o o
Total Cigarettes Smoked per Hour
Figure 3-15. Diagram for calculating the respirable suspended particle (RSP) mass from ETS
emitted into any occupied space as a function of the smoking rate and removal rate (N). The
removal rate is equal to the sum of the ventilation or infiltration rate (nv) and removal rate by
surfaces (N) times the mixing factor m. The calculated ETS-related RSP mass determined from
this figure serves as an input to Figure 3-16 to determine the ETS-related RSP mass concentration
in any space in ug/m3. Smoking (diagonal lines) are given as cigarettes smoked per hour. Mixing
is determined as a fraction and nv and n8 are in air changes per hour (ach). All three parameters
have to be estimated or measured. Calculations were made using the equilibrium form of the
mass-balance equation and assume a fixed emission rate of 26 mg/m3 of RSP.
Shaded area shows the range of RSP emissions that could be expected for a residence with one
smoker smoking at a rate of either 1 or 2 cigarettes per hour for the range of mixing, ventilation,
and removal rates occurring in residences under steady-state conditions. Reprinted from NRC
(1986).
3-46
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
5 10 20 50 100 200 500 1000 3000 (0000
Total RSP Emitted (mg)
Figure 3-16. Diagram to calculate the ETS-associated respirable suspended particle (RSP) mass
concentration in a space as a function of total mass of ETS-generated RSP emitted (determined
from Figure 3-15) and the volume of a space (diagonal lines). The concentrations shown assume a
background level in the space of zero. The particle concentrations shown are estimates during
smoking occupancy. The dashed horizontal lines (A, B, C, and D) refer to National Ambient Air
Quality Standards (health-related) for total suspended particulates established by the U.S.
Environmental Protection Agency. A is the annual geometric mean. B is the 24-hour value not to
be exceeded more than once a year. C is the 24-hour air pollution emergency level. D is the 24-
hour significant harm level. Shaded area shows the range of concentrations expected (from
Figure 3-15) for a range of typical volumes of U.S. residences and rooms in these residences.
Reprinted from NRC (1986).
3-47
05/15/92
-------
-------
DRAFT—DO NOT QUOTE OR CITE
4. HAZARD IDENTIFICATION I: LUNG CANCER IN ACTIVE SMOKERS,
LONG-TERM ANIMAL BIOASSAYS, AND GENOTOXICITY STUDIES
4.1. INTRODUCTION
Numerous epidemiologic studies have conclusively established that the tobacco smoke
inhaled from active smoking is a human lung carcinogen (U.S. DHHS, 1982; IARC, 1986). A clear
dose-response relationship exists between lung cancer and amount of exposure, without any
evidence of a threshold level. It is, therefore, reasonable to theorize that exposure to
environmental tobacco smoke (ETS) might also increase the risk of lung cancer in both smokers
and nonsmokers. As documented in the previous chapter, the chemical compositions of
mainstream smoke (MS) and ETS are qualitatively similar, and both contain a number of known
and suspected human carcinogens. In addition, both MS and ETS have been shown to be
carcinogens in animal bioassays (Wynder and Hoffman, 1967; Grimmer et al., 1988) and
genotoxins in in vitro systems (IARC, 1986). Furthermore, as the previous chapter also describes,
exposure assessments of indoor air and measurements of nicotine levels in nonsmokers confirm
that passive smokers are exposed to and absorb appreciable amounts of ETS that might result in
notable lung cancer risk.
This chapter reviews the major evidence for the lung carcinogenicity of tobacco smoke
derived from human studies of active smoking and the key supporting evidence from animal
bioassays and in vitro experiments. The evidence from the few animal and mutagenicity studies
pertaining specifically to ETS is also presented. The majority of this information has already been
well documented by the U.S. Department of Health and Human Services (U.S. DHHS) (1982) and
the International Agency for Research on Cancer (IARC) (1986). The current discussion mainly
extracts and summarizes some of the important issues and principal studies described in those
excellent reports.
In view of the abundant and consistent human evidence establishing the carcinogenic
potential of active smoking to the lung, the bulk of this chapter focuses on the human data.
Although EPA's carcinogen risk assessment guidelines (U.S. EPA, 1986a) suggest an extensive,
review of all evidence pertaining to carcinogenicity, we believe that the wealth of human cancer
studies on both MS and ETS provide the most appropriate database from which to evaluate the
lung cancer potential of ETS. Thus, the animal evidence and genotoxicity results are given only
limited attention here. Similarly, a discussion of the mutagenicity data for individual smoke
components would be superfluous in the context of the overwhelming evidence from other, more
4-1
5/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
pertinent sources and is not included. Extensive reviews of these data can be found in the U.S.
DHHS (1982) and IARC (1986) publications. Claxton et al. (1989) provide an assessment of the
genotoxicity of various ETS constituents.
4.2. LUNG CANCER IN ACTIVE SMOKERS
Studies of active smoking in human populations from many countries provide direct and
incontrovertible evidence for a dose-related, causal association between cigarette smoking and
lung cancer. This evidence includes time trends in lung cancer mortality rates associated with
increasing cigarette consumption, high relative risks for lung cancer mortality in smokers of both
sexes observed consistently in numerous independent retrospective and prospective studies, and
dose-response relationships demonstrated with respect to smoking intensity and duration and for
all four major histological types of lung cancer.
4.2.1. Time Trends
While the overall cancer death rate in the United States has been fairly stable since 1950,
the lung cancer death rate has increased drastically for both males and females (Figures 4-1 and
4-2). Age-adjusted lung cancer mortality rates in men have increased from 11 per 100,000 in
1940 to 73 per 100,000 in 1982, leveling slightly to 74 per 100,000 in 1987 (Garfinkel and
Silverberg, 1991). In women, lung cancer mortality rates have risen from 6 per 100,000 in the
early 1960s to 28 per 100,000 in 1987 (Garfinkel and Silverberg, 1991).
The striking time trends and sex differences seen in lung cancer mortality rates correlate
with historical smoking patterns. Increases in lung cancer death rates parallel increases in
cigarette consumption with a roughly 20-year lag time, accounting for the latency period for the
development of smoking-induced lung cancer. Males started smoking cigarettes in large numbers
during the years around World War I, whereas females did not begin smoking in appreciable
numbers until World War II. Cigarette consumption per capita (based on the total population age
18 and older) in the United States rose from 1,085 in 1925 to a high of 4,148 in 1973. In the past
two decades, cigarette consumption has decreased to 2,888 in 1989 (Garfinkel and Silverberg,
1991). This decline correlates with the leveling off of lung cancer mortality rates in recent years.
4-2
5/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
4.2.2. Dose-Response Relationships
More than 50 independent retrospective studies have consistently found a dose-related
association between smoking and lung cancer (U.S. DHHS, 1982). Eight major prospective studies
from five countries corroborate this association:
• American Cancer Society (ACS) Nine-State Study (white males)
(Hammond and Horn, 1958a,b)
• Canadian War Veterans Study
(Best et al., 1961; Lossing et al., 1966)
• British Doctors Study
(Doll and Hill, 1964a,b; Doll and Peto, 1976; Doll et al., 1980)
• American Cancer Society (ACS) 25-State Study
(Hammond, 1966; Hammond and Seidman, 1980)
• U.S. Veterans Study
(Kahn, 1966; Rogot and Murray, 1980)
• California Labor Union Study
(Weir and Dunn, 1970)
• Swedish Study (sample of census population)
(Cederlof et al., 1975)
• Japanese Study (total population of 29 health districts)
(Hirayama, 1967, 1975a,b, 1977, 1978, 1982, 1985)
Details of the designs of these studies are summarized in Table 4-1. These eight studies
together represent more than 17 million person-years and more than 330,000 deaths. Lung cancer
mortality ratios from the prospective studies are presented in Table 4-2. Combining the data from
the prospective studies results in a lung cancer mortality ratio of about 10 for male cigarette
smokers compared to nonsmokers.
This strong association between smoking and lung cancer is further enhanced by very
strong and consistent dose-response relationships. A gradient of increasing risk for lung cancer
mortality with increasing numbers of cigarettes smoked per day was established in every one of
the prospective studies (Table 4-3). Lung cancer mortality ratios for male smokers who smoked
more than 20 cigarettes daily were generally 15 to 25 times greater than those for nonsmokers.
Marked increases in lung cancer mortality ratios were also seen in all the lowest dose categories.
Males who smoked fewer than 10 cigarettes per day had lung cancer mortality ratios 3 to 10 times
4-3
5/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
greater than those for nonsmokers. There is no evidence of a threshold level for the development
of smoking-induced lung cancer in any of the studies.
Dose-response relationships with respect to the duration of smoking have also been well
established. From the British male physicians study, Peto and Doll (1984) calculated that the
excess annual incidence rates of lung cancer after 45, 30, and 15 years of cigarette smoking were
in the approximate ratio of 100:20:1 to each other. The California and Swedish studies also
demonstrated an increasing risk of lung cancer in men with longer smoking duration (Table 4-4).
Four of the prospective studies examined lung cancer mortality in males by age at
initiation of smoking and found increasing risk with younger age (Table 4-5). Some of the studies
also investigated smoking cessation in men and observed a decrease in lung cancer risk with
increasing number of years since quitting smoking (Table 4-6). The Cancer Prevention Study II, a
study of 1,200,000 people in all 50 states, reveals a similar trend for women who quit smoking
(Figure 4-3). The occurrence of higher lung cancer mortality ratios in the groups with only a few
years since cessation as compared to current smokers (Table 4-6 and Figure 4-3) is attributable to
the inclusion of recent ex-smokers who were forced to stop smoking because they already had
smoking-related symptoms or illness (U.S. DHHS, 1990). The demonstration of increasing lung
cancer risks the younger the age of smoking commencement and decreasing risks with time since
smoking cessation establishes the initiation and promotion capabilities of tobacco smoke.
Additional dose-response relationships have been derived from consideration of the types
of tobacco products used. Pipe and cigar smokers, who inhale less deeply than cigarette smokers,
have lower risks of lung cancer than cigarette smokers (Table 4-7). Furthermore, the American
Cancer Society 25-State Study found decreased risks for lung cancer in males and females who
smoked cigarettes with lower tar and nicotine content compared to those who smoked cigarettes
with higher tar and nicotine content (Table 4-8), although these decreased risks are still
substantially higher than the risk to nonsmokers. Similarly, it has been established that smokers of
filtered cigarettes have relatively lower lung cancer risks than smokers of nonfiltered cigarettes
(Table 4-9). Filters reduce the amount of tars, and hence a portion of the carcinogenic agents, in
the MS inhaled by the smoker. Passive smokers, however, do not share in any benefit derived
from cigarette filters (see Chapter 3) and may, in fact, be exposed to more ETS if smokers of
filtered cigarettes smoke a greater number of cigarettes to compensate for any reduction in
nicotine uptake resulting from the filters (U.S. DHHS, 1986).
4-4
5/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
4.2.3. Histological Types of Lung Cancer and Associations With Smoking
A number of epidemiologic studies have also examined the association between various
histological types of lung cancer and smoking. The results of some of these investigations are
summarized in Table 4-10. Problems in interpreting the results of such studies include
differences in the nomenclature, criteria, and verification of tumor classification; inadequacy of
some specimens, and the small size of many of the patient groups, resulting in unstable risk
estimates, particularly in women. There are four major histological types of lung cancer:
squamous-cell carcinoma, small-cell carcinoma, adenocarcinoma, and large-cell undifferentiated
carcinoma. Sometimes two broad categories—Kreyberg Group I, containing squamous-cell and
small-cell carcinomas, and Kreyberg Group II, containing all other epithelial lung cancers,
including adenocarcinomas and large-cell undifferentiated carcinomas—are used for classification.
The majority of the studies demonstrate an Increase in the risk for lung cancer with increasing
amount smoked for all four major histological groups in both males and females. The slope of the
gradient for adenocarcinomas, however, is shallower than the slopes for the other types.
4.2.4. Proportion of Risk Attributable to Active Smoking
Table 4-11 presents data on the proportion of lung cancer deaths attributable to smoking
in various countries. Differences by sex and between countries largely correlate with differences
in the proportion of smokers within these populations and the duration and intensity of cigarette
usage. In the early 1960s, 50% of U.S. men and 30% of U.S. women smoked, although these
proportions have been declining in recent years (Garfinkel and Silverberg, 1991).
In the United States, deaths from lung cancer currently represent one quarter of all cancer
deaths. The American Cancer Society predicts there will be 143,000 lung cancer deaths in 1991
(Garfinkel and Silverberg, 1991). Over 85% of this lung cancer mortality is estimated to be
attributable to tobacco smoking. In other words, the overwhelming majority of lung cancer
deaths, which are a significant portion of all cancer deaths, result from smoking. The strong
association between smoking and lung cancer and the dose-response relationships, with effects
observable at low doses and no evidence of a threshold, make it highly plausible that passive
smoking also causes lung cancer in humans.
4.3. LIFETIME ANIMAL STUDIES
The human evidence for the carcinogenicity of tobacco smoke is corroborated in
experimental animal bioassays. The main animal evidence is obtained from inhalation studies in
4-5
5/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
the hamster, intrapulmonary implantations in the rat, and skin painting in the mouse. There are
no lifetime animal inhalation studies of ETS; however, the carcinogenicity of ETS condensates has
been demonstrated in intrapulmonary implantations and skin painting experiments.
Negative responses in short term animal studies (e.g., 60 to 90 days) are not reliable
indicators of the carcinogenic potential of a compound because of the long latency period for
cancer development. Long-term animal studies at or near the maximum tolerated dose level are
used to ensure an adequate power for the detection of carcinogenic activity (U.S. EPA, 1986a).
4.3.1. Inhalation Studies
Although evidence of the carcinogenicity of cigarette smoke originated in humans,
attempts were made to develop an inhalation model for smoking in experimental animals in order
to study the carcinogenicity of various tobacco products. Such inhalation studies are difficult to
conduct, however, because laboratory animals are reluctant to inhale cigarette smoke and will
adopt shallow breathing patterns in response to aerosols and irritants. Furthermore, rodents are
obligatory nose-breathers, and the anatomy and physiology of the respiratory tract and the
biochemistry of the lung differ between rodents and humans. Because of these distinctions,
laboratory animals and humans are likely to have different deposition and exposure patterns for
the various cigarette smoke components in the respiratory system. For example, rodents have
extensive and complex nasal turbinates where significant particle deposition could occur,
decreasing exposure to the lung.
The Syrian golden hamster has been the most useful animal inhalation model found so far
for studying smoking-induced carcinogenesis. It is more tolerant of tobacco smoke than mice and
rats and is relatively resistant to respiratory infections. The hamster also has a low background
incidence of spontaneous pulmonary tumors and is, in fact, refractory to the induction of lung
cancers by known carcinogenic agents. The inhalation of tobacco smoke by the hamster does,
however, induce carcinomas of the larynx. In one study (Dontenwill et al., 1973), three groups of
80 male and 80 female Syrian golden hamsters were exposed for 10 minutes to air-diluted
cigarette smoke (1:15) once, twice, or three times daily, 5 days per week, for their lifetimes. Pre-
invasive carcinomas of the upper larynx were detected in 11.3%, 30%, and 30.6% of the animals,
respectively, and invasive carcinomas were found in 0.6%, 10.6%, and 6.9%, respectively. No
laryngeal tumors were observed in control animals. In another experiment, exposure for 59 to 80
weeks to a 11% or 22% cigarette smoke aerosol twice daily for 12 minutes resulted in laryngeal
carcinomas in 3 of 44 and 27 of 57 animals, respectively, providing some evidence of a dose-
4-6
5/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
response relationship for the induction of carcinoma of the larynx by cigarette smoke (Bernfeld et
al., 1979). Bernfeld et al. suggest that the greater deposition of tar per unit of surface area in the
larynx compared to the lung may explain the high yield of laryngeal cancers and lack of lung
tumors in this animal model.
4.3.2. Intrapulmonary Implantations of Cigarette Smoke Condensates
Because of the difficulties with inhalation studies of cigarette smoke, some in vivo studies
examine the carcinogenicity of cigarette smoke condensate (CSC) collected from smoking
machines. CSC assays may not, however, reveal all of the carcinogenic activity of actual cigarette ,
smoke, since these condensates lack most of the volatile and semivolatile components of whole
smoke. In lifetime rat studies, intrapulmonary implants of CSC in a lipid vehicle cause a dose-
dependent increase in the incidence of lung carcinomas (Stanton et al., 1972; Dagle et al., 1978).
ETS condensates have also demonstrated carcinogenicity when implanted into rat lungs
(Grimmer et al., 1988). (Actually, only sidestream smoke was examined, but this constitutes
roughly 85% of ETS [Fielding, 1985].) Sidestream smoke (SS) emitted by a smoking machine was
separated into condensate fractions containing the semivolatiles, the polycyclic aromatic
hydrocarbon (PAH)-free particulates and the PAHs with two or three rings, or the PAHs with
four or more rings. These fractions were implanted into female Osborne-Mendel rats, following
the.procedure of Stanton et al. (1972), at a dose level of one cigarette per animal. At the end of
the lifetime study, none of the 35 rats in each of the untreated control, vehicle control, or
semivolatile-rexposed groups had lung carcinomas. In the group exposed to the fraction containing
PAH-free particulates and PAHs with two or three rings, there was 1 lung carcinoma in 35
animals. In the group exposed to the fraction comprising PAHs with four or more rings, there
were 5 lung carcinomas in 35 rats. An additional group that was exposed to a dose of 0.03 mg
benzo[a]pyrene (BaP) per rat exhibited 3 lung carcinomas in 35 animals. The condensate fraction
containing BaP and the other PAHs with four or more rings from the SS generated by a single
cigarette contains about 100 ng of BaP. Assuming a linear, nonsynergistic dose-response
relationship, this would suggest that less than 1% of the total carcinogenicity of that CSC fraction
can be attributed to the BaP present in the smoke.
4.3.3. Mouse Skin Painting of Cigarette Smoke Condensates
In addition, numerous studies have shown that when CSC suspended in acetone is
chronically applied to mouse skin, significant numbers of the mice develop papillomas or
4-7
5/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
carcinomas at the site of application (e.g., Wynder et al., 1957; Davies and Day, 1969). Mouse
skin studies have also demonstrated that CSC has both tumor-initiating and tumor-promoting
capabilities (Hoffman and Wynder, 1971).
One mouse skin painting study examined the carcinogenicity of ETS condensate (Wynder
and Hoffman, 1967). Cigarette tar from SS deposited on the funnel of a smoking machine was
suspended in acetone and administered to mouse skin. Fourteen of thirty mice developed skin
papillomas, and 3 of 30 developed carcinomas. In a parallel assay in the same study, a suspension
of MS condensate applied to deliver a comparable amount of condensate to the skin of 100 mice
yielded benign skin tumors in 24 and malignant tumors in 6 of the mice. This suggests that the
condensate of SS has greater mouse skin tumorigenicity per unit weight than that of MS.
4.4. GENOTOXICITY
Supportive evidence for the carcinogenicity of tobacco smoke is provided by the
demonstration of genotoxicity in numerous short-term assays. Extensive reviews of these studies
can be found in IARC (1986) and DeMarini (1983), and only the highlights are presented here. A
few studies deal with whole smoke, but most examine CSC. Tobacco smoke is genotoxic in
virtually every in vitro system tested, providing overwhelming supportive evidence for its
carcinogenic potential.
In Salmonella typhimurium, for example, Basrur et al. (1978) found that whole smoke and
smoke condensates from various types of tobacco were mutagenic in the presence of a metabolic
activating system. SS (Ong et al., 1984) and extracts of ETS collected from indoor air (Lofroth et
al., 1983; Alfeim and Randahl, 1984; Lewtas et al., 1987; Ling et al., 1987; Lofroth et al., 1988)
also exhibit mutagenic activity in this bacterium. Claxton et al. (1989) found that SS accounted
for approximately 60% of the total S. typhimurium mutagenicity per cigarette—40% from the ETS
particulates and 20% from the ETS semivolatiles. The highly volatile fraction, from either MS or
SS, was not mutagenic.
Similarly, cigarette smoke produced mitotic gene conversion, reverse mutation, and
reciprocal mitotic recombination in fungi (Gairola, 1982). In addition, CSCs induce mutations,
sister chromatid exchanges, and cell transformation in various mammalian cells in culture.
Putnam et al. (1985) demonstrated dose-dependent increases in sister chromatid exchange
frequencies in bone-marrow cells of mice exposed to cigarette smoke for 2 weeks.
4-8
5/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
4.5. SUMMARY AND CONCLUSIONS
The unequivocal causal association between tobacco smoking and lung cancer in humans
with dose-response relationships extending down to the lowest observed exposures, as well as the
corroborative evidence of the carcinogenicity of both MS and ETS provided by animal bioassays
and in vitro studies, clearly establish the plausibility that ETS is also a human lung carcinogen.
Furthermore, biomarker studies verify that passive smoking results in detectable uptake of
tobacco smoke constituents by nonsmokers, affirming that ETS exposure is a public health
concern (Chapter 3).
Active smoking induces squamous-cell carcinomas, small-cell carcinomas, large-cell
carcinomas; and adenocarcinomas in humans, all in a dose-related manner. Lung cancer mortality
rates have increased dramatically over the past 60 years in males and, more recently, in females,
with increasing cigarette consumption. High relative risks for lung cancer, associated with the
number of cigarettes smoked per day, have been demonstrated in countless studies, with no
evidence of a threshold level of exposure. Dose-response relationships have also been established
with respect to duration of smoking. Lung cancer risk increases the younger the age at initiation
of smoking and decreases the longer the time since cessation of smoking. These latter trends,
coupled with the evidence from mouse skin painting studies, show that tobacco smoke has both
tumor-initiating and tumor-promoting capabilities.
Inhalation studies in hamsters confirm that tobacco smoke is carcinogenic to the
respiratory tract. In addition, mouse skin painting experiments and intrapulmonary implantations
in rats have demonstrated the carcinogenicity of condensates from both MS and ETS. Numerous
genotoxicity tests contribute supporting evidence for the carcinogenic potential of cigarette smoke
and smoke condensates. The mutagenicity of ETS and its extracts has also been established. As
discussed in Chapter 3, MS and ETS are qualitatively similar in composition, and both contain a
number of known and suspected human carcinogens.
In fact, these observations alone—the dose-related association between tobacco smoking
and lung cancer in humans, which extends to the lowest reported doses; the chemical similarity
between MS and ETS; and the confirmation of the carcinogenicity of MS and ETS in animal and
in vitro experiments—are sufficient to establish weight-of-evidence for the carcinogenicity of
ETS to humans. According to EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA,
1986a), a Group A (known human) carcinogen designation is used "when there is sufficient
evidence from epidemiologic studies to support a causal association between exposure to the
4-9
5/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
agents and cancer." The Guidelines establish "three criteria (that) must be met before a causal
association can be inferred between exposure and cancer in humans:
1. There is no identified bias that could explain the association.
2. The possibility of confounding has been considered and ruled out as explaining the
association.
3. The association is unlikely to be due to chance."
Given the strong dose-related associations with high relative risks consistently observed across
numerous independent studies from several countries and the biological plausibility provided by
ancillary evidence of the genotoxicity and animal carcinogenicity of MS and by knowledge of the
existence of specific carcinogenic components within MS, confounding, bias, and chance can all
be ruled out as possible explanations for the observed association between active smoking and lung
cancer. Therefore, under the EPA carcinogen classification system, MS would be a Group A
(known human) carcinogen, and, due to the similarity in chemical composition between MS and
ETS and the known human exposure to ETS (Chapter 3), ETS would also be classified as a known
human carcinogen.
In addition, however, there exists a whole body of evidence dealing specifically with
human exposure to ETS. Substantial epidemiologic evidence demonstrates increased risks of lung
cancer in nonsmokers exposed to actual ambient levels of ETS. Therefore, unlike with many
environmental hazards where extrapolation from high-dose animal bioassays or high-level,
generally occupational, human exposures must be used to estimate the human risk at
environmental levels of exposure, the health risk of ETS exposure can be examined directly from
the epidemiologic data. The epidemiologic evidence for the human lung carcinogenicity
associated specifically with ETS is the subject of Chapter 5.
4-10
5/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 4-1. Main characteristics of major cohort studies on the relationship between smoking and
cancer
Study
ACS
Nine-state
study
Canadian
veterans
study
British
doctors
study
ACS
25-state
study
*
U.S.
veterans
study
Sample size;
Initial samples; -
in brackets,
Year of population for
enrollment follow-up
1952 204,547 men
[187,783]
1955-1956 207,397
subjects
(aged 30+)
' [92,000]
1951 34,440 men
(aged 20+)
6,194 women
(aged 20+)
1959-1960 1,078,894 subjects
First follow-up:
440,558 men,
562,671 women
(aged 35-84);
second follow-up:
358,422 men,
483,519 women
1954 293,958 men
(aged 3 1-84)
[248,046]
Source Of
information on Duration of
smoking follow-up
(proportion of and no. of
respondents) deaths
Self-administered 44 months
questionnaire 11,870 deaths
Self-administered 6 years
questionnaire 9,491 deaths
(57% in men;
respondents) 1 ,794 deaths
in women
Self-administered 20 years
questionnaire 10,072 deaths
(69%
respondents)
Self -administered 22 years
questionnaire 1,094 deaths
(60%
respondents)
Self-administered 4.5 + 5 years
questionnaire 26,448 deaths
in men;
16,773 deaths
in women
Self -administered 16 years
questionnaire 107,563 deaths
(85%
respondents)
Completeness
of follow-up
for mortality
98.9%
NA
99.7%
99%
97.4% in
women
97.9% in men
in first
follow-up
Almost 100%
ascertainment
of vital status;
97.6% of death
certificates
retrieved
(continued on the following page)
4-11
5/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 4-1. (continued)
Study
Californian
study
Swedish
study
Japanese
study
Sample size;
Initial samples;
in brackets,
Year of population for
enrollment follow-up
1954-1957 68,153 men
(aged 35-64)
1963 27,342 men,
27,732 women
(aged 18-69)
1965 122,261 men,
142,857 women
(aged 40+)
Source of ;
s information on. Duration of
smoking follow-up
(proportion of and no, of
t espondents) % deaths
Self-administered 5-8 years
questionnaire 4,706 deaths
Self-administered 10 years
questionnaire 5,655 deaths
(89% (2,968 autopsies)
respondents)
Interview 16 years
(95% of 5 1,422 deaths
population in
area)
Completeness
for mortality
NA
NA
Total
NA * not available.
Source: IARC, 1986.
4-12
5/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 4-2. Lung cancer mortality ratios—prospective studies
Population
British
doctors study
Swedish
study
Japanese
study
ACS 25 -State
study
U.S. veterans
Canadian
veterans
ACS 9-state
study
California males
in 9 occupations
Size
34,000 males
6,194 females
27,000 males
28,000 females
122,000 males
143,000 females
358,000 males
483,000 females
290,000 males
78,000 males
188,000 males
68,000 males
Number
of Deaths Moflsmokers
441
27
55
8
940
304
2,018
439
3,126
331
448
368
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
1.00
Cigarette
Smokers
14.0
5.0
7.0
4.5
3.76
2.03
8.53
3.58
11.28
14.2
10.73
7.61
Source: U.S. DHHS, 1982.
4-13
5/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 4-3. Lung cancer mortality ratios' for men and Women, by current number of cigarettes
smoked per day—prospective studies
Population
ACS 25-state
study
British
doctors
study
Swedish study
Japanese study
(all ages)
U.S. veterans
study
ACS 9-state
study
Canadian
veterans
California
males
in nine
occupations
, " ..'.... TV/ten..
Cigarettes * , '
smoked per Day^
Nonsmoker
1-9
10-19
20-39
40+
Nonsmoker
1-14
15-24
25+
Nonsmoker
1-7
8-15
.... l 16+
Nonsmoker
1-19
20-39
40+
Nonsmoker
1-9
10-20
21-39
S: 40
Nonsmoker
1-9
10-20
20+
Nonsffloker
1-9
10-20
20+
Nonsmoker
about i pk
about 1 pk
about H pk
, Wnrnftti
Mortality
'" ratios
1.00
4.62
8.62
14.69
18.71
1.00
7.80
12.70
25.10
1.00
2.30
8.80
1.3.70
•1.00
3.49
5.69
6.45
1.00
3.89
9.63
16.70
23.70
1,00
8.00
10.50
23.40
1.00
9.50
15.80
17.30
1.00
3.72
9.05
9.56
Cigarettes
,stnokfed per day
Nonsmoker
1-9
10-19
20-39
40+
Nonsmoker
1-14
15-24
25+
Nonsmoker
1-7
8-15
16+
Nonsmoker
<20
20-29
Mortality
ratios
1.00
1.30
2.40
4.90
7.50
1.00
1.28
6.41
29.71
1.00
1.80
11.30
.' •
1.00
1.90
4.20
Source: U.S. DHHS, 1982.
•4-14
5/15/92
-------
DRAFT—Dp NOT QUOTE OR £ITE
Table 4-4. Relationship between risk of lung gancer and duration of. smoking in men, based on
available information from cohort studies ,
Reference
Weir & Dunn (1970)
Cederlof et al.
(1975)
Duration of smoking
(years)
1-9
10-19
20+
nonsmokers
1-29
>30
nonsmokers
. Standardized
mortality ratio
(n,o. of observed' '
-deaths)' "- " ^' '
1.13
6.45
8.66
1.0
1.8(5)
7.4(23)
1.0(7)
' Approximate annual
' excess death rate
- (%)
0.002 (0.001)
0.09 (0.05)
0.12 (0.08)
0
0.01 (0.008)
0.1 (0.06)
0
The mortality ratio among nonsmokers was assumed to be 15.6/100,000 per year, as in the
American Cancer Society 25-state study. Figures in parentheses were computed by the, IARG
working group, applying the British doctors' mortality rate among nonsmokers (10.0/100,000
per year).
Source: IARC, 1986.
4-15
5/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 4-5. Lung cancer mortality ratios for males, by age of smoking initiation—prospective
studies
Study
ACS 25-state
study
Japanese
study
U.S. veterans
Swedish
study
Age of
smoking initiation
in years
Nonsmoker
25+
20-24
15-19
Under 15
Nonsmoker
25+
20-24
Under 20
Nonsmoker
25+
20-24
15-19
Under 15
Nonsmoker
19+
17-18
Under 16
Mortality
ratio
1.00
4.08
10.08
19.69
16.77
1.00
2.87
3.85
4.44
1.00
5.20
9,50
14.40
18.70
1.00
6.50
9.80
6.40
Source: U.S. DHHS, 1982
4-16
5/15/92
-------
DRAFT—DtO ,NOT, QUOTE OR CITE
Table 4-6. Relationship between risk of lung cancer and number of years since stopping
smoking, in men, based on available information from cohort studies
Reference
ACS
25-state study
(Hammond, 1966)
Swedish study
(Cederlof et al.,
1975)
British doctors
study (Doll & Peto,
1976)
Rogot & Murray (1980)
No. of years since
stopping smoking
1-19 cig./day
Current smokers
<1
1-4
5-9
10+
Nonsmokers
20+ cig./day
Current smokers
<1
1-4
5-9
10+
Nonsmokers
<10
>10
Nonsmokers
Current smokers
1-4
5-9
10-14
15+
Nonsmokers
Current smokers
<5
5-9
10-14
15-19
20+
Nonsmokers
Mortality' ratio
(no. of observed deaths)
6.5 (80)
7.2 (3)
4.6 (5)
1.0(1)
0.4 (1)
1.0 (32)
13.7 (351)
19.1 (33)
12.0 (33)
7.2 (32)
1.1 (5)
1.0(32)
6.1 (12)
1.1 (3)
1.0 (7)
15.8 (123)
16.0 (15)
5.9 (12)
5.3 (9)
2.0 (7)
1.0(7)
11.3(2609)
18.8 (47)
-7.5 (86)
-5.0 (100)
-5.0(115)
2.1 (123)
1.0 NA
NA = not available.
Source: IARC, 1986.
4-17
5/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 4-7. Relative risks of lung cancer in some large cohort studies among men smoking
cigarettes and other types of tobacco
Study
ACS Nine-state
study1
Canadian
veterans
study
ACS25-state
study1
Swedish study1
Smoking category
Never smoked
Occasionally only
Cigarettes only
Cigars only
Pipes only
Cigarettes + other
Cigars + pipes
Nonsmokers
Cigarettes only
Cigars only
Pipe only
Ex-smokers
Never smoked
Cigarettes only
Cigars only
Pipes only
Cigarettes + other
Cigars + pipes
Nonsmokers
Cigarettes only
Cigarettes + pipe
Pipe only
Cigars only
Ex-smokers
Relative
risk
1.0
1.5
9.9
1.0
3.0
7.6
0.6
1.0
14.9
2.9
4.4
6.1
1.0
9.2
1.9
2.2
7.4
0.9
1.0
7.0
10.9
7.1
9.2
6.1
Death, rate
per 100,000
12.8
19.2
27.2
13.1
38.5
97.7
7.3
12
111
22
27
89
11
, No, of
cases
15
8
249
7
18
148
3
7
325
2
18
18
49
719
23
21
336
11
7
28
27
31
6
12
(continued on the following page)
4-18
5/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 4-7. (continued)
Study
British doctors
study
U.S. veterans
study1
Norwegian
study1
Smoking category
Nonsmokers
Current smokers .
Cigarettes only
Pipes and/or cigars only
Cigarettes + other
Ex-smokers
Nonsmokers
Cigarettes
Cigarettes only
Cigars only
Pipes only
Ex-cigarette smokers
Nonsmokers
Cigarettes
Cigarettes only
Pipes or cigars only
Ex-smokers
Relative
risk
1.0
10.4
14.0
5.8
8.2
4.3
1.0
11.3
12.1
1.7
2.1
4.0
1.0
9.7
9.5
2.6
2.8
Death rate
per 100,000
10
104
140
58
82
43
No; of
cases
2609
£*\j\j y
1095
J. \JStJ
41
32
J £*
517
J A. 1
7
88
70
12
11
figures given in original report.
Source: IARC, 1986.
4-19
5/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 4-8. Age-adjusted lung cancer mortality ratios for males and females, by tar and nicotine
(T/N) in cigarettes smoked
High T/N1
Medium T/N
Low T/N
Males
1.00
0.95
0.81
- Females
1.00
0.79
0.60
"The mortality rate for the category with highest risk was made 1.00 so that the relative reductions
in risk with the use of lower T/N cigarettes could be visualized.
Source: U.S. DHHS, 1982.
4-20
5/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 4-9. Relative risk for lung cancer by type of cigarette smoked (filter vs. nonfilter), in men,
based on cohort and case-control studies
Reference
Hawthorne & Fry (1978)
Rimington (1981)
Bross & Gibson (1968)
Wynder et al. (1970)
Dean et al. (1977)
Type of study
Cohort
Cohort
Case-control
Case-control
Case -control
Relative risk
0.8
0.7
0.6
0.6
0.5
Source: IARC, 1986.
4-21
5/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
i
o
DO
Comments
*3
*3
0
1
. "°
IZ w
CO
O
S'f
Nonsmokei
1.0 (RR)
en
(D
"3
OS
«M
O «g
15
k
OQ
.
"3
«f:
*o o>
QC
^5
V
"o
0 ^
ci "^
H
?7
w ir>
O
£
U">
V
^ i2 ^ ^
CM ^
CO •••*
oo -rr
CO CM
2 "-1
vo r** oo co
o' — J o CM'
t-» o in «-j
Tt rt O' «-<
ON CM OO OO
CM CO CO
OO
Sfesfa
l«-t
>«< HH
op op
fe S
J3 ,0
>» >.
(D D
£ £
V
»*
0
*~*
VH c3
"1 "§> wlS
Us" § 1
ZO K 8
+
*—'
jl
•o
.US
£ i.
***""* ^5
^ d -i«
M rt v
o o
r-' oo
oo
^s
o **j
CM" ^o
1
1 1
ti d^
•o 5
<0
oo
*O 10
o — <
1?
S£ '
"3
0
d
oo
^r
CM
T3 00
^4 CM
ol^J
O en
° O
^. °
rt f> -H
II"
0
a «
0
S
i
X
w
>n oo r-
•^ co o
o' o o
co o CN
CO CM -H
o o o'
CM O CO
CM — ; O
o o o
ON w-> co
00 0
00 0
cS
's "§ °
1 1 5 .a 1 '§
§ 1 8 a S g
a .s j- -^ .s °
rt ^ ^H S* t> 55
|3|sM
,_,
•pH
ffi ^^
*O ON
Q C
0)
00
03
D.
00
a
• rH
1
o
•d
-------
DRAFT—DO NOT QUOTE OR CITE
1
a-
u
,3-
CO
43
>I
H
•a
,3
*— H
1
tq
§
CO
VK
_*J O
^1 |
5^73 oo
1 .§ -3
«-§2
CD 03
g •§ .0
£ H s
CO ^
to o3
*i T3
co *""^
a o
co sS
3 o
00 03
"v
.2 jj ! S5
'•3
tJ M
S^
»SM
. I O
rti ^ S
N W w
1 *•
1 M
c > o
32 s
V} /* ta
O OO
\Tt I-;
t--' 00
t- ON
-H' CN
"^ — <
CN -<'
ir> tN
en w-i
0 0
oo ON
t~ 10
O O
•a
grt S *^
a * s
o 73 .5
| o| |
| g 1 |
173
2
r^ /*"s
ill
O
d
1
0
^
a
^ ^^^^ ^ ^^
to en CN 2 CN
2 ~n '*-<'** CN ir>
3 ^ CN S -! — '
S S
O 'S
z a
CO
_2
•o
^r a
^^ —
co *^5
CO rt
c§ c3
O = CO
oi
S
O
•S co c
Js -2 "co 2
lJli.fi
1 1 § 1 1 1
M:DO"O<
Ul
il
'
^>
1
g
a
o
!>\
1
7^7*
^
|>s^
1
C/3
1
CO
.2?
"o-
CO
1
c!
^ 0
O o*"
C -
•^
P O
/-<
^
t
•n- o
*^y
O ^
M'5
0 ^
i 6
-1 2
CO °
O 0
55 2
o o 10 o o ^
CN O en O O . >«
O CN o O O CN
CN CN O O O ^f
co *o *o ^^ CN CN
•-^ en CN '™H
CN ^O ^* ^O *~^' C7N
VO ^^ ON oo 0,1^-
en i^ vo •* oo oo
, CN
en eN O 'Tf oo ^O
\f^ ^_, oo *O t^" ^O
O CN — ' CN CN f-
"
§ S
• | CO
5o CT\r~TfcN-Hen
Ot,., ^H — i vo CN Tt vo
rt
1
C* rt "O
^-< *3 a -2
"53 Jg Q .S
g 1 S '| § ..
Illl'-gJ ,
*J-c>
CH s^^
S -a -
•^-< ^_,
I
• IH
4-J
o
o
C5
H
4-23
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
ci
o
o
I
0)
es
CD
8
B
a
*3
at
Histological Type
o
o
1
<£
o
ei
„
|3 ,-.,->,-.
•^ ^ ^ ^
2 o oo o
S 0 VO 0
X™*' ^w' N^*-
o m ON
o
IH
d> w
•Q *-*
6 g 0 ON ON
l£ "-
^4 "
8 S S2H
PH >>
0 g
C 8
Undifferentiated carci
Adenocarcinoma
Squamous-cell carcino
<* S3
Ss£
g gS;
Q HCi
rt
C/3
s
.2?
'3
it
>>
R
ffi
^ O ^^ oo oo ^r
ON cN ^f ^f G\ vi
^. s**/' ^^^ 's.-^
j^ O\ r/i rt\ ^^ ViD rf\
fl5 t/\ ^^ f\i
ffi
P^ jyj gQ C? ^3 ^f CO ^T O]
*~^* fli 45 45 **** ^^ ^^ ®® ^*r* ^^
|«. V"? C C ^^ ^*^ ^^ ^*^ ^^ 1^^
1 ^^
I-H
21 ff «
* * 8
>. >> -M
1) Q> C3
t-c l-c o
Kyi wV r\
•4-1
||
£3 -^^
^•a
o
o TJ , -H o r~ o
05 rt ^J ^5 _4 C3 -H
rt ^ (N
^ 0?
CS kH
ciS O
o^ >•
o •<->•§
— "'J tg "7 O O <3 C>
1-2.1
2 |^
3 S3 *° o t-
g ^ 5 7 i o i i
Q o Z — i
>d ci
ill-
lJl"i|
8IIS18
a § -o ^>4, §
S'i||l|
C/a o !•? PH C/2 <<
'S * ON
^ CD C^
4-24
05/15/92
_
-------
DRAFT-DO NOT QUOTE OR CITE
1
w
-< 0 o tj-
-
11 §§!
eo ° «-"' 53 5
.5 A
C
-------
DRAFT—DO NOT QUOTE OR CITE
S~\
1
o
o
*
e
^
—
•A
O
«»H
a"
eu
O
a
O
/— ^
CO
j"j
£
5
o
a
CO
'o
ti
_©
*•§
tj
Q
^^
^
>/;
^Cfl
..^
a>
rt
oi
a
1
S
o
ON
^f
o
'^-
ON
CO
o
CO
ON
-
o\
\fl
Tf
^H
CO
ON
•t*
CM
CM'
VJ
ON
CO
ON
£
O
!Q
03
§
1
u
o
3
|
O
O x;s
^"3^
^«0
o
^
r~
,*
r-
tN
CM'
t~^
X t~
ON CM
CO CO
0 CM
0 0
r— oo
0 O
TT -H
^^ r*^
vd r-'
CM
o ^r
vd vd
CM
a
a
o
•g^
§1
•0-3
S 3
Undifferentia
Poorly dif fere
carcinoma
vd
oo
ON
^H
•*
<
7.
o
I
4-26
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 4-11. Lung cancer death attributable to tobacco smoking in certain countries
,' - - Crude Rate in ;
, , Persons Aged 35+
Country
Canada
Men
Women
England & Wales
Men
Women
Japan
Men
Women
Sweden
Men
Women
USA
Men
Women
Year
1978
1978
1981
1981
1981
1981
1981
1981
1979
1979
Expected
No* of Deaths in
Deaths1 Momiftolcers2'
6,435
1,681
26,297
8,430
16,638
6,161
-
1,777
654
72,803
25,648
556
487
1,576
1,663
2,868
2,593
301
281
5,778
5,736
Observed ,
142.8
34.0
228.5
63.3
64.8
21.0
85.0
28.0
166.7
50.0
'''„ '
In Non-
Smokers AC3
-- - ,'-,-' •••,;*
11.8
9.9
13.3
12.4
10.7
8.9
14.0
12.3
12.7.
11.1
5,762
1,194
24,720
6,767
13,184
3,568
1,476
373
67,024
19,912
AP4
'V ' ':
0.9
0.71
0.94
0.80
0.83
0.58
0.83
0.57
0.92
0.78
the Global Epidemiological Surveillance and Health Situation Assessment data bank
of WHO.
Calculated by IARC, 1986. Slightly overestimates number of expected deaths.
3AC, number of cases attributable to smoking.
4AP, proportion of cases attributable to smoking.
Source: IARC, 1986.
4-27
5/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Rate per 100,000 male population
au
70
60
50
40
30
20
10
0
A
D
s
U
1!
"V_
t
s
/
^
ge-Adjusted Cancer
eath Rates* for
elected Sites, Males
nited States,
930-1986
**x.
\.
/""""
/
*
>.
\t
-A
/
"•.-•*,
-----
J.
*~~~^
~s—**
'^~-
/
-~_ „
\
•"— .
7
* /
\..
— --•
/
1
—.s~
x*,
~^v
/
/
'V*'*"
••s^-rt-.T
/..
— 'V
.s\
"~" '««*•.•,
y^-
""""•—.
-N,
i
/
i
•a
1930
1940
1950
--Esophagus —
• Lung. —••
Pancreas -—
1960 1970 1980 1990
—•* Prostate — • •— Colon & Rectum
•• — Bladder Leukemia
—Stomach -----Liver
Figure 4-1. Age-adjusted cancer death rates* for selected sites, males, United States, 1930-1986.
*Adjusted to the age distribution of the 1970 U.S. census population.
Source: U.S. DHHS, 1989.
4-28 5/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Rate per 100,000 female population
80
70
60
so!
40
30
20
1,0
g
A
D
-Si
u
1!
. —
_, "—
^
-,,
^
je-Ac
32th I
slecte
nited
}30-1
— .'—
•• ••
SN*-..
•*.^
1930 19
- — .
justed Cancer
fates* for
d Sites, Female
States,
986
i^«
*"•
^z~>
V
^
\
"x
BSSS2
JS,
"**»«••**
-"V
'"» — *T^
-v/-
\-<
""Z*?*
>
:«
"I
X
r
•xx<
^
• W...V,.
r
7
•*.xv
5~s:i;
T.-.^ui.-
7
-
\
40 1950 i960 : 1970 1980 -• 1990
Lung Ovary —Colon & Rectum
Breast — Uterus -±- • • -— Leukemia
Figure 4-2. Age-adjusted cancer death rates* for selected sites, females, United States, 1930-1986.
* Adjusted to the age distribution of the 1970 U.S. census population.
Source: U.S. DHHS, 1989. :
4-29 5/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
32.4
21.2
20.3
13.6
10.3 H
-Illll. A
11.4
Never Current s2 3-5 6-10 11-15 16+ Never Current s2 3-5 6-10 11-15 16 +
Smoked Smokers Smoked Smokers
Years of Cessation
Years of Cessation
Smoked 1-20 Cigarettes a Day
Smoked 21 or More Cigarettes a Day
Figure 4-3. Relative risk of lung cancer in ex-smokers, by number of years quit, women Cancer
Prevention Study II
Source: Garfinkel and Silverberg, 1991.
4-30
5/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
5. HAZARD IDENTIFICATION II: INTERPRETATION OF
EPIDEMIOLOGIC STUDIES ON ETS AND LUNG CANCER
5.1. INTRODUCTION
The Centers for Disease Control attributed 434,000 U.S. deaths in 1988 to smoking (CDC,
1991a). Major disease groups related to smoking mortality include lung cancer, chronic
obstructive pulmonary disease, coronary heart disease, and stroke, with smoking accountable for
an estimated 87%, 82%, 21%, and 18% of total deaths, respectively. Lung cancer alone accounted
for about 25% to 30% of the total smoking mortality with some 100,000 deaths. The age-
standardized annual lung cancer mortality rates for 1985 are estimated at 12 per 100,000 for
females and 15 per 100,000 for males who never smoked but 130 per 100,000 for female and 268
for male cigarette smokers, a relative risk of 10.8 and 17.4, respectively (Garfinkel and Silverberg,
1991).
Chapter 4 discusses the biological plausibility that passive smoking may also be a risk
factor for lung cancer because of the qualitative similarity of the chemical constituency of
sidestream smoke, the principal source of environmental tobacco smoke (ETS), and mainstream
smoke taken in during the act of "puffing" on a cigarette, and because of the apparent non-
threshold nature of the dose-response relationship observed between active smoking and lung
cancer. Although the relative risk of lung cancer from passive smoking would undoubtedly be
much smaller than that for active smoking, the ubiquity of ETS exposure (Chapter 3) makes
potential health risks worth investigating. This chapter analyzes the data from the large number
of epidemiologic studies on ETS and lung cancer. There is sufficient empirical evidence derived
from human experience under real-life conditions to assess the lung cancer hazard of ETS without
the attendant uncertainties of extrapolation of risk across species (e.g., from controlled animal
experiments, or from high dose to low dose, as required from human data obtained from
atypically high exposure levels). Virtually all of the 31 studies available classify never-smoking
women as "exposed" or "unexposed" to ETS based on self- or proxy-reported smoking in the
subject's environment, usually according to whether or not a woman is married to a smoker.
Consequently, the data are best suited for estimation of the relative risk of lung cancer mortality
between the exposed and unexposed groups and determination of whether a difference in lung
cancer risk between the classifications is sufficiently large to be detectable with epidemiologic
data. The use of a dose-surrogate such as spousal smoking and dichotomization of persons as
5-1
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
exposed or unexposed is not as well suited for characterization of population risk, although
estimates can be constructed.
Epidemiologic evidence of an association between passive smoking and lung cancer first
appeared 10 years ago in a prospective cohort study in Japan (Hirayama, 198la) and a case-control
study in Greece (Trichopoulos et al., 1983). Both studies concluded that the lung cancer incidence
and mortality in nonsmoking women was higher for women married to smokers than for those
married to nonsmokers. Although there are other sources of exposure to ETS, particularly outside
the home, the assumption is that women married to smokers are exposed to more tobacco smoke,
on average, than women married to nonsmokers. These two studies, particularly the cohort study
from Japan, evoked considerable critical response. They also aroused the interest of public health
epidemiologists, who initiated additional studies. ,
At the request of two Federal agencies—the U.S. Environmental Protection Agency
(Office of Air and Radiation) and the Department of Health and Human Services (Office of
Smoking and Health)~the National Research Council (NRC) formed a committee on passive
smoking to evaluate the methods for assessing exposure to ETS and to review the literature on the
health consequences. The committee's report (NRC, 1986) addresses the issue of lung cancer risk
in considerable detail and includes summary analyses of the evidence from 10 case-control and 3
cohort (prospective) studies. It concludes, "Considering the evidence as a whole, exposure to ETS
increases the incidence of lung cancer in nonsmokers."
The NRC committee was particularly concerned about the potential bias in the study
results caused by the fact that current and former smokers may have incorrectly self-reported as
lifelong nonsmokers (never-smokers). Using reasonable assumptions for misreported smoking
habits, the committee determined that a plausible range for the true relative risk is 1.15 to 1.35,
with 1.25 the most likely value. When these relative risks are also corrected for background
exposure to ETS to make the risk relative to a baseline of zero ETS exposure, the resultant
estimate is 1.42, with a plausible range of 1.24 to 1.61.
Two other major reports on passive smoking have appeared: the Surgeon General's report
on the health consequences of passive smoking (U.S. DHHS, 1986) and the report on methods of
analysis and exposure measurement related to passive smoking by the International Agency for
Research on Cancer (IARC, 1987). The Surgeon General's report concludes:
The absence of a threshold for respiratory carcinogenesis in active smoking, the
presence of the same carcinogens in mainstream and sidestream smoke, the
demonstrated uptake of tobacco smoke constituents by involuntary smokers, and
the demonstration of an increased lung cancer risk in some populations with
5-2
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
exposures to ETS lead to the conclusion that involuntary smoking is a cause of lung
cancer.
The IARC committee emphasized issues related to the physicochemical properties of ETS,
the toxicological basis for lung cancer, and methods of assessing and monitoring exposure to ETS.
Included in the 1987 IARC report is a citation from the summary statement on passive smoking of
a previous IARC report that the epidemiologic evidence available at that time (1985) was
compatible with either the presence or absence of lung cancer risk. Based on other considerations
related to biological plausibility, however, it concludes that passive smoking gives rise to some risk
of cancer. Specifically, the report (IARC, 1986) states:
Knowledge of the nature of sidestream and mainstream smoke, of the materials
absorbed during "passive smoking," and of the quantitative relationships between
dose and effect that are commonly observed from exposure to carcinogens ...
leads to the conclusion that passive smoking gives rise to some risk of cancer.
In the 5 years since those reports, the number of studies available for analysis has more
than doubled. There are now 31 epidemiologic studies available from eight different countries,
listed in Table 5-1. Twenty-seven employ case-control designs, denoted by the first four letters
of the first author's name for convenient reference, and four are prospective cohort studies,
distinguished by the designation "(Coh)." Six case-control studies, FONT (USA), JANE (USA),
KALA (Greece), LIU (China), SOBU (Japan), and WUWI (China), have been published as
recently as 1990. The small cohort study from Scotland (Gillis et al., 1984) has been updated and
is now included under the name HOLE(Coh); another small cohort study on Seventh-Day
Adventists in the United States, an unpublished dissertation, is included as BUTL(Coh). The
abstracts for a second case-control study by Kabat and Wynder and a new one by Stockwell and
colleagues are included in the critical analysis in Appendix A, but insufficient information is
available to include their results.
Because of coincidental timing, the 1986 reports of the Surgeon General and the NRC
review approximately the same epidemiologic studies available for review. More specifically, the
NRC report includes 10 of the studies shown in Table 5-1: AKIB, CHAN, CORR, GARF,
KABA, KOO, LEE, PERS, and TRIG; WU was available but not included because the crude data
were not reported. (Crude data consist of the number of exposed and unexposed subjects among
lung cancer cases and controls, where a subject is typically classified as exposed to ETS if married
to a smoker.) The NRC also excluded an earlier version of the KOO study and the studies by
Knoth et al. (1983) (no reference population was given), Miller (1984) (did not report on lung
cancers separately), and Sandier et al. (1985) (included very few lung cancers). Aside from WU,
these studies are also omitted from this report for the same reasons.
5-3
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
The subscripts on study names in Table 5-1 refer to the "tier" number assigned to it.
Following the statistical analysis of all studies (Sections 5.2 and 5.3), each study is examined
individually for sources of bias and confounding that might affect validity of its results for
assessing ETS and lung cancer, and it is given a tier number from 1 to 4 accordingly (Sections
5.4.2 to 5.4.4). Pooled estimates of relative risk by country are then recalculated by tiers,
beginning with the studies considered most valid (Tier 1) and adding Tiers 2, 3, and 4 successively
(Section 5.4.5) to see how sensitive the outcome is to the choice of studies selected in this manner
(KATA has no tier number because the odds ratio cannot be calculated). Overall, the data
analysis consists of two parts, the first dealing solely with quantitative uncertainty taken into
account by statistical methods and the second including the equally important but more subjective
sources of uncertainty related to study design, methodology, and applicability to the topic of ETS
and lung cancer.
The ETS studies are grouped by country in Table 5-2, which indicates the time period of
data collection in each study, sample size, and prevalence of ETS exposure for each study. The
geographical distribution of the current epidemiologic evidence is diverse. By country, the
number of studies and its percentage of the total number of studies over all countries is as follows:
China (4, 13%), England (1, 3%), Greece (2, 6%), Hong Kong (4, 13%), Japan (6, 19%), Scotland
(1, 3%), Sweden (2, 6%), United States (11, 35%). (One of the studies from Japan, KATA, does
not appear in most of the tables because the odds ratio cannot be calculated.) The studies differ
by size, however, which has to be taken into account in analysis. There are two large cohort
studies, GARF(Coh) and HIRA(Coh), conducted in the United States and Japan, respectively, and
two very small ones, BUTL(Coh) and HOLE(Coh), from the United States and Scotland,
respectively. There are two exceptionally large case-control studies—FONT and WUWI of the
United States and China; the first was designed specifically to assess the association between ETS
and lung cancer, whereas the second has broader exploratory objectives.
Additional characteristics of the case-control studies are summarized in Table 5-3.
The table headings are largely self-explanatory, aside perhaps from "ETS sample matched," which
refers to whether design matching applies to the ETS subjects (the never-smokers used for
ETS/lung cancer analysis). As indicated under "Matched variables," controls are virtually always
matched (or at least similar) to cases on age and usually on several other variables as well that the
researcher suspects may confound results. The matching often refers to a larger data set than just
the ETS subjects, however, because many studies included smokers and investigated a number of
issues in addition to whether passive smoking is associated with lung cancer. When the data on
5-4
05/15/92
-------
DRAFT-DO NOT QUOTE OR CITE
ETS subjects are extracted from the larger data set, matching is not retained unless smoking status
was one of the matching variables. While matching is commonly used as a method to control
confounding, there are effective ways available to control confounding in the analysis of the data.
For studies that include an "adjusted analysis" (i.e., a statistical method such as poststratification or
logistic regression that adjusts the ETS association for potential confounders), the estimated
relative risk from that adjusted analysis is compared with the outcome from the crude data alone
in Section 5.2.1. The variables taken into account in adjusted analyses differ across studies,
depending on study designs and potential confounding addressed by the authors. (Note: "Relative
risk" is used to mean estimate of the true [but unknown] relative risk. For case-control studies,
the estimate used is the odds ratio. For editorial convenience, relative risk is used for both case-
control and cohort studies.)
The selection of the most appropriate relative risk estimate to be used from each study is
addressed in Section 5.2.1. In Section 5.2.2, each chosen relative risk estimate is adjusted
downward to account for bias expected from some smokers misrepresenting themselves as
nonsmokers. This topic has been a contentious issue in the literature for several years, with claims
that this one source of systematic upward bias may account entirely for the excess risk observed in
epidemiologic studies. Recent detailed investigation of this topic by Wells and Stewart
(unpublished) make that claim unlikely (Appendix B). They found that a reasonable correction
for bias, calculated on a study-by-study basis, is positive but small. Following this methodology,
this report makes reductions in the relative risk estimates at the outset for each study individually
prior to statistical inference or pooling estimates from studies from the same country. This is in
contrast to the NRC report (1986), which makes the same downward adjustment to all studies
(applied to an overall estimate of relative risk obtained after pooling study estimates). The
estimates adjusted for smoker misclassification bias are the basis for statistical inference in Section
5.3. The statistical inference approaches consist of both estimation, with confidence intervals, and
hypothesis testing, which includes testing for an effect of ETS exposure and for an upward dose-
response trend. Section 5.4 considers potential sources of bias and confounding and extends the
data interpretation to take these into account. Conclusions are then drawn for hazard
identification (i.e., whether ETS is causally associated with increased lung cancer mortality).
Chapter 6 of this report addresses U.S. population risk of lung cancer from ETS.
5-5
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
5.2. RELATIVE RISKS USED IN STATISTICAL INFERENCE
5.2.1. Selection of Relative Risks
Two considerations largely affect -the: choice of relative risk (RR): (1) whether potential
confounders are taken into account and (2) the source and place of ETS exposure used. The
alternatives (not yet adjusted for smoker misclassification) are shown by study in Tables 5-4 and
5-5 with the ones selected for analysis in this report in boldface type. Table 5-4 lists the RRs and
their confidence intervals, along with explanatory footnotes, and Table 5-5 provides information
on source and place of exposure and on the adjusted analysis. Because most studies included
spousal smoking, and interstudy comparisons may be useful, spousal smoking was the preferred
ETS surrogate except for LAMW and SOBU. In LAMW, spousal smoking data are limited to cases
with adenocarcinoma; in SOBU, the data for cohabitants are separate from data for spousal
smoking and much of the ETS exposure appears to result from the cohabitants. Only data for
broader exposure to ETS than spousal smoking alone were collected in BUFF, CHAN, SVEN, and
HOLE(Coh).
After exposure source and place are taken into account in the choice of RR values in
Table 5-5, an adjusted RR is considered preferable to a crude RR unless the study review in
Appendix A indicates a problem with the adjustment procedure. Of the 31 studies, 20 provide
both an adjusted and crude RR, where an "adjusted estimate" is the result of a statistical
procedure that takes potential confounding factors into account, usually by stratification or
logistic regression. Based on the decision rule just described, our choice of RR is the smaller of
the crude and adjusted values in 14 of the 20 studies providing both estimates. In several studies,
RR values in addition to those shown in Table 5-5 might be considered (see Table 5-6). They
were not found to be the best choices, however, for comparison between studies.
5.2.2. Downward Adjustment to Relative Risk for Smoker Misclassification Bias
There is ample evidence that some percentage of smokers, which differs for current and
former smokers, misrepresent themselves as never-smokers (or sometimes the wording of a
questionnaire may not be explicit enough to distinguish former smokers from never-smokers) (see
Appendix B). It has been argued that the resultant misclassification of some smokers as
nonsmokers results in upward bias of the relative risk for lung cancer from ETS exposure (i.e., the
observed RR is too large). The essence of the supporting argument is based on smoking
concordance between husband and wife—a smoker is more likely than a nonsmoker to have been
married to a smoker. Consequently, the smoker misclassified as a nonsmoker is more likely to be
5-6
05/15/92
-------
DRAFT--DO NOT QUOTE OR CITE
in the ETS-exposed classification as well. Because smoking causes lung cancer, a misclassified
smoker has a greater chance of being a lung cancer case than a nonsmoker. The net effect is that
an observed association between ETS exposure and lung cancer among people who claim to be
never-smokers may be partially explainable by current or former active smoking by some
subjects.
The potential for bias due to misreported smoking habits appears to have been noted first
by Lee (see discussion in Lehnert, 1984), and it has been emphasized by him in several articles
(e.g., Lee 1986, 1987a, 1987b). In Lee (1987b), it is argued that smoker misclassification may
explain the entire excess lung cancer risk observed in self-reported never-smokers in
epidemiologic studies. Lee's estimates of bias due to smoker misclassification appear to be
overstated, for reasons discussed in Appendix B.
The NRC report on ETS (NRC, 1986) devotes considerable attention to the type of
adjustment for smoker misclassification bias. It follows the construct of Wald and coworkers, as
described in Wald et al. (1986); Wald was the author of this section of the NRC report. An
illustrative diagram for the implicit true relative risk of lung cancer from exposure to ETS in
women from spousal smoking is shown in.Figure 2 of Wald et al. (1986). A similar example is in
Table 12-5 of the NRC report.
Both Lee's and Wald's .work adjust an overall relative risk estimate, pooled over several
studies, downward, rather than address each individual study, with its own peculiarities,
separately. Furthermore, statistical analysis over the studies as a whole is conducted first, and
then an adjustment is made to the overall relative risk estimate. The recent work of Wells and
Stewart (Appendix B) on this subject makes an adjustment to each individual study separately.
Consequently, the pertinent adjustment factors that vary by study and type of society can be
tailored to each study and then applied to the observed data prior to any statistical analysis. The
latter procedure is applied in this report.
The methodology to adjust for bias due to smoker misclassification and the details of its
application to the ETS studies are in Appendix B. The results of the adjustment and estimate of
bias are given in Table 5-7. In general, the biases are low in East Asia, or in any traditional
society such as Greece, where female smoking prevalence is low and the female smoker risk is
low. Some of the calculated biases are slightly less than unity when carried to three decimal
places. This may result from the assumption in the calculations that there is no passive smoking
effect on current smokers. ,
5-7
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
5,3. STATISTICAL INFERENCE
5.3.1. Introduction
Table 5-8 lists the values of several statistical measures by study for spousal smoking (see
boldface entries in Table 5-5 for details). Their meaning will be described before proceeding to
interpretation of the data, even though the concepts discussed may be familiar to most readers.
The p-values refer to a test for effect and a test for trend. In the former, the null hypothesis of
no association (referred to as "no effect" of ETS exposure on lung cancer risk) is tested against the
alternative of a positive association (see Appendix E). The test for trend applies to a null
hypothesis of no association between RR and exposure level against the alternative of a positive
association. When data are available on more than two levels of intensity or duration of ETS
exposure, typically in terms of the husband's smoking habit (e.g., cig./day or years of smoking),
then a test for trend is a useful supplement in testing for an effect, as well as indicating whether a
dose-response relationship is likely.
The entries under "Power" in Table 5-8 are calculated for the study's ability to detect a
true relative risk of 1.5 and a decision rule to reject the null hypothesis of no effect when p < 0.05
(see Dupont and Plummer [1990], for methods to calculate power). The power is the estimated
probability that the null hypothesis would be rejected if the true relative risk is 1.5 (i.e., that the
correct decision would result; the power would be larger if the true relative risk exceeds 1.5).
Using the estimates of power for the U.S. studies in Table 5-8 for illustration, the estimated
probability that a study would fail to detect a true relative risk of 1.5 (equal to 1-power, the
probability of a Type II error [discussed in the next paragraph] when the true relative risk is 1.5)
is as follows: FONT, 0.07; GARF(Coh), 0.08; GARF, 0.40; JANE, 0.56; BUFF, 0.83; CORR, 0.78;
WU, 0.79; HUMS, 0.80; KABA, 0.83; BUTL(Coh), 0.82; BROW, 0.85. Thus, 7 of the 11 U.S.
studies have only about a 20% chance of detecting a true relative risk as low as 1.5, when taken
alone. Sources of bias effectively alter the power in the same direction as the bias (e.g., a
downward bias in RR will increase the expected p-value, i.e., reduce significance, in a test for
effect). Of the potential sources of bias discussed by study in Appendix A; the predominant
direction of influence on the observed RR, when identifiable, appears to be in the direction of
unity, thus affecting power adversely. The RRs have already been reduced to adjust for smoker
misclassification, the only systematic source of upward bias that has been established.
Studies of all sizes, large and small, are equally likely to make a false conclusion if ETS is
not associated with lung cancer risk (Type I error). However, smaller studies are less likely to
detect a real association when there is one (Type II error). This imbalance comes from using the
5-8
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
significance level of the test statistic to determine whether to reject the null hypothesis. If the
decision rule is to reject the hypothesis when the p-value is smaller than some prescribed value
(e.g., 0.05), then the Type I error rate is 0.05, but the Type II error rate increases. When a study
with low power fails to reject the null hypothesis of no effect, it is not very informative because,
if the power is low,, that outcome may be nearly as likely when the null hypothesis is false as when
it is true. When detection of a small relative risk is consequential, pooling informational content
of suitably chosen studies empowers the application of statistical methods.
The heading in Table 5-8 that remains to be addressed is "Relative weight," to be referred
to simply as "weight." When the estimates of relative risk from selected studies are combined, as
for studies within the same country as shown in the table, the logarithms of the RRs are weighted
inversely proportional to their variances (see Appendix E and Footnote 2 of Table 5-8). These
relative weights are expressed as percentages summing to 100 for each country in Table 5-8.
Study weight and power are positively associated, which is explained by the significant role of
study size to both. Consequently, studies weighted most heavily (because the standard errors of
the RRs are low) also tend to be the ones with the highest power (most likely to detect an effect
when present).
5.3.2. Outcomes by Study and Country
5.3.2.1. Tests for Association
The p-values of the test statistics for the hypothesis of no effect (i.e., RR = 1) are shown
in Table 5-8. Values of the test statistics (the standardized log odds ratio; see Appendix E) are
plotted in Figure 5-1. Also shown in Figure 5-1 for reference are the points on the horizontal
axis corresponding to p-values of 0.5, 0.2, 0.1, 0.05, 0.01, and 0.001. For example, the area under
the curve to the right of the vertical line labeled p = 0.01 is 0.01 (1%), so it is apparent from
Figure 5-1 that three studies had significance levels p < 0.01 (more specifically, 0.001 < p < 0.01.
The size of the symbol (upside-down triangle) used for a study is proportional in area to the
relative weight of that individual study, but of current interest is the location and not the size of
the symbol. If the null hypothesis is true, then the plotted values would arise from a standard
normal distribution, shown in the figure (points to the left of zero indicate that the RR is less than
1 and points to the right of zero indicate that RR is greater than 1). If the points lie more toward
the right side of the normal curve than would be likely to occur by chance alone, then the
hypothesis of no effect is rejected in favor of a positive association between ETS exposure and
lung cancer. If one constructs five intervals of equal probability (i.e., intervals of equal area
5-9
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
under the standard normal curve), the expected number of observations in each interval is six
(these five intervals are not shown on Figure 5-1). The observed numbers in these intervals,
however, for intervals from left to right are 3, 3, 1, 7, and 16, an outcome that is significant at p <
0.005, by the chi-squared goodness-of-fit test. At the points on the standard normal curve
corresponding to p-values 0.5, 0.4, 0.3, 0.2, 0.1, and 0.05, the probability that a number of
outcomes as large as that actually observed would occur by chance is less than 0.005 at all points.
Consequently, the hypothesis of no effect is rejected on statistical grounds, and that conclusion is
not attributable to a few extreme outcomes that might be aberrant in some way.
Figure 5-2 displays the U.S. studies alone (see Appendix E for calculation of the test
statistics). Figure 5-3 corresponds to Figure 5-1 except that the test statistics for the hypothesis
of no effect (i.e., RR = 1) for the significance levels shown apply to a single overall estimate of
RR for each country, formed by statistically pooling the outcomes from the studies within each
country. The areas of the symbols for countries are also in proportion to statistical weight as
given in Table 5-8. It is implicitly assumed that studies within a country, and the subpopulations
sampled, are sufficiently homogeneous to warrant to combine their statistical results into a single
estimate for the country (see S. Greenland [1987] for a discussion of applications of meta-analysis
to epidemiology). The calculational method employed weights the observed RR from each study
within a country inversely proportional to its estimated variance (see Appendix E). The relative
study weights are shown in Table 5-8. Each symbol in Figures 5-1, 5-2, 5-3, and 5-4 has been
scaled so that its area is proportional to the weight of the outcome represented, relative to all other
outcomes shown in the same figure.
Greece, Hong Kong, and Japan, which together comprise a total weight of 39%, are each
statistically significant at p < 0.01 against the null hypothesis of no increase in relative risk
(RR = 1). When the United States is included, the total weight is 73%, and each of the four
countries is significant at p < 0.02. The four studies combined into the group called Western
Europe are not large. Together they represent 5% of the total weight, and their combined odds
ratio (1.17) is slightly above 1 but not statistically significant (p = 0.21). In contrast, China is
weighted quite high (22%), the p-value is large (0.66), and the odds ratio is less than 1 (0.95),
strongly indicating no evidence of an increase in RR due to ETS. This is largely because China is
very heavily influenced by WUWI (relative weight of 60%), which is a very large case-control
study. However, this apparent inconsistency in WUWI may be due to the presence of indoor
smoke from cooking and heating which may mask any effect from passive smoking. A similar but
more extreme situation is found in LIU, conducted in a locale where indoor heating with smokey
5-10
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
coal (an established risk factor for lung cancer) and inadequate venting are common. The indoor
environments of the populations sampled in WUWI and LIU make detection of any carcinogenic
hazard from ETS unlikely, and thus render these studies to be of little value for that purpose (see
discussions of WUWI and LIU in Appendix A). Without WUWI or LIU, the combined results of
the two remaining studies in China, GAO and GENG, are significant at p = 0.03, as shown in
Table 5-8 and pictured in Figure 5-4.
5.3.2.2. Confidence Intervals
Confidence intervals for relative risk are displayed by study and by country in Table 5-8
(see Appendix E for method of calculation). The 90% confidence intervals by country are
illustrated in Figure 5-5. (Note: 90% confidence intervals are used for correspondence to a right-
tailed test of the hypothesis of no effect at a 5% level of significance.) The area of the symbol
(solid circle) locating the point estimate of relative risk within the confidence interval is
proportional to study weight. Symbol size is used as a device to draw attention to the shorter
confidence intervals, which tend to be based on more data than the longer ones. The confidence
intervals for countries jointly labeled as Western Europe are in Table 5-8, except for Sweden
which contains two studies, PERS and SVEN. For those two combined, the odds ratio (OR) is
1.19 (90% C.I. = 0.81-1.74). The confidence interval for China without LIU or WUWI (i.e.,
including only GAO and GENG) is displayed in Figure 5-6.
In descending order, the relative risks in Figure 5-6 are for Greece, Hong Kong, Japan,
China, the United States, and Western Europe. Values in the interval (1.43, 1.71) are contained in
the 90% confidence intervals of the first four countries (Greece, Hong Kong, Japan, and China),
where the observed relative risks range from 2.00 down to 1.36. The region in common to the
confidence intervals for the two remaining countries or groups of countries, United States and
Western Europe, is (1.04, 1.35), the interval for the United States alone. The observed relative
risks are close (1.19 and 1.17). If the United States and Western Europe are combined, the RR is
1.18 (90% C.I. = 1.05-1.34). The estimated relative risks from exposure to spousal smoking differ
between countries, with Greece and the Asian countries near the high end of the scale and the
Western countries, United States and Western Europe, at the low end. However, the relative risks
only pertain to ETS exposure from spousal smoking which may be a higher proportion of total
ETS exposure in some countries than in others. This emphasizes the importance of taking into
account exposure and background (nonspousal) ETS as used, which is considered in the estimation
of population risk for the United States in Chapter 6.
5-11
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
5.3.2.3. Tests for Trend
When epidemiologic data for the "exposed" group are available for two or more exposure
levels plus the control group, a test for trend can be used to test for a dose-response relationship.
A dose-response relationship increases support for a causal association by diminishing the
likelihood that the results can be explained by confounding. Furthermore, when low exposure
levels have little effect on the observed RR but the RR does increase with increasing exposure, a
trend test may be able to detect an association that would be masked in a test for effect. This is
especially likely to occur when dealing with a weak association or crude surrogate measures for
exposure (i.e., greater potential for misclassification), both of which are difficulties in studies of
ETS and lung cancer.
As discussed in Chapter 3, ETS is a dilute mixture. Furthermore, questionnaire-based
assessment of exposure to ETS is a crude indicator of actual lifetime exposure, and spousal
smoking is an incomplete surrogate for exposure because it does not consider ETS from other
sources, such as the workplace. Under these circumstances, there is considerable potential for
exposure misclassification, which is compounded when the exposed group is further divided into
level-of-exposure categories. Division into exposure-level categories also reduces the power to
statistically determine a real effect by decreasing the number of subjects in an exposure group.
This is especially problematic in small studies. These inherent difficulties with the ETS database
would tend to diminish the possibility of detecting dose-response relationships. Therefore, the
inability to demonstrate a dose-response trend is not considered evidence against causality; rather,
if a statistically significant trend can be detected despite these potential obstacles, it provides
evidential support of a causal association.
Table 5-9 presents the dose-response data and trend test results for females currently
available from the studies of ETS and lung cancer discussed in this report. Exposure is measured
by intensity (e.g., cig./day smoked by the husband), duration (e.g., number of years married to a
smoker), or a combination of both (e.g., number of pack-years—packs per day x years of smoking
by the husband). The p-vaiues reported in the table are for a test of no trend against the one-
sided alternative of an upward trend (i.e., increasing RR with increasing exposure). (Note: The
results for tests of trend are taken from the study reports. Unless the report specified that a one-
sided alternative was used, the reported p-value was halved to reflect the outcome for the one-
sided alternative of RR increasing with exposure. Where the data are available, the p-values
reported by the individual study's authors have been verified here by application of the Mantel-
Haenszel test [Mantel and Haenszel, 1963].)
5-12
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Wu-Williams and Samet (1990) previously reviewed the dose-response relationships from
the epidemiologic studies on ETS then available. They determined that 12 of 15 studies were
statistically significant for the trend test for at least one exposure measure. The probability of this
proportion of statistically significant results occurring by chance in this number of studies is
virtually zero (p < 10"13). Intensity of spousal smoking was the most consistent index of ETS
exposure for the demonstration of a dose-response relationship.
Our assessment of the dose-response data is similar and provides essentially the same
results for a slightly different set of studies. Table 5-10 summarizes the p-values of the trend
tests for the various ETS exposure measures from the studies presented in Table 5-9. The
exposure measure most commonly used was intensity of spousal smoking. Seven of the eleven
studies that reported dose-response data based on cigarettes per day showed statistical significance
at the p < 0.05 level for the trend test. Again, the probability of this many statistically significant
results occurring by chance in this number of studies is virtually zero (p < 10"6). The trend test
results for the other exposure measures were consistent, in general, with those based on cigarettes
per day (three of six studies using total years of exposure were significant, as were two of two
studies using pack-years).
Overall, 10 of the 15 studies with test for trend are statistically significant for one or more
exposure measures. These results are especially compelling in view of the fact that dividing the
data into smaller exposure categories decreases the power to detect a real effect. No possible
confounder has been hypothesized that correlates with ETS exposure and could explain the
increasing incidence of lung cancer with increasing exposure to ETS in so many independent
studies from different countries.
By country, the number of studies with significant results for upward trend is: China, 1 of
2; Greece, 2 of 2; Hong Kong, 1 of 2; Japan, 3 of 3; Sweden, 0 of 1; and United States, 3 of 5. Of
particular interest, two of the U.S. studies, GARF and CORR, are statistically significant for a
test of trend, providing evidence for an association between ETS exposure and lung cancer even
though neither was significant in a test for effect. This occurs because in both cases, the data
supporting an increase in RR are largely at the highest dose level. It might be that relatively high
exposure levels are necessary to detect an effect in the United States, as would be expected if
spousal smoking is a weaker surrogate for total ETS exposure in this country, a possibility
mentioned previously. The U.S. study by Fontham et al. (1991), a well-conducted study and the
largest case-control study of ETS and lung cancer to date, with the greatest power of all the U.S.
studies to detect an effect, was statistically significant with a p-value of 0.04 for the trend test
5-13
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
with pack-years as the exposure measure. When the analysis was restricted to adenocarcinomas,
the majority of the cases, tests for trend were statistically significant by both years (p = 0.02) and
pack-years (p - 0.01).
5.3.2.4. Statistical Conclusions
Two types of tests have been conducted: (1) a test for effect, wherein subjects must be
classified as exposed or unexposed to ETS, generally according to whether the husband is a
smoker or not, and (2) a trend test, for which exposed subjects are further categorized by some
level of exposure, such as the number of cigarettes smoked per day by the husband, duration of
smoking, or total number of packs smoked. Results are summarized in Table 5-11, with countries
in the same order as in Table 5-8. Studies are noted in boldface if the test of effect or the trend
test is significant at 0.1, or if, as in PERS, the odds ratio at the highest exposure is significant. In
9 of the 11 studies in Greece, Hong Kong, or Japan, at least one of the tests is significant at 0.1.
In 8 of these 11 studies, at least one of the tests is significant at 0.05. For the United States and
Western Europe, the corresponding numbers of studies are only 6 and 5 of 15. For the studies
within the first group of countries (Greece, Hong Kong, and Japan), the median power is 0.43,
and only 1 of the 10 studies (10%) has power less than 0.25 (INOU). By contrast, the median
power for the U.S. and Western Europe together is 0.21, and 10 of the 15 studies (67%) have
power less than 0.25. Significance is meaningful in a small study, but nonsignificance is not very
informative because there is little chance of detecting an effect when there is one. Consequently,
there are several studies in the United States-Western Europe group that provide very little
information. One of the four studies in China is significant, at both the 0.1 and 0.05 levels. Two
of the three nonsignificant studies in China (LIU and WUWI) are not very informative on ETS for
reasons previously described.
For the U.S. and Western Europe studies, 3 of the 5 with power greater than 0.25 are
shown in boldface (FONT, GARF, and PERS), indicating at least suggestive evidence of an
association between ETS and lung cancer, compared to only 3 of 10 with power under 0.25
(CORR, HUMB, and WU). The test of effect is suggestive for CORR and HUMB (p-values of
0.09 and 0.10, respectively), and CORR is positive for trend (p-value of 0.02) with an observable
upward dose-response pattern. FONT is significant for effect (p-value = 0.04) and trend
(p-value s 0.04) with observable dose-response (both p-values are for all cell types). Neither
GARF nor PERS are significant for effect, but both are significant in other tests (GARF, p-value
of 0.03 for trend; PERS, p-value of 0.02 at the high dose). The significance in CORR, GARF,
5-14
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
and PERS appears to result from an increase In the observed relative risk at the highest exposure
level. Overall, the evidence of an association is stronger in the United States and Western Europe
than appears from the test for effect alone.
To summarize, there is substantial statistical evidence that exposure to ETS from spousal
smoking is associated with increased lung cancer mortality in Greece, Hong Kong, Japan, and the
United States. The association for Western Europe appears similar to that in the United States,
but not as much statistical evidence has accumulated there. The usefulness of statistical
information from studies in China is limited, so no firm conclusions are drawn from the studies
there. The statistical evidence is also conclusive from the individual studies, without combining
studies within each country to gain power to detect an effect. The number of significant
outcomes in either the test for effect, or the test for trend, in Table 5-11 is not attributable to
chance alone. Tests for effect and for trend are jointly supportive of the same conclusion.
Adjustment on an individual study basis for potential bias due to smoker misclassification results
in slightly lower relative risk estimates but does not affect the overall conclusions.
5.4. EXTENDED DATA INTERPRETATION
5.4.1. Introduction
Whereas Section 5.3 examined the epidemiologic data by individual study and by pooling
all studies by country, this section analyzes the data in three additional ways. First, it assesses the
impact of six potential confounders on the results (Section 5.4.2). Then, in Section 5.4.3, this
report examines the possible sources of bias and other uncertainty-related design features inherent
in case-control studies to determine whether there are any systematic sources of bias (other than
smoker misclassification bias addressed in Section 5.2) that might affect the observed results. The
third extended analysis approach judges the comparative quality of the individual studies
according to how well they have been able to control for these potential biases and confounders
and categorizes each study into one of four tiers (Section 5.4.4). This separation of studies into
tiers is used in the statistical analysis presented in Section 5.4.5 to determine whether the studies
with higher utility provide different conclusions.
The element of chance has been taken into account by the statistical methods previously
applied. It remains to consider potential sources of bias and confounding and whether an
association between ETS exposure and lung cancer may be causally related. Validity is the most
relevant concern for hazard identification. Generalizability of results to the national population
(depending on "representativeness" of the sample population, treated in the text) is important for
5-15
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
the characterization of population risk, but no more so than validity. As stated by Breslow and
Day (1980), "In an analysis, the basic questions to consider are the degree of association between
risk for disease and the factors under study, the extent to which the observed associations may
result from bias, confounding and/or chance, and the extent to which they may be described as
causal."
Confounding requires the presence of a non-ETS cause of lung cancer associated with ETS
exposure. Candidate confounders included in the ETS studies are reviewed in the next section.
Attention is then turned to methodological issues of data classification, collection, and analysis
that may produce bias or inadequate control for confounding. Potential bias and confounding in
each study is discussed vis-a-vis its statistical outcome in Table 5-11, based on the detailed
reviews in Appendix A. In addition, each study is assigned to one of four tiers, depending on the
review. Tier 1 studies are those of greatest utility for investigating a potential association between
ETS and lung cancer. Other studies are assigned to Tiers 2, 3, and 4 as confidence in their utility
diminishes. Tier 4 is reserved for studies we would exclude from analysis for ETS, for various
reasons specified in the text. The summary RR for each country is then recalculated for studies
in Tier 1 alone and for Tiers 1-2, 1-3, and 1-4 (the last category corresponds to Table 5-11). This
exercise provides some idea of the extent to which the summary RR for a country depends on the
choice of studies. The outcome is used to assess the epidemiologic weight-of-evidence for hazard
identification. The concluding section of this chapter draws on the previous statistical analysis
and the material in this chapter to formulate conclusions regarding the association of ETS
exposure with lung cancer and the evidence supporting causality.
Our objective is to consider the influence of sources of uncertainty on the statistical
measures summarized in Table 5-11, although there are limitations to such an endeavor. For
example, not controlling for a potential confounder such as age in the statistical analysis, which
should be done whether or not the study design is on age, may require reanalyzing data not
included in the study report. Potential sources of bias are just that—potential—and their actual
effect may be impossible to evaluate (e.g., selection bias in case-control studies). Although
numerous questions of interest cannot be answered unequivocally, or even without a measure of
subjective judgment, it is nevertheless worthwhile to consider issues that may affect interpretation
of the quantitative results. The issues of concern are largely those of epidemiologic investigations
in general that motivate the conscientious investigator to implement sound methodology.
Statistical uncertainty aside, the outcomes of studies that fare well under close examination inspire
more confidence and thus deserve greater emphasis than those that do poorly.
5-16
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Preliminary to the next sections, some relevant notes on epidemiologic concepts are
excerpted from two IARC volumes entitled Statistical Methods in Cancer Research (Breslow and
Day, 1980 and 1987), dealing with case-control and cohort studies, respectively, which are
excellent references. In the interest of brevity, an assortment of relevant passages is simply
quoted directly from several locations in the references (page numbers and quotation marks have
been omitted to improve readability). Some readers may wish to skip to the next section; those
interested in a more fluid, cogent, and thorough presentation are referred to the references.
Bias and confounding. The concepts of bias and confounding are most easily
understood in the context of cohort studies, and how case-control studies relate to
them. Confounding is intimately connected to the concept of causality. In a
cohort study, if some exposure E is associated with disease status, then the
incidence of the disease varies among the strata defined by different levels of E.
If these differences in incidence are caused (partially) by some other factor C, then ,
we say that C has (partially) confounded the association between E and the disease. ,
If C is not causally related to disease, then the differences in incidence cannot be
caused by C, thus C does not confound the disease/exposure association.
Confounding in a case-control study has the same basis as in a cohort
study .. . and cannot normally be removed by appropriate study design alone. An
essential part of the analysis is an examination of possible confounding effects and
how they may be controlled.
Bias in a case-control study, by contrast, [generally] arises from the
differences in design between case-control and cohort studies. In a cohort study,
information is obtained on exposures before disease status is determined, and all
cases of disease arising in a given time period should be ascertained. Information
on exposure from cases and controls is therefore comparable, and unbiased
estimates of the incidence rates in the different subpopulations can be constructed.
In case-control studies, however, information on exposure is normally obtained
after disease status is established, and the cases and controls represent samples
from the total. Biased estimates of incidence ratios will result if the selection
processes leading to inclusion of cases and controls in the study are different
(selection bias) or if exposure information is not obtained in a comparable manner
from the two groups, for example because of differences in response to a
questionnaire (recall bias). Bias is thus a consequence of the study design, and the
design should be directed towards eliminating it. The effects of bias are often
difficult to control in the analysis, although they will sometimes resemble
confounding effects and can be treated accordingly.
To summarize, confounding reflects the causal association between
variables in the population under study, and will manifest itself similarly in both
cohort and case-control studies. Bias, by contrast, is not a property of the
underlying population. It results from inadequacies in the design of case-control
studies, either in the selection of cases or controls or from the manner in which the
data are acquired.
On prospective cohort studies. One of the advantages of cohort studies over case-
control studies is that information on exposure is obtained before disease status is
ascertained. One can therefore have considerable confidence that errors in
5-17
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
measurement are the same for individuals who become cases of the disease on
interest, and the remainder of the cohort. The complexities possible in
retrospective case-control studies because of differences in recall between cases
and controls do not apply. [Regarding the success of a cohort study, the] follow-
up over time ... is the essential feature. . . . The success with which the follow-up
is achieved is probably the basic measure of the quality of the study. If a
substantial proportion of the cohort is lost to follow-up, the validity of the study's
conclusions is seriously called into question.
On case-control studies... . despite its practicality, the case-control study is not
simplistic and it cannot be done well without considerable planning. Indeed, a
case-control study is perhaps the most challenging to design and conduct in such a
way that bias is avoided. Our limited understanding of this difficult study design
and its many subtleties should serve as a warning—these studies must be designed
and analyzed carefully with a thorough appreciation of their difficulties. This
warning should also be heeded by the many critics of the case-control design.
General criticisms of the design itself too often reflect a lack of appreciation of the
same complexities which make these studies difficult to perform properly.
The two major areas where a case-control study presents difficulties are in
the selection of a control group, and in dealing with confounding and interaction
as part of the analysis.. . . these studies are highly susceptible to bias, especially
selection bias which creates non-comparability between cases and controls. The
problem of selection bias is the most serious potential problem in case-control
studies. . . . Other kinds of bias, especially that resulting from non-comparable
information from cases and controls are also potentially serious; the most common
of these is recall... bias which may result because cases tend to consider more
carefully than do controls the questions they are asked or because the cases have
been considering what might have caused their cancer.
5.4.2. Potential Confounders
In addition to standard demographic risk factors (e.g., age) that are frequently either
adjusted for or controlled for by study design, a number of other variables have been considered
as potential risk factors for lung cancer and thus confounders of the ETS-lung cancer association.
In the following discussion, relevant findings from the ETS studies are summarized for six general
categories: (1) personal history of lung disease, (2) family history of lung disease, (3) heat sources,
(4) cooking with oil, (5) occupation, and (6) diet. Table 5-12 provides an overview of results in
these categories. Two shortcomings are common in the statistical inference of nonspousal ETS
factors: failure to control for the potential confounding effects of other factors, including ETS
exposures other than from spousal smoking, and failure to adjust significance levels for multiple
comparisons. Multiple tests on the same data increase the chance of a false positive (i.e., outcomes
appear to be more significant than warranted due to the multiple comparisons being made on the
same data).
5-18
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
5.4.2.1. History of'Lung Disease
Results regarding history of lung disease have been reported in eight of the reviewed ETS
studies, but with little consistency. Tuberculosis (TB), for example, is significantly associated
with lung cancer in GAO (OR = 1.7; 95% C.I. = 1.1 - 2.4) but not in SHIM (OR = 1.1, no other
statistics), LIU or WU (no ORs provided). Chronic bronchitis, on the other hand, is
nonsignificant in GAO (OR = 1.2; 0.8 - 1.7), SHIM (OR = 0.8), KABA, and WU, but is highly
significant in LIU (OR = 7.37; 2.40 - 22.66 for females; OR = 7.32; 2.66 - 20.18 for males) and
mildly so in WUWI (OR = 1.4; 1.2 - 1.8). (Notably, both the LIU and WUWI populations were
exposed to non-ETS sources of household smoke.) Consideration of each lung disease separately,
as presented, ignores the effect of multiple comparisons described above. For example, GAO
looked at five categories of lung disease. If that were taken into account, the confidence interval
for TB would no longer indicate significance. No discussion of the multiple comparisons effect
was found in any of the references, which should at least be acknowledged.
Broadening our focus to examine the relationship of lung cancer to history of lung disease
in general does little to improve consistency. GENG reports an adjusted OR of 2.12
(1.23 - 3.63) for history of lung disease, GAO's disease-specific findings are consistently positive,
and WUWI reports three positive associations out of an unknown number assessed. SHIM and
WU, however, consistently found no effect except marginally for silicosis (perhaps better
construed as an occupational exposure surrogate) in SHIM and for childhood pneumonia in WU.
LIU found a significant association only for chronic bronchitis and KABA only for pneumonia.
Interpretation is hampered by the lack of numerical data for factors that were not statistically
significant in KABA, LIU, and WU. Even with such data, however, interpretation is hampered
by the absence of control for key potential confounders in many of the studies (e.g., age in GENG
and LIU). Only one study (WU) attempted to control for a history variable (childhood
pneumonia), which reportedly did not alter the ETS results. The importance of prior lung disease
as a potential confounder in studies of ETS is thus unclear, but it does not appear to distort results
one way or the,other.
5.4.2.2. Family History of Lung Disease
Only a few of the studies addressed family history of lung disease as a potential risk factor
for lung cancer. GAO found no significant association between family history of lung cancer and
subjects' disease status (e.g., parental lung cancer OR = 1.1; 95% C.I. = 0.6 - 2.3), and positive
family histories were very rare (e.g., 1.0% among mothers of either cases or controls). In contrast,
5-19
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
WUWI reports a significant association with history of lung cancer in first-degree relatives (OR =
1.8; 1.1 - 3.0), which occurred in about 4.5% of the cases. The presence of TB in a household
member (OR « 1.6; 1.2 - 2.1) is also significant, even after adjustment for personal smoking and
TB status. The rarity of family-linked lung cancer in these populations makes accurate assessment
difficult and also reduces the potential impact on results of any effect it may have. Its study in
populations where such cancer is more common would be more appropriate. The household TB
outcome may be the result of multiple comparisons and/or confounding, particularly in view of
the weaker (nonsignificant) outcome noted for personal TB status.
5.4.2.3. Heat Sources for Cooking or Heating
Household heating and cooking technologies have received considerable attention as
potential lung cancer risk factors in Asian ETS studies. Most studies have focused on fuel type.
Kerosene was specifically examined in three studies. All three found positive associations—
CHAN and LAMW for kerosene cooking, and SHIM for kerosene heating—but none of the
associations were statistically significant, and the SHIM relationship held only for adult and not
for childhood exposure. Five studies specifically examined coal. GENG evaluated use of coal for
cooking and found a significant positive association. Use of coal for household cooking or heating
prior to adulthood is significantly associated with lung cancer in WU's study of U.S. residents, but
no results for adulthood are mentioned. Recent charcoal stove use showed a positive (OR = 1.7)
but not significant association in SHIM. Separate analyses of five coal-burning devices and two
non-coal-burning devices by WUWI found positive though not always significant associations for
the coal burners. In contrast, SOBU found no association between use of unventilated heating
devices—including mostly kerosene and coal-fueled types but also some wood and gas burners—
and lung cancer (OR = 0.94 for use at age 15, 1.09 at age 30, 1.07 at present). Results for wood or
straw cooking were specifically reported in three studies. SOBU found a significant association
for use of wood or straw at age 30 (OR = 1.89; 95% C.I. = 1.16 - 3.06) but only a weak
relationship at age 15. GAO found no association with current use of wood for cooking (OR =
1.0; 0.6 - 1.8), and WUWI mentions that years of household heating with wood, central heating,
and coal showed nonsignificant trends (negative, negative, and positive, respectively).
Overall, studies that examined heating and cooking fuels generally found evidence of an
association with lung cancer for at least one fuel, which was usually but not always statistically
significant. Such relationships appeared most consistently for use of coal and most prominently in
WUWI and LIU. Neither study found a significant association between ETS and lung cancer, nor
5-20
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
did either address whether coal use was associated with ETS exposure. The presence of non-ETS
sources of smoke within households, however, may effectively mask detection of any effect due to
ETS (as noted by the authors of WUWI). Evidence of effects of other fuel types and devices is
more difficult to evaluate, particularly because many studies do not report results for these
factors, but kerosene-fueled devices seem worthy of further investigation.
5.4.2.4. Cooking With Oil
Cooking with oil was examined by GAO and WUWI, both conducted in China; with
positive associations for deep-frying (OR ranges of 1.5-1.9 and 1.2-2.1, respectively, both
increasing with frequency of cooking with oil). GAO also reports positive findings for stir-
frying, boiling (which in this population often entails addition of oil to the water), and smokiness
during cooking and found that most of these effects seemed specific for users of rapeseed oil.
These results may apply to other populations where stir-frying and certain other methods of
cooking with oil are common. Neither study, however, addressed whether cooking with oil is
associated with ETS and thus may confound the effect attributed to ETS.
5.4.2.5. Occupation
Seven studies investigated selected occupational factors, with five reporting positive
outcomes for one or more occupational variables. The outcomes appear somewhat inconsistent,
however. SHIM found a strong and significant relationship with occupational metal exposure (OR
= 4.8) and a nonsignificant one with coal, stone, cement, asbestos, or ceramic exposure, while
WUWI found significant positive relationships for metal smelters (OR = 1.5), occupational coal
dust (OR = 1.5), and fuel smoke (OR = 1.6) exposure. Textile work is positively associated with
lung cancer in KABA and negatively in WUWI. BUFF divided occupations into nine categories
plus housewife and found eight positive and one negative associations relative to housewives, but
only one ("clerical") is significant. GAO, on the other hand, found no association with any of six
occupational categories, while GENG found a significant association for an occupational exposure
variable that encompassed textiles, asbestos, benzene, and unnamed other substances (OR = 3.1;
1.58 - 6.02). WU reported "no association between any occupation or occupational category,"
although there was a nonsignificant excess among cooks and beauticians. Finally, BUTL(Coh)
found an increased RR for wives whose husbands worked in blue collar jobs (> 4; never-smoker).
HIRA(Coh) did not present findings for husband's occupation as a risk factor independently but
reported that adjustment for this factor did not alter the study's ETS results. Few studies
5-21
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
attempted to adjust ETS findings for occupational factors—SHIM found only modest effects of
such adjustment for occupational metal exposure, despite an apparent strong independent effect
, !
for this factor, and GENG found only minimal effect of occupational exposure on active smoking
results but did no adjustment of ETS results. Overall, multiple comparisons, confounding by
other factors (e.g., socioeconomic status, age), and the rarity of most specific occupational
exposure sources probably account for the inconsistent role of occupation in these studies.
5.4.2.6. Dietary Factors
Investigations related to diet have been reported in six of the ETS studies, with mixed
outcomes. The fundamental difficulty lies in obtaining accurate individual values for key
nutrients of interest such as /3-carotene. The relatively modest size of most ETS study populations
adds further uncertainty in attempts to detect and assess any dietary effect that, if present, is
likely small. In those studies where dietary data were collected and adjusted for in the analysis of
ETS, diet has had no significant effect. Nevertheless, diet has received attention in the literature
as a potential confounder, or source of bias, for ETS (e.g., Koo, 1988; Koo et al., 1988; Sidney,
1989; Butler, 1990, 1991; Marchand et al., 1991), so a more detailed and specific discussion is
provided in this section.
Diet is of interest for a potential protective effect against lung cancer, unlike more typical
potential confounders that cause lung cancer. If nonsmokers unexposed to passive smoke have a
lower incidence of spontaneous (unrelated to tobacco smoke) lung cancer incidence due to a
protective diet, then the effect would be upward bias in the RR for ETS. However, for diet to
explain fully the significant association of ETS exposure in Greece, Hong Kong, Japan, and the
United States, which differ by diet as well as other lifestyle characteristics, it would need to be
shown that in each country: (1) there is a diet protective against lung cancer from ETS exposure,
(2) diet is inversely associated with ETS exposure, and (3) the association is strong enough to
produce the observed relationship between ETS and lung cancer. Diet may modify the magnitude
of any lung cancer risk from ETS (conceivably increase or decrease risk, depending on dietary
components), but that would not affect whether ETS is a lung carcinogen.
The literature on the effect of diet on lung cancer is not consistent or conclusive, but
taken altogether there may be a protective effect from a diet high in /?-carotene, vegetables, and
possibly fruits. Also there is some evidence that low consumption of these substances may
correlate with increased ETS exposure, although not necessarily for all study areas. The
calculations made by Marchand et al. (1991) and Butler (1990, 1991) are largely conjectural, being
5-22
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
based only on assumed data. Therefore, we examined the passive smoking studies themselves for
empirical evidence on the effect of diet and whether it may affect ETS results.
It was found that nine of the studies have data on diet, although only five of the them use
a form of analysis that assesses the impact of diet on the ETS association. None of those five
(CORR, HIRA[Coh], KALA, SHIM, and SVEN) found that diet made a significant difference. In
the four studies where data on diet were collected but not controlled for in the analysis of ETS,
three (GAO, KOO, and WUWI) are from East Asia and one (WU) is from the United States. Koo
(1988a), who found strong protective effects for a number of foods, has been one of the main
proponents of the idea that diet may explain the passive smoking lung cancer effect. To our
knowledge, however, she has not published a calculation examining that conjecture in her own
study where data were collected on ETS subjects. In WU, a protective effect of /3-carotene was
found, but the data include a high percentage of smokers (80% of the cases for adenocarcinoma,
86% for squamous cell) and the number of never-smokers is small.
The equivocal state of the literature regarding the effect of diet on lung cancer is also
apparent in the nine ETS studies that include dietary factors, summarized in Table 5-13. Note
that GAO found an adverse effect from /3-carotene and no one found it protective. HIRA and
KOO found opposite effects from fish while SHIM found no effect. Fruit was found to be
protective by KALA and KOO but adverse by SHIM and WUWI. Retinol (based on consumption
of eggs and dairy products) was found to be protective by KOO but adverse by GAO and by
WUWI.
In view of the results summarized in Tables 5-12 and 5-13, the actual data of ETS studies
do not support the suspicion that diet introduces a systematic bias in the ETS results. Indeed, it
would be difficult to show otherwise. Dietary intake is difficult to assess; dietary habits vary
within countries and enormously between countries, making it difficult to attribute any effect on
lung cancer to a particular food group; lifestyle characteristics and consumption of food and
beverage with possibly an adverse effect may be associated, either positively or negatively, with
the food group under consideration. The potential for bias and particularly for confounding is
high; an effect on lung cancer rate is probably small, requiring sizable samples and meticulous
design for reliable results; exposure to other known risk factors for lung cancer needs to be
carefully controlled. It would, of course, be helpful to identify dietary factors that may affect
lung cancer, positively or negatively, because that information could usefully contribute to public
health. To affect interpretation of ETS results, however, it would need to be established also that
5-23
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
consumption of the dietary factor of interest is highly correlated with ETS exposure in study
populations where ETS exposure is linked with increased incidence of lung cancer.
5.4.2.7. Summary on Potential Confounders
In summary, an examination of potential confounding factors finds no factor that can
explain the association between lung cancer and ETS exposure observed by independent
investigators across several countries that vary in social and cultural behavior, diet, and the
presence of other potential confounders. On the other hand, the high levels of indoor air pollution
from other sources (e.g., smokey coal) that occur in some parts of China and show statistical
associations with lung cancer in the studies of WUWI and LIU may mask any ETS effect in those
studies.
5.4.3. Potential Sources of Bias and Other Uncertainty
Some of the major areas contributing to study limitations and uncertainty are shown in
Table 5-14 (in two parts, 14A and 14B) with an indication of which studies may be affected.
Although each study has its own individual strengths and weaknesses, this table provides an
overview and index, of sorts, to the detailed reviews in Appendix A for the topics shown. The
table headings are broadly categorized under classification (selection and classification of subjects
and data), collection (sources and methods of data collection), and analysis (methods and topics
related to data analysis). The subheadings are described below. The likely direction of the bias is
indicated when apparent.
"ETS Subjects" refers to classification of candidates as nonsmokers, generally never-
smokers. A study is included under this heading if the restriction on prior smoking is not explicit,
is perhaps too lenient (e.g., may have smoked regularly for up to 6 months), or males are included
with females in the analysis (JANE only). Former smoking is a source of upward bias (i.e., away
from the null hypothesis). "ETS Exposure" refers to the criteria for classifying ETS subjects as
exposed or unexposed. Studies under this heading typically do not distinguish as sharply as they
might, resulting in downward bias (i.e., toward the null hypothesis). ("Errors of measurement in
the exposure variables .. . reduce the apparent risk, unless the errors are linked in some unusual
manner to confounding variables," Breslow and Day, 1980, p. 114).
Inclusion of misdiagnosed cases also produces a downward bias. Studies for which the
methods of diagnosis (or confirmation, if performed) of cell type are either not indicated or are
not designated exclusively by histology or cytology are listed under "Cases" in Table 5-14A. The
5-24
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
distribution of diagnosis methods is shown in Table 5-15. BROW, WU, and TRIG are included
under cases because they are restricted by cell type for which the bias is unknown.
As indicated above, the selection of the control group is one of the most important and
difficult tasks in a case-control study. Quoting Breslow and Day (1980), "There is no one type of
control group suitable for all studies andj it must be acknowledged, there are no firm criteria for
what is an acceptable group. . . . The characteristics and source of the case series must heavily
influence the type of control selected if comparability of the two series is to be achieved, that is,
if selection bias is to be avoided." The source of controls is shown in Table 5-3. Entries under
"Controls" in Table 5-14A may have questionable comparability of cases and controls or just not
be matched by design on any variables (see individual studies in Appendix A). In numerous
studies, the data for ETS subjects were drawn from a larger study that includes active smokers
and was usually matched on several variables (Table 5-3). If smoking habit was not a matching
variable, however, then the ETS data alone are not matched (indicated under "ETS sample
matched" in Table 5-3). Matching is a way of equalizing confounding variables, although they
still should be taken into account in the analysis. When matching is not implemented, then
confounding variables can only be controlled for in the analysis.
In Table 5-14A, "Representativeness" refers to whether the ETS subjects are reasonably
representative of the target population, which is the general public for purposes of this report.
Lack of representativeness is not always a negatiye characteristic, but it needs to be taken into
account. The implications are usually more relevant for characterizing population risk than for
hazard identification.
The headings under "Collection" in Table 5-14B are more self-explanatory. It is noted in
the table whether a "Self-Questionnaire" was used, which may be a source of bias of undetermined
direction. Among the studies under "Response and Follow-up," two (GENG and KATA) provide
no information. Good follow-up is essential to cohort studies, as indicated in the notes above.
Follow-up is lacking in some respect in three of the four cohort studies, all of which are in
westernized countries, although GARF(Coh) is the only large one. The direction of potential bias
is not evident.
"Proxy Response" in Table 5-14B identifies studies where information from surrogate
respondents, typically, next-of-kin, was used, as necessary, or no information was provided.
Subject response is generally considered more reliable than proxy response. Proxy response
percentages are shown for cases and controls in Table 5-3. The results of GARF might suggest
upward bias; those of JANE, downward. In JANE, type of respondent is a matching variable, and
5-25
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
controls are "healthy." In GARF, by contrast, proxy responses are from cases only (889 of cases),
and the controls are cancer patients themselves. Unfortunately, it is not clear what factors may
affect bias from proxy response, which direction the bias may take, or how serious it may be.
The "Analysis" classification in Table 5-14B refers to methods of data analysis. How were
unmarried women treated with regard to exposure to ETS, where the "Unmarrieds" heading refers
to women who are not currently married or have not been continuously married to the same
husband for an extended period? Studies vary in the degree to which exposure history is taken
into account, and some introduce assumptions regarding exposure of women who have never been
married (e.g., equivalent to being married but not exposed to ETS). A study is listed under
"Unmarrieds" if information is lacking or if unmarrieds are included with assumptions regarding
exposure. In several instances both apply or some information is lacking (e.g., LAMT treats single
women as unexposed to ETS, but it is not clear how exposure was handled for widows and
divorcees; 60 of the subjects in KOO are widows, but their distribution between cases and controls
is unknown). The "Unmarrieds" topic is closely related to the "ETS Subjects" and "ETS Exposure"
subheadings under "Classification" in Table 5-14A. The direction of bias need not be consistent
across entries in this category.
An adjusted analysis takes into account potential confounders. In particular, variables that
have been used for matching in the design should be incorporated in the analysis as potentially
confounding variables. This follows because the matching factors must be considered a priori as
ones for which stratification would be necessary, that is, as confounding variables (Breslow and
Day, 1980). We would expect matching on age, at least. Because the ETS data alone are not
matched on any variable in many studies (Table 5-3), a form of analysis that adjusts for potential
confounding appears particularly relevant. For the studies with both a crude and adjusted analysis
(Table 5-5), the adjustments have only a small to moderate effect in either direction. A study is
included under "Adjusted Analysis" in Table 5-14B if no adjusted analysis was conducted or if we
had a problem with the method used (details are in Appendix A). The three entries under "Trend
Analysis" in Table 5-14B are listed because the outcomes are of questionable interest, either due to
small sample size or because of the method or interpretation of data analysis.
It is concluded from the above discussion that there are no additional sources of bias (other
than smoker misclassification, which was adjusted for in Section 5.2) that would systematically
cause higher observed relative risk estimates. Therefore, this report concludes that the observed
association between ETS and lung cancer cannot be explained by bias. In fact, the association is
5-26
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
apparent despite the existence of the downward bias that results from nearly universal ETS
exposure, which has not yet been corrected for.
5.4.4. Potential Effects on Individual Studies
This section compares the utility of individual studies for determination whether there is
an association between ETS and lung cancer. For selected studies in Table 5-11, largely those
with power greater than 0.20, principal characteristics related to potential bias and confounding
are discussed. For studies with a significant outcome, indicated by boldface type in Table 5-11,
attention is focused on influences that would cause the relative risk to be overstated; otherwise,
influences toward unity are of greater interest. Independent of the quantitative results, the utility
of a few studies for ETS is questionable because of limited information, low quality, or
inadequate control for other (non-ETS) household risks of lung cancer (most prominently, indoor
smoke). The tier number assigned to a study on the basis of the critical analysis in Appendix A is
presented at the end of each discussion.
AKIB
Extensive use of proxy respondents (Table 5-3) and poor response rate may have led to
poor quality of exposure data and/or selective response. These could have contributed to the
observed association between spousal .smoking and lung cancer if (1) proxies of lung cancer cases
were more inclined to "remember" a history of spousal smoking than were those of controls or (2)
persons suspecting an ETS-lung cancer link were more likely to participate in the study. The
possible lack of subject-interviewee blinding potentiates both possibilities. With regard to (1),
however, type of informant was reported to have no significant effect on the results, and the
typical influence of poor exposure data is in the direction of no effect. ETS subjects are atomic
bomb survivors (Table 5-3), which affects representativeness but not hazard identification, to our
awareness. Finally, reliance on diagnoses based only on radiological or clinical evidence in nearly
one-half (43%) of the cases (Table 5-15) could have led to substantial misclassification of disease
status, probably creating a bias toward the null. (Tier 2)
CHAN
A number of factors may have had a bearing on the nonsignificant negative association
this study observed between ETS exposure at home or work and lung cancer. The measure of
ETS exposure utilized is nonspecific and subjective, lumping together home, work, childhood, and
5-27
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
adulthood exposure, which is determined primarily by adult at-home exposure. No histological or
cytological diagnosis was performed in 18% of cases, and there is no indication that secondary
tumors were excluded (Table 5-15). These characteristics contribute to potential error in disease
or exposure classification, both sources of downward bias. Orthopedic-ward controls may not
have been appropriate or comparable to cases, particularly insofar as the total control population
appeared to contain an elevated proportion of smokers compared to the general population in the
area; they might thus be more ETS exposed as well, thereby producing a negative bias. Treatment
of unmarrieds is unknown, and nearly one-half of the eligible cases were not included, raising the
possibility of selection bias; the probable direction of effect of these factors is unknown. Finally,
only a crude analysis is presented. Although cooking fuel, residence, and occupation did not
appear significantly associated with lung cancer, no attempts were made to control for potential
confounders. The uncertainty due to potential confounding, bias, and inadequate statistical
methods, along with neglect of basic epidemiologic principles needed for credibility, render this
study of little value for evaluation of ETS and lung cancer. (Tier 4)
CORK
Several potential biasing factors need to be considered with regard to the positive (but not
statistically significant) association between ETS and lung cancer noted in this study. The
comparability of cases and controls is uncertain; for the most part, the direction of effect of this
i
noncomparability of cases and controls is equally uncertain, but the fact that 15% of the controls
had cardiovascular disease would presumably lead to some downward bias in association. Not
considering former smoking status or duration of exposure in some spousal smoking analyses is
conducive to exposure misclassification, with consequent upward and downward bias,
respectively. Proxies were used more frequently for cases (24%) than for controls (11%), at least
in the parent study (Table 5-3), but exclusion of proxy respondents was reported to have no effect
on the results for spousal smoking. Treatment of widows and divorcees is unclear, and
misclassification of these women as "unexposed" despite smoking by former husbands, // it
occurred, could bias downward. More substantial potential to distort the results is presented by an
inadequate approach to confounding. While stratification on race or respondent type reportedly
had no effect on results, a dichotomizatiori into women over and under 60 years of age was the
only attempt to control for age or other factors. This marginal control for age and lack of control
for socioeconomic status (SES) and other potential confounders leaves open the possibility that
5-28
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
such factors may have significantly contributed to the observed association—either positively or
negatively. (Tier 3)
FONT
The positive associations between four separate sources of adult ETS exposure and lung
cancer, all of which reach statistical significance when restricted to adenocarcinoma, cannot be
readily explained by bias and confounding. This study was designed to investigate the ETS-lung
cancer relationship specifically and goes to great lengths to minimize potential bias and
confounding. Reliance on proxies for some subjects' exposure information apparently
introduced no problem, because exclusion of proxy responses reportedly did not alter the results.
ETS subjects may have smoked up to 6 months, however, and if that amount of smoking has an
affect on lung cancer risk, it could be a source of mild upward bias, depending on the number of
subjects with smoking histories. Diet, cooking and heating practices, and occupation were not
directly controlled for (such analyses are pending in an expanded version of this study), leaving
open the potential for confounding with either upward or downward effects. These factors,
however, might well co-vary with age, race, geographic area, income, and education, and results
were adjusted for this combination of variables (Table 5-5). In addition, adjusted results were
virtually independent of whether colon cancer or general population controls were used, providing
general evidence against selection or recall bias (Table 5-5). And although it remains possible that
lung cancer cases and their proxies tended to overestimate their ETS exposure relative to colon
cancer patients, it is unlikely that such recall bias would be specific to adenocarcinoma. (Tier 1)
GAO
The small nonsignificant association of lung cancer incidence with spousal smoking, and
the lack of association with total ETS exposure, may have been influenced by several factors.
First, only duration (not intensity of ETS exposure) was considered, and spousal exposure of less
than 20 years' duration was disregarded. Substantial misclassification of exposure is likely, a
source of downward bias. Second, ETS controls and cases were unmatched and their
comparability is unclear; the controls in the combined ETS and non-ETS population contained a
higher proportion of individuals in the oldest age group than did the cases, but analyses were
adjusted for age. Third, 19% of these cases were diagnosed solely by radiological or clinical
means (Table 5-15), increasing the likelihood of disease misclassification and the attendant bias
toward the null. Finally, spousal smoking (though not all-source ETS) analyses were also
5-29
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
adjusted for education, but there was no adjustment or other control for a number of Variables
(e.g., several cooking habits and previous respiratory diseases) found to be significantly and/or
more strongly associated than ETS with lung cancer in the data. Confounding by such factors
could have biased results in either direction. Thus, results were probably biased downward by
exposure classification problems and possibly also affected in unknown degrees by apparent noh-r
ETS risk factors. (Tier 3)
GARF
The main drawback of this study lies in its heavy reliance on proxy respondents (88% of
cases, unknown percentage of controls—Table 5-3). The statistically significant association
between high ETS exposure from spouse or related cohabitant and lung cancer disappears when
proxy respondents are excluded, but it is unclear whether this is the result of eliminating upward
bias operating among proxy respondents or simply a random consequence of eliminating most of
the study population. A systematic recall bias among proxy respondents is less likely given the
reported blinding of interviewers and subjects to the study hypothesis and the use of colorectal
cancer controls. Confounding due to dietary, heating, or cooking practices was not assessed and
thus could potentially have produced bias in either direction, but it is unclear why such bias
would have operated selectively on cases where offspring rather than patients themselves were
interviewed. (Tier 2) .•;•-.
GARF(Coh)
Although an unusually large study with good power, this cohort investigation suffers from
a number of limitations that in aggregate greatly mitigate its ability to detect a putative ETS-lung
cancer association and could account for its lack of significant findings. Data on exposure are
limited to husband's self-reported current smoking habits in 1959, and no information on former
spouses or other ETS sources was collected. Tremendous potential for distortion and
misclassification of relevant ETS exposure thus exists, providing a likely bias toward the null
hypothesis (it is possible that wives of smokers had a greater tendency to become active smokers
during follow-up, thus producing an increase in risk of lung cancer, but the authors state that
change in smoking status during follow-up was rare, so this effect, if extant, should be minor). In
addition, diagnostic confirmation and exclusion of secondary lung cancers was carried out for less
than one-third of the cases (the actual percentage was not stated), creating opportunities for
disease misclassification and another likely resultant bias toward the null. These and other factors
5-30
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
(see,review in Appendix A) could have reduced the observed association—which approached
statistical significance—below the nominal significance level and/or distorted the dose-response
pattern. (Tier 3)
GENG. • • - -. ..• . • . ; .••. :
The positive and statistically significant association, complete with clear dose-response
pattern, between spousal smoking and lung cancer is difficult to assess due to the dearth of details
supplied in the short published report. It is not clear that former personal smoking habits were
considered, for example, presenting a potential for upward bias arising from an association of
former smoking with spousal smoke exposure. Information on comparability of cases and
controls, response rates, utilization of proxies, and treatment of martial status are lacking,
precluding evaluation of these potential sources of error. No potential confounding factors are
taken into account, despite the observation of substantial associations between several non-ETS
exposures and lung cancer. Given the lack of design and methodological detail and absence of
efforts to address confounding, the influence of bias and confounding cannot be assessed beyond
noting that it may be substantial. (Tier 4) ,
HIRA(Coh) ;
The statistically significant positive association and trend for lung cancer with spousal
smoking and some other ETS exposures observed in the HIRA study cannot readily be explained
by bias or confounding. Classification based on self-reported status at baseline interview, without
regard to possible changes in status over time, could cause the RR to be overstated, but the bias
(expected effect) is in the other direction for error in exposure measurement. The use of death
certificates, with potential inaccuracies leading to misclassification of primary lung cancer, is also
a source of downward bias. The age adjustment is handled poorly in the analysis of cohort data,
utilizing the husband's rather than the subject's age, but that is corrected in the nested case-
control analyses, with similar results. More sophisticated methods of survival analysis could be
implemented, but that is unlikely to affect the evidence significantly.
Related observations of interest include the following: increased emphysema, asthma, and
paranasal sinus cancer in women exposed to spousal smoking. Among consumption of foods
(greenryellow vegetables, fish, meat, milk, and soybean paste soup), meat consumption is more
common among wives of smoking husbands. Consumption of fish is associated with higher lung
cancer incidence; meat consumption is suggestively associated with lower incidence. The analysis
5-31
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
for dietary habits controls for wife's age and husband's occupation but not for possible
confounding between food types. (Tier 2)
INOU
The odds ratio of 2.55 is quite high, even for Japan, and is based on a relatively small
number of subjects. Unfortunately, the sparse details provided regarding the study's design and
execution make assessment nearly impossible. For example, no data are provided on means of
diagnosis or confirmation, if any, and there is no indication that secondary cancers were excluded
(Table 5-15); reliance on proxy respondents may have reached 100% (Table 5-3), and there is no
indication of blinding or consideration of former (versus current) smoking status. These and other
potential problems could have substantially biased the results upward—or downward. The high
uncertainty associated with this study, in part due to very limited information about it, renders it
potentially misleading to the point that it may be preferable to omit it from the analysis. (Tier 4)
JANE
The results of this large and largely well-executed study are enigmatic. Overall, spousal
smoking is not positively associated with lung cancer, whereas all-cohabitant smoking is somewhat
associated and childhood smoking is substantially associated. Yet when observation is limited to
those subjects with at least some exposure, estimated lung cancer risk increases with increasing
exposure; further, while proxy respondents yield substantially negative overall spousal smoking-
lung cancer associations, nonproxies yield consistently more positive results. Confounding by
smoking habits of other cohabitants could have decreased the association seen for spousal smoking.
The study's treatment of unmarrieds is unknown and thus theoretically could have affected
results, such effects usually being a downward bias through exposure misclassification.
Supporting that possibility is the high prevalence of exposed controls (80%), suggesting that the
requirement for classification as ETS-exposed may be too lenient, a source of downward bias.
That aside, the only glaring flaw in the study's conduct is the pooling of male and female subjects,
with no consideration of gender in the analyses. It is clear that this could lead to a weaker
observed association than analyses restricted to females if smoking by the spouse is generally more
intense and/or contributes a greater proportion of total ETS exposure for women than for men, as
seems likely. But it is unclear why this would drive the risk estimates for spousal smoking, but
not for exposure from all-cohabitant or childhood ETS, below unity or produce the observed
5-32
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
response curve. Several speculative mechanisms are suggested in the more extensive study review
(see Appendix A), but the study's results cannot be clearly attributed to a negative bias. (Tier 2)
KALA
Interviewer bias is the only obvious potential "flaw" in this high-quality study that could
have substantially contributed to the observed positive association between spousal smoking and
lung cancer. Adjustment for interviewer, however, along with age and education (Table 5-5)
strengthened the observed association (unadjusted OR = 1.60, adjusted OR = 1.92), arguing
against interviewer bias as the source of the results.
Related observations of interest pertain to analyses evaluating 16 food groups.
Consumption of fruit appears to have a protective effect; retinol is marginal; no effect was found
for 0-carotene. (Tier 1)
KOO
This study's analyses of various measures of ETS exposure yielded predominantly positive
but uniformly nonsignificant associations with lung cancer. The study's modest sample size—and
hence modest power (0.43)—could have contributed to the failure to achieve statistical
significance, but other factors may have influenced the results as well. The complexity and
assumptions of several of the approaches used to quantify exposure are questionable (e.g.,
simultaneous exposures were not added and lifetime average exposure was estimated by dividing
current exposure level by age). Such approaches may have increased misclassification, probably
with a bias toward the null, and in combination with small numbers within exposure strata may
have contributed to the downward dose-response pattern over the range of ETS exposure (the
dose-response patterns observed appeared to be sensitive to the measure of exposure used).
Comparability of the cases and controls utilized in the ETS analyses is uncertain (unmatched; no,
demographics), so bias in either direction is possible. A number of potential confounders were
included in the analyses (e.g., age, place of residence, public versus private housing), so
significant confounding by standard age or SES-related factors is unlikely. There were no data
available on diet or cooking habits that have shown evidence of substantial though inconsistent
associations with lung cancer in some other studies conducted in Hong Kong. Bias in any
direction could potentially have arisen from this source, but adjustment for age, education, and
residence—factors probably co-varying to some degree with diet and cooking habits—resulted in
5-33
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
stronger ETS-lung cancer associations, thus hinting that potential confounding by diet or cooking
would probably introduce a downward bias, if any. (Tier 2)
LAMT
Assessment of the statistically significant positive association between spousal smoke
exposure and lung cancer observed in this study includes several potential sources of bias and
confounding. Although cases and controls appeared highly comparable as initially assembled, they
were not matched on smoking, and thus the nonsmokers used in the ETS analysis may not have
been as comparable, leading to potential bias of indeterminate direction. While exclusion of single
women reportedly did not alter results, treatment of widows and divorcees was unclear and thus
could have potentially led to exposure misclassification and another indeterminate bias. Finally,
and most important, no attempt to control for major potential confounders, including the
fundamental factor of age, was undertaken (except for the gender restriction). Often, adjustment
for such factors has little effect on results, but such an outcome cannot simply be assumed, and
thus the potential for significant upward (or downward) shifts in association due to confounding
exists. (Tier 2)
LAMW
Despite only modest power (0.39), this study found a statistically significant association
between ETS exposure and lung cancer. Comparability of cases and controls is unclear because
the ETS population was unmatched and not demographically characterized, leaving bias of
indeterminate direction. Similarly unpredictable but potentially more important is the effect of
possible confounding factors that were not addressed in the analyses. Neither kerosene fume nor
incense exposure was significantly associated with lung cancer, but no other potential
confounders—-including age—were investigated. Thus, strong effects in either direction due to
confounding by basic risk factors cannot be ruled out, although the potential effects of other
possible sources of bias are preponderantly negative. (Tier 3)
LIU
The small but not statistically significant negative association between passive smoking and
lung cancer observed may have been shaped by bias and confounding. Uncertain case-control
comparability, scarcity of histopathological diagnosis, lack of diagnostic verification and exclusion
of secondary tumors (Table 5-15), and a nonspecific and nonquantitative ETS measure probably
5-34
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
reduce accuracy of disease and exposure assessment and bias results toward the null. Lack of
control for nearly all relevant potential confounders could distort the observed association in any
direction. The ubiquitous exposure to smokey coal combustion products, which often reaches
extreme levels in the study area, makes evaluation of a comparatively minor exposure like ETS
particularly problematic—even for active smoking, only a weak association of ever-versus never-
smoking with lung cancer is detected. Due to these problems, this study's findings regarding ETS
are not very meaningful. (Tier 4)
PERS
It appears likely that the potential sources of upward bias actually have little effect,
leaving several possible sources of downward bias. Although data on smoking status in 1963-64
were collected directly from subjects, data on ETS were derived from follow-up questionnaires
distributed in 1984. Because only cases diagnosed by 1980 were included in the study, proxy
respondents must have been used for most lung cancer cases, although no actual numbers are
supplied. The nearly identical results obtained for controls matched on vital status (and hence
requiring proxy respondents) and for nonstatus-matched controls indicate that a systematic bias in
responses of proxies versus nonproxies was not responsible for the observed results. Preferential
recall of ETS exposure by relatives of lung cancer cases remains a possibility, particularly as
controls were drawn from the general population. Observed associations were specific for
squamous and small cell cancer only, however, while recall bias would be expected to affect all
lung cancer types equally. Assessment of exposure was compromised by basing spousal exposure
of remarried women on only their longest marriage and classifying all unmarried women as
unexposed, sources of downward bias. Reliance on self-administered questionnaires may also
reduce data accuracy, also a downward bias. It is reported that occupation, radon exposure, and
urban location were not important factors. (Tier 2)
SHIM
Downward bias may be a factor. The inconsistent although largely positive associations
noted between various sources of ETS exposure and lung cancer in this study may have been
affected by several sources of bias and confounding. Lack of consideration of tobacco products
other than cigarettes, no differentiation between past and current exposures, and an unstated (and
thus possibly imprecise) definition of exposure may have led to misclassification of exposure
status and consequent downward bias. Failure to exclude smoking-related diseases from the
5-35
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
hospital control group and use of a self-administered questionnaire may also contribute to
downward bias. There was no control for age or other potential confounders such as marital
status, nor for occupational metal exposure, heating fuel type, or medical history of silicosis,
despite observation of substantial though not statistically significant associations of these factors
with lung cancer. The direction of bias possibly introduced by these uncontrolled factors cannot
be determined.
Results of related interest include the findings that occupational exposure to iron or other
metals is significantly associated with lung cancer (OR = 4.8); for exposure to coal, stone, cement,
asbestos, or ceramics, the OR is 3.3, but it is not statistically significant. No effect was found for
the eight food groups evaluated. A personal medical history of silicosis is suggestive (OR = 2); a
history of chronic bronchitis, asthma, or tuberculosis is not. Recent use of a kerosene or coal
(charcoal) stove for household heating may be of interest (OR =1.6 and 1.7, respectively). It is
clear that some variables were adjusted for others, in particular, the group classifications of
sources of ETS in the home. With so many potential non-ETS factors addressed, however, it
would appear unlikely that all possible confounders could be controlled. (Tier 3)
SOBU
Upward bias or confounding stemming from the sources of uncertainty in this study are a
possibility. Cases and controls are unmatched, with apparent differences that may affect
comparability. Controls tend to be younger and more educated—variables that are taken into
account by an adjusted analysis. Education and age, however, may reflect differences in lifestyle
and socioeconomic status that could be biasing. In this study, exposure from spousal smoking
alone is analyzed separately from smoking only from other cohabitants, with the latter significant
and the former nonsignificant (see Table 5-5; the results for the latter appear in Table 5-11). It is
possible that the higher OR for smoking cohabitants could be confounded by some factor related
to lifestyle, although it is not clear how. The results for ETS need to be adjusted for use of wood
and straw as cooking materials, a possible risk factor for lung cancer (OR = 1.9). Possible
inaccuracies in exposure and other risk factor assessments resulting from use of a self-
administered questionnaire are further features that may have compromised results. That might
increase nondifferential misclassification, however, biasing the relative, risk toward unity.
Related observations of interest include the following: significant association with lung
cancer for women who had used wood or straw as cooking fuels at age 30 (OR = 1.9). Other
5-36
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
sources of indoor heating (by gas, kerosene, coal, charcoal, and wood stoves without chimneys)
were not significant nor was the use of charcoal foot warmers. (Tier 3)
SVEN
The small size of the nonsmoking female population in this study leads to nonsignificant
results even for the OR of 2.1 observed for women exposed to ETS both at home and at work. The
OR for exposure at home or work, but not both, is only 1.2. These results could thus be attributed
to chance, but the dose-response pattern argues against this interpretation. The nonquantitative
and rather rough measures of exposure used would most likely have lessened the observed
association rather than biasing the ORs upward. All interviews with cases were face to face, but
42% of the controls were interviewed by telephone. If persons interviewed by telephone were less
likely to report or recall ETS exposure than those interviewed face to face, this could explain the
observed association. The researchers report that results were similar regardless of whether
hospital or general population controls were used, however, and presumably the telephone
interviews were predominantly those for general population controls. Although treatment of
unmarried subjects is not mentioned, bias toward the null due to exposure misclassification from
this source is unlikely because no analyses limited to spousal exposure were conducted. (Tier 2)
TRIC
The statistically significant positive association between spousal smoking and lung cancer
observed in this study is not readily attributable to general sources of bias and confounding. The
inclusion of smokers who quit over 20 years before the study tends to bias results toward no
association, unless occurrence of a strong correlation between such former smoking and having a
smoking spouse contributed an upward bias. The frequency of diagnoses based only on
radiological and/or clinical means, coupled with lack of diagnostic confirmation or exclusion of
secondary tumors (Table 5-15), potentiates disease misclassification and thereby bias toward the
null. Possible lack of blinding could have led to some (upward) interviewer bias. And while
standard demographic factors associated with lung cancer were not associated with spousal
smoking, diet and cooking and heating practices were not specifically addressed, leaving some
potential for confounding (either positive or negative) by these factors. A constellation of positive
biases sufficient to offset their negative counterparts and produce associations of the magnitude
(e.g., ORs in excess of 2) observed, however, seems unlikely. (Tier 2)
5-37
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
wu
This study found several nonsignificant positive associations between ETS exposure and
lung cancer. Cases were restricted to adenocarcinoma and squamous cell carcinoma, but because
these are the predominant cell types anyway, it should not have significantly compromised the
results. Most analyses employed adjustment for personal smoking habits within a population of
smokers and nonsmokers; comparability of cases and controls within this population appeared
excellent. The comparability of nonsmoking cases and controls is unclear; a lack of comparability
in key areas (e.g., age) could have influenced results. Both smoking-adjusted and never-smoker-
restricted approaches yielded positive associations, however. A very high proportion (44%) of
identified cases were not included in the study, but this was largely due to avoidance of proxy
respondents. Excluded cases were demographically similar to those included, arguing against a
selection bias. Handling of former marriages was not described, leaving open a possibility for
upward or downward bias from this source. The matched-pairs analysis by its nature eliminated
the possibility of confounding due to age or neighborhood. Because the analysis attempted to
adjust for the effect of active smokers, instead of removing the data on active smokers, isolation
of the effects of ETS on lung cancer from those of personal smoking habits is subject to question.
An analysis restricted to nonsmokers was unmatched and thus subject to potential confounding by
age and/or neighborhood. Both approaches produced similar results, however, making such
confounding less likely.
More problematic is the failure to control for dairy product and egg intake despite its
substantial association with lung cancer. The nature of the connection (if any) between this factor
and ETS exposure, and hence the magnitude and direction of bias it would introduce, is not clear.
No clear attribution of the results to any of the above sources is possible, because their potential
effects could have been in any direction. In analyses of non-ETS factors, childhood pneumonia is
significantly associated with lung cancer (only cases of adenocarcinoma and squamous cell
carcinoma were included in the study), but history of six other lung diseases were not. Also
significant is heating or cooking with coal during the preadult years and diets low in 0-carotene,
dairy products, or eggs. No significant associations were noted for vitamin A intake or for
occupation. (Note: It is clear that some variables were adjusted for others, but with so many
potential non-ETS factors addressed, it would appear unlikely that all possible confounders could
be controlled. Multiple comparisons may cause some results to appear more significant than
warranted.) (Tier 2)
5-38
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
WUWI
Despite excellent power, this study found a slightly negative association between ETS
exposure and lung cancer. Classification of subjects with histories of up to 6 months of active
smoking as never-smokers may be a source of upward bias, while the lack of histological diagnosis
for 26% of the cases (Table 5-15) may have increased disease misclassification, a source of
downward bias. Controls were rather loosely matched (and only on age) to cases, so comparability
could have been a problem of indeterminate direction of effect. This potential is mitigated by
adjustment for age^ education, and area in the analyses, which also largely controls for
confounding by key demographic-associated factors. The factor that makes this study of little
value for ETS exposure, however, is the presence of other household factors found to be
substantially associated with lung cancer, indoor smoke from nontobacco sources in particular. If
ETS is a risk factor for lung cancer, its relative risk is reasonably small and unlikely to be detected
against a background of competing and probably stronger exposures (e.g., indoor coal combustion
products). The authors note the limitation of their exploratory study for inference on ETS and
lung cancer: "Perhaps in this study population the effect of environmental tobacco smoke was
obscured by the rather heavy exposures to pollutants from coal-burning Kang, other indoor
heating sources, and high levels of neighborhood air pollution." Although it is informative vis-a-
vis its principal objectives, this study is not very useful for assessing ETS exposure and lung
cancer. (Tier 4)
5.4.5. Analysis by Tier and Country
The assignment of studies to tiers is summarized in Table 5-16. Tier 1 contains the studies
judged to be of highest utility for addressing the potential relationship of ETS and lung cancer,
based on the material in the last sections and the detailed reviews in Appendix A. Overall, only
three studies are in the highest tier, while 12, 10, and 3 studies are in Tiers 2, 3, and 4,
respectively. Studies in Tier 4 are not recommended for, the reasons described in the previous
section. The statistical weight for Tiers 1 and 2 pooled together for each country is shown in
Table 5-17 as a percentage of the total for corresponding tiers over all countries, except for China
where the weight for Tier 3 is used in the absence of studies in Tiers 1 and 2. Emphasis on
studies through Tier 2 is somewhat arbitrary—aside from the United States, the power of Tier 1
alone is too small. GAO is the only study in China that was not placed in Tier 4, but there is little
basis to assume that this single study from Shanghai should be "representative" of a vast country
like China.
5-39
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 5-17 presents adjusted relative risk estimates, 90% confidence intervals, and
significance levels (one-sided) from studies pooled by country and by tier. The results of using
only the higher utility studies (Tier 1 and Tier 2) are generally similar to those generated by using
all studies, with the higher ranked studies yielding slightly higher estimates of relative risk for
five of the six country groups. The pooled estimates from the higher utility studies are all greater
than one and are statistically significant for four of the five higher tiered (Tiers 1 and 2) country
groups—Greece, Hong Kong, Japan, and the United States, in that order. The pooled results from
the three Western European Tier 1 and 2 studies show about the same association as those of the
six U.S. studies but, with less power, are not statistically significant.
Analysis by tiers provides a methodology for weighing studies both qualitatively and
quantitatively. Qualitatively, it allows one to emphasize the better designed and conducted
studies, thought to provide better data for analysis of an ETS effect. The addition of studies of
lower utility to the analysis, however, has only a small effect. In view of that outcome, and the
results and discussion in Sections 5.4.2 and 5.4.3, this analysis indicates that confounding and bias
in these studies have little effect on the overall results. In summary, it is concluded that the
association of ETS and lung cancer observed from the analysis of the 31 epidemiology studies in
eight different countries is not due to chance alone and is not attributable to bias or confounding.
5.5. CONCLUSIONS FOR HAZARD IDENTIFICATION
5.5.1. Criteria for Causality
According to EPA's Guidelines for Carcinogen Risk Assessment (U.S. EPA, 1986a), a
Group A (known human) carcinogen designation is used "when there is sufficient evidence from
epidemiologic studies to support a causal association between exposure to the agents and cancer."
The Guidelines establish "three criteria (that) must be met before a causal association can be
inferred between exposure and cancer in humans:
1. There is no identified bias that could explain the association.
2. The possibility of confounding has been considered and ruled out as explaining the
association.
3. The association is unlikely to be due to chance."
As indicated in Section 5.3, the overall results observed in the 31 epidemiology studies are highly
unlikely to be due to chance. The discussion and analyses in Section 5.4 conclude that the
association cannot be explained by bias or confounding either.
5-40
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
In addition, the evidence for a causal association between ETS and lung cancer is
evaluated according to seven specific criteria for causality developed by an EPA workshop to
supplement the Guidelines (U.S. EPA, 1989). These criteria are similar to the original and classical
recommendations of Hill (1953, 1965). The seven recommended (but not official) criteria from
the EPA workshop, which vary between essential and desirable, are listed below (U.S. EPA, 1989).
A causal interpretation is enhanced for studies to the extent that they meet the
criteria described below. None of these actually establishes causality; actual proof
is rarely attainable when dealing with environmental carcinogens. The absence of
any one or even several of the others does not prevent a causal interpretation.
Only the first criterion (temporal relationship) is essential to a causal relationship:
with that exception, none of the criteria should be considered as either necessary or
sufficient in itself. The first six criteria apply to an individual study. The last
criterion (coherence) applies to a consideration of all evidence in the entire body of
knowledge.
1. Temporal relationship: The disease occurs within a biologically reasonable
time frame after the initial exposure to account for the specific health
effect.
2. Consistency: When compared to several independent studies of a similar
exposure in different populations, the study in question demonstrates a
similar association which persists despite differing circumstances. This
usually constitutes strong evidence for a causal interpretation (assuming the
same bias or confounding is not also duplicated across studies).
3. Strength of association: The greater the estimate of risk and the more
precise, the more credible the causal association.
4. Dose-response or biologic gradient: An increase in the measure of effect is
correlated positively with an increase in the exposure or estimated dose. If
present, this characteristic should be weighted heavily in considering
causality. However, the absence of a dose-response relationship should not
be construed by itself as evidence of a lack of a causal relationship.
5. Specificity of the association: In the study in question, if a single exposure
is associated with an excess risk of one or more cancers also found in other
studies, it increases the likelihood of a causal interpretation.
6. Biological plausibility: The association makes sense in terms of biological
knowledge. Information from toxicology, pharmacokinetics, genotoxicity,
and in vitro studies should be considered.
7. Coherence: Coherence exists when a cause-and-effect interpretation is in
logical agreement with what is known about the natural history and biology
of the disease. A proposed association that conflicted with existing
knowledge would have to be examined with particular care. (This criterion
has been called "collateral evidence" previously.)
5-41
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
5.5.2. Assessment of Causality
We consider the extent to which the criteria for causality are satisfied for the ETS studies.
Regarding temporal relationship, ETS exposure classification is typically based on the marital
history of a subject, which varies, or on the status at the beginning of a prospective cohort study.
There are seven exceptions where exposure classification is based only on current status or
duration of exposure appears not to have been taken into account (see reviews of studies entered
under "ETS Exposure" in Table 5-14). None of the seven (BUFF, CHAN, GAO, INOU, KABA,
LEE, and SHIM) are in Tier 1 or 2. This criterion appears to be adequately satisfied for the
studies in Tiers 1 and 2.
If ETS causes lung cancer, then the true relative risk is small for detection by
epidemiologic standards and may differ between countries as well. However, by considering the
totality of the evidence, it is determined that the large accumulation of epidemiologic evidence
from independent sources in different locales and circumstances, under actual exposure
conditions, is adequate for conclusiveness. Having accpunted for variable study size, adjusted for
a possible systematic spousal bias due to smoker misclassification, and considered potential bias,
confounding, and other sources of uncertainty on a study-by-study basis, consistency of a
significant association is clearly evident for the summary statistical measures for Tiers 1 and 2 in
Greece, Japan, Hong Kong, and the United States. The combined countries from Western Europe
are similar in outcome for the United States, although significance is not attained. There is too
much obscurity and uncertainty attached to the studies in China for adequate data interpretation.
The relative risks for each country are obtained by pooling estimates from the
epidemiologic studies conducted in the country. The strength of observation is limited by the true
value of the relative risk, which is small. Statistical significance is attained, however, for the
pooled studies of the United States and most other countries. The data were obtained from actual
conditions of environmental exposure so imprecision is not increased by extrapolation of results
from atypically high exposure concentrations, a common situation in risk analysis. Additionally,
all studies were individually corrected for systematic bias from smoker misclassification at the
outset, and qualitative characteristics of the studies were carefully reviewed to emphasize the
results from the better studies. The outcome for the United States is heavily influenced by the
large NCI study (FONT) that was specifically designed and executed to avoid methodologic
problems that might undermine the accuracy or precision of the results.
Of the 14 studies reporting test for upward trend, 8 are statistically significant at 0.05
(nine at 0.06). The outcome is similar for studies in Tiers 1 and 2 only—6 of 10 are significant,
5-42
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
which would occur by chance alone with probability less than 0.0001. In four of those sii
significant studies, the observed RR values increase monotonically with dose (i.e., the RR
increases with each increase in exposure level). This evidence of dose-response is very supportive
of a causal interpretation because it would be an unlikely result of any operative sources of bias or
confounding. .
Specificity does not apply to ETS. Although ETS has been assessed for the same endpoint
(lung cancer) in all studies, the occurrence of lung cancer is not specific to ETS exposure. Data
on histological cell type are not conclusive. The study by Fontham and colleagues (1991) suggests
that adenocarcinoma may be more strongly related to ETS exposure than other cell types.
Adenocarcinoma, however, does not appear to be etiologically specific to ETS.
Biomarkers such as cotinine/creatinine levels clearly indicate that ETS is taken up by the
lungs of nonsmokers (see Chapter 3). The similarity of carcinogens identified in sidestream and
mainstream smoke, along with the established causal relationship between lung cancer and
smoking in humans with high relative risks and dose-response relationships in four different lung
cell types down to low exposure levels, provide biological plausibility that ETS is also a lung
carcinogen (Chapter 4). In addition, animal models and genotoxicity assays provide corroborating
evidence for the carcinogenic potential of ETS (Chapter 4). The epidemiologic data provide
independent empirical verification of the anticipated risk of lung cancer from passive smoking
and also an estimate of the increased risk of lung cancer to never-smoking women. The coherence
of results from these three approaches and the lack of significant arguments to the contrary
strongly support causality as an explanation of the observed association between ETS exposure and
lung cancer.
5.5.3. Conclusion for Hazard Identification
Based on the assessment of all the evidence considered in Chapters 3, 4, and 5 of this
report and in accordance with the EPA .Guidelines and the causality criteria above for
interpretation of human data, this report concludes that ETS is a Group A human carcinogen, the
EPA classification "used only when there is sufficient evidence from epidemiologic studies to
support a causal association between exposure to the agents and cancer" (U.S. EPA, 1986a).
5-43
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 5-1. Epidemiologic studies on ETS and lung cancer in this report
Study1
AKIB2
BROW3
BUFF3
CHAN<
CORRj
FONT,
GAO3
GARF2
GENG4
HUMB2
INOU4
JANE2
KABA2
KALA,
KATA2
KOO2
LAMT2
LAMW3
LEE3
LIU,
PERS2
SHIM2
Country
Japan
USA
USA
Hong Kong
USA
USA
China
USA
China
USA
Japan
USA
USA
Greece
Japan
Hong Kong
Hong Kong
Hong Kong
England
China
Sweden
Japan
"^Within country
Hiroshima
Colorado
Texas
Louisiana
Five metro areas
Shanghai
New Jersey, Ohio
Tianjin
New Mexico
Kanajawa
New York
New York
Athens
Xuanwei
Nagoya
,.
References
Akiba et al. (1986)
Brownson et al. (1987)
Buffler et al. (1984)
Chan and Fung (1982)
Correa et al. (1983)
Fonthametal. (1991)
Gao et al. (1987)
Garfinkel et al. (1985)
Gengetal. (1988)
Humble et al. (1987)
Inoue and Hirayama (1988)
Janerich et al. (1990)
Kabat and Wynder (1984)
Kalandidi et al. (1991)
Katadaetal. (1988)
Koo et al. (1987)
Lam et al. (1987)
Lam (1985)
Lee etal. (1986)
Liu etal. (1991)
Pershagen et al. (1987)
Shimizu et al. (1988)
5-44
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 5-1. (continued)
Study1
SOBU3
WU2
WUWI4
BUTL(Coh)2
GARF(Coh)3
HIRA(Coh)2
HOLE(Coh),
Country
Japan
USA,
China
USA
USA
Japan
Scotland
Within country
Osaka '
California
California
Paisley Renfrew
References
Sobue(1990)
Wuetal. (1985)
Wu- Williams and Samet
(1990)
Butler (1988)
Garfinkel (1981)
Hirayama (1984)
Hole et al. (1989)
1 Subscripts refer to this report's ratings of studies for utility of studying the association of ETS
and lung cancer, where "1" is highest. Studies with"4" are judged to be of little value. The
ratings are subjective, based on qualitative features described in the reviews (Appendix A). It
was not possible to describe meaningful rule to follow in setting a rating. Nevertheless, it is
useful to have some indication of which studies have taken greater care to avoid or control for
potential bias and confounding. The ratings are carried as subscripts throughout the text so the
reader can replace them with his own opinions if desired. The ratings are not intended as
numerical weighting factors.
2 KATA has no tier number because the odds ratio cannot be calculated.
5-45
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 5-2. Studies by location, time, size, and ETS exposure
Country
Greece
Greece
Hong Kong
Hong Kong
Hong Kong
Hong Kong
Japan
Japan
Japan
Japan
Japan
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
Study
KALA
TRIG
CHAN
KOO
LAMT
LAMW
AKIB
HIRA(Coh)
INOU
SHIM
SOBU
BROW
BUFF
BUTL(Coh)
CORR
FONT
GARF
GARF(Coh)
HUMB
JANE
KABA
WU
Accrual*
~ Period
1987-89
1978-80
1976-77
1981-83
1983-86
1981-84
1971-80
1965-81
1973-83
1982-85
1986-88
1979-82
1976-80
1976-82
1979-82
1985-88
1971-81
1959-72
1980-84
1982-84
1961-80
1981-82
Cases
90
40
84
86
199
604
94
22
90
144
19
41
22
420
134
20
191
24
29s
Size1
Controls
116
149
139
136
335
1444
270
47
163
731
47
196
133
7806
402
176,739 -
162
191
25
62*
ETS
Cases
71
73
60
59
58
624
78
82
58
56
21
80
64
70
67
75
777
54
*
Exoosurfe (Wl*
Controls
60
52
53
49
45
444
70
1f\
64
56
54
15
84
46
636
61
56
807
60
*
5-46
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 5-2. (continued)
Country
W. Eurooe
Scotland
England
Sweden
Sweden
China
China
China
China
Study
HOLE(Coh)
LEE
PERS
SVEN
GAO
GENG
LIU
WUWI
Accrual*
Period
1972-85
1979-82
1961-80
1983-85
1984-86
1983
1985-86
1985-87
Cases
32
67
34
246
54
54
417
"Size2"
Controls
1*7Q A
,/o4
66
*
174
375
93
202
602
ETS
Cases
69
49
71
77
63
83
49
Exposure (%Y...
Controls
71
/ J
68
*
66
74
44
87
55
1 Time during which cases occurred.
2 Number of subjects included in ETS analyses; where numbers differ for spousal smoking and
other exposures, those for spousal smoking are given.
3 Spousal smoking unless otherwise noted.
4 Adenocarcinoma only. Data for all cell types was only available for general passive smoke
exposure, which showed 77% of 75 cases and 56% of 144 controls exposed.
5 Figure pertains to "spouse pairs" cohort, which is of principal interest regarding ETS; a
subgroup of this cohort comprised the "ASHMOG" cohort.
* Figure is for population controls; study also included 351 colon cancer controls (66% exposed).
7 General ETS exposure; ORs but no exposure prevalences presented for spousal smoking.
8 Adenocarcinoma only. Analyses for other cell types included smokers while adjusting for
smoking status.
* Data not available.
5-47
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 5-3. Case-control studies of ETS: characteristics
Percent proxy x;
response1
Study Ca Co , >'
AKIB 90 88
BROW 69 39
BUFF 82 76
CHAN * *
CORR * *
FONT 34 0-
1025
GAO 0 *
GARF 88 *
GENG * *
HUMB * *
INOU 100 100
JANE 3319 3319
Female age*
Ca
70.2
35-95
66.3
30-79
39-70
*
20-79
35-69
!>40
£65
£85
*
67.119
, Source of
Co controls -
* Atomic bomb
* survivor
population
68.2 Cancer cases4
30-79 Cancer cases*
39-70 Orthopedic
patients
* Hospital
patients8
20-79 Cancer cases;
general
population
35-69 General
population
S40 Cancer cases9
£65 *
£85 General
population
* Cerebrovas-
cular disease
deaths
68.1 19 New York
State Dept. of
Motor
Vehicles
Matched
variables
Age, sex,
residence,
med. exam
participation7
Age, sex
Age, sex
Matched but
variables
unspecified
Age (± 5),
sex, race
Age, (for
cancer
controls) race
Age (± 5)
Age (± 5),
hospital
Age (±2),
sex, race,
marital status
Age (± 10),
sex, ethnicity
Age, year
of death
(± 2.5),
district
Age, sex,
county,
smoking
history
" "saniple
matched
Yes
No3
No3
No3
No3
Yes
No3
Yes
No3
No3
No3
Yes
5-48
05/15/92
-------
Table 5-3. (continued)
DRAFT—DO NOT QUOTE OR CITE
Percent proxy
response1
Study Ca 'Co
KABA * *
KALA 0 0
KATA 0 0
KOO * *
LAMT * *
LAMW * *
LEE 3813 38
LIU 0 0
PERS *ls *
SHIM * *
SOBU 0 0
SVEN 0 0
TRIG * *
Female age2
Ca Co'
61.6 53.9
;>35 s»35
67.8 *
* *
* *
67.5 66
35-74 35-74
52 52
*16 *
59 58
35-81 35-81
60 56
66.3
62.8 62.3
Source of
controls
Patients10
Orthopedic
Patients
Non-cancer
patients
"Healthy"11
"Healthy"12
Hospitalized
orthopedic
patients
Patients14
General
population?
*17
Patients18
Patients
General
population
Hospitalized
orthopedic
patients
Matched
variables
Age (± 5),
sex, race,
hospital
Sex
Age (±2),
sex
Age (± 5),
residence,
housing
Age (±5),
residence
Age, socio-
economic
status,
residence22
Age, sex,
hospital
location, time
of interview
Age (± 2),
sex, village
Age (± 1),
sex
Age (± 1),
hospital,
admission
date
None
Age
Age,
occupation,
education22
BTS
sample
matched
Yes
Yes
Yes
No3
No3 •
No3
• No3-5
Yes
Yes
Yes
No
No3
No3
5-49
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 5-3. (continued)
Study
WU
WUWI
Percent proxy
response1
Ca Co
* *
0 0
Female
Ca
<76
55.920
age2
Co
<76
55.420
,, .,
Source of
controls
Neighbor-
hood12
General
population
' •* % r -.
Matched
variables
Age (± 5),
sex, race
Sex, age21
'" ETS
sample
, matched
No3
No3
10
11
12
13
14
15
16
17
18
19
20
21
22
23
"Ca" and "Co" stand for "cases" and "controls," respectively.
Single values are the average or median. Paired values are the range.
Not matched on personal smoking status (e.g., smoker/nonsmoker).
Persons with cancers of bone marrow or colon in Colorado Control Cancer Registry.
Ongoing study modified for passive smoking.
Population-based and decedent comparison subjects selected from state and Federal records.
Participation in RERF biennial medical examination program.
Assorted ailments.
Colorectal cancer.
Diseases not related to smoking.
Selected from a healthy population.
Living in neighborhood of matched case.
Applies only to the 143 patients in the followup study.
Excluding lung cancer, chronic bronchitis, ischemic heart disease, and stroke.
No overall percentages given.
Two control groups: 15 to 65 and 35 to 85 for both cases and controls in groups 1 and 2,
respectively.
Two control groups were randomly chosen from the cohort under study.
Patients in the same or adjacent wards with other diseases.
Includes males and females and long-term ex-smokers.
Entire study population, including smokers.
Frequency matched by 5-year age group to age distribution of cases reported in study area 2
years prior to initiation of study.
"Similar" but not actually matched.
0% for general population and 10% for colon cancer controls.
* Data not available.
5-50
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 5-4. Estimated relative risk of lung cancer from spousal ETS by epidemiologic study
(crude and adjusted for cofactors)
Case-eoatrol
AKIB
BROW
BUFF
CHAN
CORR
FONT26
GAO
GARF
GENG
HIRA6
HUMB
INOU
JANE
Never-Smokers
Crude RR!>" ., ^
1.52
(0.96, 2.41)
1.52"
(0.49, 4.79)
-1.82"
(0.45, 7.36)5
0.81n
(0.39, 1.66)
0.75
(0.48, 1.19)
2.0719
(0.94, 4.52)
1.37
(1.10, 1.69)
1.21
(0.94, 1.56)
1,32
(1.08, 1.61)
1.19
(0.87, 1.63)
1.31
(0.93, 1.85)
2.16
(1.21,3.84)
1.533
(1.10, 2.13)
2.34
(0.96, 5.69)
2.55"
(0.90, 7.20)
0.86
(0.57, 1.29)
-
Adj.RR1'3^
1.5
(1.0, 2.5)
*
1.6818
(0.39, 6.90)5
*
*
*
1.29
(1.03, 1.62)
1.28
(0.98, 1.66)
*
1.343'4
1.7021
(0.98, 2.94)5
*
1.643
*
2.2
(0.9, 5.5)
2.543-7
*
0.93/0.448
5-51
05/15/92
-------
Table 5-4. (continued)
DRAFT—DO NOT QUOTE OR CITE
Case-control
KABA*2
KALA
KATA
KOO
LAMT
LAMW
LEE
LIU
PERS
SHIM
SOBU
SVEN
TRIG
WU
WUWI
BUTL
(Coh)
-
Crude RRU)4
0.79
(0.30, 2.04)
1.629
(0.99, 2.65)
1.41
(0.78, 2.55)
*16
1.55
(0.98, 2.44)
1.65
(1.22,2.22)
2.51"
(1.49, 4.23)
1.03
(0.48, 2.20)
0.74
(0.37, 1.48)
1.28
(0.82, 1.98)
1.08*
(0.70, 1.68)
1.069
(0.79, 1.44)
1.77
(1.29,2.43)
1.26U
(0.65, 2.48)
2.08M
(1.31, 3.29)
1.41*
(0.63, 3.15)
0.79
(0.64, 0.98)
2.4S24
Never-Smokers
Adj. RR>'V,
*
1.92
(1.02, 3.59)s
*
. : *
1.64
*
*
0.75/1. 6010
0.77
(0.35,1.68)
1.2
(0.7, 2.1)s
*
1.139
(0.78, 1.63)5
1.57
(1.07, 2.31)5
1.4"
*
1.2
(0.6, 2.5)5
0.7
2.02
(0.48, 8.56)s
5-52
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 5-4. (continued)
Case-control ,
GARF
(Coh)
HIRA
(Coh)
HOLE25
(Coh)
••
f^^v+jt .** TE^1 tJ ** W"
' LffUue JK.K,
«
1.38
(1.03, 1.87)
2.27
(0.40, 12.7)
Never-Smokers
Adj.RR1**1*
1.173
(0.85, 1.61)5
1.61
1.99
(0.24, 16.7)5
1 Parentheses contain 90% confidence limits, unless noted otherwise. When not represented in
the original studies, the crude ORs and their confidence limits were calculated (or verified) by
the reviewers wherever possible. Boldface indicates values used for analysis in text of this
report. Odds ratios are shown for case-control studies; relative risk for cohort studies.
2 Calculated by a statistical method that adjusts for other factors (see Table 5-3), but not
corrected for smoker misclassification.
3 Composite measure formed from categorical data at different exposure levels.
4 For Gao, data are given as (number of years lived with a smoker, adj. OR): (< 20,1.0), (20-29,
1.1), (30-39, 1.3), (40+, 1.7).
5 95% confidence interval.
6 Case-control study nested in the cohort study of Hirayama. OR for ever-smokers is taken
from cohort study (shown in table below). This case-control study is not counted in any
summary results where HIRA(Coh) is included.
7 For Inoue, data are given as (number of cig./day smoked by husband, adj. OR): (< 19,1.58),
(20+, 3.09). ,
8 From subject responses/from proxy responses.
9 For the first value, "ETS-exposed" means the spouse smokes; for the second value, "ETS-
exposed" means a member of the household other than the spouse smokes.
From subject responses/from spouse responses.
Exposure at home and/or at work.
Exposure to regularly smoking household member(s). Differs slightly from published value of
0.78, wherein 0.5 was added to all exposure cells.
OR reported in study is 2.25, in contrast to the value shown that was reconstructed from the
confidence intervals reported in the study; no reply to inquiry addressed to author had been
received by press time.
ORs for never-smokers applies to exposure from spousal smoking, unless indicated otherwise.
Raw data for Wu are from Table 11 of Surgeon General (1986). Data apply to adenocarcinoma
only.
Odds ratio is not defined because number of unexposed subjects is zero for cases or controls.
Table entry is for exposure to smoking spouse, cohabitants, and/or coworkers; includes lung
cancers of all cell types. The OR for spousal smoking alone is for adenocarcinoma only: 2.01
(90% C.I. = 1.20, 3.37).
Adenocarcinoma only. Data and OR value communicated from author (Brownson).
Excludes bronchioalveolar carcinoma. Crude OR with bronchioalveolar carcinoma included is
reported to be 1.77, but raw data for calculation of confidence interval are not provided.
to
12
14
5-53
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 5-4. (continued)
20
21
22
Known adenocarcinomas and alveolar carcinomas were excluded, but histolbgical diagnosis was
not available for many cases. Data are from Trichopoulos et al. (1983).
Estimate for husband smoking 20 cig./day.
For second KABA study (see addendum in study description of KABA), preliminary
unpublished data and analysis based on ETS exposure in adulthood indicate 68% of never-
smokers are exposed and OR - 0.90 (90% C.I. = 0.51, 1.58), not dissimilar from the table entry
shown.
From crude data, estimated to be: exposed cases 52, exposed controls 91, unexposed cases 38,
unexposed controls 72.
RR is based on person-years of exposure to spousal smoking. "Prevalence" in those units is
20%.
RR values under never-smoker are for lung cancer mortality. For lung cancer incidence, crude
RR is 1.51 (90% C.I. = 0.41, 5.48) and adj. RR is 1.39 (95% C.I. = 0.29, 6.61).
24 The first, second, and third entries are calculated for population controls, colon cancer
controls, and both control groups combined, respectively. For adenocarcinoma alone, the
corresponding ORs, both crude and adjusted, are higher by 0.15-0.18.
Note: Values used for inference in this report are shown in boldface.
* Data not available.
25
5-54
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 5-5. Effect of statistical adjustments for cofactors on risk estimates for passive smoking16
Ca$0-C0ntt
Study
AKffi
BROW
BUFF
CHAN
CORR
FONT
GAO
GARF
GENG
HIRA
HUMB
INOU
JANE
KABA
KALA
KOO
01 Exposure
Source? Place5
Sp
Sp
A
Co
A
Sp
M(C)
Sp
SP
Sp
A
Sp
Sp
.. Sp
Sp
Sp
Sp
SP
A(C)
Sp
Sp
oc
SP
Co
A
A
P
H
A
A
A
A
A
A
A
H
A
A
A
A
A
A
H
A
A
H
A
H
Crude
RR*
1.52
1.52
L82
0.81
0.75
2.0713
1.6613
L3717
1.2118
1.19
*
1.31
2.16
1.539
1.53
2.34
2.55
0.86
*
0.79
1.62
1.41
1.55
1.34
Adj
RR6
1.5
*
1.68
*
*
1.3613
1.2917
1.2818
1.349
0.9
1.70
*
1.649
1.50
2.2
2.549
0.93/0.4415
1.09/2.075
*
1.92
1.64
1.68
Adjustment
Factors)1
A,L,0
*
A,I,0
*
*
Sm
A,E,I,L,R
A,E,I,L,R
A,E
A
A,SES,H,Yd
*
A,F,Oh,
F
A,R
A,N
A,L,R
A,R
*
A,E,Ir
A,E3,Yc
A,E,B,Yc
Adj "
Techniqoe*
LR
*
LR
*
*
R
LR
LR
R
LR
R
*
S
S
R
S
M,S
*
LR
*
LR
LR
5-55
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 5-5. (continued)
Case-Control
Study
LAMT
LAMW
LEE
LIU
PERS
SHIM
SOBU
SVEN
TRIG
WU
WUWI
BUTL (Coh)
GARF (Coh)
HERA (Coh)
HOLE (Coh)
Exposure
Source?
Sp
SP
A
Sp
Co
Co
Sp
Sp
SP
Sp
oc
A
Sp
SP
Sp
Co
Sp
Sp
Sp
Co
Place3
A
*
#
A
H
A
A
A
H
A
A
H,W
A
A
P
P
A
A
A
A
Crude
RR6
1.65
2.017
2.5114
1.38
0.75
[1.03
0.80
0.74
1.28
1.28
1.08
1.06
1.77
1.1/1.810
(1.26)
2.08
1.41"
0.79
0.78
2.45
*
1.38
2.27
Adj
- RR*
*
*
*
1.608
0.75
1.00]
0.879
0.77
1.2
1.479
*
1.13
1.57
1.2/2. 110
(1.4)
*
1.2
0.7
0.7
2.02
1.27/1. 10"
1.17
1.37/1. 04"
1.61
1.99
Adjustment
FactorCs)1
*
*
*
A
A
C
A,V
A
*
A,E
A,E
A
#
A,L
As
A,E,L
A,E,L
A
A
A,E,L,R,Oh
Ah
A,SES
Adj
Technique4
*
*
*
s
s
LR
M
s
*
S
s
s
*
M
LR
LR
LR
S
S
S
S
S
1 Adjustment factors: A = age of subject; Ah = age of husband; As = age started smoking; B =
number of live births; C = cooking habits; D = diet; E = education; F = fish consumption; G =
5-56
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 5-5. (continued)
gender; H = hospital; I = income; Ir = interviewer; L = location; M = marital status; O =
occupation (of subject); Oh = occupation (of husband); R = racial or ethnic group; SES =
socioeconomic status; Sm = active smoking; V = vital status; Yc = years since exposure ceased;
Yd = year of diagnosis; Yi = year of interview.
2 Source: A = Anyone; (C) = childhood; Co = cohabitants); M = mother; OC = cohabitants)
other than spouse; Sp = spouse; W = coworker(s).
3 Place: A = anywhere; H = home/household; P = proximity of subjects; W = workplace.
4 LR = logistic regression; R = regression; M = matched analysis; S = stratified.
5 1-24 smoker-years/ >_ 25 smoker-years.
6 OR for case-control studies; RR for cohort studies.
7 Adenocarcinoma only.
8 First value is for smoking information provided by patient's spouse, second for information
provided by patient herself, [third utilizes available data from either source(s) with subject
classified as exposed if either source so indicates].
Composite measure formed from categorical data at different exposure levels.
Exposed at home but not at work or vice versa/Exposed both at home and at work followed by
weighted average of exposed strata.
Crude OR from Table 11 of Surgeon General (1986); note that Adj OR from WU is not restricted
to never-smokers and analysis includes only adenocarcinoma.
Spouse smokes 1-20 cig./day/spouse smokes >_ 20 cig./day. The composite RR is 1.17.
13 Bronchioalveolar carcinoma excluded. Spousal smoking OR = 1.77 with bronchioalveolar
carcinoma excluded; no corresponding value reported for maternal smoking.
All cell types.
15 Cases and controls matched on A, L, and N; first value is from subject; second value is from
proxy sources.
16 Values used for inference in this report are shown in boldface.
17 Population controls, all cell types (crude and adjusted ORs for adenocarcinoma alone are 1.52 and
1.47, respectively).
18 Colon cancer controls, all cell types (crude and adjusted ORs for adenocarcinoma alone are 1.35
and 1.44, respectively).
* Data not available.
10
11
12
14
5-57
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 5-6. Alternative estimates of lung cancer relative risks associated with active and passive
smoking
Active/
Study Passive
BUTL Act
(Coh)
BUFF1 Pass
FONT15 Pass
HUMB2 Pass
KOO3 Pass
PERS4 Act
SHIM5 Pass
HIRA6 Act
(Coh)
HOLE7 Act
(Coh)
Alternative
Controls
ETS Exposure Exp.(SS) Estimate
N.A.9
Household members
regularly smoking for 33+
years
Spousal smoking,
all types
Spousal cigarette smoking2
Home and/or workplace
exposure over lifetime3
N.A.9
Total household ETS
exposure5
N.A.9
N.A.9
1410 Adj. RR 4.0"
71 Crude OR 0.95
(0.38, 2.40)
63 Crude OR 1.5212
(1.19, 1.96)
Adj. OR 1.47
66 Crude OR 1.3513
(1.02, 1.80)
Adj. OR 1.44
64 Crude OR 1.4714
(1.15, 1.87)
No Adj. OR
57 Crude OR 1.8
(0.6, 5.4)
Adj. OR 1.7
64 Crude OR 1.36
(0.83, 2.21)
Adj. OR 1.86
3710 Crude OR 4.2
77 Crude OR 1.36
4410 Adj. RR 3.79
5610 Adj. RR 4.2
Comparison
Estimate8
*
0.81
1.37
1.29
1.21
1.28
1.32
*
2.3
2.2
1.34
1.64
*
1.08
2.67
*
1 Values in Tables 5-4 and 5-5 include household smoking for any duration. Lung cancer may have
a long latency period, however, so the extended exposure may be of interest.
2 Values in Tables 5-4 and 5-5 include spousal smoking of cigars and pipes.
3 Value in Table 5-5 is for household cohabitant smoke exposure during adulthood.
5-58
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 5-6. (continued)
4 Estimate is based on papers by Cederlof et al. (1975) and Floderus et al. (1988) describing larger
populations on which Pershagen study was based.
5 Composite estimate from crude ORs for exposure from husband, parents, and father-in-law.
Values in Tables 5-4 and 5-5 consider only spousal smoke exposure.
6 Compares active smokers with never-smokers unexposed to ETS, thus providing a reference group
more truly unexposed to tobacco smoke. The value in Table 5-4 is the more conventional
comparison of ever-smokers with never-smokers, regardless of passive smoking status.
7 Estimate is from adjusted RR for both sexes combined with assumption that female RR is 75% of
male RR.
8 Nearest equivalent from Tables 5-4 or 5-5.
9 Not applicable because alternative estimate is for active smoking.
10 Percent ever-smokers.
11 Rough estimate based on data hi Fraser et al. (1991). The prevalence of female ever-smoking is
estimated from KALA and TRIG, similar conservative societies.
12 Population controls only.
13 Colon cancer controls only.
14 Control groups combined.
15 As in Table 5-4 except for adenocarcinoma alone.
* Data not available.
5-59
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 5-7. Estimated correction for smoker misclassification
Case
Control
AKffi
BROW
BUFF
CHAN
CORR
FONT
GAO
GARF
GENG
HIRA
HUMB
INOU
JANE
KABA
KALA
KATA
KOO
Uncorrected1
1.52
(0.49, 4.79)
0.81
(0.39, 1.66)
0.75
(0.48, 1.19)
2.07
(0.94, 4.52)
1.29
(1.03, 1.62)
1.31
(0.93, 1.85)
1.53
(1.10,2.13)
2.2
(0.9, 5.5)
0.86
(0.57, 1.29)
0.79
(0.30, 2.04)
*
1.55
(0.98, 2.44)
Never-Smokers RR7
Corrected5
(2)
1.5
(1.0, 2.5)
1.50
(0.48, 4.72)
0.70
(0.34, 1.43)
0.74
(0.47, 1.17)
1.90
(0.86, 4.15)
1.26
(1.01, 1.58)
1.19
(0.87, 1.63)
1.24
(0.88, 1.76)
2.16
(1.21, 3.84)
1.52
(1.10,2.12)
1.98
(0.8, 5.0)
2.55
(0.90, 7.20)
0.78
(0.51, 1.16)
0.74
(0.28, 1.90)
1.92
(1.13, 3.23)4
*
1.54
(0.98, 2.43)
a>/a>
1.00
1.01
1.16
1.01
1.09
1.03
1.00
1.06
1.00
(0.995)
1.01
1.11
1.00
(0.996)
1.10
1.07
1.00
*
1.01
ORtJsed^
2.38
4.30
7.06
3.48
12.40
8.0
2.54
6.0
2.77
3.20
16.3
1.66
8.0
5.90
3.32
*
2.77
5-60
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 5-7. (continued)
Case
Control
LAMT
LAMW
LEE
LIU
PERS
SHIM
SOBU
SVEN
TRIG
WU
WUWI
BUTL
(Coh)
GARF
(Coh)
HIRA
(Coh)
HOLE
(Coh)
'
Uncorreeted1 .
(1)
1.65
(1.21, 2.21)
1.03
(0.48, 2.20)
1.2
(0.7, 2.1)*
1.08
(0.70, 1.68)
1.26
(0.65, 2.48)
1.41
(0.63, 3.15)
0.79
(0.64, 0.98)
2.02
(0.48, 8.56)4
1.17
(0.85, 1.61)4
1.38
(1.03, 1.87)
1.99
(0.24, 16/7)4
Never-Smofcers RE*
Corrected5
(2)
1.64
(1.21,2.21)
2.51
(1.49, 4.23)
1.01
(0.47,2.15)
0.77
(0.35, 1.68)
1.17
(0.75, 1.87)4
1.07
(0.7, 1.67)
1.57
(1.13, 2.15)4
1.19
(0.62, 2.35)
2.08
(1.31, 3.29)
1.31
(0.58, 2.92)
0.78
(0.63, 0.96)
2.01
(0.61, 6.73)4
1.15
(0.88, 1.51)
1.37
(1.02, 1.86)
1.97
(0.34, 11. 67)4
.
Bias*
(l)/(2)
1.01
1.00
(0.996)
1.02
1.00
1.03
1.01
1.00
1.06
1.00
1.08
1.01
1.00
1.02
1.01
1.01
Ever-Smofcers
ORXI$eda
3.77
4.12
4.61
-
*
4.2
2.8
2.81
6.00
2.81
4.38
2.24
4.0
3.5
3.20
4.26
1 Adjusted OR in Table 5-4 is used unless the confidence interval is unknown or the study review
(Appendix A) is critical of the method(s) used.
2 The crude OR for ever-smokers in Table 5-4 is used in the calculations for the corrected value
(Appendix B), when available. Ever-smoker ORs for GARF, JANE, PERS, and SHIM are
5-61
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 5-7. (continued)
approximated from the data of other studies for suitable location and time period. The ever-smoker
OR for BUTL(Coh) [LEE] is based on data in Fraser (1991) [Alderson et al. (1985)].
3 Values shown are miscalculated ratio, 1). Calculated ratios less than 1 are shown in parentheses.
* 95% confidence interval.
5 Corrected (2) (estimate and confidence interval) equals uncorrected (1) times ratio [(2)/(l)]. All
corrected 95% confidence intervals have been converted to 90% confidence intervals.
6 Adjusted RR value hi Table 5-6.
7 OR for case-control studies; RR for cohort studies.
5-62
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 5-8. Statistical measures by individual study and pooled by country, corrected for smoker
misclassiflcation1
Location
Greece
Greece
Greece
HK
HK
HK
HK
HK
Japan
Japan
Japan
Japan
Japan
Japan
USA
USA
USA
USA
USA
Study '
KALA
TRIG
ALL
CHAN
KOO
LAMT
LAMW
ALL
AKIB
HIRA
(Coh)
INOU
SHIM
SOBU
ALL
BROW
BUFF
BUTL
(Coh)
CORR
FONT8
Relative
Weight2
(*>
43
57
5
20
20
45
15
15
15
35
3
16
30
19
1
3
1
3
35
f '< •f i s
o-value
Power3
0.39
0.45
0.43
0.43
0.73
0.39
0.42
0.75
0.17
0.377
0.66
0.15
0.17
0.18
0.22
0.93
Effect4
0.02
0.01
<0.01
>0.5
0.06
<0.01
<0.01
<0.01
0.05
0.04
0.07
0.39
0.01
0.01
0.28
>0.5
0.17
0.09
0.04
Trend5
0.04
<0.01
*
0.16
<0.01
*
0.03
<0.01
<0.03
*
*
*
*
*
0.01
0.04
RR*
1.92
2.08
2.00
0.74
1.54
1.64
2.51
1.48
1.50
1.37
2.55
1.07
1.57
1.43
1.50
0.70
2.01
1.90
1.26
} '
Confidence
Interval
90%
(1.13,3.23)
(1.31, 3.29)
(1.42, 2.83)
(0.47, 1. 17)
(0.98, 2.43)
(1.21, 2.21)
(1.49, 4.23)
(1.21, 1.81)
(1.00, 2.50)
(1.02, 1.86)
(0.90, 7.20)
(0.70, 1.67)
(1.13, 2.15)
(1.20, 1.71)
(0.48, 4.72)
(0.34, 1.43)
(0.61, 6.73)
(0.86, 4.15)
(1.01, 1.58)
5-63
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 5-8. (continued)
Location
USA
USA
USA
USA
USA
USA
USA
Scotland
Eng.AVales
Sweden
Sweden
W. Europe
China
China
China
China
China
China
Study x
GARF
GARF
(Coh)
HUMB
JANE
KABA
WU
ALL
HOLE
(Coh)
LEE
PERS
SVEN
ALL
GAO
GENG
LIU
WUWI
ALL
GAO and
GENG
Relative
Weight2' ,
^ W) ' J
15
25
2
10
2
3
34
100
100
68
32
5
28
8
4
60
22
'owei3
0.607
0.92
0.20
0.447
0.177
0.21
0.09
0.20
0.457
0.24
0.66
0.32
0.18
0.897
p-value
Effect4 Trenill^
0.15 <0.02
0.19 *
0.10 ns
>0.5 *
>0.5 *
0.29 ns
0.02
0.26 *
0.50 *
0.22 0.12
0.32 *
0.21
0.19 0.29
0.01 <0.05
>0.5 *
>0.5 *
>0.5
0.03
•.
1.24
1.15
1.98
0.78
0.74
1.31
1.19
1.97
1.01
1.17
1.19
1.17
1.19
2.16
0.77
0.78
0.95
1.36
Confidence
Interval
(0.88, 1.76)
(0.88, 1.51)
(0.81, 4.95)
(0.51, 1.16)
(0.28, 1.90)
(0.58, 2.92)
(1.04, 1.35)
(0.34, 11.67)
(0.47, 2.15)
(0.75, 1.87)
(0.62, 2.35)
(0.84, 1.64)
(0.87, 1.62)
(1.21, 3.84)
(0.35, 1.68)
(0.63, 0.96)
(0.80, 1.12)
(1.03, 1.79)
1 * means information is not available.
2 A study's relative weight (wt) is 1/var flog(OR)), divided by the sum of those terms for all studies
included, tunes 100 (to express as a percentage). Study weights shown for whole countries are
with WUWI included for China.
5-64
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 5-8. (continued)
A priori probability of significant (p < 0.05) test of effect when true relative risk is 1.5.
One-sided p-value for test of RR = 1 versus RR> 1.
p-value for upward trend, p-values from studies reporting only the significance level for trend
were halved to reflect a one-sided alternative, i.e., upward trend.
Adjusted for smoker misclassification. OR used for case-control studies; RR for cohort studies.
Calculated for matched study design.
For population control group only, all cases.
5-65
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 5-9. Case-control and cohort studies: exposure response trends for females
Study
AKIB
(cig./day)
AIOB
(years)
CORR
(pack-yrs.)
FONT5
(years)
FONT16
(years)
FONT5
(pack-yrs.)
FONT16-5
(pack-yrs.)
GAO
(tot. yrs.)3
GARF
(cig./day)
Case^
21
29
22
12
21
20
29
22
8
5
9
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
99
93
107
76
44
29
17
26
Cont*
82
90
54
23
82
30
81
59
72
38
23
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
57
63
78
48
157
90
56
44
Exposure4
1-19
20-29
^30
0
1-9
20-39
^40
0
1-40
S»41
0
1-15
16-30
>30
0
1-15
16-30
>30
0<15
15-39
40-79
;>so
0<15
15-39
40-79
^80
0-19
20-29
30-39
2>40
0
1-9
10-19
;>20
RR14
1.0
1.3
1.5
2.1
1.0
2.1
1.5
1.3
1.00
1.18
3.52
1.00
1.19
1.14
1.25
1.00
1.33
1.40
1.43
1.00
0.96
1.13
1.25
1.33
1.00
1.03
1.26
1.49
1.70
1.0
1.1
1.3
1.7
1.00
1.15
1.08
2.11
ex*
(0.7, 2.3)2
(0.8, 2.8)2
(0.7, 2.5)2
(1.0, 4.3)2
(0.8, 2.7)2
(0.7, 2.5)2
(0.44, 3.20)
(1.45, 8.59)
(0.88, 1.61)
(0.82, 1.59)
(0.91, 1.72)
(0.93, 1.89)
(0.96, 2.05)
(0.99, 2.09)
(0.72, 1.29)
(0.81, 1.59)
(0.86, 1.81)
(0.68, 2.58)
(0.73, 1.46)
(0.85, 1.87)
(0.98, 2.27)
(0.82, 3.49)
(0.7, 1.8)
(0.8, 2.1)
(1.0, 2.9)
(0.8, 1.6)
(0.8, 1.5)
(1.1,4.0)
p-Trf
0.03
0.24
0.01
0.07
0.02
0.04
0.01
0.29
<0.02
5-66
05/15/92
-------
Table 5-9. (continued)
DRAFT—DO NOT QUOTE OR CITE
Study
GENG
(cig./day)
GENG
(years)
HUMB
(cig./day)
INOU
(cig./day)
KALA
(cig./day)
KALA
(years)
KOO
(cig./day)
LAMT5
(cig./day)
LAMT16
(cig/day)
Case
*
*
*
*
*
*
*
*
*
*
*
*
*
*
26
34
22
8
26
15
15
17
17
32
17
25
12
84
22
56
20
53
17
37
15
Conk
*
*
*
*
*
*
*
*
*
*
*
*
*
*
46
39
22
9
46
21
20
15
16
67
15
35
19
183
22
66
21
92
12
28
9
Exposure4
0
1-9
10-19
2:20
0
<20
20-39
2:40
0
1-20
2:21
0-4
5-19
2:20
0
1-20
21-40
41 +
0
<20
20-29
30-39
2:40
0
1-10
11-20
2:21
0
1-10
11-20
2:21
0
1-10
11-20
2:21
RR1*
1.00
1.40
1.97
2.76
1.00
1.49
2.23
3.32
1.0
1.8
1.2
1.00
1.58
3.09
1.00
1.54
1.77
1.57
1.00
1.26
1.33
2.01
1.88
1.00
2.33
1.74
1.19
1.00
2.18
1.85
2.07
1.00
2.46
2.29
2.89
c,i/ ,,
(1.1, 1.8)
(1.4, 2.7)
(1.9,4.1)
(1.15, 1.94)
(1.54, 3.22)
(2.11, 5.22)
(0.6, 5.6)2
(0.3, 5.2)2
(0.4, 5.7)2
(1.0, 11.8)2
(0.88, 2.70)
(0.93, 3.35)
(0.64, 3.85)
(0.56, 2.87)
(0.58, 3.03)
(0.86, 4.67)
(0.82, 4.33)
(0.9, 5.9)
(0.8, 3.8)
(0.5, 3.0)
(1.14,4.15)
(1.19, 2.87)
(1.07, 4.03)
(1.09, 5.54)
(1.26, 4.16)
(1.18, 7.07)
P-1W
<0.058
<0.05"
ns
<0.03
0.08
0.04
0.16
0.01
0.01
5-67
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 5-9. (continued)
Study
PERS7
(cig./day)
TRIG9
(cig./day)
WU"
(years
exposed as adult)
GARF12 (Coh)
(cig./day)
HIRA13 (Coh)
(cig./day)
Case
34
26
7
24
24
14
*
*
*
65
39
49
37
99
64
*
*
*
Conk
*
*
*
109
56
25
*
*
*
*
*
*
21,895
44,184
25,461
Exposure4
0
1-15
£16
0
1-20
£21
0
1-30
£31
0
1-19
£20
0
l_19io
£20
0
1-15
>15
RR14
1.0
1.0
3.2
1.00
1.95
2.55
1.0
1.2
2.0
1.00
1.27
1.10
1.00
1.41
1.93
1.0
0.8
1.8
C.L1 p-TmwP
0.12
(0.6, 1.8)
(1.0, 9.5)
0.01
(1.13, 3.36)
(1.31, 4.93)
ns
*
*
*
(0.85, 1.89)
(0.77, 1.61)
0.01
(1.03, 1.94)
(1.35, 2.74)
*
*
*
9
10
It
12
13
14
Confidence intervals are 95% unless noted otherwise.
90% confidence interval.
Years lived with a smoking husband.
Smoking by spouse unless otherwise specified.
All histologies.
Very limited number of cases (6 total) and no C.I. information available.
Low exposure level is for husband smoking up to 15 cigarettes per day or one pack (50 g) of pipe
tobacco per week, or smoking any amount during less than 30 years of marriage. High exposure
level is for husband smoking more than 15 cigarettes per day or one pack of pipe tobacco per
week during 30 years of marriage or more.
Neither crude data nor a test for trend is included in reference articles. The relative risk at each
exposure category is significant alone, however, at p < 0.05.
Data from Trichopoulos et al. (1983), with RRs corrected (see letter, Trichopoulos, 1984).
Includes former smokers of any exposure level.
Years of exposure to spousal smoke plus years of exposure to workplace smoke; adenocarcinomas
only.
Value under RR is mortality ratio of observed to expected lung cancer deaths. Value under
"Case" is number of observed lung cancer deaths.
Standardized for age of subject (Hirayama, 1984). Values under "Case" are numbers of lung
cancer deaths; values under "Cont." are total population.
OR for case-control studies; RR for cohort studies.
5-68
05/15/92
_
-------
DRAFT—DO NOT QUOTE OR CITE
Table 5-9. (continued)
15
16
p-value for upward trend, p-values from studies reporting only the significance level for trend
were halved to reflect a one-sided alternative (i.e., upward trend). Values below 0.01 are shown
as 0.01.
Adenocarcinomas only.
* Data not available.
5-69
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 5-10. Reported p-values of trend tests for ETS exposure by study1
AKffi
CORR
FONT
GAO
GARF
GENG
HUMB
INOU
KALA
KOO
LAMT
PERS
TRIG
WU
GARF(Coh)
HIRA(Coh)
HOLE(Coh)
,
Intensity
(cig/day)
0.03
*
*
*
*
<0.02
<0.054
ns
<0.03
0.08
0.16
<0.01
<0.013
0.12
<0.01
*
*5
<0.01
*5
., , ^end.Test Results, ...,
. - Duration,
(total years)
0.24
*
0.072
<0.023
0.29
*
<0.054
*
*
0.04
*
*
*
*
ns
*
#
*
Cumulative
(pack-years)6
*
0.01
0.04
<0.01
*
*
*
*
*
*
*
*
*
*
*
*
*
*
1 Detailed data presented in Table 5-9.
2 All cell types.
3 Adenocarcinoma only.
4 Same footnote as for GENG, Table 5-9.
5 Trend results presented without p-values or raw data—see Table 5-9.
6 A "pack-year" is equivalent to one pack/day for 1 year.
* Data not available.
5-70
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 5-11. P-values of tests for effect and for trend by individual study1
Country
Greece
Greece
Hong Kong
Hong Kong
Hong Kong
Hong Kong
Japan
Japan
Japan
Japan
Japan
USA
USA
USA
USA
USA
USA
" Study
KALA
TRIC
CHAN
KOO
, LAMT
LAMW
AKIB
HIRA(Coh)
INOU
SHIM
SOBU
BROW
BUFF
BUTL(Coh)
CORR
FONT
GARF
Power
0.39
0.45
0.43
0.43
0.73
0.39
0.42
0.75
0.17
0.37
0.66
0.15
0.17
0.18
0.22
0.93
0.60
Test
Effect
Trend
Effect
Trend
Effect
Effect
Trend
Effect
Trend
Effect
Effect
Trend
Effect
Trend
Effect
Trend
Effect
Effect
Effect
Effect
Effect
Effect
Trend
Effect
Trend
Effect
Trend
p-value?
0.02
0.04
0.01
<0.01
>0.50
0.06
0.16
<0.01
<0.01
0.01
0.05
<0.03
0.04
<0.01
0.07
0.03
0.39
0.01
0.28
>0.50
0.17
0.09
0.01
0.043
0.043
0.15
<0.02
5-71
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 5-11. (continued)
Country
USA
USA
USA
USA
USA
W. Europe
Scot
Eng
Sweden
Sweden
China
China
China
China
.Study
GARF(Coh)
HUMB
JANE
KABA
WU
Hole(Coh)
LEE
PERS
SVEN
GAO
GENG
LIU
WUWI
Power
0.92
0.20
0.44
0.17
0.21
0.09
0.20
0.45
0.24
0.66
0.32
0.18
0.89
Te$t
Effect
Effect
Trend
Effect
Effect
Effect
Effect
Effect
Effect
Trend
Effect
Effect
Trend
Effect
Trend
Effect
Effect
*«*#
0.19
0.10
n.s.
>0.50
>0.50
0.29
0.26
0.50
0.22
0.12
0.32
0.19
0.29
0.01
<0.05
>0.50
>0.50
1 Test for effect — H0: no increase in lung cancer incidence in never-smokers exposed to spousal
ETS; HA: an increase. Test for trend — H0: no increase in lung cancer incidence as exposure to
spousal ETS increases; HA: an increase, p-values less than 0.1 are in boldface.
2 Smallest p-value is used when there is more than one test for trend.
3 For all cell types, p-values for adenocarcinoma alone were smaller.
5-72
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 5-12. Other risk-related factors for lung cancer evaluated in selected studies
Category
Possible Hisfc Factor Mixed Outcome
No Evidence
Personal or family
history
Heat source for
cooking or heating
Cooking with oil
Diet
Beta-Carotene
Occupation
WU(USA)
GENG(Ch)
LlU(Ch)
WU(USA)
WUWI(Ch)
GENG(Ch)
GAO(Ch)
LIU(Ch)
WUWI(Ch)
GAO(Ch)
WU(USA)
WUWI(Ch)
SHIM(Jap)
GENG(Ch)
BUTL(USA)
BUFF(USA)
SHIM(Jap)
GAO(Ch)
SOBU(Jap)
KALA(Gr)
HIRA(Jap)
LAMW(HK)
SHIM(Jap)
WUWI(Ch)
KALA(Gr)
GAO(Ch)-harmful
WU(USA)
GAO(Ch)
5-73
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
f
i
•as
'.
11
U
1
£
8?
SI
is
^ S
a*
Pattern
Passive
the tren
CO «
!3 £
«
.
-a-a
-
'
*
trolled for age, years of schoolin
energy intake. No confounding
the passive smoking effect and
between that of fruits and that o
isk increased to 2.11 when
umption.
r
ont
tal
ee
, or
ve
co
r
ri
ns
ers. C
, and to
ed
f fr
F
r fr
b
fruits,
assi
uit c
o
er
rv
es.
fo
er-sm
rview
obser
ffect
able
ted f
s
Nev
inte
was
the ef
veget
adiu
•§
vor-csoro
\ f> f— «
.£ ^
00
03
0-HMD
ooooo
000
I i i i i
OOc
^^'O'O'I-H'
1 1 1 1
ooooo
55'3
1.1
?
•
I
-------
DRAFT-DO NOT QUOTE OR CITE
1
52
•A
I
5-75
05/15/92
-------
0)
1
DRAFT—DO NOT QUOTE OR CITE
1
•I
*§
w
4lfi
.
^•Ssa -M
s-gSi
.f
1
?0
ll£sE-
»»S!«
II
II
K
cso^-vb
0000
i|
4-* QJ
l
O\
O
ON
OO
ON
K
1
ON
%
H
8
oo $
•aS «5
/S? K «
VH &
5> GI
S
^^TO
ops
« o-S
a 8 -a «
5-76
05/15/92
-------
DRAFT-DO NOT QUOTE OR CITE
Table 5-14A. Study limitations and sources of uncertainty
Sfcidy
AKIB
BROW
BUFF
CHAN
CORR
FONT
GAO
GARF
GENG
HUMB
INOU
JANE
KABA
KALA
KATA
KOO
LAMT
LAMW
LEE
LIU
PERS
SHIM
SOBU
SVEN
TRIG
WU
WUWI
BUTL (Coh)
'
ETS ' JETS
Subjects Exposure
X
X
X
X
X
X
X
X
X X
X X
X
X
X
X
X
X X
X
X
X
- Classification
Cases
X
X
X
X
X
X
X
X
X
X
X
"
Jtepresen-
Controls tativeness
X X
X
X X
X X
X
X
X
X
X
X
X
X
X
X
X
X
X
x
X
X
X
5-77
05/15/92
-------
DRAFT-DO NOT QUOTE OR CITE
Table 5-14A. (continued)
Study
GARF (Coh)
HIRA (Coh)
HOLE (Coh)
t! S
ETS
Subject
X
-
ETS
Exposure
X
X
- Classification '
Cases Controls
X
X
X
Represen-
tativeness
X
X
5-78
05/15/92
-------
DRAFT-DO NOT QUOTE OR CITE
Table 5-14B. Study limitations and sources of uncertainty2
Study
AKIB
BROW
BUFF
CHAN
CORR
FONT
GAO
GARF
GENG
HUMB
INOU
JANE
KABA
KALA
KATA
KOO
LAMT
LAMW
LEE
LIU
PERS
SHIM
SOBU
SVEN
TRIG
WU
WUWI
BUTL
(Coh)
GARF
(Coh)
HIRA (Coh)
HOLE
(Coh)
Seif-
Qaestionnaire
X
X
X
X
X
X
Collection
Response
and
Follow-up
X
X
X
X
X
X
X
X
X
X
Proxy
Response Cfnmameds
X
X X
x x
X X
X X
X X
X
X X
X
X X
X
X
X X
X
X X
X
X
X
X
Analvsi&
Adjusted Tr0nd
Analysis Analysis
X X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X X
X
X
5-79
05/15/92
-------
DRAFT-DO NOT QUOTE OR CITE
Table 5-15. Diagnosis, confirmation, and exclusion of lung cancer cases
Study
MOB2
BROW
BUFF2'3
CHAN2'3
CORR2
FONT
GAO2'6
GARF6
GENG2
HUMB8-9
INOU
JANE2
KABA
KALA
KATA
KOO ^
LAMT
LAMW
LEE
LIU10
PERS
SHIM
SOBU
Diagnosis/Confirmation* {%)
Radio./
Histology Cytology Clinical
53 4 43
inn
inn
82
97
100
43 38 19
100
85 4
9!\
* * *
99 1
100
48 38
100
94
inn
mn
* * *
n.._ .. R1
83 16
100
inn
Other/
tfnspec.
0
18
3
10
11
17
• *
14
6
*
0
1
Excluded
Secondary JX5
Y
Y
Y
N
Y
Y
Y
Y
N
Y
N
Y
Y
Y
N
Y
Y
Y
N
! N
Y
Y
Y
5-80
05/15/92
-------
DRAFT-DO NOT QUOTE OR CITE
Table 5-15. (continued)
•*
Study s Histology
SVEN2 70
TRIG2 28
WU 100
WUWI2 42
BUTL (Con)4
GARF (Coh) *
HIRA (Coh) *
HOLE (Coh)7 *
Diagnosjs/ConfirmatiQn5 (%)
,/ ,"
Radio,/ Other/ Excluded
Cytology Clinical UnspecL Secondary LC5
29 1
37 35
32 26
100
* *
* *
* *
Y
N
Y
Y
Y
N
N
N
1 Figures apply to confirmation of original diagnosis when conducted.
2 Not restricted to never-smokers (contains former smokers or ever-smokers).
3 Inconsistency in article. May be 100% histology.
4 Includes one former smoker.
5 Y (for yes) if specifically indicated; otherwise, N.
Diagnostic information was reviewed for study.
' Death certificate diagnosis checked against Scottish cancer registry records.
8 Includes males.
9 Available histologic specimens (17 cases) reviewed by pathologists. Poor agreement between
review diagnoses and original cancer registry diagnoses (8 of 17 cases). Only reviewed cases,
however, are presented in article.
10 Includes male ever- and never-smokers and one female ever-smoker (control).
* Data not available.
5-81
05/15/92
-------
DRAFT-DO NOT QUOTE OR CITE
Table 5-16. Classification of studies by tier
Country
Greece
Greece
Hong Kong
Hong Kong
Hong Kong
Hong Kong
Japan
Japan
Japan
Japan
Japan
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
USA
Study Tier 1 Tier 2 Tier 3 - Tier 4
KALA X
TRIG X
KOO X
LAMT X
LAMW X
CHAN X
AKffi X
HIRA(Coh) X
SHIM X
SOBU X
INOU , X
FONT X
BUTL(Coh) X
GARF X
HUMB X
JANE X
WU X
BROW X
BUFF X
CORR X
GARF(Coh) X
KABA X
5-82
05/15/92
-------
DRAFT-DO NOT QUOTE OR CITE
Table 5-16. (continued)
Country ,
W. Europe
Scot
Sweden
Sweden
England
China
China
China
China
Study Tier 1
HOLE(Coh) X
PERS
SVEN
LEE
GAO
GENG
LIU
WUWI
Tie* 2 Tier 3 Tier 4
X
X
X
X
X
X
X
5-83
05/15/92
-------
DRAFT-DO NOT QUOTE OR CITE
Table 5-17. Summary data interpretation by country4
Through Weight?
Tier1 (%)
1
2 9
2 17
3
4
2 17
3
4
1
2 40
3
1
2 8
3
3 11
4
Country3
Greece
Greece
Hong Kong
Hong Kong
Hong Kong
Japan
Japan
Japan
USA
USA
USA
W. Europe
W. Europe
W. Europe
China
China
Studies
Added
KALA
TRIC
KOO, LAMT
LAMW
CHAN
AKIB, HIRA(Coh)
SHIM, SOBU
INOU
FONT
BUTL(Coh), GARF,
HUMB, JANE, WU
BROW, BUFF,
CORR, GARF(Coh),
KABA
HOLE(Coh)
PERS, SVEN
LEE
GAO
GENG, LIU, WUWI
RR
1.92
2.00
1.61
1.75
1.48
1.44
1.41
1.43
1.26
1.19
1.19
1.99
1.22
1.17
1.19
0.95
90% ex
(1.13, 3.23)
(1.42, 2.83)
(1.25, 2.06)
(1.39, 2.19)
(1.21, 1.81)
(1.12, 1.85)
(1.17, 1.68)
(1.20, 1.71)
(1.01, 1.58)
(1.02, 1.40)
(1.04, 1.35)
(0.34, 11.67)
(0.84, 1.76)
(0.84, 1.64)
(0.87, 1.62)
(0.80, 1.12)
p-val«e
0.02
0.0005
0.0009
0.00002
0.0008
0.008
0.0009
0.0005
0.04
0.04
0.02
0.26
0.19
0.22
0.19
0.70
1 Each line contains the studies in the previous tiers plus those added.
2 Percent of total weight for Tier 2. Total of 102% is due to rounding.
3 W. Europe consists of England, Scotland, and Sweden.
4 Use of Tier 2, shown hi boldface, is recommended. Tier 4 is not recommended.
5-84
05/15/92
-------
DRAFT-DO NOT QUOTE OR CITE
TESTS OF THE HYPOTHESIS THAT RR = 1
BY STUDY
p = 0.01
p = 0.001
Figure 5-1. Test statistics for hypothesis RR = 1, all studies
TESTS OF THE HYPOTHESIS THAT RR = 1
USA
p = 0.5
0.2
p =* 0.01
p = 0.001
JANE SCAB* 0 WU HUU3
BUfT BROW CORK
C*RF(Coh)
BUTLWoh)
rONT
Figure 5-2. Test statistics for hypothesis RR = 1, USA only
5-85
05/15/92
-------
DRAFT-DO NOT QUOTE OR CITE
TESTS OF THE HYPOTHESIS THAT RR = 1
BY COUNTRY
p = 0.5
0.2
0.01
p = 0.001
CHINA W.EUR USA JAPAN
GREECE
HK
Figure 5-3. Test statistics for hypothesis RR = 1, by country
TESTS OF THE HYPOTHESIS THAT RR = 1
BY COUNTRY (Chino w/o WUWI & LIU)
p = 0.5
p-0.2
p = 0.01
p = 0.001
W.EUR CHINA JAPAN
USA GREECE
HK
Figure 5-4. Test statistics for hypothesis RR = 1, China w/o WUWI and LIU
5-86
05/15/92
-------
DRAFT-DO NOT QUOTE OR CITE
90% CON
GREECE
HONG KONG
JAPAN
USA
W. EUROPE |
CHINA ,_»
FIDENCE INTERVALS FOR RR
BY COUNTRY
1— O— 1
— *
0.0 0.5 10 IIS 2.0 2.5 3.0
figure 5-5. 90% confidence intervals, by country
90% CON
BYCOU
GREECE
HONG KONG
JAPAN
USA
W. EUROPE |
CHINA
1 , | .._._
FIDENCE INTERVALS FOR RR
NTRY (Chino w/o WUWI
-------
-------
DRAFT—DO NOT QUOTE OR CITE
6. POPULATION RISK OF LUNG CANCER FROM PASSIVE SMOKING
6.1. INTRODUCTION
The preceding chapter addressed the topic of hazard identification and concluded that
environmental tobacco smoke (ETS) exposure is causally associated with lung cancer. If an effect
is large enough to detect in epidemiologic studies investigating the consequences of ETS exposure
at common exposure levels, the individual risk associated with exposure is considered to be high
compared to most environmental contaminants assessed. Of course, the number of lung cancer
deaths attributable to ETS exposure for a whole population, such as the United States, depends on
the number of persons exposed as well as the individual risk. Studies of cotinine/creatinine
concentrations in nonsmokers indicate that ETS is ubiquitous. For example, in urinary bioassays
of 663 nonsmokers, Cummings et al. (1990) found that over 90% had detectable levels of cotinine.
Among the 161 subjects who reported no recent exposure to ETS, the prevalence of detectable
cotinine was still about 80%. Although the average cotinine level for all those tested may be
below the average for subjects exposed to spousal ETS, as studied in this report, it indicates
uptake of ETS to some extent by a large majority of nonsmokers (see also Chapter 3).
Consequently, exposure to ETS is a public health issue that needs to be considered from a national
perspective.
This chapter derives U.S. lung cancer mortality estimates for female and male never-
smokers and long-term (5+ years) former smokers. Section 6.2 discusses prior approaches to
estimating U.S. population risk. Section 6.3 presents this report's estimates. First, the parameters
and formulae used are defined (Section 6.3.2), and then lung cancer mortality estimates are
calculated from two different data sets and confidence and sources of uncertainty in the estimates
are discussed. Section 6.3.3 derives estimates based on the combined relative risk estimates of the
11 U.S. studies from Chapter 5. Section 6.3.4 bases its estimates on the data from the single
largest U.S. study, that of Fontham et al. (1991). Finally, Section 6.3.5 discusses the sensitivity of
the estimates to changes in various parameter values. ETS-attributable lung cancer mortality rates
for each of the individual studies from Chapter 5 are presented in Appendix D.
6.2. PRIOR APPROACHES TO ESTIMATION OF POPULATION RISK
Several authors have estimated the population risk of lung cancer from exposure to ETS
previously. Two approaches have been used almost exclusively. One approach analyzes the
overall epidemiologic evidence available from case-control and cohort studies, as done in this
report; the other estimates a dose-response relationship for ETS exposure extrapolated from active
6-1
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
smoking, based on "cigarette-equivalents" determined from a surrogate measure of exposure
common to passive and active smoking. A recent review of risk assessment methodologies in
passive smoking may be found in Repace and Lowrey (1990).
6.2.1. Examples Using Epidemiologic Data
The National Research Council report (NRC, 1986) is a good example of the epidemiologic
approach. An overall estimate of relative risk (RR) of lung cancer for never-smokers exposed to
both and spousal smoking and background ETS versus those exposed only to background ETS is
obtained by statistical summary across all available studies. Two "corrections" are then made to
the estimate of RR to correct for the two sources of systematic bias. The first correction accounts
for expected upward bias from former smokers and current smokers who may be misclassified as
never-smokers; this correction results in a decrease in the RR estimate. The second correction is
an upward adjustment to the RR taking into account the risk from background exposure to ETS
(experienced by a never-smoker whether married to a smoker or not) to obtain estimates of the
excess lung cancer risk from all sources of ETS exposure (spousal smoking and background ETS)
relative to the risk in an ETS-free environment. Population risk can then be characterized by
estimating the annual number of lung cancer deaths among never-smokers attributable to all
sources of ETS exposure. This calculation requires the final corrected estimates of relative risk
(one for background ETS only and one for background plus spousal smoking), the annual number
of lung cancer deaths (LCDs) from all causes in the population assessed (e.g., never-smokers of
age 35 and over), and the proportion of that population exposed to spousal smoking. The entire
population is assumed to be exposed to some average background level of ETS; although, in fact,
the population contains some individuals with high exposure and others with virtually no
exposure.
The NRC report combines data for female and male never-smokers to obtain an overall
observed RR estimate of 1.34 (95% C.I. = 1.18-1.53), but this estimate is most heavily influenced
by the abundant female data. (The female data alone generate a combined RR estimate of 1.32
[95% C.I. = 1.18-1.52], while the male data produce an RR estimate of 1.62 [95% C.I. = 0.99-
2.64].) To adjust for potential misclassification bias, the NRC uses the construct of Wald and
coworkers. The technical details of the adjustment are contained in Wald et al. (1986) and to a
lesser degree in^the NRC report. After correcting the overall observed RR estimate of 1.34
downward for an expected positive (upward) bias from smoker misclassification, the NRC
concludes that the relative risk is about 1.25, and probably lies between 1.15 and 1.35. Correction
for background sources (i.e., nonspousal sources of ETS) increases the NRC estimate of RR for an
6-2
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
"exposed" person (i.e., exposed to ETS from spousal smoking) to 1.42 (range of 1.24 to 1.61); the
change is due only to implicit redefinition of RR to mean risk relative to zero-ETS exposure
instead of relative to nonspousal sources of ETS. Under this redefinition, the RR for an
"unexposed" person (i.e., unexposed to spousal ETS) versus a truly unexposed person (i.e., in a
zero-ETS environment) becomes 1.14 (range of 1.08 to 1.21). The NRC report further estimates
that about 21% of the lung cancers in nonsmoking women and 20% in nonsmoking men may be
attributable to exposure to ETS (NRC, 1986, Appendix C); these estimates, however, are based on
RRs corrected for background ETS but not for smoker misclassification. Applying these
percentages to estimates of 6,500 LCDs in never-smoking women and 3,000 LCDs in never-
smoking men in 1988 (American Cancer Society, personal communication), the number
attributable to ETS exposure is 1,365 and 600, respectively, for a total of about 2,000 LCDs
among never-smokers of both sexes.
Robins (NRC, 1986—Appendix D [included in the NRC report but neither endorsed nor
rejected by the committee]) explores three approaches to assessment of lung cancer risk from
exposure to ETS, each with attendant assumptions clearly stated. A related article by Robins et al.
(1989) contains most of the same information. Method 1 is based solely on evaluation of the
epidemiologic data applying two assumptions: (1) correction of relative risk for background
exposure to ETS independent of age, and (2) the excess relative risk in a nonsmoker is
proportional to the lifetime dose of ETS. In this method, Robins uses a weighted average RR of
1.3. After correcting this RR for background ETS exposure, age-adjusted population-attributable
risks are calculated for females and males separately. Adjusting Robins' results to 6,500 annual
LCDs in female never-smokers and 3,000 LCDs in male never-smokers, for comparison purposes,
yields estimates of 1,870 female LCDs and 470 male LCDs attributable to ETS. Method 2 uses an
overall relative risk value based on epidemiologic data, but also makes some assumptions to appeal
to results of Day and Brown (1980) and Brown and Chu (1987) on lung cancer risk in active
smokers. Again, adjusting Robins' estimates to 6,500 female LCDs and 3,000 male LCDs, the
range of excess LCDs attributable to ETS is 1,650 to 2,990 for never-smoking females and 420 to
1,120 for never-smoking males. Method 3 is a "cigarette-equivalents" approach and is discussed in
Section 6.2.2.
The Centers for Disease Control has published an estimate of 3,825 (2,495 female and
1,330 male) deaths in nonsmokers from lung cancer attributable to passive smoking for the year
1988 (CDC, 1991a), with reference to the NRC report of 1986. Those figures are the mid-range of
values for males and females from method 2 of Robins in Appendix D of the NRC report (NRC,
1986).
6-3
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Blot and Fraumeni (1986) published a review and discussion of the available epidemiologic
studies about the same time as the Surgeon General's and the NRC reports appeared. The set of
studies considered by Blot and Fraumeni are almost identical to those included in the NRC report,
except for omission of one cohort study (Gillis et al., 1984), and inclusion of WU, the case-control
study excluded by the NRC because the raw data were unpublished. An overall relative risk
estimate calculated from the raw data for females yields 1.3 (95% C.I. = 1.1-1.5). When the results
are combined for high exposure categories, the overall relative risk estimate is 1.7 (1.4-2.1).
Wells (1988) provides a quantitative risk assessment that includes several epidemiologic
studies subsequent to the NRC and Surgeon General's reports of 1986 (NRC, 1986; U.S. DHHS,
1986). Like the NRC report, the epidemiologic data for both women and men are considered, for
which Wells provides separate estimates of overall relative risk and attributable risk. Wells
calculates an overall relative risk of 1.44 (95% C.I. = 1.26-1.66) for females and 2.1 (1.3-3.2) for
males. Following the general approach of Wald et al. (1986), the misclassification percentage for
ever-smokers is assumed to be 5% (compared to 7% for Wald et al.). Rates are corrected for
background exposure to ETS, except in studies from Greece, Japan, and Hong Kong, where the
older nonsmoking women are assumed to experience very little exposure to ETS outside the home.
A refinement in the estimation of population-attributable risk is provided by adjusting for age at
death (which also appears in the calculations of Robins [NRC, Appendix D]). The calculation of
population-attributable risk applies to former smokers as well as never-smokers, which is a
departure from Wald et al. and the NRC report. The annual number of LCDs attributable to ETS
in the United States is estimated to be 1,232 (females) and 2,499 (males) for a total of 3,731.
About 3,000, however, is thought to be a best current estimate (Wells, 1988). (In addition to the
estimates of ETS-attributable LCDs, Wells uses the epidemiological approach to derive estimates
of ETS-attributable deaths from other cancers (11,000) and from heart disease (32,000).)
Saracci and Riboli (1989), of the International Agency for Research on Cancer, review the
evidence from the three cohort studies and 11 of the case-control studies (Table 4-1). The authors
follow the example of the NRC and Wald et al. with respect to the exclusion of studies, and add
only one additional case-control study (Humble et al., 1987). The overall observed relative risk
for the studies, 1.35 (1.20-1.53), is about the same as that reported by the NRC, 1.34 (1.18-1.53).
It is not reported how the overall relative risk was calculated.
Repace and Lowrey (1985) suggest two methods to quantify lung cancer risk associated
with ETS. One method is based on epidemiologic data but, unlike in the previous examples
discussed, Repace and Lowrey use a study comparing Seventh-Day-Adventists (SDAs) (Phillips et
al., 1980a,b) with a demographically and educationally matched group of non-SDAs who are also
6-4
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
never-smokers to obtain estimates of the relative risk of lung cancer mortality, in what they
describe as a "phenomenological" approach. The SDA/non-SDA comparison provides a basis for
assessing lung cancer risk from ETS in a broader environment, particularly outside the home, than
the other epidemiologic studies. It also serves as an independent source of data and an alternative
approach for comparison. Information regarding the number of age-specific LCDs and person-
years at risk for the two cohorts is obtained from the study. The basis for comparison of the two
groups is the premise that the non-SDA cohort is more likely to be exposed to ETS than the SDA
group due to differences in lifestyle. Relatively few SDAs smoke, so an SDA never-smoker is
probably less likely to be exposed at home by a smoking spouse, in the workplace, or elsewhere, if
associations are predominantly with other SDAs. One of the virtues of this novel approach is that
it contributes to the variety of evidence for evaluation and provides a new perspective on the
topic.
Phillips et al. reported that the non-SDA cohort experienced an average lung cancer
mortality rate equal to 2.4 times that of the SDA cohort. Using 1974 U.S. Life Tables, Repace and
Lowrey calculate the difference in lung cancer mortality rates for the two cohorts by 5-year age
intervals and then apply this value to an estimated 62 million never-smokers in the United States
in 1979 to obtain a number of LCDs attributable to ETS annually. The result, 4,665, corresponds
to a risk-rate of about 7.4 LCDs per 100,000 person-years. In an average lifespan of 75 years,
that value equates to 5.5 deaths per 1,000 people exposed. The second method described by
Repace and Lowrey is a "cigarette-equivalents" approach and is discussed in Section 6.2.2.
Wigle et al. (1987) apply the epidemiologic evidence from the SDA/non-SDA study
(Phillips et al., 1980a,b) to obtain estimates of the number of LCDs in never-smokers due to ETS
in the population of Canada. The estimated number of deaths from lung cancer attributable to
passive smoking is calculated separately for males and females, using age-specific population
figures for Canada and the age-specific rates of death from lung cancer attributable to ETS
estimated by Repace and Lowrey (1985). A total of 50 to 60 LCDs per year is attributed to
spousal smoking alone, with 90% of them in women. Overall, involuntary exposure to tobacco
smoke at home, work, and elsewhere may cause about 330 LCDs annually.
6.2.2. Examples Based on Cigarette Equivalents
The cigarette-equivalents approach assumes that the dose-response curve for lung cancer
risk from active smoking also applies to passive smoking, after extrapolation of the curve to lower
doses and conversion of ETS exposure into an "equivalent" exposure from active smoking,
determined from a surrogate measure of exposure common to passive and active smoking.
6-§
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Relative cotinine concentrations in body fluids (urine, blood, or saliva) of smokers versus
nonsmokers and tobacco smoke particulates in sidestream smoke (SS) and mainstream smoke (MS)
have commonly been used for this purpose. The lung cancer risk of ETS is assumed to equal the
risk from active smoking at the rate determined by the cigarette-equivalents. For example,
suppose the average cotinine concentration in exposed never-smokers is 1% of the average value
found in people who smoke 30 cigarettes per day. The lung cancer risk for a smoker of (0.01)30 =
0.3 cigarettes per day is estimated by low-dose extrapolation from a dose-response curve for
active smoking, and that value is used to describe the lung cancer risk for ETS exposure. This
general explanation describes the nature of the approach; however, authors vary in their
constructed solutions and level of detail. The basic assumption of cigarette-equivalents
procedures is that the lung cancer risks in passive and active smokers are equivalently indexed by
the common measure of exposure to tobacco smoke, i.e., a common value of the surrogate measure
of exposure in an active and a passive smoker would imply the same lung cancer risk in both.
This assumption may not be tenable, however, as MS and SS differ in the relative composition of
carcinogens and other components identified in tobacco smoke and in their physicochemical
properties in general; the lung and systemic distribution of chemical agents common to MS and SS
are affected by their relative distribution between the vapor and particle phases, which differs
between MS and SS and changes with SS as it ages; and active and passive smoking also differ in
characteristics of intake, for example intermittent (possibly deep) puffing in contrast to normal
(shallow) inhalation, which may affect deposition and systemic distribution of various tobacco
smoke components as well.
Several authors have taken issue with the validity of the cigarette-equivalents approach.
For example, Hoffmann et al. (1989), in discussing the longer clearance times of cotinine from
passive smokers than from active smokers, conclude "The differences in the elimination time of
cotinine from urine preclude a direct extrapolation of cigarette-equivalents to smoke uptake by
involuntary smokers." A recent consensus report of an IARC panel of experts (Saracci, 1989)
states "Lacking knowledge of which substances are responsible for the well established
carcinogenic effect of MS, it is impossible to accurately gauge the degree of its similarity to ETS
in respect to carcinogenic potential." The Surgeon General's report devotes a three-page section to
the concept of cigarette-equivalents, quantitatively demonstrating how they can vary as a measure
of exposure (U.S. DHHS, 1986). It concludes "These limitations make extrapolation from
atmospheric measures to cigarette-equivalents units of disease risk a complex and potentially
meaningless process." [On a lesser note, it has generally been assumed that the dose-response
relationship for active smokers is reasonably well characterized. Recent literature raises some
6-6
05/15/92
-------
DRAFT-DO NOT QUOTE OR CITE
questions on this issue (Moolgavkar et al, 1989; Gaffney and Altshuler, 1988; Freedman and
Navidi, 1987a, 1987b; Whittemore, 1988.)]
Citing cigarette-equivalents calculated in other sources, Vutuc (1984) assumes a range of
0.1 to 1.0 cigarettes per day for ETS exposure. Relative risks for nonsmokers are calculated for
10-year age intervals (40 to 80) based on the reported relationships of dose, time, and lung cancer
incidence in Doll and Peto (1978). Relative risks for smokers of 0.1 to 1.0 cigarettes per day give
a range in relative risk from 1.03 to 1.36. The author concludes "As it applies to passive smokers,
this range of exposures may be neglected because it has no major effect on lung cancer incidence."
Vutuc assumes that his figures apply to both males and females. If an exposure fraction of 75% is
assumed for both males and females, the range of relative risks given correspond to a range for
population-attributable risk. If the number of LCDs among never-smokers in the United States in
1988 is about 6,500 females and 3,000 males (personal communication from the American Cancer
Society), then the number of LCDs in never-smokers attributable to ETS is estimated to range
from 240 to 2,020 (140 to 1,380 for females alone). So Vutuc's figures are consistent with several
hundred excess LCDs among never-smokers in the United States. These estimates are from our
extension of Vutuc's analysis, however, and are not the claim of the author.
Repace and Lowrey (1985) describe a cigarette-equivalents approach as an alternative to
their "phenomenological" approach discussed in Section 6.2.1. One objective is to provide an
assessment of exposure to ETS from all sources that is more inclusive and quantitative than might
be available from studies based on spousal smoking. They consider exposure to ETS both at home
and in the workplace, using a probability-weighted average of exposure to respirable suspended
particulates (RSP) in the two environments. Exposure values are derived from their basic
equilibrium model relating ambient concentration of particulates to the number of burning
cigarettes per unit volume of air space and to the air change rate. From 1982 statistics of lung
cancer mortality rates among smokers and their own previous estimates of daily tar intake by
smokers, the authors calculate a lung cancer risk for active smokers of 5.8 x 10"6 LCDs/year per
mg tar/day per smoker of lung cancer age. The essential assumption linking lung cancer risk in
passive and active smokers is that inhaled tobacco tar poses the same risk to either on a per unit
basis. Extrapolation of risk from exposure levels for active smokers to values calculated for
passive smokers is accomplished by assuming that dose-response follows the one-hit model for
carcinogenesis. An estimated 555 LCDs per year in U.S. nonsmokers (never-smokers and former
smokers) are attributed to ETS exposure (for 1980). The ratio of total LCDs in 1988 to 1980 is
approximately 1.37 (Repace, 1989). With that population adjustment factor, the approximate
6-7
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
number of LCDs attributable to ETS among nonsmokers is closer to 760 for 1988 (including
former smokers).
Method 3 of Robins (NRC, 1986, Appendix D—again, included in the NRC report but not
specifically endorsed by the committee) extrapolates from data on active smoking, along with
several assumptions. Applying his results to 6,500 females and 3,000 males, the range of excess
LCDs in never-smokers due to ETS is 550 to 2,940 for females and 153 to 1,090 for males.
Russell and coworkers (1986) use data on urinary nicotine concentrations in smokers and
nonsmokers to estimate exposure and risk from passive smoking. The risk of premature death
from passive smoking is presumed to be in the same ratio to premature death in active smokers as
the ratio of concentrations of urinary nicotine in passive to active smokers (about 0.007).
Calculations are made using vital statistics for Great Britain and then extrapolated to the United
States. The latter estimate, 4,000+ deaths/year due to passive smoking, is for all causes of death,
not just LCDs.
Arundel et al. (1987) attributes only five LCDs among female never-smokers to ETS
exposure. The corresponding figure for males is seven (both figures are adjusted to 6,500 females
and 3,000 males). The expected lung cancer risk for never-smokers is estimated by downward
extrapolation of the lung cancer risk/mg of particulate ETS exposure for current smokers. The
authors' premise is that the lung carcinogenicity of ETS is entirely attributable to the partieulate
phase of ETS, and the consequent risk in passive smoking is comparable to active smoking on a
per mg basis of particulate ETS retained in the lung. If the vapor phase of ETS were also
considered, the number of LCDs attributable to ETS would likely increase (e.g., see Wells, 1991).
6.3. THIS REPORTS ESTIMATES OF LUNG CANCER MORTALITY ATTRIBUTABLE TO
ETS IN THE UNITED STATES
6.3.1. Introduction and Background
This report uses the epidemiologic approach because of the abundance of human data from
actual environmental exposures. Furthermore, the assumptions are fewer and more valid than for
the cigarette-equivalents approach. The report generally follows the epidemiologic methodology
used by the NRC and others (Section 6.2.1) with three important differences. The first difference
is that the NRC combined the data on females and males for its summary relative risk estimate.
This report uses only the data on females because there are likely to be true sex-based differences
in relative risk due to differences in exposure to background ETS and differences in background
(i.e., nontobacco-smoke-related) lung cancer risk. Furthermore, the vast majority of the data are
for females. The second difference is that the NRC combined study estimates of relative risk
6-8
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
across countries for its summary relative risk estimate; this report combines relative risk estimates
only within countries, and then bases the U.S. population risk assessment on the U.S. estimate
only. As discussed in Chapter 5, there are apparently true differences in the observed relative risk
estimates from different countries, which might reflect lifestyle differences, differences in
background lung cancer rates in females, exposures to other indoor air pollutants, and differences
in exposure to background levels of ETS. Therefore, for the purposes of U.S. population risk
assessment, it is appropriate to use the U.S. studies; and there are far more studies currently
available, so there is less need to combine across countries. The third difference is that the NRC
corrected its overall estimate of relative risk downward for smoker misclassification bias. In this
report the individual study estimates are corrected for smoker misclassification bias at the outset,
i.e., prior to any analysis, using the particular parameters appropriate for each separate study
(Appendix B).
Estimates of ETS-attributable population mortality are calculated from female lung cancer
mortality rates (LCMRs), which are themselves derived from summary relative risk estimates
either from the 11 U.S. studies combined (Section 6.3.3) or from the Fontham et al. (1991) study
alone (Section 6.3.4), along with other parameter estimates from prominent sources (Section 6.3.2).
The LCMRs in this instance are defined as the number of lung cancer deaths in 1985 per 100,000
of the population at risk. The LCMR in U.S. women under age 35 is minuscule, so only persons
of age 35 and above are considered at risk. Although these LCMRs are expressed as a mortality
rate per 100,000 of the population at risk, as derived they are applicable only to the entire
population at risk and not to any fraction thereof that might, for example, have a different
average exposure or age distribution.
The LCMR for the subpopulation and exposure scenario to which the epidemiolbgic
studies apply most directly—never-smoking females exposed to spousal ETS—is estimated first.
That estimate is then incremented to include exposure to nonspousal ETS for all never-smoking
females. For the ETS-attributable population mortality estimates, these LCMRs are applied to
never-smoking males and former smokers at risk, as well as to the females at risk for which the
rates were specifically derived. The most reliable component of the total estimate constructed for
the United States is the estimate for the female never-smokers exposed to spousal ETS. The other
components require additional assumptions, which are described. As the number of assumptions
increases, so does the uncertainty of the estimates. Thus, the total estimate of lung cancer risk to
U.S. nonsmokers of both sexes is comprised of component estimates of varying degrees of
certainty.
6-9
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
One might argue that smokers are among those most heavily exposed to ETS, since they
are in close proximity to sidestream smoke (the main component of ETS) from their owii
cigarettes and are also more likely than never-smokers to be exposed to ETS from other smokers.
The purpose of this report, however, is to address respiratory health risks from ETS exposure in
nonsmokers. In current smokers, the added risk from passive smoking is relatively insignificant
compared to the self-inflicted risk from active smoking.
6.3.2. Parameters and Formulae for Attributable Risk
Several parameters and formulae are needed to calculate attributable risk. These are
presented in Table 6-1, with the derivations explained below.
The size of the target population, in this case the number of women in the United States of
age 35+ in 1985, is denoted by N, with N = Nt + N2, where Nj = the number of ever-smokers and
N2 - the number of never-smokers. The total number of LCDs from all sources, T, is apportioned
into components from four attributable sources: (1) nontobacco-smoke-related causes, the
background causes that would persist in an environment free of tobacco smoke; (2) background
ETS, which refers to all ETS exposure other than that from spousal smoking; (3) spousal ETS; and
(4) ever-smoking. The risk from nontobacco-smoke-related causes (source 1) is a baseline risk
(discussed below) assumed to apply equally to the entire target population (never-smokers and
ever-smokers alike). The ever-smoking component of attributable risk (source 4) refers to the
incremental risk above the baseline in ever-smokers (this report does not partition the incremental
risk in ever-smokers further into components due to background ETS and spousal ETS, except for
long-term [5+ years] former smokers). The background ETS component (source 2) is the
incremental risk above the baseline in all never-smokers from exposure to nonspousal sources of
ETS. The spousal ETS component (source 3) is the additional incremental risk in never-smokers
exposed to spousal smoking.
The calculational formulae also require values for the parameters Pl (prevalence of ever-
smokers), P2 (proportion of never-smokers exposed to spousal smoking), RRj (average lung cancer
risk for ever-smokers relative to the average risk for never-smokers in the population), and RR2
(lung cancer risk of never-smokers exposed to spousal ETS relative to never-smokers not exposed
to spousal ETS). Additional parameters (RRU, Z, RR01, RR02» and RR03) are introduced or
developed below.
The "baseline" risk is defined as the term in the denominator of a risk ratio. For example,
in RRj the baseline risk is the lung cancer risk in a population of never-smokers with P2 exposed
to spousal ETS and 1-P2 not exposed to spousal ETS. The conversion of RRt to the same baseline
6-10
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
risk as RR2 (the risk of never-smokers not-exposed to spousal ETS but still exposed to
nontobacco-smoke-related causes and to background ETS), is given by
RRn = RRj(P2RR2 + 1 - Pj).
(6-1)
To convert relative risks to the baseline risk of lung cancer from nontobacco-smoke-related causes
only (i.e.i excluding background ETS in the baseline) requires some assumptions. Let RRo2 denote
the conversion of RR2 to this new baseline. It is assumed that: (1) the excess risk of lung cancer
from ETS exposure is proportional to ETS exposure; and (2) the ratio of ETS exposure from
spousal smoking plus other sources to exposure from other sources alone, denoted by Z, is known
and Z > RR2 > 1 (For the values used in this document this relation is true. See also the discussion
in Section 8.3). Under these assumptions, it is readily verified that
(6-2)
Determination of a value for Z from data on cotinine concentrations (or cotinine/cfeatinine) is
discussed below. The conversion of RRt to the same zero-ETS baseline risk as RR02 follows from
multiplying expression (6-1) by RR02/RR2, i.e.,
RR01 = RR1(P2RR02 + (1-P2)RR02/RR2).
(6-3)
The terms RRoi and RR<)2 are the lung cancer risks for ever-smokers and for never-smokers
exposed to spousal ETS, respectively, relative to the risk for never-smokers in a zero-ETS
environment. The risk of never-smokers not exposed to spousal ETS (but exposed to background
ETS and nonsmoking causes) relative to the zero-ETS baseline risk is
(6-4)
The population-attributable risk of lung cancer in the total population (Levin, 1953) for a
source (risk factor) is a ratio. The numerators of the ratios for sources of tobacco smoke are:
current/former active smoking in ever-smokers,
(6-5)
6-11
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
background ETS plus spousal ETS in never-smokers exposed to both,
(6-6)
background ETS in never-smokers not exposed to spousal ETS,
1-P2XRR02/RR2-1).
(6-7)
The denominator for each term is their sum plus one, i.e.,
Ex(6-5) + Ex(6-6) + Ex(6-7) + 1
(6-8)
where Ex(6-5) refers to expression (6-5), etc. The population-attributable risk for remaining
causes of lung cancer (nontobacco-smoke-related background causes) is
l/Ex(6-8).
(6-9)
nontobacco-smoke-related causes:
ever-smoking:
spousal ETS:
background ETS:
Multiplying the population attributable risk for a source by the total number of LCDs
yields the number of LCDs attributable to that source. Alternatively, the source-attributable LCD
estimates can be derived by first calculating LCMRs. LCMRs are obtained for each source as
follows:
LCMRnt = 105Ex(6-9)T/N.
LCMRnt(RR01-l).
LCMRnt(RR02-RR03).
LCMRnt(RR03-l).
Then the number of LCDs attributable to a source is estimated by multiplying the LCMR for that
source by the total population at risk from that source.
We now consider parameter values for N, T, Pl5 P2, RRl5 and Z to be used with the value
1.19 for RR2, the pooled estimate of RR2 from the 11 U.S. studies (Table 5-17), for the
population risk assessment in Section 6.3.3. The value used for RR2 is then changed to 1.26, the
estimate from the Fontham et al. study in the United States, and a new value of Z is constructed
from the cotinine data in that study, for the alternative population risk assessment calculations in
Section 6.3.4. The female population in 1985 of age 18+ years of age is approximately 92 million
(U.S. DHHS, 1989, Chap. 3). Detailed census data by age for 1988 indicates that the proportion of
women 35+ years of age in the female population of age 18+ is 0.63 (U.S. Bureau of the Census,
1990). Applying that proportion to the 1985 population gives approximately 58 million women of
6-12
05/15/92
-------
DRAFT-DO NOT QUOTE OR CITE
aged 35+ in 1985, the value used for N. There were approximately 38,000 female lung cancer
deaths in the United States in 1985 (U.S. DHHS, 1989), which is used as the value for T.
Using figures from the Bureau of the Census and the 1979/80 National Health Interview
Survey, Arundel et al. (1987) estimate the number of women of age 35+ by smoking status,
obtaining a value of 0.443 as the fraction of ever-smokers. The National Center for Health
Statistics (as reported in U.S. DHHS, 1989) provides the proportion of the female population by
smoking status (never, former, current) for 1987. When applied to figures from the Bureau of the
Census (1990) for the female population by age group available for 1988, the same fractional value
(0.443) is obtained. These sources suggest that the proportion of ever-smokers in the female
population has been fairly constant between 1980 and 1987, so Pt will be given the value 0.443.
Multiplying N by Pj gives an estimate of Nj = 25.7 million ever-smokers, leaving N2 = 32.3
million never-smokers.
RRt applies to ever-smokers, which consist of current and former smokers. The relative
risks of current and former female smokers of age 35+ for the period 1982-1986 are estimated at
11.94 and 4.69, respectively, from data in the American Cancer Society's Cancer Prevention Study
II (CPS-II; as reported in U.S. DHHS, 1989). For 1985, the composition of ever-smokers is 63.4%
current smokers and 36.6% former smokers (CDC, 1989). Using those percentages to weight the
RRs for ever-smokers and former smokers gives 9.26, which will be used as the value of RRt.
The proportion of never-smokers exposed to spousal ETS in epidemiologic studies
typically refers to married persons, so we need to consider how to treat unmarried persons as well
in order to set a value for P2. The American Cancer Society's CPS-II (reported in Stellman and
Garfinkel, 1986) percentages for marital status of all women surveyed (not just never-smokers)
are: married, 75.3; divorced, 5.1; widowed, 14.6; separated, 0.8; and single, 4.2. Our estimates of
risk apply to married female never-smokers, about 75% of female never-smokers, so it is
necessary to consider exposure to ETS in the remaining 25% of unmarried never-smokers.
Cummings (1990) obtained urinary cotinine levels on a total of 663 self-reported never-
smokers and former smokers. The cotinine levels were slightly higher in males than in females
(9.6 and 8.2 ng/mL, respectively), and slightly more than one-half of the subjects were females.
The average cotinine level was 10.7 ng/mL for married subjects if the spouse smoked and 7.6
ng/mL otherwise. The average cotinine levels reported by marital status are: married, 8.3 ng/mL;
never married, 10.3 ng/mL; separated, 11.8 ng/mL; widowed, 10.4 ng/mL; and divorced, 9.2
ng/mL. The study, in which 7% of the subjects were of age 18 to 29, and 47% were of age 60 to
84, does not claim to be representative. Nevertheless, the results suggest that in terms of ETS
exposure, an unmarried never-smoker is probably closer, on average, to a never-smoker married
6-13
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
to a smoker (an exposed person) than to a never-smoker married to a nonsmoker (an unexposed
person). This observation is also consistent with the findings of Friedman et al. (1983).
The proportion of never-smoking controls exposed to spousal smoking varies .among
studies in the United States. If we exclude studies of uncertain representativeness (entered in
Table 5-14A under "representativeness"), the median value for the remaining studies is 0.6. From
the evidence on ETS exposure to unmarried female never-smokers, it is reasonable to assume that
their exposure to ETS, on average, is at least as large as the average background level plus 60% of
the average exposure from spousal smoking. For the calculations needed from these figures, this
assumption is equivalent to treating unmarried and married female never-smokers alike, in terms
of exposure to ETS (i.e., 60% exposed at a level equivalent to spousal smoking plus background
and 40% exposed at the background level only). Consequently, the value P2 = 0.6 is assumed to
apply equally to married and unmarried female never-smokers.
The NRC report of 1986 uses Z = 3 for the ratio of ETS exposure from spousal smoking
plus other sources to ETS exposure from nonspousal sources alone. That value was primarily
based on data from Wald and Ritchie (1984), for men in Great Britain, although Lee (1987b) had
reported a value of 3.3 for women in Great Britain. The results of Coultas et al. (1987) were also
considered, wherein a value of 2.35 was observed for saliva cotinine levels in a population-based
survey of Hispanic subjects in New Mexico. More recent data suggest that a lower value of Z
may be more accurate for the United States. The study of 663 volunteers in Buffalo, New York,
reported by Cummings et al. (1990), observed a value of 1.55 based on mean urinary cotinine
levels among married females (n = 225; Cummings, 1990). A study by Wall et al. (1988)
containing 48 nonsmokers observed a ratio of mean cotinine levels of 1.53. A survey of municipal
workers at a health fair found a cotinine ratio of 2.48 for the 112 women surveyed, but the
comparison is between women who shared living quarters with a smoker and those who did not
(Haley et al., 1989). The 10-country collaborative cotinine study conducted by IARC, (Riboli,
1990) collected urinary cotinine samples from nonsmoking women in four groups totaling about
100 each—married to a smoker (yes, no) and employed (yes, no)—including two locations, Los
Angeles and New Orleans, in the continental United States. The ratios of average
cotinine/creatinine concentrations for women married to a smoker to women not married to a
smoker range from 1.75 to 1.89 in New Orleans, when the percentage of women employed is
assumed to be between 25% and 75%. The data from Los Angeles contain an abnormally high
mean for women who are employed and also married to a smoker (a mean of 14.6 based on only
13 observations, compared to the other three means for Los Angeles of 2.1, 4.5, and 6.6), so only
the two means for unemployed women (married to a smoker and married to a nonsmoker) were
6-14
05/15/92
-------
DRAFT-DO NOT QUOTE OR CITE
used. The resultant ratio of cotinine/creatinine concentrations is 1.45. Data from the Fontham et
al. (1991) study of lung cancer and ETS exposure in five U.S. cities yield a Z of 2.0 based oh mean
urinary cotinine levels in 239 never-smoking women (data provided by Dr. Elizabeth Fontham).
Cotihine data exhibit variability both within and between subjects, as well as between
studies due to different experimental designs, protocols, and geographical locations (see also
Chapter 3). The Z values from recent U.S. studies mostly range between 1.55 and 2.0. A value of
1.75 for Z appears reasonable based on the available U.S. data and will be used in Section 6.3.3
along with the combined RR estimate from 11 U.S. studies (Chapter 5) to calculate ETS-
attributable lung cancer mortality estimates. Z = 2.0 and Z = 2.6, which is based on median
cotinine levels, will be used in Section 6.3.4 for alternative calculations of lung cancer mortality
based on the results of the Fontham et al. (1991) study. The sensitivity of the lung cancer
mortality estimates to changes in Z and other parameters is discussed in Section 6.3.5.
6.3.3. U.S. Lung Cancer Mortality Estimates Based on Results of Combined Estimates from
11 U.S. Studies
This section calculates ETS-attributable U.S. lung cancer mortality estimates based on the
combined relative risk estimate (RR2 = 1.19) derived in Chapter 5 for the 11 U.S. studies.
Alternatively, the estimate from just the combined Tier 1 and Tier 2 studies (RR2 =1.19 from 6
of the 11; see Table 5-17) could have been used since these six studies were assessed as having the
greater utility in terms of evaluating the lung cancer risks from ETS; however, the results would
be virtually the same since all that differs from these combinations is the confidence interval on
the relative risk. It was therefore decided to use the data from all the U.S. studies for the
purposes of the population risk assessment.
6.3.3.1. U.S. Lung Cancer Mortality Estimates for Female Never-Smokers
The parameter values presented in Section 6.3.2 are assumed along with RR2 = 1.19. For
Z = 1.75, RR02 = 1.59 (from expression 6-2, denoted hereafter as Ex(6-2); see also Table 6-1).
Given those parameter values, the formulae in Section 6.3.2 yield the estimated lung cancer
mortality for U.S. women in 1985 by smoking status (ever-smoker, never-smoker exposed to
spousal ETS, and never-smoker not exposed to spousal ETS) and source (nontobacco-smoke-
related causes, background ETS in never-smokers, spousal ETS in never-smokers, and ever-
smoking), as displayed in Table 6-2. The lung cancer mortality rate from nontobacco-smoke-
related causes (LCMRnt) is estimated to be 9.4 per 100,000, and is assumed to apply equally to all
persons in the target population, regardless of smoking status. The excess LCMR in never-
6-15
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
smokers from exposure to background ETS is 3.2, with an additional 2.4 if exposed to spousal
ETS. The excess LCMR in ever-smokers, which includes whatever effect exposure to ETS has on
ever-smokers as well as the effect from active smoking, is 120.8.
In rounded figures, 5,470 (14.4%) of the 38,000 LCDs in U.S. women of age 35+ in 1985
are unrelated to smoking (active or passive). The remaining 32,530 lung cancer deaths (85.6% of
the total) are attributable to tobacco smoke: 31,030 in 25.7 million ever-smokers and 1,500 in 32.3
million never-smokers. These 1,500 ETS-attributable LCDs in never-smokers account for about
one-third of all lung cancer deaths in female never-smokers. Of the 1,500 LCDs, about 1,030
(69%) are due to background ETS, and 470 (31%) are from spousal ETS. In summary, the total
38,000 LCDs from all causes is due to: nontobacco-smoke-related causes, 5,470 (14.4%),
occurring in ever-smokers and never-smokers; ever-smoking, i.e., the effects of past and current
active smoking as well as ETS exposure, 31,030 (81.7%), occurring in ever-smokers; background
ETS, 1,030 (2.7%), and spousal ETS, 470 (1.2%), occurring in never-smokers. In other words,
ever-smoking causes about 81.7% of the lung cancers in women of age 35+; exposure to ETS from
all sources accounts for some 3.9%; and causes unrelated to tobacco smoke are responsible for the
remaining 14.4%. The LCDs in never-smokers attributable to ETS equal about 5% (1,500/31,030)
of the total attributable to ever-smoking. Part of the mortality attributed to ever-smoking here,
however, is due to ETS exposure in former smokers, to be taken into account in Section 6.3.3.3.
6.3.3.2. V.S. Lung Cancer Mortality Estimates for Male Never-Smokers
There are 11 studies of exposure to ETS and lung cancer in males. The studies and their
respective relative risks are AKIB, 1.8; BROW, 2.2; BUFF, 33+ years exposure, 1.6; CORR, 2.0;
HUMB, 4.2; KABA, 1.0; LEE, 1.3; HIRA(Coh), 2.25; HOLE(Coh), 3.5; plus the data in Kabat
(1990), 1.2; and Varela (1987, Table 13 scaled down to 50 years of exposure), 1.2. (Data for
BROW, BUFF, and HUMB were supplied via personal communication from Drs. Brownson,
Buffler, and Humble). A weighted average of the passive smoking risk (RR2) from these 11
Studies is about 1.6. For the seven U.S. studies, BROW, BUFF, CORR, HUMB, KABA, Kabat
(1990), and Varela (1987), the weighted average RR is about 1.4, but this value is heavily
weighted (about 66%) by the Kabat (1990) and Varela studies, neither of which was used in the
analysis of the female data. The combined risk for the five U.S. studies not including Kabat
(1990) and Varela is about 1.8, but they are all small, low-weight studies. In any case the
observed relative risks for males appear to be at least as great as those for females.
When an attempt is made to correct the observed male risks for smoker misclassification,
however, using the procedures outlined in Appendix B and the community survey-based
6-16
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
misclassification factors for males (1.6% for current regular smokers, 15% for current occasional
smokers, and 5.9% for former smokers), it is found that for most of these cohorts, the number of
smokers misclassified as never-smokers either exceeds the relatively small number of observed
never-smokers, or is so great as to drive the corrected relative risk substantially below unity. This
implies that the misclassification factors from the community surveys are too high to accurately
correct the risks in the epidemiologic studies. Until better misclassification data on males are
available, no real sense can be made of the male passive smoking relative risks.
Under these circumstances, it was decided to apply the incremental LCMRs for spousal
and nonspousal ETS exposure in female never-smokers to male never-smokers. The incremental
LCMRs were used instead of the relative risk estimates because relative risk depends on the
background risk of lung cancer (from nontobacco-related causes) as well as the risk from ETS,
and background lung cancer risk may differ between females and males. From Section 6.3.3.1, the
LCMR from spousal ETS exposure was 2.4 per 100,000 at risk, and the LCMR from nonspousal
ETS exposure was 3.2 per 100,000. The 1985 male population age 35 and over is 48 million (U.S.
DHHS, 1989) of whom 27.2% (private communication from Dr. Ronald W. Wilson of the U.S.
National Center for Health Statistics), or 13.06 million, were never-smokers. Of these, 24%
(Wells, 1988), or 3.13 million, were spousally exposed. Applying the female ETS LCMRs, 3.13
million x 2.4/100,000 = 80 deaths in males from spousal ETS exposure and 13.06 million x
3.2/100,000 = 420 deaths from nonspousal exposure, for a total of 500 ETS-attributable LCDs
among never-smoking males. These estimates based on female LCMRs are believed to be
conservatively low because males generally have higher exposure to background ETS than females.
This would lead to lower Z values and subsequently higher estimates of deaths attributable to
background (nonspousal) ETS sources. In conclusion, confidence in these estimates for male
never-smokers is not as high as in those for female never-smokers.
6.3.3.3. U.S. Lung Cancer Mortality Estimates for Long-Term (5+Years) Former Smokers
There is a scarcity of data on the relative risks of lung cancer for former smokers exposed
to ETS. With former smokers, it would be difficult to know how much of the observed lung
cancer mortality is attributable to nontobacco-smoke-related causes, how much is due to ETS
exposure, and how much is accounted for by prior smoking. Consequently, observational data on
the number of lung cancers in former smokers are not utilized. Instead, long-term former
smokers are assumed to have the same lung cancer mortality rate from exposure to ETS as never-
smokers. Assuming that the residual excess risk of lung cancer from active smoking largely
diminishes in about five years, this analysis treats former smokers who have quit for less than
6-17
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
5 years the same as current smokers and those who have quit for longer periods the same as never-
smokers. Varela (1987) studied the relative risk for lung cancer from ETS exposure in 242 long-
term (10+ years) former smokers. He found that for total household exposure to ETS there was
very little difference between the relative risks for these long-term former smokers and the
never-smokers (see, for example, his Tables 12 and 13). There is still some uncertainty in the
application of these assumptions because the risk to long-term former smokers may not, in fact,
be the same as the risk to never-smokers. For example, ETS may have an additional promotional
effect on former smokers because of their previous exposures to high concentrations of
carcinogens from active smoking.
Female ever-smokers comprise about 44.3%, or 25.7 million, of the total U.S. female
i
population, age 35 and over, of 58 million. Long-term (5+ years) former smokers comprise about
34% of these ever-smokers (U.S. DHHS, 1990) or about 8.7 million. Using a 2.2 concordance
factor for former smokers married to ever-smokers versus never-smokers married to never-
smokers (See Appendix B), it is estimated that about 77% of the former smokers, or about 6.7
million, would be spousally exposed compared with the 60% for the never-smokers. Thus, based
on the LCMRs derived for female never-smokers, the expected number of ETS-attributable LCDs
for female long-term former smokers would be 6.7 million x 2.40/100,000 =160 deaths from
spousal exposure and 8.7 million x 3.20/100,000 = 280 deaths from nonspousal exposure, for a
total of 440.
Male ever-smokers comprise 72.8% of the U.S. male population, age 35 and over, of 48
million equal to 35 million of whom about 43% (derived from data in U.S. DHHS, 1990, page 60,
Table 5), or about 15 million, are 5+ year quitters. Of the never-smoking males, 24% were
married to smokers (Section 6.3.3.2). Again using a 2.2 concordance factor for former smokers, it
is estimated that 41% of the 15 million former smoking males, or 6.2 million, would be married to
ever-smokers. Applying the female never-smoker LCMRs from Section 6.3.3.1, 6.2 million x
2.40/100,000 - 150 deaths from spousal ETS exposure and 15 million x 3.20/100,000 = 480 deaths
from nonspousal ETS exposure for a total of 630 ETS-attributable LCDs among male long-term
former smokers.
Table 6-3 displays the resultant estimates for LCDs attributable to background ETS and
spousal ETS, for never-smokers and for former smokers who have quit for at least five years, by
sex. The LCMRs for background ETS and spousal ETS, assumed to be independent of smoking
status and sex, are the same as derived in Section 6.3.3.1 for females never-smokers (3.2 and 2.4,
respectively). Background ETS accounts for about 2,200 (72%) and spousal ETS for 860 (28%) of
the total due to ETS. Of the 3,060 ETS-attributable LCDs, about two-thirds are in females
6-18
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
(1,930; 63%) and one-third in males (1,130; 37%). More females are estimated to be affected
because there are more female than male never-smokers. By smoking status, two-thirds are in
never-smokers (2,000; 65%) and one-third in former smokers who have quit for at least 5 years
(1,060; 35%).
The numbers shown in Table 6-3 depend, of course, on the parameter values assumed for
the calculations. The sensitivity of the totals in Table 6-3 to alternative parameter values is
addressed in Section 6.3.5. First, however, tables equivalent to Tables 6-2 and 6-3 are developed
based on the FONT study alone, for comparison.
6.3.4. U.S. Lung Cancer Mortality Estimates Based on Results of the Fontham et al. (1991)
Study (FONT)
The estimate of RR2 (1.19), the risk of lung cancer to female never-smokers with spousal
ETS exposure relative to the risk for female never-smokers without spousal ETS exposure, used in
Section 6.3.3, is based on the combined outcomes of the 11 U.S. epidemiologic studies from
Chapter 5 (see Table 5-17). In this section the quantitative population impact assessment is
repeated with FONT, the single U.S. study with Tier 1 classification (Section 5.4.4), as the source
of the estimates of RR2 and Z (constructed from urine cotinine measures), with the remaining
parameter values left unchanged. While a single study has lower power and larger confidence
intervals on the relative risk estimate than can be obtained by combining the various U.S. studies,
using the specific data from a single study decreases the uncertainties inherent in combining
results from studies that are not fully comparable. FONT is the only study of passive smoking and
lung cancer that collected cotinine measurements, thus providing estimates for RR2 and Z from a
single study population. The total number of lung cancers attributable to total ETS exposure is
particularly sensitive to those two parameters (discussed in Section 6.3.5).
The NCI-funded Fontham et al. study (1991) is a large, well-conducted study designed
specifically to investigate lung cancer risks from ETS exposure (see also the critical review in
Appendix A). It addresses some of the methodological issues that have been of concern in the
interpretation of results regarding lung cancer and passive smoking: smoker misclassification, use
of surrogate respondents, potential recall bias, histopathology of the lung tumors, and possible
confounding by other factors (see also Sections 5.3, 5.4.2 and 5.4.3). Cases and controls were
drawn from five major cities across the United States (Atlanta, New Orleans, Houston, Los
Angeles, and San Francisco) and, hence, should be fairly representative of the general U.S.
population, at least of moderate climate urban areas. Furthermore, the results of the study are
consistent across the five cities.
6-19
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
In spite of the care incorporated into the FONT design to avoid smoker misclassification
bias, some might still exist; thus, the adjusted relative risk of 1.29 reported in FONT is "corrected"
slightly to 1.26 in this report. The parameter P2, the proportion of never-smokers exposed to
spousal ETS, was assigned the value 0.60 in the preceding section. In FONT, the observed
proportion of spousal-exposed controls is 0.60 (0.66) for spousal use of cigarettes only (any type of
tobacco) among colon-cancer controls and O.S6 (0.63) in population controls. Consequently, the
previous value of 0.60 is retained. Of the 669 FONT population controls, whose current cotinine
levels are considered the most representative of typical ETS exposure, there were 59 living with a
current smoker and 239 whose spouses never smoked. (The other 371 were nonsmoking women
who either no longer lived with a smoking spouse or whose spouse was a former smoker.) The
mean cotinine level for never-smoking women with spouses who are current smokers (n = 59) is
15.90 ± 16.46; the mean level for the other 239 was 7.97 (± 11.03). The ratio is 15.90/7.97, giving
Z « 2.0 (data provided by Dr. Elizabeth Fontham). The median is a measure of central tendency
that is less sensitive to extremes, so the ratio of median cotinine levels is also considered
(Z » 11.4/4.4 » 2.6). Results for both values of Z are displayed in Tables 6-4 and 6-5, which
correspond to Tables 6-2 and 6-3, respectively, of the previous sections for direct comparison.
The results of Section 6.3.2 are based on RR2 =1.19 (combined U.S. study results) and
Z « 1.75 (from studies on cotinine levels). In this section, RR2 and Z are both increased (RR2 to
1.26 and Z to 2.0 and 2.6). The change in RR2 increases the estimated number of LCDs from
background and spousal ETS, while increasing Z decreases the figure for background ETS and has
no effect on the number for spousal ETS (see Tables 6-2 and 6-4). Relative to the total for ETS
in the last section (3,060), the net effect is an increase of 8% to 3,300 at Z = 2.0 and a decrease of
19% to 2,480 when Z - 2.6. Subject to the accuracy of the parameter values assumed, these two
analyses support an estimate in the neighborhood of 3,000 total lung cancer deaths in never-
smokers and former smokers (quitters of 5+ years) from exposure to ETS in the United States for
1985.
The 3,000 figure is a composite value from estimates of varying degrees of uncertainty.
The confidence for the never-smoker estimates is highest. Comparing Tables 6-2 and 6-4 for
never-smokers, the lung cancer estimates for never-smoking females from exposure to spousal
ETS (470 - 610) are based on the direct evidence from epidemiologic studies and require the
fewest assumptions. Adding in a figure .for exposure to background ETS in never-smoking
females (640 - 1,030) is subject to the assumptions and other uncertainties attached to the estimate
of the parameter Z. The relative risk from'ETS exposure, which depends on the risk from
background sources of lung cancer as well as the risk from ETS, may differ in females and males.
6-20
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Consequently, the absolute risk (LCMR) in females was assumed to apply to males, adding 360 -
510 to the total (80 - 100 for spousal ETS and 260 - 420 for background ETS;: Table 6-3 and 6-5).
Males, however, are thought to have higher background exposures to ETS than females, so this
assumption is likely to underestimate the ETS-attributable lung cancer mortality in males. Also,
there is uncertainty attached to the number of male never-smokers and the proportion exposed to
ETS, which affect the figures for males.
The confidence in the estimates for former smokers is less than in those for never-
smokers. The estimates are probably low since they assume that ETS-attributable rates in never-
smokers and former smokers are the same. Figures for lung cancer mortality from ETS in former
smokers, for the same categories as never-smokers (i.e., females and males, background and
spousal ETS ) account for an additional 870 - 1,160 (totals of 310 - 410 for spousal ETS and 470 -
760 for background ETS, for both sexes). These figures for former smokers are summed from
appropriate entries in Tables 6-3 and 6-5 (Tables 6-2 and 6-4 do not make them explicit; they
are accounted for in the entry for lung cancer attributable to ever-smoking).
FONT is the largest study and therefore the dominant influence in the combined relative
risk from the 11 U.S. studies (RR2 = 1.19),. so the outcomes being compared here with those in
Section 6.3.3 are not independent. Similarly, the Z-value of 1.75 used with RR2 = 1.19 in the first
analysis is subjectively based on the outcomes of several U.S. cotinine studies, including the
FONT cotinine results. It is already apparent that the estimate of total lung cancer mortality
attributable to ETS is sensitive to the values of Z and RR2. Uncertainties associated with the
parameter values assumed and the sensitivity of the estimated total ETS-attributable LCDs to the
various parameter values are examined next. ....'.
6.3.5. Sensitivity to Parameter Values
The estimates for ETS-attributable lung cancer mortality are clearly sensitive to the
studies, methodology, and choice of models used, and previous methodologies have been presented
in Section 6.2. Even for this current model, however, estimates will vary with different input
values. Specifically, the estimates depend on the parameter values assumed for the total number
of lung cancer deaths from all sources (T), the population size (N), the proportion of ever-smokers
in the population (P^, the proportion of never-smokers exposed to spousal ETS (P2), the risk of
ever-smokers relative to never-smokers (RRj), the risk of never-smokers exposed to spousal ETS
relative to unexposed never-smokers (RR2), and- the ratio of ETS exposure from spousal smoking
and background (i.e., nonspousal) sources to background sources alone (Z).
6-21
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
The effects of changing several of the parameters is readily discernible. A change in T/N
produces a proportional change in the same direction for all estimates of attributable mortality. A
change in Pj creates a proportional change in the same direction in all mortality figures for ever-
smokers and a change in the opposite direction proportional to 1 - Pj in all estimates for never-
smokers. The parameter values assumed for these three parameters are from the sources described
in the preceding text and are assumed to be acceptably accurate. The value of P2 is assumed to be
0.6, but values between 0.5 and 0.7 are easily credible. At either of those extremes, there is a 17%
change in the lung cancer mortality due to spousal smoking, which only amounts to 80 for the first
analysis (Table 6-2) and 100 for the second one (Table 6-4). The impact of changing RRj, RR2,
or Z on the total lung cancer mortality attributable to ETS from the first analysis is displayed in
Table 6-6 for RRt from 8 to 11, RR2 between 1.04 and 1.35 (extremes of the 90% confidence
intervals for the 11 U.S. studies; Table 5-17) and for Z in the range 1.5 to 3.0.
For RRj in the interval (8,11), the total lung cancer mortality from ETS ranges from about
2,600 to 3,500, a 14% change in either direction relative to the comparison total of 3,060. The
extremes are much greater over the range of values considered for RR2 (1.04 to 1.35). At the low
end, where the excess relative risk from spousal ETS is only 4%, there is an 77% decrease in the
total lung cancer mortality to 700. The percentage change is roughly equivalent in the opposite
direction when the excess relative risk is at the maximum value 35%, for a total of 5,190. The
total is also sensitive to the value of Z. A decrease of only 0.25 from the comparison value of 1.75
increases the total by 36% to 4,160. A 36% decrease occurs at 2.5, leaving a corresponding
estimate of 1,950. At Z = 3.0, the total drops further to 1,680, a 45% decrease. Clearly, there are
fairly large swings in the estimated total number of lung cancer deaths attributable to ETS from
varying RR2 or Z by itself.
Varying more than one parameter value simultaneously may have a compounding or
canceling effect on the total lung cancer mortality due to ETS. For example, at the following
values of RR2, the range of percentage changes from the total of 3,060 ETS-attributable lung
cancer deaths for values of Z in the interval 1.50 to 3.0 are shown in parentheses: RR2 = 1.04 (-
69%, -88%), RR2 - 1.15 (+10%, -56%), RR2 = 1.25 (+73%, -30%), and RR2 = 1.35 (+131%, -7%).
The total ETS-attributable LCD estimates range from 380 (at RR2 = 1.04, Z = 3.0) to 7,060 (at
RR2 » 1.35, Z » 1.5). Without considering the additional variability that other parameters might
add, it is apparent that the estimated lung cancer mortality from ETS is sensitive to the parameters
RR2 and Z, and that the uncertainty in these parameters alone leaves a fairly wide range of
possibilities for the true population risk.
6-22
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
While various extreme values of these parameters can lead to the large range of estimates
noted, the extremities of this range are less likely possibilities for the true population risk because
the parameters RR2 and Z are not actually independent and would be expected to co-vary in the
same direction, not in the opposite direction as expressed by the extreme values. For example, if
the contributions of background ETS to total exposure decrease, Z would increase, and the
observable relative risk from spousal exposure, RR2, would be expected to increase as well.
Furthermore, most of the evidence presented in this report suggests that a narrower range of both
RR2 and Z are appropriate. Thus, while variations are possible, this report concludes that the
estimate of approximately 3,000 ETS-attributable LCDs based on the 11 U.S. studies is a
reasonable one, with the population risk analysis based on FONT providing a fairly reliable
range—2,500 to 3,300 ETS-attributable LCDs.
6.4. SUMMARY AND CONCLUSIONS ON POPULATION RISK
Having concluded in the previous chapter that ETS is a cause of lung cancer in humans
and belongs in EPA Group A of carcinogens, this chapter assesses the magnitude of that health
impact in the U.S. population. The ubiquity of ETS in a typical individual's living environment
results in the respiratory uptake of tobacco smoke to some degree in a very high percentage of the
adult population, conservatively upwards of 75% based on the outcome of cotinine/creatinine
studies in nonsmokers. Compared to observations on active smokers, urinary cotinine in
nonsmokers is small, on the order of a few percent, and there is considerable variability in inter-
individual metabolism of nicotine to cotinine. Some authors have used the relative cotinine levels
in active and passive smokers to estimate the probability of lung cancer in nonsmokers, by
extrapolating downward on a dose-response curve for active smokers. This "cigarette-equivalents"
approach requires several assumptions, e.g. that the dose-response curve used for active smokers is
reasonably accurate and low-dose extrapolation of risk for active smokers is credible, that cotinine
is proportional (and hence a substitute for) whatever is used for "dose" in the dose-response curve,
and that the risk calculated in this way applies equally to active and passive smokers with
equivalent cotinine measures. The effect of differences in physico-chemical properties of
mainstream smoke and sidestream smoke (the principal component of ETS), in lung dosimetry
between active and passive smoking, and in exposure patterns (related to concentration and
duration of exposure) are not fully understood, but the current state of knowledge casts doubts on
the validity of these assumptions.
The remaining approach to population risk extrapolates to the general population from the
epidemiologic evidence of increased relative risk of lung cancer in never-smoking women married
6-23
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
to smokers. To extrapolate exposure and consequent risk to other sources of ETS exposure,
cotinine levels of never-smokers exposed to spousal ETS are compared with those of never-
smokers exposed only to other sources of ETS (background), and it is assumed that excess risks of
lung cancer from ETS exposures, using cotinine levels as a surrogate measure, are proportional to
current ETS exposure levels. The use of current cotinine data to estimate ETS exposure in
nonsmokers seems reasonable because cotinine levels correlate quite well with questionnaire
response on ETS exposure. However, the total estimate of population risk is sensitive to
uncertainty in making these assumptions and variability in the use of cotinine measures.
This report uses the modeling approach based on direct ETS epidemiologic evidence
because the assumptions are fewer and more valid than for the "cigarette-equivalents" approach,
and the abundance of human data from actual environmental exposures makes this preferred
approach feasible. The total number of lung cancer deaths in U.S. females from all causes is
partitioned into components attributable to nontobacco-smoke-related causes (background causes
unrelated to active or passive smoking), background ETS (also called nonspousal ETS), spousal
ETS, and ever-smoking. Two sets of calculations are made for the U.S. female population of age
35+ in 1985 based on parameter values from national statistics and estimates from the
epidemiologic studies on ETS and lung cancer. They differ in the values assumed for two
parameters in the formulae for attributable risk: RR2, the relative risk of lung cancer for never-
smokers exposed to spousal smoke, and Z, the ratio of cotinine concentrations in never-smokers
exposed to spousal ETS to those exposed to background ETS only. The first analysis uses the
pooled estimate of RR2 from the 11 U.S. studies from Chapter 5, and a subjective value of Z
based on the outcomes of independent U.S. cotinine studies (RR2 =1.19 and Z = 1.75). The
second analysis uses the estimates of RR2 and Z from the large, high-quality Fontham et al. study
(1991), the sole U.S. study that collected cotinine data for its study population (RR2 = 1.26 with
mean Z » 2.0 and with median Z = 2.6).
The estimated lung cancer mortality in never-smoking women from ETS (background and
spousal ETS) is 1,500 in the first analysis and 1,630 (1,250) in the second analysis for Z = 2.0
(2.6). When estimates for never-smoking males and former smokers (5+ year quitters) of both
sexes are added, the corresponding totals are 3,060 and 3,300 (2,480). All of these figures are
based on calculations in which unknown parameter values are replaced with numerical estimates
which are subject to uncertainty, and departures in either direction cannot be precluded as
unrealistic possibilities for the correct population risks. Nonetheless, because of the large data
base utilized and the extensive analysis performed, there is a high degree of confidence in the
estimates derived for female never-smokers. The figures for male never-smokers and former
6-24
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
smokers of both sexes are subject to more uncertainty since more assumptions were necessary for
extrapolation from the epMemiologic results. The estimates for male never-smokers, in particular,
may be on the low side because males are generally exposed to higher levels of background ETS
than females. In summary, our analyses support a total of approximately 3,000 as an estimate for
the annual U.S. lung cancer deaths in nonsmokers attributable to ETS exposure, with 2,500 to
3,300 comprising a reasonable range of values. .- ••••-.
Despite some unavoidable uncertainties, we believe these estimates of ETS-attributable
lung cancer mortality to be fairly reliable, if not conservatively low with respect to the male
nonsmoker component. First, the weight-of-evidence that ETS is a human lung carcinogen is
very strong. Second, the estimates are based on a large amount of data from various studies of
human exposures to actual environmental levels of ETS. They do not suffer from a need to
extrapolate from an animal species to humans or from high to low exposures, as is nearly always
the case in environmental quantitative health risk assessment. Thus, the confidence in these
estimates is judged to be medium to high. In summary, the evidence demonstrates that ETS has a
very substantial and serious public health impact.
6-25
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
sS
u>
U)
II
W
_. fc.
s «
w to "S
•S <« 2
U CU m
3
U
O O U
*3 «n o
•- 3 -S
.S P _,
vo
^
CO
1
' *•*
•s
<«
0-
O
H
$
W
^g
£^
[INATOR
^
i
W
Q
1 i
OT 0
i CX
1 X
|W ^
o u
i *"
0 °
* Exposure
o
8
P
o
oo
00
CO
II
"ON
_•
II
S
&
CO
-^
II
S
P4
-
l—n
4-»
d
o
it
sf
2
^
II
r4
^
|
•
U
^
cs
II
i
i
i
SB]+[ETSS]
W
7?
£
52
I
g
£
1
O
Tf
•o
rr
CO
Rt J*j
1*»1
UH ^W
00"* co
d || >
lNt>
l£«
oo ••* d
i ^ g
> ll
eo S fe
>y CO ^
•^ 2 n
I_ O. o
O X <«
«« pa rr
S-o °^
••3 g =
&o
oo So o
OT o 00
^«s
-!<
S^.a
O CS o
t-i CO •*-?
Jl 1"®
I'-S"1"
•§•« -'
0^ v
43 O
i
CO
O
CO
1
•S-
2?
ii
vq t""1
o ^
11 *»
§ o
B &
11
o
SH W
00 E"1
rt T^
co^l
co ^^ 1-1
•t Q. 00
o oo
II g
u-> "3
CO co
00
CO
II
+
•g
oo
03
"S rt s
us
+
rTi
CO
(U
F«-l
rt
>
IH
Paramete
Population
I Estimates of
T3
d
d
CO
d
•2
'd
CD
0
CO
00
oj
d
of women i
Total number
ii
Z,
.2
'1
00
II
.2
1
CO
(N
CO
d
's
CN
52
S
1
CO
Ui
CO
rt
"3
roportion) o
•g
CO
1
1
o
•*-*
•o
CO
s
a
X
CO
NS women
Prevalence (pi
Proportion of
it
oT
ii
oT
PS
CO
onsmoker
_d,
_g
CO
"o3
.cotinine le'
o
43
CM
0
H
N
d
o
S
W
•a
d
Q
*-)
00
^
c^
43
O
*i
•s
CO
O
D.
X
CO
CO
to (nonsmokei
00
o
10
00
ON
00
CO
•4^
rt
•S
W
CO
o
1
00
J3
3
i2
H
H
6-26
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
*\
«
w
(J5
TS
*s
^
•§5
Wl ^3
4) d
l«5
i-j
2
«3
£ o
Gj 4)
>> "3
S
^* •*••
*"** 2
5 ^
o •-
S £
cu '^'
8|
a 4>
J5 "S
o
9> Di
13 a>
S €
if
•^ oo
^
vo
|
^
'
&
M "
i5
1
f&4
1
fan
e
»4
^
'
a
o
H
tso
-S
o
s •• S
C/3
T
/"•N >,
!«, (^
13
*72
B
. ^3
i^*v ja^
S' C/3
T>
O
Jj>
<^ ^/ jJ/J. (jjj)
*$s
'd
S
•w
•<** J3 ,>
s!
ll
ii
^
ll
f-^1
ss
o
ro
S
c\J
•
O
^
X— N /-^v
-H vq
f— < *— 1
O 0
T **G
-^' v_/
0 O 0
CN m ts
r^ 00^ -^
-4" -H1 of
(N OO ^N
°5 ' **} ^
c4 ON «-!
-H I— 1 CN
CO
s. § a
o
o
oo"
{^~
SS
0
CO
2-
^
ts"
o
^T
^^
fi
O
<=£
1— 1
^
^
•— . '
o
r—
*o
O
o
od
1
ob
*^i
o
S
en
1
t-i
V
Q
*J
(D
2
•o
«
rt
1
CD
£A
I
en
O
^8
d^
cu o
g^
c d
.§1
d-0
g s
E3
00 0
H &,
w^
8 8
_o
N o)
rt ^
•s.5
r o
6 -S »
1 ° «
£«<«•"
g S ° S3
•^ s a
* 2 c S
»^H ej M co
-^ C o> o
iPi
§l!i
3W-S »
'Percent of grand to
^S: never-smokers;
Background source;
"This figure attribut
6-27
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
oo—<
I
1
u
VO
6-28
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
B
o
>2
Wl
w *
S 2
-2 .3
| |
tl
•O 4>
if
tu S3
^^ -^5 »H
8 .Tfi
^t/3 «
.M,_*4>
•a^ *
1C ^ + '
O 1H VI
S M*S
a -3 &
u a «>
i« ^
u 00 X
UfiON a
§rt J
.3 c/T £
« 5 «.
2^ «
a w •*
TS T3 a
S w B
2 -tS U9
n *PM
« « S
T3& S
Sh Ita
O O
w «S «S
to
•
«
VO
W
J.
M
<
H
^3 O
w 1
11
1 w
jJ5 •^*s P?-
O
1
WJ
• ' S II
O
c
i m
CO
2
1 e%^
^ "s~
SW*' *r ilS
X>H
O f^
ll
- II
, i
W)
o S
II
x-s
0 b o r^
^00 O W3
o 2 o o
~i § tN OO
^ <~J, fO -*
^
n- oo
0 O O 0
•- 1 (N) fN O
^" NO f**} *~^
-
«N 00
(N ON °i T !
_ »> — M
O d) O
-------
DRAFT—DO NOT QUOTE OR CITE
•s
ti
o
o
vo
o
oo
tN
o
>n
en
o
oo
(N
o
O
tN
OO
oo
(S
vo'
I
CO
£
S2
0
s
n
CO
o
VO
en
00
CD
(1)
8 S
O CO
a "
OJ 4>
B S
••§«§
fiT.
rt W
S Bj"
w S
3
of
•-
a v
8 «
II
6-30
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
o
x
+*
o
1-1
€5'
<3
I
-
'
1
QO
>s
o
w
8 |
w
**•»
£ igj
£w
1
ft Is
C- ffi W
1
§t H*
I Tj <*v
- * s^ « i
ipj' o o 13 'S ts
w
s isf
£<*£-
1 1
g-0
oo
11
y^^v "~*s.
O 'N
oi "-H
oo ^o
^5 ^i
3o-
^-N
2 5
0 0
^O ^O
^5 ^^ ^5 ^^
^H ^0 ^H 00
o ^ o o o o
»— < »*N i— ^ »*^ (N| *N
CS oo ON
CT\ co ^o
-H -H CM
o 8
§ § 8
o
o
oo"
X- X ' \
CM'
^ /11
I
o§
6-31
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
a
'.2
1
en
i
a g
H
a *i
g
W g
i i
o o
*-< o
8 M
.ar
• ct
fli O
M O
o ^=>
y eo
ra IH
*O 4)
s
•8.S S
V
1
u
vo
V
S
" «.
" '
6-32
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
3 » 2
-•5 8
*- * >>
*>
« a
-------
DRAFT—DO NOT QUOTE OR CITE
o -
CO
§
o
o
0 O 00
00 tN. O "
II
CS
II
j3 t3t/j
i .^ i it,
a a>
« d
Ull
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
TABLE 6-6. Effect of single parameter changes on lung
cancer mortality due to ETS in never-smokers
and former smokers who have quit 5+ years
Parameter
Change
None4
Z = 1.50
1.75
2.00
2.25
2.50
2.75
3.00
RR2 = 1.04
1.05
1.10
1.15
1.19
1.20
1.25
1.30
1.35
RR, = 8.00
8.50
9.00
9.26
9.50
10.00
10.50
11.00
Background*
2,210
3,310
2,210
1,660
1,320
1,100
950
830 .
510
630
1,220
1,780
2,210
2,310
2,820
3,290
3,750
2,510
2,380
2,260
2,210
2,160
2,060
2,020
1,890
LCM Due to ETS
Spousal2
850
8.50
850
850
850
850
850
850
190
240
470
690
85(9
890
1,080
1,270
1,440
970
920
870
850
830
800
780
730
Total
3,060
4,160
3,060
2,510
2,170
1,950
1,800
1,680
700
870
1,690
2,470
3,060
3,200
3,900
4,560
5,190
3,480
3,300
3,130
3,060
2,990
2,860
2,800
2,620
Percent
Change*
0
+36
0
-18
-29
-36
-41
-45
-77
-72
-45
-19
0
+5
+27
+49
+70
+14
+8
+3
0
-2.
-7
-9
-14
'69,100,000 at risk.
235,400,000 at risk.
3Percent change from total shown in boldface. (The outcome
from Tables 6-2 and 6-3, using the 11 U.S. studies).
4Z=1.75, RR2=1.19, RR,=9.26.
6-35
05/15/92
-------
-------
DRAFT—DO NOT QUOTE OR CITE
7. PASSIVE SMOKING AND RESPIRATORY DISORDERS OTHER THAN CANCER
7.1. INTRODUCTION
In 1984, a report of the Surgeon General identified cigarette smoking as the major cause
of chronic obstructive lung disease in the United States (U.S. DHHS, 1984). The same report
stated that there is conclusive evidence showing that smokers are at increased risk of developing
respiratory symptoms such as chronic cough, chronic phlegm production, and wheezing (U.S.
DHHS, 1984). More recently, longitudinal studies have demonstrated accelerated decline in lung
function in smoking adults (Camilli et al., 1987). In children and adolescents who have recently
taken up smoking, several cross-sectional studies have found statistically significant increases in
the prevalence of respiratory symptoms (cough, phlegm production, and dyspnea [i.e., shortness of
breath]) (Seely et al., 1971; Bewley et al., 1973). Longitudinal studies have also demonstrated that,
among young teenagers, functional impairment attributable to smoking may be found after as
little as 1 year of smoking 10 or more cigarettes per week (Woolcock et al., 1984).
From a pathophysiologic point of view, smoking is associated with significant structural
changes in both the airways and the pulmonary parenchyma (U.S. DHHS, 1984). These changes
include hypertrophy and hyperplasia of the upper airway mucus glands, leading to an increase in
mucus production, with an accompanying increased prevalence of cough and phlegm. Chronic
inflammation of the smaller airways leads to bronchial obstruction. However, airway narrowing
may also be due to the destruction of the alveolar walls and the consequent decrease in lung
elasticity and development of centrilobular emphysema (Bellofiore et al., 1989). Smoking may also
increase mucosal permeability to allergens. This may result in increased total and specific IgE
levels (Zetterstrom et al., 1981) and increased blood eosinophil counts (Halonen et al., 1982).
The ascertained consequences of active smoking on respiratory health, and the fact that
significant effects have been observed at relatively low-dose exposures, leads to an examination
for similar effects with environmental tobacco smoke (ETS). Unlike active smoking, involuntary
exposure to ETS (or "passive smoking") affects individuals of all ages, and particularly infants and
children. An extensive analysis of respiratory effects of ETS in children suggests that the lung of
the young child may be particularly susceptible to environmental insults (NRC, 1986). Exposures
in early periods of life during which the lung is undergoing significant growth and remodeling
may alter the pattern of lung development and increase the risk for both acute and chronic
respiratory illnesses.
Acute respiratory illnesses are one of the leading causes of morbidity and mortality during
infancy and childhood. One-third of all infants have at least one lower respiratory tract illness
7-1
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
(bronchitis, bronchiolitis, croup, or pneumonia) during the first year of life (Wright et al., 1989),
whereas approximately one-fourth have these same illnesses during the second and third years of
life (Gwinn et al., 1991). The high incidence of these, potentially severe illnesses has an important
consequence from a public health viewpoint: Even small increases in risk due to passive exposure
to ETS would considerably increase the absolute number of cases in the first 3 years of life (see
Chapter 8). In addition, several studies have shown that lower respiratory tract illnesses occurring
early in life are associated with a significantly higher prevalence of asthma and other chronic
respiratory diseases and with lower levels of respiratory function later in life (reviewed
extensively by Samet and collaborators [1983]).
This chapter reviews and analyzes epidemiologic studies of noncancer respiratory system
effects of passive smoking, starting with possible biological mechanisms (Section 7.2). The
evidence indicating a relationship between exposure to ETS during childhood and acute
respiratory illnesses (7.3), middle ear disease (7.4), chronic respiratory symptoms (7.5), asthma
(7.6), sudden infant death syndrome (7.7), and lung function impairment (7.8) is evaluated.
Passive smoking as a risk factor for noncancer respiratory illnesses and lower lung function in
adults is also analyzed (7.9). Finally, a health hazard assessment and population impact is
presented in the next chapter.
7.2. BIOLOGICAL MECHANISMS
7.2.1. Plausibility
It is plausible that passive smoking may produce effects similar to those known to be
elicited by active smoking. However, several differences both between active and passive forms
of exposure and among the individuals exposed to them need to be considered.
The concentration of smoke components inhaled by subjects exposed to ETS is small
compared with that from active smoking. Therefore, effect will be highly dependent on the
nature of the dose-response curve (NRC, 1986). It is likely that there is a distribution of
susceptibility to the effects of ETS that may depend on, among other factors, age, gender, genetic
predisposition, previous respiratory history, and concomitant exposure to other risk factors for the
particular outcome being studied. The ability to ascertain responses to very low concentrations
also depends on the reliability and sensitivity of the instruments utilized.
Breathing patterns for the inhalation of mainstream smoke (MS) and ETS differ
considerably; active smokers inhale intensely and intermittently, and usually hold their breath for
some time at the end of inspiration. This increases the amount of smoke components that are
deposited and absorbed (U.S. DHHS, 1986). Passive smokers inhale with tidal breaths and
7-2
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
continuously. Therefore, patterns of particle deposition and gas diffusion and absorption differ
considerably for these two types of inhalation.
There are also important differences in the physicochemical properties of ETS and MS (see
Chapter 3). These have been extensively reviewed earlier by the National Research Council
(NRC, 1986) and the Surgeon General (U.S. DHHS, 1986). ETS is a combination of exhaled MS,
sidestream smoke (that is, the aerosol that is emitted from the burning cone between puffs), smoke
emitted from the burning side of the cigarette during puffs, and gases that diffuse through the
cigarette paper into the environment. This mixture may be modified by reactions that occur in
the air before involuntary inhalation. This "aging" process includes volatilization of nicotine,
which is present in the particulate phase in MS but is almost exclusively a component of the vapor
phase of ETS. Aging of ETS also entails a decrease in the mean diameter of its particles from 0.32
(im to 0.1-0.14 pm, compared to a mean particle diameter for MS of 0.4 /tm (NRC, 1986).
Individual and socioeconomic susceptibility may be important determinants of possible
effects of ETS on respiratory health. A self-selection process almost certainly occurs among
subjects who experiment with cigarettes, whereby those more susceptible to the irritant and/or
sensitizing effects of tobacco smoke either never start or quit smoking (the so-called "healthy
smoker" effect). Infants, children, and nonsmoking adults may thus include a disproportionate
number of susceptible subjects when compared to smoking adults. In addition, recent studies have
clearly shown that, as incidence and prevalence of cigarette smoking has decreased, the
socioeconomic characteristics of smokers have also changed. Among smokers, the proportion of
subjects of lower educational level has increased in the last 20 years (Pierce et al., 1989). The
female-to-male ratio has also increased (Fiore et al., 1989), and this is particularly true for young,
poor women, in whom incidence and prevalence of smoking has increased (Williamson et al.,
1989). It is thus possible that exposure to ETS may be most prevalent today among precisely those
infants and children that are known to be at a high risk of developing respiratory illnesses early in
life.
7.2.2. Effects of Exposure In Utero and During the First Months of Life
A factor that may significantly modify the effect of passive smoking (particularly in
children) is exposure to tobacco smoke components by the fetus during pregnancy. This type of
exposure differs considerably from passive smoking; in fact, the fetus (including its lungs) is
exposed to components of tobacco smoke that are absorbed by the mother and that cross the
placental barrier, whereas passive smoking directly affects the bronchial mucosa and the alveolus.
It is difficult to distinguish between the possible effects of smoking during pregnancy and those
7-3
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
of ETS exposure after birth. Some women may quit smoking during pregnancy, only to resume
after pregnancy is over. Most mothers who smoke during pregnancy continue smoking after the
birth of their child (Wright et al., 1991), and among those who stop smoking after birth, the
influence on that decision of events occurring shortly after birth (such as respiratory illnesses in
their child) cannot be excluded. Recall bias may also influence the results of retrospective studies
claiming differential effects on lung function of prenatal and postnatal maternal smoking habits
(Yarnell and St. Leger, 1979).
To attempt to circumvent these problems, researchers have studied infant lung function
shortly after birth (the youngest group of infants reported was 2 weeks old [Neddenriep et al.,
1990]), with the implication that subsequent changes encountered could be attributed mainly to
ETS exposures. However, the possibility that even brief exposures to ETS may affect the lungs at
a highly susceptible age may not be discarded. Maternal smoking during pregnancy needs to be
considered, therefore, as a potential modifier of the effect of passive smoking on respiratory
health, particularly in children.
Exposure to compounds present in tobacco smoke may affect the fetal and neonatal lung
and alter lung structure much like these same compounds do in smoking adults. Neddenriep and
coworkers (1990) studied 31 newborns and reported that those whose mothers smoked during
pregnancy had significant increases in specific lung compliance (i.e., lung compliance/lung
volume) at 2 weeks of age when compared with infants of nonsmoking mothers. The authors
concluded that exposure to tobacco products detrimentally affects the elastic properties of the
fetal lung. Although these effects could also be attributed to postnatal exposure to ETS, it is
unlikely that such a brief period of postnatal exposure would be responsible for these changes
affecting the lung parenchyma (U.S. DHHS, 1986).
There is evidence for similar effects in animal models of prenatal lung development.
Collins and associates (1985) exposed pregnant rats to MS during day 5 to day 20 of gestation.
They found that pups of exposed rats showed reduced lung volume, reduced number of lung
saccules, and reduced length of elastin fibers in the lung interstitium. This apparently resulted in
a decrease in lung elasticity: For the same inflation pressure, pups of exposed mothers had
significantly higher weight-corrected lung volumes than did pups of unexposed mothers. Vidic
and coworkers (1989) exposed female rats for 6 months (including mating and gestation) to MS.
They found that lungs of their 15-day-old pups had less parenchymal tissue, less extracellular
matrix, less collagen, and less elastin than found in lungs of control animals. This may explain the
increased lung compliance observed by Collins et al. (1985) in pups exposed to tobacco smoke
products in utero.
7-4
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Hanrahan and coworkers (1990) reported that infants born to smoking mothers had
significantly reduced levels of forced expiratory flows. They studied 80 mother/child pairs and
found significant correlations between the cotinine/creatinine ratio in urine specimens obtained
during pregnancy in the mother and maximal expiratory flows and tidal volumes at a
postconceptional age of 50 weeks or less in their children. They concluded that exposure due to
prenatal smoking diminishes infant pulmonary function at birth and, by inference, airway size.
These authors also measured maximal flows during tidal breathing in their subjects. At rather low
lung volumes, such as those present during tidal breathing, airway size and maximal flows are
both a function of lung elasticity. These results may thus be due to both a specific alteration of
the infant's airways and an increased lung compliance in infants whose lungs are small relative to
the infant's length.
It has also been suggested that the increased IgE levels observed in adult smokers may also
be present among fetuses whose mothers smoke during pregnancy. Magnusson (1986) reported
that cord serum levels of IgE and IgD were significantly higher for neonates whose mother
smoked during pregnancy, particularly if the neonates had no parental history of allergic
disorders. Cord serum levels of IgD (but not of IgE) were increased for neonates whose fathers
smoked, and this effect was independent of maternal smoking. A more recent study on a larger
sample (over 1,000 neonates) failed to find any significant difference in cord serum IgE levels
between infants (N = 193) of mothers who smoked during pregnancy and those (N = 881) of
mothers who did not (Halonen et al., 1991).
It has also been recently reported that the pulmonary neuroendocrine system may be
altered in infants whose mothers smoke during pregnancy. The pulmonary neuroendocrine
system, located in the tracheobronchial tree, consists of specialized cells (isolated or in clusters
called "neuroepithelial bodies") that are closely related to nerves. In humans, these cells increase
in number significantly during intrauterine development, reach a maximum around birth, and
then rapidly decline during the first 2 years of life. Their function is not well understood, but the
presence of potent growth factors and bronchoconstrictive substances in their granules suggests
that they play an important role in growth regulation and airway tone control during this period
of lung development (Stahlman and Gray, 1984). Chen and coworkers (1987) reported that
maternal smoking during pregnancy increases the size of infant lung neuroepithelial bodies and
decreases the amount of core granules present in them. Wang and coworkers (1984) had
previously reported that mother mice receiving tap water with nicotine during pregnancy and
during lactation had offspring with increased numbers of neuroepithelial bodies at 5 days of age
when compared to baby mice whose mothers were not exposed. Baby mice exposed to nicotine
7-5
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
only during pregnancy had neuroepithelial bodies of intermediate size with respect to these two
groups, whereas those exposed only during lactation had neuroepithelial bodies of normal size. By
age 30 days, only baby mice exposed to nicotine during both pregnancy and lactation had
neuroepithelial bodies that were larger than those of control animals. .
Activation of the pulmonary neuroendocrine system is not limited to ETS exposure; it is
activated by active smoking as well. Aguayo and collaborators (1989) reported that
bronchoalveolar lavage fluids obtained from healthy smokers have increased levels of
bombesin-like pep tides, which are a normal component and a secretion product of human lung
neuroendocrine cells (Cutz et al., 1981).
In summary, effects of maternal smoking during pregnancy on the fetus are difficult to
distinguish from those elicited by early postnatal exposure to ETS. Animal studies suggest that
postnatal exposure to tobacco products enhances the effects of in utero exposure to these same
products.
7.2.3. Long-Term Significance of Early Effects on Airway Function
By altering the structural and functional properties of the lung, prenatal exposure to
tobacco smoke products and early postnatal exposure to ETS increase the likelihood of more
severe complications during viral respiratory infections early in life. Martinez and collaborators
(1988a) measured lung function before 6 months of age and before any lower respiratory illness in
124 infants. They found that infants with the lowest levels for various indices of airway size were
3-9 times more likely to develop wheezing respiratory illnesses during the first year of life than
the rest of the population. The same authors (Martinez, 199la) subsequently showed that, in these
same infants with lower initial levels of lung function, recurrent wheezing illnesses were also more
likely to occur during the first 3 years of life. A similar study performed in Australia (Young et
al., 1990) confirmed that infants who present episodes of cough and wheeze during the first
6 months of life have lower maximal expiratory flows before any such illnesses develop.
The increased likelihood of pulmonary complications during viral respiratory infections in
infants of smoking parents has important long-term consequences for the affected individual.
There is considerable evidence suggesting that subjects with chronic obstructive lung diseases have
a history of childhood respiratory illnesses more often than subjects without such diseases
(reviewed by Samet and coworkers [1983]). Burrows and collaborators (1988) found that active
smokers without asthma (N » 41) who had a history of respiratory troubles before age 16 years
showed significantly steeper declines in FEVt (as a percentage of predicted) after the age of 40
than did nonasthmatic smokers without such a history (N = 396). Although these results may have
7-6
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
been influenced by recall bias, they suggest that lower respiratory tract illnesses during a period of
rapid lung development may damage the lung and increase the susceptibility to potentially
harmful environmental stimuli.
There is no information available on the degree of reversibility of changes induced by
exposure to ETS during early life. Longitudinal studies of lung function in older children have
shown, however, that diminished levels of lung function are found in children of smoking parents
at least until the adolescent years (see below).
7.2.4. Exposure to ETS and Bronchial Hyperresponsiveness
Bronchial hyperresponsiveness consists of an enhanced sensitivity of the airways to
pharmacologic or physical stimuli that normally produce no changes or only small decreases in
lung function in normal individuals. Subjects with bronchial hyperresponsiveness have significant
drops in airway conductance and maximal expiratory flows after inhalation of stimuli such as cold
air, hypertonic saline, nebulized distilled water, methacholine, or histamine. Bronchial
hyperresponsiveness is regarded as characteristic of asthma (O'Connor et al., 1989) and may
precede the development of this disease in children (Hopp et al., 1990). It has also been
considered as a predisposing factor for chronic airflow limitation in adult life (O'Connor et al.,
1989).
Recent studies of large population samples have shown that active smokers have increased
prevalence of bronchial hyperresponsiveness (Woolcock et al., 1987; Sparrow et al., 1987; Burney
et al., 1987) when compared with nonsmokers. This relationship seems to be independent of other
possible determinants of bronchial hyperresponsiveness (O'Connor et al., 1989). However, one
large study of almost 2,000 subjects from a general population sample failed to find a significant
relationship between smoking and prevalence of bronchial hyperresponsiveness (Rijcken et al.,
1987). The subjects involved in the latter study were younger and were therefore exposed to a
smaller average cumulative pack-years of smoking than were the subjects of studies in which a
positive relationship was found. This suggests that the relationship may be evident only among
individuals with a high cumulative exposure.
Epidemiologic studies have demonstrated that exposure to ETS is associated with an
increased prevalence of bronchial hyperresponsiveness in children. Murray and Morrison (1986),
in a cross-sectional study, reported that asthmatic children of smoking mothers were four times
more likely to show increased responsiveness to histamine than were asthmatic children of
nonsmoking mothers. O'Connor and coworkers (1987), in a study of a general population sample,
found a significant association between maternal smoking and bronchial hyperresponsiveness (as
7-7
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
assessed with eucapnic hyperpnea with subfreezing air) among asthmatic children, but not among
nonasthmatic children (Weiss et al., 1985). Martinez and coworkers (1988b) reported a fourfold
increase in bronchial responsiveness to carbachol among male children of smoking parents when
compared to male children of parents who were both nonsmokers. A smaller (and statistically not
significant) increase in bronchial responsiveness was reported in girls. These authors also found
that the effect of parental smoking was stronger in asthmatic children, and results were still
significant after controlling for this factor in a multivariable analysis. Because only a small
proportion of mothers in this population smoked during pregnancy, the effect was considered to
be associated mainly with exposure to ETS in these children. Lebowitz and Quackenboss (1990)
showed that odds of having bronchial reactivity (as assessed by the diurnal variability in maximal
expiratory flow rate) were 3.6 times as high among 18 children aged < 15 years who lived with
smokers of > 20 cigarettes per day than among 62 children of the same age who lived with
nonsmokers (95% CI - 1.2-10.6). Children living with smokers of 1 to 20 cigarettes per day had a
prevalence of bronchial reactivity that was similar to that of children living with nonsmokers.
There is, therefore, evidence indicating that parental smoking enhances bronchial
responsiveness in their children. The mechanism for this effect and the possible role of atopy in it
(see below) are unknown. The doses required to enhance bronchial responsiveness in children
exposed to ETS are apparently much lower than those required to elicit similar effects among
adult active smokers. A process of self-selection, by which adults who are more sensitive to the
effects of tobacco smoke do not start smoking or quit smoking earlier, may explain this finding.
Variations in bronchial responsiveness with age may also be involved (Hopp et al., 1985).
7.2.5. ETS Exposure and Atopy
Atopy has been defined epidemiologically as the presence of immediate hypersensitivity to
at least one potential allergen administered by skin prick test. Atopy is an immediate form of
hypersensitivity to antigens (called allergens) that is mediated by IgE immunoglobulin. Allergy (as
indicated by positive skin test reactivity to allergens, high levels of circulating IgE, or both) is
known to be present in almost all cases of childhood asthma. Recent epidemiologic studies have
indicated that an IgE-mediated reaction may be necessary for the occurrence of almost all cases of
asthma at any age (Burrows et al., 1989).
Although genetic factors appear to play a major role in the regulation of IgE production
(Meyers et al., 1987; Hanson et al., 1991), several reports have indicated that active smoking
significantly increases total serum IgE concentrations and may thus influence the occurrence of
allergy (Gerrard et al., 1980; Burrows et al., 1981; Zetterstrom et al., 1981; Taylor et al., 1985).
7-8
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Active smokers have also been found to have higher eosinophil counts and increased prevalence of
eosinophilia when compared to nonsmokers (Kauffmann et al., 1986; Halonen et al., 1982; Taylor
et al., 1985). The physical and chemical similarities between MS and ETS have prompted the
investigation of a possible role of passive smoking in allergic sensitization in children.
Weiss and collaborators (1985) first reported a 2.2-fold increased risk of being atopic in
children of smoking mothers. Martinez and coworkers (1988b) confirmed that children of
smoking parents were significantly more likely to be atopic than were children of nonsmoking
parents, and reported that this association was stronger for male children. They also found a
rough dose-response relationship between the number of cigarettes smoked by parents and the
intensity of the skin reactions to a battery of allergens. Ronchetti and collaborators (1990)
extended these findings in the same population sample of Martinez and coworkers. They found
that total serum IgE levels and eosinophil counts were significantly increased in children of
smoking parents, and the effect was related to both maternal and paternal smoking.
It is relevant to note that, due to the so-called "healthy smoker effect," children of smokers
should be genetically less sensitive than children of nonsmokers, because the latter are likely to
include a disproportionate number of allergic subjects who are very sensitive to the irritant effects
of smoke. As a consequence, the atopy-inducing effects of ETS may be substantially
underestimated. .
In summary, there is convincing evidence that both maternal smoking during pregnancy
and postnatal exposure to ETS alter lung function and structure, increase bronchial
responsiveness, and enhance the process of allergic sensitization. These changes elicited by
exposure to tobacco products may predispose children to lower respiratory tract illnesses early in
life, and to asthma, lower levels of lung function, and chronic airflow limitation later in life.
Most of these same effects have been described for active smoking in adults. These smoke-
induced changes are, therefore, known biological mechanisms for the increased prevalence of
respiratory diseases associated with ETS exposure described later in this chapter.
Exposure to tobacco smoke products during pregnancy and to ETS soon after birth may be
the most important preventable cause of early lung and airway damage leading to both lower
respiratory illness in early childhood and chronic airflow limitation later in life.
7.3. EFFECT OF PASSIVE SMOKING ON ACUTE RESPIRATORY ILLNESSES IN
CHILDREN
A review of the literature that examined the effects of exposure to ETS on the acute
respiratory illness experiences of children was contained in the Surgeon General's report on the
7-9
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
health consequences of involuntary smoking (U.S. DHHS, 1986) and in the report on
environmental tobacco smoke by the NRC (1986). Table 7-1 shows the studies referenced in these
two reports.
The Surgeon General's report concluded that "the results of these studies show excess acute
respiratory illness in children of parents who smoke, particularly in children under 2 years of age"
(page 44) and that "this pattern is evident in studies conducted with different methodologies and
in different locales" (page 44). It estimated that the increased risk of hospitalization for severe
bronchitis or pneumonia ranged from 20% to 40% during the first year of life. The report stated
that "young children appear to be a more susceptible population for the adverse effects of
involuntary smoking than older children and adults" (page 44). Finally, the report suggested that
"acute respiratory illnesses during childhood may have long-term effects on lung growth and
development, and might increase the susceptibility to the effects of active smoking and to the
development of chronic lung disease" (page 44).
The 1986 NRC report observed that "all the studies that have examined the incidence of
respiratory illnesses in children under the age of 1 year have shown a positive association between
such illnesses and exposure to ETS. The association is very unlikely to have arisen by chance"
(page 208). It pointed out that "some of the studies have examined the possibility that the
association is indirect by allowing for confounding factors . . . and have concluded that such
factors do not explain the results. This argues, therefore, in favor of a causal explanation" (page
208). The report concluded that "bronchitis, pneumonia, and other lower-respiratory-tract
illnesses occur up to twice as often during the first year of life in children who have bne or more
parents who smoke than in children of non-smokers" (page 217).
7.3.1. Recent Studies on Acute Lower Respiratory Illnesses
Several recent studies not referenced in the Surgeon General's Report or in the NRC
report have addressed the relationship between parental smoking and acute lower respiratory
illnesses in children (see Table 7-2).
Chen and coworkers (1986) studied 1,058 infants out of 1,163 infants born in a given
period in two neighborhoods in Shanghai, People's Republic of China. Information on hospital
admissions from birth to 18 months, smoking habits of household members, parental education,
and social and living conditions was obtained by use of a self-administered questionnaire
completed by the parents when the child reached 18 months of age. Hospital admissions were
divided into those due to respiratory illness and those from all other conditions. None of the
mothers in the study smoked. There was no statistically significant association between exposure
7-10
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
to ETS and admission to the hospital for any condition other than respiratory illnesses. Compared
to nonsmoking households, the risk of being admitted to a hospital for respiratory illnesses was
17% higher when 1 to 9 cigarettes were smoked daily by household members (95% C.I. = 0.6-2.3),
and was 89% higher when > 9 were smoked daily by household members (95% C.I. = 1.1-3.4). The
authors controlled for the effects of crowding, chronic respiratory illness in the family, father's
education, type of feeding, and birthweight.
Chen and coworkers (1988) subsequently studied 2,227 out of 2,315 children born in the
last quarter of 1983 in Chang-Ning District, Shanghai, People's Republic of China. There were
no smoking mothers in this population. The authors reported a significant linear relationship of
total daily cigarette consumption by family members with incidence density of hospitalization for
respiratory illness and with cumulative incidence of bronchitis and pneumonia in the first 18
months of life. The relationship was stronger for the 1- to 6-month period than for the 7- to
18-month period: When compared to households whose members did not smoke at home, the risk
of being hospitalized for respiratory illness during the 1- to 6-month interval was three times as
high (95% C.I. = 1.6-5.7) in households whose members smoked > 9 cigarettes at home, whereas
comparison of the same two types of household showed that the risk of being hospitalized for
respiratory illness during the 7- to 18-month interval was only 1.8 times as high (95%
C.I. = 1.0-3.2) in the smoking household. The relationship was also stronger among
low-birthweight infants. Results were independent of sex, birthweight, feeding practices, nursery
care, paternal education, family history of chronic respiratory diseases, and use of coal for
cooking.
In a different publication based on the same data from the 1988 study, Chen (1989)
reported that the effects of passive smoking were stronger in artificially fed infants than in
breast-fed infants. When compared to breast-fed infants of nonsmoking families, the risk of
being hospitalized for respiratory illness in the first 18 months of life was 1.6 times as high for
breast-fed infants of smoking families (> 19 cig./day), whereas the same risk was 3.4 times as high
among non-breast-fed infants of smoking families.
The studies by Chen and coworkers (1986, 1988, 1989) were retrospective in nature and
thus not immune to possible biases generated by the fact that the occurrence of the outcome event
may enhance reporting or recall of the conditions considered as risk factors. However,
conclusions are strengthened by the finding that admissions for nonrespiratory illnesses were
unrelated to passive smoking in the study in which the relationship was assessed (Chen, 1986) and
by the fact that the finding remained significant after adjusting for known confounders.
7-11
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Breese-Hall and coworkers (1984) studied 29 infants hospitalized with confirmed
respiratory syncytial virus (RSV) bronchiolitis before age 2, 58 controls hospitalized for acute
nonrespiratory conditions, and 58 controls hospitalized for acute lower respiratory illnesses from
causes other than RSV. Cases and controls were matched for age, sex, race, month of admission,
and form of payment for hospitalization. information on smoking habits in the family was
obtained at the time of each patient's admission. Cases were 4.8 times as likely as controls (95%
C.T. « 1.8-13.0) to have one or more household members who smoked 5 or more cigarettes per day.
However, there was no significant difference in the prevalence of cigarette smoking in the
households of subjects with respiratory illnesses caused by RSV and those not caused by RSV.
This Was attributable to the fact that the controls with respiratory illnesses not caused by RSV
were also much more likely to live with smokers of 5 or more cigarettes per day than; were
controls with nonrespiratory illnesses (OR = 2.7, 95% C.I. = 1.3-5.7). Little information is given
about enrollment and refusals; thus, it is not possible to know if selection bias may have
influenced the results. Also, other possible confounders such as socioeconomic level were not
taken into account when matching cases to controls or when data were analyzed.
McConnochie and Roghmann (1986a) compared 53 infants drawn from the patient
population of a group practice in Rochester, New York, who had physician-diagnosed
bronchiolitis before age 2 years, with 106 controls from the same practice who did not have lower
respiratory illnesses during the first 2 years of life and who were matched with cases for sex and
age. Parental interviews were conducted when the child had a mean age of 8.4 years. Parents
were asked about family history of respiratory conditions and allergy, socioeconomic status,
passive smoking, home cooking fuel, home heating methods, and household pets. Passive smoking
was defined as current and former smoking of "at least 20 packs of cigarettes or 12 oz of tobacco
while living in the home with the subject." Current and former smoking was scored equally,
based on the assumption that the report of either reflected passive smoking in the first 2 years of
life. Frequency of paternal smoking was not increased among children who had bronchiolitis.
Cases were 2.4 times (95% C.I. = 1.2-4.8) as likely to have smoking mothers as were controls. The
association was stronger in families with older siblings (OR = 8.9); however, a multiplicative test
for this interaction did not reach statistical significance. The authors studied 63% of eligible cases
and 34% of eligible controls. Although the reasons for exclusion from both groups are detailed,
selection bias cannot be completely excluded, and the authors give no information about maternal
smoking habits among excluded subjects. Also, overreporting of smoking by parents who were
aware of their child's history of bronchiolitis may have introduced biases due to differential
7-12
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
misclassification. However, the results were consistent across groups classified according to
family history of asthma or allergy, social status, presence of older siblings, and crowding.
Ogston and coworkers (1987) conducted a prospective study of 1,565 infants of
primigravidae enrolled antenatally in the Tayside Morbidity and Mortality Study in New Zealand.
Information on the father's smoking habits and on the mother's smoking habits during pregnancy
was obtained at the first antenatal interview and from a postnatal questionnaire. A summary
record was completed when the child was 1 year of age and included a report of the child's
respiratory illnesses (defined as "infections of the upper or lower respiratory tract") during the
first year of life derived from observations made by health visitors during scheduled visits to see
the child. The authors used a multiple logistic regression to control for the possible effects of
maternal age, feeding practices, heating type, and father's social class on the relationship between
parental smoking and child health. Of the 588 children of nonsmokers in this sample, 146 (24.8%)
had respiratory illnesses during the first year of life. Paternal smoking was associated with a 43%
increase (95% C.I. = 4.7%-96.1%) in the risk of having respiratory illnesses in the first year of life,
and this was independent of maternal smoking. The risk of having a respiratory illness was 82%
higher (95% C.I. = 25.6%-264.4%) in infants of smoking mothers than in infants of nonsmoking
parents. Smoking by both parents did not increase the risk of having respiratory illnesses beyond
the level observed in infants with smoking mothers and nonsmoking fathers. It is difficult to
compare this study with other reports on the same issue because the authors could not distinguish
between upper and lower respiratory tract illnesses.
Anderson and coworkers (1988) performed a case-control study of 102 infants and young
children hospitalized in Atlanta, Georgia, for lower respiratory tract illnesses before age 2 and 199
age- and sex-matched controls. The unadjusted relative odds of having any family member
smoking cigarettes were 2.0 times as high (p < 0.05) among cases as among controls (confidence
interval was not calculable from the reported data). The effect disappeared, however, after
controlling for other factors (prematurity, history of allergy in the child, feeding practices,
number of persons sleeping in the same room with the child, immunization of the child in the last
month) in a multivariable logistic regression analysis. No information is provided in this report
about maternal and paternal smoking separately, and the number of cigarettes smoked at home by
each family member was not recorded either. Also, almost 30% of all target cases declined
participation in the study, and no information was available on smoking habits in the families of
these children. No information is given about number of refusals among controls.
Woodward and collaborators (1990) obtained information about the history of acute
respiratory illnesses in the previous 12 months on 2,125 children aged 18 months to 3 years whose
7-13
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
parents answered a questionnaire mailed to 4,985 eligible families in Adelaide, Australia. A
"respiratory score" was calculated from responses to questions regarding 13 different upper and
lower respiratory illnesses. A total of 1,218 parents (57%) gave further consent for a home
interview. From this total, parents of 258 cases (children whose respiratory score fell in the top
20% of scores) and 231 "controls" (children whose scores were within the bottom 20% of scores)
were interviewed at home. When compared to controls, cases were twice as likely to have a
mother who smoked during the first year of life (95% C.I. = 1.3-3.4). This effect was independent
of parental history of respiratory illnesses, other smokers in the home, use of group child care,
parental occupation, and level of maternal stress and social support. The authors found no
differences in the way smokers and nonsmokers perceived or managed acute respiratory illnesses
in their children. Based on this finding, they ruled out that such differences could explain their
findings. They also reported that feeding practices strongly modified the effect of maternal
smoking; among breast-fed infants, cases were 1.8 times as likely to have smoking mothers as
were controls (95% C.I. = 1.2-2.8), whereas among non-breast-fed infants, cases were 11.5 times
as likely to have smoking mothers as were controls (95% C.I. = 3.4-38.5).
Wright and collaborators (1991) studied the relationship between parental smoking and
incidence of lower respiratory tract illnesses in the first year of life in a cohort of 847 white,
non-Hispanic infants from Tucson, Arizona, who were enrolled at birth and followed
prospectively. Lower respiratory illnesses were diagnosed by the infants' pediatricians. Maternal
and paternal smoking was ascertained by questionnaire. For verification of smoking habits, the
researchers measured cotinine in umbilical cord serum of a sample of 133 newborns who were
representative of the population as a whole. Cotinine was detectable in umbilical cord sera of all
infants whose mothers reported smoking during pregnancy and in 7 of 100 cord specimens of
infants whose mothers said they had not smoked during pregnancy. There was a strong
relationship between cotinine level at birth and the amount that the mother reported having
smoked during pregnancy.
Children whose fathers smoked were no more likely to have a lower respiratory tract
illness in the first year of life than were children of nonsmoking fathers (31.3% vs. 32.2%,
respectively). The incidence of lower respiratory tract illnesses was 1.5 times higher (95%
C.I. - 1.1-2.2) in infants whose mothers smoked as in infants whose mothers were nonsmokers.
This relationship became stronger when mothers who were heavy smokers were separated from
light smokers; 45.0% of children born to mothers who smoked > 20 cigarettes per day had a lower
respiratory illness, compared to 32.1% of children whose mothers smoked 1 to 19 cigarettes per
day and 30.5% of children of nonsmoking mothers (p < 0.05). The authors tried to differentiate
7-14
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
the effects of maternal smoking during pregnancy from those of postnatal exposure to ETS but
concluded that the amount smoked contributed more to lower respiratory tract illness rates than
did the time of exposure. The authors also found that maternal smoking had a significant effect
on the incidence of lower respiratory tract illnesses only for the first 6 months of life; the risk of
having a first lower respiratory illness between 6 and 12 months was independent of maternal
smoking habits. A logistic regression showed that the effect of maternal smoking was independent
of parental childhood respiratory troubles, season of birth, day-care use, and room sharing.
Feeding practices, maternal education, and child's gender were unrelated to incidence of lower
respiratory illnesses in this sample and were not included in the regression. The analysis also
showed a significant interaction between maternal smoking and day-care use; the effects of
maternal smoking were significant when the child did not use day care (OR = 2.7; 95% C.I. =
1.2-5.8) but were weaker and did not reach significance among infants who used day care (OR =
1.9; 95% G.I. = 0.9-4.0). The authors suggested that day-care use may protect against lower
respiratory illnesses by reducing exposure to ETS.
7.3.2. Summary and Discussion on Acute Respiratory Illnesses
Both the literature referenced in the Surgeon General's report (U.S. DHHS, 1986) and the
NRC report (1986) and the additional, more recent studies considered in this report provide strong
evidence demonstrating that children who are exposed to ETS in their home environment are at
considerably higher risk of having acute lower respiratory tract illnesses than are unexposed
children. Increased risk associated with ETS exposure has been found in different locales, using
different methodologies, and in both inpatient and outpatient settings. The effects are
biologically plausible (see Section 7.2). Several studies have also reported a dose-response
relationship between degree of exposure (as measured by number of cigarettes smoked in the
household) and risk of acute respiratory illnesses. This also supports the existence of a causal
explanation for the association.
The majority of studies found that the effect was stronger among children whose mothers
smoked than among those whose fathers smoked. This is further evidence in favor of a causal
explanation, because infants are generally in closer, more frequent, contact with their mothers.
There are now also fairly convincing data showing that the increased incidence of acute
respiratory illnesses cannot be attributed exclusively to in utero exposure to maternal smoke. In
fact, Chen and coworkers (1986, 1988, 1989) reported increased risk of acute respiratory illnesses
in Chinese children living with smoking fathers and in the total absence of smoking mothers. This
effect could also be attributed either to in utero exposure to the father's smoke or to an effect on
7-15
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
the father's sperm. This seems unlikely, however, because no such effects of parental smoking
during pregnancy have been described in similar studies performed in western countries.
Furthermore, Woodward and coworkers (1990) found that children of smoking mothers were
significantly more prone to acute respiratory illnesses even after mothers who smoked during
pregnancy were excluded from the analysis. This clearly suggests the existence of direct effects of
ETS exposure on the young child's respiratory health that are independent of in utero exposure to
tobacco smoke products.
There is also convincing evidence that the risk is inversely correlated with age; infants
aged 3 months or less are reported to be 3.3 times more likely to have lower respiratory illnesses if
their mothers smoke 20 cigarettes per day or more than are infants of nonsmoking mothers
(Wright et al., 1991). Increases in incidence of 50% to 100% (relative risks of 1.5-2.0) have been
reported in older infants and young children. The evidence for an effect of ETS is less persuasive
for school-age children, although trends go in the same direction as those reported for younger
children. This may be due to a decrease in illness frequency, to physiological development of the
respiratory tract or immune system with age, or to a decreased contact between mother and child
with age.
Reasonable attempts have been made in most studies to adjust for a wide spectrum of
possible confounders. The analyses indicate that the effects are independent of race, parental
respiratory symptoms, presence of other siblings, socioeconomic status or parental education,
crowding, maternal age, child's sex, and source of energy for cooking. One study (Graham et al.,
1990) also showed that the effect of ETS exposure on proneness to acute respiratory illnesses in
infancy and early childhood was also independent of several indices of maternal stress, lack of
maternal social support, and family dysfunction. Other factors, such as breastfeeding, decreased
birthweight, and day-care attendance, have been shown to modify the risk.
Some sources of bias may have influenced the results, but it is highly unlikely that they
explain the consistent association between acute lower respiratory illness and ETS exposure. With
one exception (Wright et al., 1991), all studies relied exclusively on questionnaires or interviews to
assess exposure. Although questions tend to be very specific, overreporting or more accurate
reporting of smoking habits by parents of affected children is possible, particularly in
case-control and retrospective studies. However, such a bias should affect both respiratory and
nonrespiratory outcomes, and at least two studies have shown no association between
nonrespiratory outcomes and ETS exposure (Chen et al., 1988; Breese-Hall et al., 1984). Selection
bias could not be excluded in some case-control studies, but satisfactory efforts were made to
avoid this source of bias in most studies.
7-16
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
7.4. PASSIVE SMOKING AND ACUTE AND CHRONIC MIDDLE EAR DISEASES
The Surgeon General's report (U.S. DHHS, 1986) and the NRC report (1986) reviewed five
studies demonstrating an excess of chronic middle ear disease in children exposed to parental
cigarette smoke (Table 7^3). Both reports conclude that the data are consistent with increased
rates of chronic ear infections and middle ear effusions in children exposed to ETS at home.
7.4.1. Recent Studies on Acute and Chronic Middle Ear Diseases
Several recent studies not referenced in the Surgeon General's report or in the NRC report
have addressed the relationship between parental smoking and middle ear illnesses in children
(Table 7-4).
Fleming and coworkers (1987) examined retrospectively risk factors for the acquisition of
infections of the upper respiratory tract in 575 children less than 5 years of age. Information on
smoking habits and on upper respiratory tract infections and ear infections in the 2 weeks prior to
interview was obtained from the child's guardian. The authors reported a 1.7-fold increase
(p = 0.01) in the risk of having an upper respiratory illness in children of smoking mothers when
compared to children of nonsmoking mothers. This effect was independent of feeding practices,
family income, crowding, day^care attendance, number of siblings aged less than 5 years, child's
age, arid race. The authors calculated that 10% of all upper respiratory illnesses in the population
were attributable to maternal smoking, a proportion that was comparable to that attributable to
day-care attendance. There was no relationship between maternal smoking and frequency of ear
infections in this population sample.
Willatt (1986) studied 93 children who were the entire group of children admitted to a
Liverpool hospital for tonsillectomy (considered an index of frequent upper respiratory or ear
infections) during a 3-month period, and 61 age- and sex-matched controls. The median age was
6.9 years (range 1.8-14.9). Parents were asked the number of sore throats in the previous 3
months and the smoking habits of all members of the household. There was a significant
relationship (p < 0.05) between number of episodes of sore throat and number of cigarettes
smoked by the mother. The effect was independent of birthweight, sex, child's age, feeding
practices, social class, crowding, and number of sore throats and tonsillectomies in other
household members. The relative odds of having a smoking mother were 2.1 times as high (95%
C.I. = 1.1-4.0) in children about to undergo tonsillectomy as in children not undergoing
tonsillectomy.
Tainio and coworkers (1988) followed 198 healthy newborns from birth to 2.3 years of age.
They recorded physician-diagnosed recurrent otitis media (defined as more than four episodes of
7-17
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
otitis media during the first 2 years or more than four episodes during the second year). Parental
smoking was more frequent (55%) among the infants with recurrent otitis media than in the
comparison group (33%; p < 0.05). The authors comment, however, that "parental smoking was
not a risk factor for recurrent otitis media," probably because there was no significant relationship
between parental smoking and recurrent otitis media using definitions of the latter that differed
from the one described above. No distinction was made in this study between the possible effects
of maternal and paternal smoking. In addition, the study sample was probably too small to obtain
reliable risk calculations.
Reed and Lutz (1988) studied 24 out of 70 eligible children who had been seen in a family
practice office for acute otitis media during a period of 4 months, and 25 out of 70 eligible
children who had been seen for other reasons. Forty-five of these children had tympanograms
performed and also had information on household smoke exposure. Prevalence of an abnormal
tympanogram (indicating the presence of middle ear effusion) was higher among children exposed
to smokers at home (OR = 4.86, 95% C.I. = 1.4-17.2). Results were independent of feeding
practices, history of upper respiratory illness in the past month, low socioeconomic status, sex,
age, and attendance at a day-care center. Only a small fraction of eligible subjects were included
in this study, and the possibility of selection bias as an explanation for the reported results cannot
be ruled out.
Hinton (1989) compared 115 children aged 1 to 12 years (mean = 5 years) admitted to a
British hospital for grommet insertion with 36 children aged 2 to 11 (mean = 6 years) with normal
ears who were taken from an orthoptic clinic. Prevalence of smoking was significantly higher in
parents of cases than in parents of controls (OR = 2.1, 95% C.I. = 1.0-4.5). Potential sources of
selection bias or selective misclassification cannot be determined from the data reported by the
author. No effort was made to control for possible confounders.
Teele and coworkers (1989) studied consecutively enrolled children being followed in two
health centers in Boston from shortly after birth until 7 years of age. Acute otitis media and
middle ear effusion were diagnosed by the child's pediatrician. Data were analyzed for 877
children observed for at least 1 year, 698 children observed for at least 3 yearSj and 498 children
observed until 7 years of age. A history of parental smoking was obtained when each child
became 2 years old. A parent was considered a smoker if he or she smoked more than one
cigarette per day. The child was considered exposed if either parent was a smoker. The authors
reported that the incidence of acute otitis media during the first year of life was 13% higher in
children of smoking parents when compared to children of nonsmoking parents (p < 0.05), but
statistical significance was no longer present after controlling for alleged confounders (site of
7-18
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
health care, season of birth, birthweight, socioeconomic status, presence and number of siblings,
room sharing, feeding practices, sibling or parental history of ear infection and allergic diseases).
Several of these variables could have not been confounders if they were not related to both
parental smoking and incidence of acute otitis media. Controlling for risk factors that are not
confounders may result in overcorrection. Parental smoking was not associated with an increased
risk for acute otitis media during the first 3 years or 7 years of life. Likewise, parental smoking
was associated with a significant increase in the number of days with middle ear effusion, but
only during the first year of life (p < 0.009), and the effect was no longer present after alleged
confounders were controlled for. The authors do not provide information on separate risks for
maternal and paternal smoking or on the incidence of acute otitis media and middle ear effusion
in children of heavy smokers.
Takasaka (1990) performed a case-control study on 201 children aged 4 to 8 in Sendai,
Japan. Sixty-seven subjects had otitis media with effusion, and the remaining 134 children were a
control group matched to cases by age, sex, and kindergarten class. The investigators found no
significant differences in prevalence of exposure to two or more household cigarette smokers
between children with and without otitis media with effusion (no information on either odds
ratios or C.I.s is given). The power of this study may have been too low to determine risk factors
for middle ear effusions reliably.
Corbo and coworkers (1989) examined 1,615 children aged 6 to 13 years who shared a
bedroom with siblings or parents in Abruzzo, Italy. Parents were asked if the child snored and the
frequency of snoring. Parents were asked about their own smoking habits; they were considered
moderate smokers if the summed total for both parents was fewer than 20 cigarettes per day and
heavy smokers if the summed total was 20 or more cigarettes per day. Prevalence of habitual
snoring in children increased slightly with the amount of cigarettes smoked by parents; children of
heavy smokers were 1.9 times as likely to be habitual snorers as children in nonsmoking
households (95% C.I. = 1.2-3.1), whereas children of moderate smokers were 1.8 times as likely to
be habitual snorers as children of nonsmoking parents (95% C.I. = 1.1-3.0). Habitual snorers were
more likely to have had a tonsillectomy, but only if their parents smoked. The authors suggested
that these results are plausible because adult smokers are also at increased risk of being habitual
snorers.
Strachan and collaborators (1989) performed tympanograms and collected saliva for
cotinine determinations in 736 children in the third primary class (ages 6i to 7i years) in
Edinburgh, Scotland. Median of salivary cotinine concentrations was 0.19 ng/mL for 405 subjects
living with no smoker, 1.8 ng/mL for 241 subjects living with one smoker, and 4.4 ng/mL for 124
7-19
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
subjects living with & two smokers. For a given number of smokers in the household, girls had
higher cotinine levels than boys and children living in rented houses (i.e., of lower socioeconomic
level) had higher cotinine levels than children living in houses owned by their parents. The
authors found a linear relation between the logarithm of the salivary cotinine concentration and
the prevalence of middle ear effusion. The authors calculated odds ratios for abnormal
tympanometry relative to children with undetectable cotinine concentrations, after adjustment for
sex, housing tenure (rented or owned), social class, crowding, gas cooking, and the presence of
damp walls. The odds ratio for a doubling of salivary cotinine concentration was 1.14 (95%
C.I. « 1.03-1.27). At a salivary cotinine concentration of 1 ng/mL, the odds ratio of having an
abnormal tympanogram was 1.7, whereas an odds ratio of 2.3 was calculated for a cotinine level of
5 ng/mL. At least one-third of all cases of middle ear effusion may have been attributable to
passive smoking.
7.4.2. Summary and Discussion of Middle Ear Diseases
There is some evidence suggesting that the incidence of acute upper respiratory tract
illnesses and acute middle ear infections may be more common in children exposed to ETS.
However, several studies have failed to find any effect. In addition, the possible role of
confounding factors, the lack of studies showing clear dose-response relationships, and the
absence of a plausible biological mechanism preclude more definitive conclusions.
Available data provide good evidence demonstrating a significant increase in the
prevalence of middle ear effusion in children exposed to ETS. Several studies in which no
significant association was found between ETS exposure and middle ear effusion were not
specifically designed to test this relationship, and, therefore, either power was insufficient or
assessment of the degree of exposure was inadequate. Also, Iversen and coworkers (1985), who
assessed middle ear effusion objectively, suggested that the risk associated with passive smoking
increased with age. This may explain the negative results of several studies based on preschool
children; the sample sizes of these studies may have been inadequate to test for increased risks of
50% or less, as would be expected in children < 6 years of age. The finding of a log-linear
dose-response relationship between salivary cotinine levels and the prevalence of abnormal
tympanometry in one study (Strachan et al., 1989) adds to the evidence favoring a causal link.
Although not all studies adjusted for possible confounders, and selection bias cannot be excluded
in the case-control studies reviewed, the evidence as a whole suggests that the association is not
likely to be due to chance, bias, or factors related to both ETS exposure and middle ear effusion.
7-20
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
The biological mechanisms explaining the association between ETS exposure and middle
ear effusion require further elucidation. Otitis media with effusion is usually attributed to a loss
of patency of the eustachian tube, which may be enhanced by upper respiratory infection,
impaired mucociliary function, or anatomic factors (see Strachan et al., 1989). It is possible that
pharyngeal narrowing by adenoidal tissue (and, consequently, eustachian tube dysfunction) may
be more common in these children. This is suggested by reports of a higher prevalence of
maternal smoking among children about to undergo or who have undergone tonsillectomy and by
an increased prevalence of habitual snoring among children of smoking parents. Impaired
mucociliary clearance has been convincingly demonstrated in smoking adults (U.S. DHHS, 1984).
No data are available on mucociliary transport in children exposed to ETS. However, ETS may
affect mucociliary clearance in children as in adults. If this were the case, and if normal
mucociliary clearance is required for rapid resolution of otitis media, exposure to ETS could result
in increased prevalence of chronic middle ear effusion.
The increased prevalence of middle ear effusion attributable to ETS exposure has very
important public health consequences. Middle ear effusion is the most common reason for
hospitalization of young children for an operation, and thus imposes a heavy financial burden to
the health-care system (Black, 1984). There is also evidence suggesting that hearing loss
associated with middle ear effusion may have long-term consequences on linguistic and cognitive
development (Maran and Wilson, 1986).
7.5. EFFECT OF PASSIVE SMOKING ON COUGH, PHLEGM, AND WHEEZING
Studies addressing the effects of passive smoking on frequency of chronic cough, phlegm,
and wheezing were reviewed both in the Surgeon General's report (U.S. DHHS, 1986) and in the
report by the NRC (1986) (see Table 7-5).
The Surgeon General's report concluded that children whose parents smoke were found to
have 30% to 80% excess prevalence of chronic cough or phlegm compared with children of
nonsmoking parents. For wheezing, the increase in risk varied from none to over sixfold among
the studies reviewed. The report noted that the association with parental smoking was not
statistically significant for all symptoms in all studies, but added that the majority of studies
showed an increase in symptom prevalence with an increase in the number of smoking household
members in the home. The report stated that the results of some studies could have been
confounded by the child's own smoking habits, but noted that many studies showed a positive
association between parental smoking and symptoms in children at ages before significant
experimentation with cigarettes is prevalent. The report concluded that "chronic cough and
7-21
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
phlegm are more frequent in children whose parents smoke compared to nonsmokers. The
implications of chronic respiratory symptoms for respiratory health as an adult are unknown and
deserve further study" (page 107). .
The NRC report concluded that "children of parents who smoke compared with children
of parents who do not smoke show increased prevalence of respiratory symptoms, usually cough
sputum and wheezing. The odds ratios for the larger studies, adjusted for the presence of parental
symptoms, were 1.2-1.8, depending on the symptoms. These findings imply that ETS exposures
cause respiratory symptoms in some children" (page 216).
7.5.1. Recent Studies on the Effect of Passive Smoking on Cough, Phlegm, and Wheezing
Several recent studies not considered either in the NRC report (1986) or in the Surgeon
General's report (U.S. DHHS, 1986) have addressed the relationship between passive smoking and
respiratory symptoms in children (Table 7-6).
McConnochie and Roghmann (1986b) studied 223 out of 276 eligible children aged 6 to 10
years without a history of bronchiolitis who were drawn from the patient population of a group
practice in Rochester, New York. Information regarding the child's history of wheezing in the
previous 2 years, socioeconomic status, family history of respiratory illnesses, and smoking in the
household was obtained by questionnaire. Information on breastfeeding was obtained by record
checks and interviews. Children whose mothers smoked were more likely to be current wheezers
than were children whose mothers did not smoke (OR = 2.2, 95% C.I. = 1.0-4.8). Neither paternal
smoking nor total household smoking had any influence on the prevalence of wheezing. When the
authors controlled for family history of respiratory allergy, direct effects of maternal smoking on
prevalence of wheezing failed to reach statistical significance. However, there was a strong
association between maternal smoking and wheezing among children with a positive family history
of respiratory allergy (OR = 4.5, 95% C.I. = 1.7-12.0), and the interaction between these terms was
highly significant in multivariable analysis, suggesting the combined importance of both genetic
factors and maternal smoking.
Park and Kim (1986) studied 3,651 children aged 0 to 14 from a randomized, clustered
sample of households in South Korea (response rate: 89%). A questionnaire was administered to
household members about their smoking habits and respiratory symptoms. Mothers answered
questions about the presence of cough in the child in the 3 months prior to interview. The authors
reported dose-response relationships between the child's cough and number of smokers in the
family, number of smokers in the same room, number of cigarettes smoked by all family
members, and number of cigarettes smoked by parents. The relationship was present in children
7-22
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
of different ages (less than 5 years, 6 to 11 years, and 12 to 14 years). The authors controlled for
parental education, socioeconomic status, birth rank, parental age, birth interval, number of
family members, and number of siblings. Family members with cough or with morning phlegm
production were significantly more likely to live with children with cough. After correcting for
these two factors, chronic cough was 2.4 times as likely in children of families whose members
smoked 1 to 14 cigarettes per day (95% C.I. = 1.4-4.3) and 3.2 times as likely in children of
families whose members smoked more than 15 cigarettes per day (95% C.I. = 1.9-5.5). However,
effects were more noticeable and only reached statistical significance in children of families
whose adult members did not have chronic cough.
Bisgaard and coworkers (1987) studied 5,953 infants out of a total of 8,423 eligible
newborns (71%) enrolled in a prospective study. At the age of 1 year, the child's mother was
interviewed regarding episodes of wheeze during the previous year and possible risk factors for
wheezing. The risk of wheezing was 2.7 times as high (95% C.I. = 1.8-4.0) in children whose
mothers smoked S 3 cigarettes per day as in children whose mothers smoked < 3 cigarettes per
day. Results were independent of social status and sex of the child. The authors decided not to
control for quarter of birth or use of day-care facilities, with the assumption that these factors did
not modify the relationship between maternal smoking and wheezing. Also, biases could have
been introduced by the fact that almost one-third of the original sample was not included in the
analysis.
Geller-Bernstein and coworkers (1987) studied 80 children aged 6 to 24 months who had
been seen as outpatients or inpatients in Israel for wheezing and who had a diagnosis of atopy.
The children were examined every 6 months during 4 years by a physician. At the end of
assessment, the authors classified children as having "recovered" if they had been symptom-free
for at least 1 (the last) year; otherwise they were classified as "persistent wheezers." "Persistent
wheezers" were more likely to have smoking parents than were "recovered" children (OR = 3.1,
95% C.I. = 1.1-8.9). This result was independent of changes in IgE levels during the study period.
The authors did not control for the possible confounding effect of parental symptoms.
Cogswell and coworkers (1987) studied 100 newborns who had at least one parent with a
history of hay fever or asthma. Ninety-two children were still being followed at 1 year of age and
73 at the age of 5 years. Children were examined periodically and whenever they had signs of
respiratory illness. At the child's first birthday, the number of those who had developed wheezing
was equally distributed between parents who did or did not smoke. By the age of 5 years,
however, 62% of parents who smoked had children who had wheezed compared with 37% in
nonsmoking families (p < 0.05). It is unlikely that these results can be explained by the
7-23
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
confounding effect of parental symptoms, because all parents were allergic by definition. It is
also quite unlikely that preferential withdrawal of nonwheezing children of smoking parents could
have biased the results.
Toyoshima and coworkers (1987) from Osaka, Japan, followed 48 out of 65 wheezy infants
and children less than 3 years old for up to 4 years. Outcome information was obtained from
charts or by telephoning the child's mother. Among 18 children who were still symptomatic 25 to
44 months after their first visit, 17 lived with smokers, compared to 13 of 22 children who lived
with smokers and who stopped having symptoms during follow-up (OR = 11.8, 95% C.I. =
1.3-105.0). Results were independent of family history of allergy, feeding practices, and
disturbances at birth. Selection bias related to the number of subjects lost for follow-up or with
missing information could have influenced the results of this study.
Tsimoyianis and collaborators (1987) evaluated the effects of exposure to ETS on
respiratory symptoms in a group of 12- to 17-year-old high school athletes (N = 193). Histories
of smoking by all household members were obtained for all subjects. Athletes exposed to ETS at
home were more likely to report cough than were unexposed athletes (p = 0.08). Frequency of
bronchitis, wheeze, and shortness of breath were similar in both groups. A greater awareness of
the smoking habits of those around them by subjects with cough cannot be excluded as an
explanation of these findings, but this source of bias cannot explain the exposure-response trends
for ETS and lung function seen in this same sample (see Section 7.7.1).
Andrae and collaborators (1988) mailed questionnaires to the parents of 5,301 children
aged 6 months to 16 years living in the city of Norrkoping, Sweden. Data were obtained from
4,990 children (94% response rate). Children with parents who smoked had exercise-induced
cough more often than did children of nonsmokers (OR = 1.4, 95% C.I. = 1.1-1.8). Exposure to
ETS interacted with living in houses with damage by dampness; children exposed to both had
more exercise-induced cough and allergic asthma when compared to those exposed to only one or
neither. Results of this cross-sectional study may have been biased by preferential reporting of
symptoms by smoking parents, although a reliability study performed in a random sample was
reported to confirm 95% of the answers regarding respiratory symptomatology. In addition, no
effort was made to control for active smoking in older children.
Somerville and coworkers (1988) enrolled 88% of 8,118 eligible children aged 5 to 11 from
England and Scotland. Data on the child's respiratory symptoms and parental smoking were
obtained from a self-administered questionnaire completed by the child's mother. After
exclusions for missing data, the proportions of children available ranged from 60.9% to 63.9% of
all subjects, depending on the variables involved. Logistic regression analysis was used to control
7-24
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
for child's age, presence of siblings, one- or two-parent families, paternal employment, social
class, maternal smoking during pregnancy, overcrowding, maternal education, maternal age,
triceps skinfold thickness, and birthweight. For Scottish children (who were only 19% of all
subjects), the authors found a significant relationship between number of cigarettes smoked at
home and "chest ever wheezy" (p < 0.01; OR not reported). Among English children, there was a
significant relationship between number of cigarettes smoked at home by mother and father
together and prevalence of a wheezy or whistling chest most nights (adjusted OR in children
whose parents smoked 20 cig./day = 1.6; 95% C.I. = 1.2-2.2). Attacks of bronchitis and cough
during the day or at night were also significantly correlated with number of cigarettes smoked by
parents in the English sample; odds ratios in children of parents who smoked 20 cigarettes per day
were 1.4 and 1.3, respectively, but no confidence intervals were reported. The authors concluded
that the effect of parental smoking on respiratory symptoms in this age group is small and requires
a large number of subjects to be detected.
Rylander and collaborators (1988) from Stockholm, Sweden, studied 67 children aged
4 to 7 years who had been hospitalized with virologically proven RSV infections before age 3.
Questionnaires were mailed to parents regarding their smoking habits and the child's history of
wheezing illnesses after the initial episode. Children who had subsequent occasional wheezing
(n = 21) were more likely to have smoking parents than those (n = 24) who had no subsequent
respiratory symptoms (OR = 4.3, 95% C.I. = 1.1-16.4). However, frequency of parental smoking
among children who had no subsequent respiratory symptoms was not significantly different from
that of children who had subsequent recurrent wheezing. The inconsistency of the results in this
study may be explained by the small number of subjects involved.
Strachan (1988) studied 1,012 out of a target sample of 1,095 schoolchildren aged 6.5 to 7.5
years in Edinburgh, Scotland. Parents answered a questionnaire on their smoking habits and on
respiratory symptoms in their children. There was no relationship between number of smokers in
the household and prevalence of wheezing in the population. Cough at night (> 3 nights in the
past month) was more likely to occur in children living with one smoker (OR = 1.6; 95% C.L =
1.1-2.6) or two smokers (OR = 2.5; 95% C.I. = 1.5-4.0) than in children living with nonsmokers.
Occurrence of "chesty colds" in children was also more frequent in households with one (OR = 1.3;
95% C.I. = 0.9-1.9) or two smokers (OR = 1.9; 95% C.I. = 1.3-3.0).
A subsequent report (Strachan et al., 1990) based on the same population sample studied
the relationship between salivary cotinine levels and respiratory symptomatology in a subset of
770 children (see also Strachan et al. [1989]), Section 5.4.1). The authors found no relationship
between cotinine levels and wheezing or frequent night cough. Frequency of chesty colds was
7-25
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
significantly correlated with quintals of salivary cotinine (p < 0.01). The authors noted that
objective markers of recent exposure to ETS may not adequately reflect exposure at some critical
period in the past. They also noted that there may be different ways of understanding the concept
of "wheezing" and proposed that this could explain the lack of association between this symptom
and both questionnaire-based and cotinine-based assessment of exposure to ETS in their sample.
Lewis and coworkers (1989) performed a case-control study of risk factors for chronic
cough in children under 6 years in Salford, United Kingdom. They enrolled 60 children referred
to a pediatric outpatient clinic with cough lasting more than 2 months or frequent episodes of
cough without wheeze. These 60 subjects were compared with controls admitted for routine
surgical procedures. Children with chronic cough were 1.7 times (95% C.I. = 0,8-3.5) as likely to
Jive with a smoker as were controls. Because of the small number of subjects and the high,
prevalence of parental smoking (> 50%), the power of this study may have been too low to allow
for meaningful conclusions.
Neuspiel and coworkers (1989) studied 9,670 out of 9,953 eligible children enrolled at
birth in Great Britain. Information on parental smoking was obtained at birth, at age 5 years, and
at age 10 years. Outcome data were obtained from maternal interviews when the children were 10
years old. Children of smoking mothers had 11% higher risk (95% C.I. = 2%-21%) of wheezing
between ages 1 and 10 than did children of nonsmoking mothers. An exposure-response
relationship was also present: Cumulative incidence was 5.2% in children whose mothers were
nonsmokers, 6.6% in children whose mothers smoked 1 to 4 cigarettes per day, 7.5% in children
whose mothers smoked 5 to 14 cigarettes per day, 8.1% in children whose mothers smoked
15 to 24 cigarettes per day, and 8.9% in children whose mothers smoked > 24 cigarettes per day.
The risk was also increased in children of mothers who did not smoke during pregnancy but were
smokers thereafter (RR = 2.2, 95% C.I. = 1.2-3.9). The association persisted after a logistic
regression model was used to control for the effect of child's sex, child allergy, paternal smoking,
parental allergy, crowding, bedroom dampness, feeding practices, gas cooking, and social status.
The increase in risk was cut approximately in half but did not disappear when additional
corrections for maternal respiratory symptoms and for a measure of maternal depression were
made. Results of this study may be explained in part by preferential reporting of wheezy illnesses
j
by smoking mothers. However, it is unlikely that the association between maternal smoking and
wheezy illnesses found in this study can be exclusively explained by uncontrolled sources of bias;
there was a striking exposure-response effect and the association persisted after controlling for
most known confounders and was independent of maternal smoking during pregnancy.
7-26
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Chan and collaborators (1989) studied 134 children aged 7 years out of 216 eligible infants
of under 2,000 g birthweight who were admitted to the neonatal unit of two hospitals in London,
England. Parents of each of these 134 cases and of 123 control schoolchildren born in the same
period but with normal birthweight completed a self-administered questionnaire on respiratory
illnesses and on social and family history. At age 7, children whose mothers smoked were at
increased risk of having frequent wheeze independent of their neonatal history (adjusted
OR = 2.7; 95% C.I. = 1.3.-5.5), although the increase only reached statistical significance for
children of normal birthweight. Prevalence of frequent cough was also more likely to occur in
children of smoking mothers (OR = 2.4, 95% C.I. = 1:3-4.6), and the association was significant
for both cases and controls studied separately. The authors performed a logistic regression to
control for possible confounders (only the low birthweight group was included). The relationship
between frequent wheeze and maternal smoking persisted among low-birthweight children after
controlling for family history of asthma, atopy, socioeconomic status, and use of neonatal oxygen.
The relationship between frequent cough and maternal smoking was no longer significant among
low-birthweight infants after controlling for the same possible confounders. For the low-
birthweight group, the authors assessed the reliability of some of the responses to their
questionnaires; there was a high correlation (r = 0.96) between the number of hospitalizations
reported by parents and those documented in the outpatient clinic of the neonatal unit that
followed the infants. The authors concluded that misclassification due to parental failure to recall
previous respiratory illnesses in the low-birthweight group was unlikely.
Krzyzanowski and collaborators (1990) studied a sample of 298 children aged 5 to 15 who
were family members of county employees enrolled in a prospective study. Parents answered a
questionnaire on their smoking habits and on respiratory symptoms in their children. Indoor
formaldehyde concentrations in the living environment were also measured. Prevalence rates of
chronic bronchitis (as diagnosed by a physician) were significantly higher in children exposed
both to ETS and to formaldehyde concentrations of over 60 parts per billion than in children with
one or none of these exposures. The authors also reported that similar effects were not seen in
adults.
Dijkstra and collaborators (1990) obtained consent for participation in their study for
1,051 out of a total of 1,314 (80%) eligible 6- to 12-year-old schoolchildren from a rural area in
The Netherlands. Parents completed a self-administered questionnaire on their smoking habits
and on respiratory symptoms in their children. Complete information was available for 775
children. When compared to children of nonsmoking households, children exposed to ETS at
home were significantly more likely to have cough on most days for at least 3 months
7-27
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
consecutively (OR = 2.5, 95% C.I. = 1.1-5.6); wheezy or whistling sounds in the chest in the last
year (OR - 1.9; 95% C.I. = 1.0-3.5), and attacks of shortness of breath with wheeze in the last year
(OR - 2.0; 95% C.L = 0.9-4.2). Exposed children were significantly more likely to have one or
more of the above symptoms than were unexposed children (OR = 2.0; 95% C.I. = 1.2-3.7).
Results were still significant after adjusting for parental respiratory symptoms and for maternal
smoking during pregnancy. The authors also measured nitrogen dioxide in the homes of all
children but found no association of the latter with respiratory symptoms.
Mertsola and coworkers (1991) followed prospectively for 3 months 54 patients aged 1 to 6
years from Turku, Finland, who had a history of recurrent attacks of wheezy bronchitis. The
parents were told to record the symptoms of the child daily and were asked to bring their child to
the hospital emergency room if the child developed signs of an acute respiratory infection.
Incidence of prolonged wheezing episodes (> 4 days) during follow-up was significantly more
likely in children exposed to ETS than in unexposed children (OR = 4.8; 95% C.I. = 1.9-12.6).
The result was independent of number of siblings, age, sex, medication, and personal history of
allergy.
7.5.2. Summary and Discussion on Cough, Phlegm, and Wheezing
Recent studies reviewed in this report that were not included either in the Surgeon
General's report (U.S. DHHS, 1986) or in the NRC report (1986) substantially confirm the
conclusions reached in those two reports. There is sufficient evidence for the conclusion that ETS
exposure at home is causally associated with respiratory symptoms such as cough, phlegm, or
wheezing in children.
The evidence is particularly strong for infants and preschool children; in this age range,
most studies have found a significant association between exposure to ETS (and especially to
maternal smoking) and respiratory symptoms in their children, with odds ratios generally ranging
between 1.2 and 2.4. Selection bias may have influenced the results of certain cross-sectional
studies; retrospective studies may also have been biased by preferential recall of their children's
symptoms by smoking parents. However, the presence of a causal relationship is strongly
supported by the consistency of the results for different geographic areas (Japan, Korea, People's
Republic of China, Europe, and North America) and by the positive findings in prospective
studies that are less subject to selection and recall biases.
In addition, efforts have been made by all researchers to control for possible confounders
and to avoid sources of bias. It is not feasible for each study to take into account all possible
factors that may affect the relationship under study; some of these factors may even be unknown
7-28
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
at present. However, all reviewed studies have controlled for at least some of the best-known
confounders (family history of respiratory illnesses, parental respiratory symptoms, socioeconomic
status, crowding, presence of other siblings, home dampness, gas cooking, maternal level of
education, perinatal problems, low birthweight, maternal age, birth rank, maternal stress or
depression). Of these possible confounders, a history of respiratory symptoms in parents has been
particularly scrutinized. The NRC report (1986) noted that bias may be introduced by parents
who have a history of respiratory illnesses for several reasons. These parents may overstate their
children's symptoms, or their children may actually have more respiratory illnesses and symptoms.
The latter possibility could be the result of intrafamily correlation of susceptibility (referred to as
familial resemblance by Kauffmann and coworkers [1989a]). Because smokers are more likely to
have respiratory symptoms, one would expect that controlling for respiratory symptoms in parents
would result in a decrease in statistical significance of the relationship between ETS and
symptoms in the child. In fact, most recent studies that have addressed the issue report that
controlling for family history of respiratory symptoms decreases but does not entirely explain the
increased risk of respiratory symptoms in young children exposed to ETS. It has been stressed, •
however, that the use of these statistical adjustment procedures may induce an underestimation of
the effect of passive smoking; this would indeed be the case if parents with symptoms (and thus
more likely to be smokers) were more prone to report symptoms in their children than were
parents without symptoms. Several studies have also found that the effect is independent of
maternal smoking during pregnancy and cannot be attributed exclusively to intrauterine exposure
to tobacco products (although the latter may potentiate the effects of postnatal exposure to ETS).
The evidence is significant but less compelling for a relationship between exposure to ETS
and respiratory symptoms in school-age children. Odds ratios for this age group are usually
between 1.1 and 2.0. Several studies have shown that, among school-age children, there are
significant differences in susceptibility to ETS exposure between individuals. There is, in fact,
evidence showing that several factors may amplify the effects of passive smoking: prematurity, a
family history of allergy, a personal history of respiratory illnesses in early childhood, and being
exposed to other environmental pollutants such as formaldehyde. In addition, long-term exposure
may have more important effects than short-term exposure. One study of 7-year-old children
(Strachan, 1988; Strachan et al., 1990) used both questionnaires regarding smoking habits in the
household and the child's saliva cotinine levels as indices of exposure to ETS. The authors found
a significant increase in the risk of having frequent cough when the questionnaire was used to
ascertain exposure, but no association between saliva cotinine levels and frequency of cough. As
the authors remarked, biochemical markers permit characterization of recent tobacco smoke
7-29
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
exposures, but they may not adequately reflect exposure at some critical period in the past.
Recent studies of intraindividual variability of cotinine levels have also suggested that it may be
misleading to assess the validity of questionnaire measures against a single determination of a
biologic marker (Coultas, 1990b; Idle, 1990). It is thus possible that associations evaluated with
salivary cotinine are likely to underestimate the true relationship between passive smoking and
respiratory morbidity (Strachan et al., 1990).
In the case of older children who may have started experimenting with cigarettes, the
confounding effects of active smoking need to be considered. Most researchers have been aware
of this problem and have attempted to control for it. A great difficulty lies in misclassification of
smokers due to underreporting. Young persons may be reluctant to admit smoking cigarettes.
Data are often obtained from parents, who may not be aware of the child's smoking.
In summary, this report concludes that ETS exposure at home causes increased prevalence
of respiratory symptoms in infants and young children. There is also good evidence indicating
that passive smoking causes respiratory symptoms in some older children, particularly in children
who have predisposing factors that make them more susceptible to the effects of ETS.
7.6. EFFECT OF PASSIVE SMOKING ON ASTHMA
Studies addressing the effects of passive smoking on frequency of asthma were directly
reviewed only in the Surgeon General's report (U.S. DHHS, 1986) and not explicitly in the report
on environmental tobacco smoke by the NRC (1986). The Surgeon General's report concluded
that epidemiologic studies of children had shown no consistent relationship between the report of
a doctor's diagnosis of asthma and exposure to involuntary smoking. The report pointed out that,
although one study had shown an association between involuntary smoking and asthma
(Gortmaker et al., 1982), others had not (Schenker et al., 1983; Horwood, 1985). This variability
was attributed to differing ages of the children studied, differing exposures, or uncontrolled bias.
The report also concluded that maternal cigarette smoking may influence the severity of asthma.
Alteration of nonspecific bronchial responsiveness was proposed as a mechanism for this latter
effect.
7.6.1. Recent Studies on the Effect of Passive Smoking on Asthma in Children
Several new cross-sectional and longitudinal studies published after the U.S. Surgeon
General's report was released have addressed the relationship between frequency, incidence, and
severity of asthma and parental cigarette smoke (Table 7-7). (Studies on the relationship between
ETS exposure and bronchial responsiveness were reviewed in Section 7.1.4.)
7-30 05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Burchfiel and coworkers (1986) studied 3,482 nonsmoking children and adolescents 0 to 19
years of age out of 4,378 eligible subjects from Tecumseh, Michigan. Subjects or their parents
(for children aged 15 years or younger) answered questionnaires on past history of asthma and
other respiratory conditions. Information on parental smoking habits was obtained from each
parent. Prevalence rates of asthma were higher among children whose parents both had smoked
during the child's lifetime than among children whose parents had never smoked. The effect was
stronger and only reached statistical significance for males (OR for boys = ,1.7, 95% C.I. = 1.2-2.5
in boys; OR for girls = 1.2, 95% C.I. = 0.8-1.9). Children with one parental smoker were not more
likely to have asthma than was the unexposed reference group. When results were stratified by
parental history of respiratory conditions, there was some reduction in the magnitude of the
parental smoking effects, but results remained significant for asthma in males. Results were also
independent of age, parental education, family size, a diagnosis of hay fever, and a history of
other allergies. Reporting bias and diagnostic bias may in part explain the relationships reported
in this study; smoking parents may be more likely to report asthma in their children, and
physicians may be more prone to diagnose asthma in children of smoking parents.
Evans and coworkers (1987) studied 191 out of 276 children aged 4 to 17 years from low-
income families who were receiving health care for physician-diagnosed asthma in New York.
Excluded children were younger and had fewer emergency room visits for asthma than those with
complete data. The authors suggested that the latter subjects had more severe asthma than the
general community population of low-income children with asthma. Emergency room visits and
hospitalizations for asthma were assessed by reviewing hospital records. Passive smoking by the
child was measured by asking one parent if he or she or anyone else in the house smoked.
Authors did not differentiate between maternal and paternal smoking; no attempt was made to
assess the degree of exposure to cigarette smoke. Eight children who were active smokers were
excluded. There was a significant correlation between number of emergency room visits and
cigarette smoke exposure (p = 0.008); the mean frequency (± SD) of annual emergency room visits
observed for children exposed to passive smoking was 3.1 ± 0.4, compared to 1.8 ± 0.3 for children
from nonsmoking households. Passive smoking had no effect on either the frequency of days with
asthma symptoms or on the annual frequency of hospitalizations. Results were independent of
ethnicity and parental employment status. The association could have been explained by lower
compliance with prescribed treatment of their children's asthma by smoking parents, but the
authors found no significant differences in compliance (as assessed by an index of asthma self-
management activities) between smoking and nonsmoking parents. The authors estimated that the
additional cost for emergency care for asthma was $92 ± $68 per family per year.
7-31
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
O'Connor and coworkers (1987) performed bronchial challenges with subfreezing air in
292 subjects 6 to 21 years of age. They were selected from 879 eligible subjects of the same age
who were participating in a longitudinal study on respiratory illnesses in East Boston. An attempt
was made to include as many subjects as possible who reported a history of asthma or wheezing on
standardized questionnaires. Therefore, the latter group of subjects were overrepresented among
those tested. The change in FEVl caused by subfreezing air was significantly higher in asthmatic
subjects whose mothers smoked at least 1 cigarette per day than in those whose mothers were
nonsmokers. This relationship was independent of age, sex, height, personal smoking, paternal
smoking, atopy, and baseline lung function. There was no relationship between maternal smoking
and response to cold air among nonasthmatics.
Murray and Morrison (1989) studied 415 nonsmoking children aged 1 to 17 years
consecutively referred to an allergy clinic in Vancouver, Canada, for asthma or recurrent
wheezing of the chest. Questionnaires were administered to the parents of all children at the time
of their first visit. Forced expiratory flows and bronchial reactivity to histamine were also
measured. An asthma symptom score was calculated for each subject based on .the severity of
asthma and the need for medication, as reported by parents. Children of smoking mothers had
significantly higher indices of asthma severity (p < 0.01) and significantly lower FEVj (84.4%
predicted vs. 77.3% predicted, p < 0.01) than did children of nonsmoking mothers. They were also
significantly more responsive to histamine than were children of nonsmoking mothers (p = 0.01).
The effect was present in both genders but was stronger for boys than for girls. Also, the effect
was stronger for older children (12 to 17 years of age) than for children aged £ 6 years of age.
The authors also reported a positive correlation between length of exposure to ETS and asthma
symptom score. It is unlikely that these results can be explained by parental overreporting
because the association between passive smoking and severity of symptoms paralleled that between
passive smoking and objective measurements of severity.
In their previously reviewed report (Section 5.1.1), Krzyzanowski and coworkers (1990)
found that children exposed to ETS and to > 60 ppb of formaldehyde had significantly higher
prevalence rates of asthma than those exposed to only one of these contaminants or to none (OR
for the latter comparison = 9.0; 95% C.I. = 2.4-34.0). No such association was seen among adult
household members. It is unlikely that this association is attributable to parental overreporting of
asthma because the authors relied on objective measurement of indoor formaldehyde
concentrations.
Sherman and collaborators (1990) reported on the results of a longitudinal study of
determinants of asthma in a sample of 770 schoolchildren enrolled in East Boston in 1974.
7-32
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Questionnaires were used to obtain data on respiratory symptoms and illnesses, cigarette smoking
history of parents and children, and household demographics. They were administered on entry
and for 11 consecutive years (1978-1988). Parents answered for children aged 9 or less, except for
questions on the child's smoking history. The authors identified risk factors for the onset of
asthma the occurrence of which antedated the time of first diagnosis of asthma. There was no
significant relationship between maternal smoking and either prevalence of asthma at the first
survey or incidence of new cases of asthma during follow-up (sex-adjusted RR = 1.1; 95% C.I. =
0.7-1.7). The authors considered it unlikely that this finding could be due to exposure levels too
low to increase the risk of asthma. However, no effort was made to assess the relationship
between incidence of asthma and number of cigarettes smoked by parents. Likewise, no effort
was made to determine the possible role of factors known to modify exposure to ETS such as
parental socioeconomic level (Strachan et al., 1989).
Weitzman and coworkers (1990) studied 4,331 children aged 0 to 5 years who were part of
the U.S. National Health Interview Survey. Children were categorized as having asthma if their
parents reported that asthma was current at the time of interview and had been present for more
than 3 months. Mothers were asked about their smoking habits during and after pregnancy. Odds
of having asthma were 2.1 times as high (95% C.I. = 1.3-3.3) among children of mothers who
smoked z 10 cigarettes per day than among children of nonsmoking mothers. The risk of having
asthma was not significantly increased in children of mothers who smoked < 10 cigarettes per day.
Use of asthma medication was also more frequent among children of mothers who smoked a 10
cigarettes per day (OR = 4.1; 95% C.I. = 1.9-8.9). Results did not change significantly after
controlling for gender, race, presence of both parents, family size, and number of rooms in the
households. No information was available on parental respiratory symptoms or socioeconomic
status. The results of this study could be partially explained by overreporting of asthma by
smoking mothers.
Oldigs and collaborators (1991) exposed 11 asthmatic children to ETS and to ambient air
for 1 hour. They found no significant difference in lung function or in bronchial responsiveness
to histamine after ETS exposure when compared to sham exposure. The study was only designed
to determine if acute exposures to ETS caused immediate effects and did not assess the changes
induced by chronic exposure to ETS.
Martinez and coworkers (1991b) studied incidence of new cases of asthma in a population
sample of 774 out of 786 eligible children aged 0 to 5 years enrolled in the Tucson study of
chronic obstructive lung disease. At the time of enrollment, standardized questionnaires about
personal respiratory history and cigarette smoking habits were answered by the child's parents.
7-33
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Surveys were performed on an approximately yearly basis, and parents were asked if the child had
been seen by a doctor for asthma in the previous year. There were 89 (11.5% of the total) new
cases of asthma during follow-up. Children of mothers with <; 12 years of formal education and
who smoked s: 10 cigarettes per day were 2.5 times as likely (95% C.I. = 1.4-4.6) to develop
asthma as were children of mothers with the same education level who did not smoke or smoked
< 10 cigarettes per day. This relationship was independent of self-reported symptoms in parents.
Decrements in lung function paralleled the increase in asthma incidence (see Section 7.7.1). No
relationship was observed between maternal smoking and asthma incidence among children of
mothers with > 12 years of formal education.
7.6.2. Summary and Discussion on Asthma
There is now sufficient evidence to conclude that passive smoking is causally associated
with additional episodes and increased severity of asthma in children who already have the
disease. Several studies have found that bronchial responsiveness is more prevalent and more
intense among asthmatic children exposed to maternal smoke. Emergency room visits are more
frequent in children of smoking mothers, and these children have also been found to need more
medication for their asthma than do children of nonsmoking mothers (see Table 7-4).
A simple bronchospastic effect of cigarette smoke is probably not responsible for the
increased severity of symptoms associated with passive smoking because acute exposure to ETS
has been found to have little immediate effect on lung function parameters and airway
responsiveness in asthmatic children. Therefore, the mechanisms by which passive smoking
enhances asthma in children who already have the disease are likely to be similar to those
responsible for inducing asthma and entail chronic exposure to relatively high doses of ETS (see
discussion below). Murray and Morrison (1988) reported that ETS exposure decreased lung
function and increased medication requirements in asthmatic children only during the cold, wet
season and not during the dry, hot season in Vancouver, Canada. These seasonal differences may
be at least partly explained by the finding by Chilmonczyk and collaborators (1990) that urine
cotinine levels of children exposed to ETS are significantly higher in winter than in summer.
These seasonal fluctuations also suggest that the effects of passive smoking on asthma severity are
reversible and that decreasing exposure to ETS could prevent many asthmatic attacks in affected
children.
New evidence available since the Surgeon General's report (U.S. DHHS, 1986) and the
NRC report (1986) also indicates that passive smoke exposure increases the number of new cases
of asthma among children who have not had previous episodes (see Table 7-7 for results and
7-34
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
references). Although most studies are based on parental reports of asthma, it is highly unlikely
that the relationship between asthma and ETS exposure is entirely attributable to reporting bias.
In fact, concordance in the relationship between ETS exposure and both questionnaires and
objective parameters such as lung function or bronchial provocation tests has been reported in
several studies. The association is also biologically plausible; the mechanisms that are likely to be
involved in the relationship between ETS exposure and asthma have been extensively discussed in
Section 7.2. The consistency of all the evidence leads to the conclusion that ETS is a risk factor
for inducing new cases of asthma. The evidence is suggestive of a causal association, but not
conclusive.
Data suggest that levels of exposure required to induce asthma in children are high; in
fact, most recent and earlier studies that classified children as exposed to ETS if the mother
smoked one cigarette or more usually failed to find any effect of ETS on asthma prevalence or
incidence. Furthermore, two recent large studies found an increase in the prevalence (Weitzman,
1990) or incidence (Martinez et al., 1991b) of asthma only,if the mother smoked 10 cigarettes per
day or more. It is also important to consider that, for any level of parental smoking, exposure to
ETS is higher in children belonging to families of a lower socioeconomic level (Strachan, 1989)
and that the relationship of maternal smoking to asthma incidence may be stronger in such
families (Martinez et al., 1991b). Concomitant exposure to other pollutants may also enhance the
effects of ETS (Krzyzanowski, 1990).
7.7. ETS EXPOSURE AND SUDDEN INFANT DEATH SYNDROME
The relationship between ETS exposure and sudden infant death syndrome (SIDS) was not
addressed in either the Surgeon General's report (U.S. DHHS, 1986) or in the NRC report (1986).
Because of the importance of this syndrome as a determinant of infant mortality and because of
the available evidence of an increased risk of SIDS in children of smoking mothers, the issue has
been added to this report (Table 7-8). .
SIDS is the most frequent cause of death in infants aged 1 month to 1 year.
Approximately 2 out of every 1,000 live-born infants (more than 5,000 in the United States alone
each year) die suddenly and unexpectedly, usually during sleep, and without significant evidence
of fatal illness at autopsy (CDC, 1989). The cause or causes of these deaths are unknown.. The
most widely accepted hypotheses suggest that some form of respiratory failure is involved with
most cases of SIDS.
In 1966, Steele and coworkers (1966) first reported that maternal smoking was associated
with an increased incidence of SIDS. They studied the hospital records of 80 infants who had
7-35
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
died of SIDS in Ontario, Canada, during 1960-1.961 and compared them with 157 controls
matched for date of birth, sex, hospital at which the child was born, and parity of the mother.
Infants of mothers who smoked 1 to 19 cigarettes per day were twice as likely (OR = 2.1; 95%
C.I. » 1.1-3.8) to die of SIDS as were infants of nonsmoking mothers. The odds ratio was 3.6 (95%
C.I. = 1.7-7.9) when infants of mothers who smoked 20 or more cigarettes per day were compared
to infants of nonsmoking mothers. The authors reported that the risk of dying of SIDS was higher
in low-birthweight infants whose mothers smoked when compared to low-birthweight infants
whose mothers did not smoke. However, they made no effort to control for other confounders
that were related both to maternal smoking and to SIDS, such as maternal age and socioeconomic
status. In addition, they made no reference to the relative roles of in utero exposure to tobacco
smoke products and postnatal ETS exposure.
Naeye and collaborators (1976) studied 59,379 infants born between 1959 and 1966 in
participating hospitals from several U.S. cities. After meticulous investigation of clinical and
postmortem material, they identified 125 of these infants (2.3 per 1,000 live births) as having died
of SIDS and compared them with 375 infants matched for place of birth, date of delivery,
gestational age, sex, race, and socioeconomic status. Infants of mothers who smoked were more
than 50% more likely (OR = 1.6; 95% C.I. = 1.0-2.4) to die of SIDS than were those of mothers
who denied smoking. When compared with the latter, infants of mothers who smoked 6 or more
cigarettes per day were 2.6 times more likely (95% C.I. = 1.7-4.0) to die of SIDS. The authors
made no attempt to distinguish between in utero exposure to tobacco smoke products and ETS
exposure after birth.
Bergman and Wiesner (1976) selected 100 well-defined cases of SIDS occurring in white
children in King County, Washington. These cases were matched for race, sex, and birth date
with 100 controls. Questionnaires were mailed to the mothers of cases and controls, but only 56
cases and 86 controls returned them. Mothers who did not respond tended to be younger and
poorer. A higher proportion of mothers of SIDS victims smoked cigarettes during pregnancy (61%
vs. 42%). Infants of mothers who smoked after delivery were 2.4 times as likely (95%
C.I. 3 1.2-4.8) to die of SIDS as were infants of nonsmoking mothers. The relationship between
postnatal exposure to ETS and SIDS was significantly stronger and only reached statistical
significance for mothers aged 25 years or less (OR = 4.4; 95% C.I. = 1.7-11.2). Infants of mothers
aged 25 years or less who smoked 20 or more cigarettes per day were 7.7 times as likely to die of
SIDS (95% C.I. » 1.7-35.4) as were infants of nonsmoking mothers.' Effects were independent of
maternal education. The authors did not try to determine the independent effects of pre- and
postnatal exposures to maternal smoking on the incidence of SIDS.
7-36
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Lewak and coworkers (1979) studied all infants who died during the first year of life and
who were enrolled in a health plan in Oakland, California. Using predefined criteria, they
classified 44 infants (2.3 per 1,000 live births) as having died of SIDS and compared them with the
rest of the population for several possible risk factors for SIDS. Mothers of infants who died of
SIDS were 4.4 times (95% C.I. = 2.1-9.2) as likely to be smokers as mothers of infants who
survived. Paternal smoking had no significant influence on SIDS frequency. The authors made no
effort to control for possible confounding factors, nor did they discriminate between the possible
roles of prenatal and postnatal exposure to tobacco smoke products.
Malloy and coworkers (1988) linked birth and death certificates to study possible risk
factors for neonatal and postneonatal mortality in over 305,000 singleton white live births in
Missouri. They identified 372 infants whose deaths were attributed to SIDS (1.2 per 1,000 live
births). Infants whose mothers smoked were 1.8 times as likely (95% C.I. = 1.4-2.2) to die of SIDS
than were infants of nonsmoking mothers. This relationship was independent of maternal marital
status, education level, age, parity, and child's birthweight. There were no data available that
would have allowed one to differentiate the effects of pre- and postnatal exposure to tobacco
smoke products.
Hoffman and collaborators (1988) reported on the results of the National Institute of Child
Health and Human Development Cooperative Epidemiological Study of Sudden Infant Death
Syndrome risk factors. They studied 800 SIDS cases and 1,600 control infants collected at six
study centers across the United States. Control infants were matched for age only (N = 800) or for
age, low birthweight, and race (N = 800). SIDS cases were 3.8 and 3.4 times as likely to have
smoking mothers as the first and second control groups mentioned earlier, respectively (p < 0.005
for both comparisons). There were no data on pre- and postnatal exposure to tobacco smoke
products.
Haglund and Cnattingius (1990) examined risk factors for SIDS in a prospective study
based on more than 279,000 Swedish infants who survived the first week of life. SIDS was
reported as the sole cause of death in 190 infants (0.7 per 1,000), and in most cases the diagnosis
was confirmed by the results .of an autopsy. Infants of mothers who smoked 1 to 9 cigarettes per
day were 1.8 times as likely (95% C.I. = 1.2-2.6) to die of SIDS as were infants of nonsmoking
mothers. Infants of mothers who were heavy smokers had an even higher risk (OR = 2.7; 95%
C.I. = 1.9-3.9) of dying of SIDS, suggesting an exposure-response relationship. These findings
were independent of birthweight, maternal age, social situation, parity, sex, and type of birth. No
information was available regarding smoking in the household by either mother or father after the
infant's birth.
7-37
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
/
Mitchell and coworkers (1991) studied SIDS cases occurring in several health districts in
New Zealand between 1 November 1987 and 31 October 1988. After careful assessment of the .
material available from necropsy, 162 infants were classified as having died of SIDS (3.6 per 1,000
live births). These cases were matched for age with three to four times as many controls. The
researchers interviewed the parents of and obtained complete information for 128 cases and 503
controls. Information on maternal smoking during pregnancy (as a yes/no variable) was obtained
from the obstetric records, whereas information on number of cigarettes smoked by the mother in
the 2 weeks preceding the interview was obtained from questionnaires. Mothers of infants who
died of SIDS were 3.3 times as likely (95% C.I. = 2.2-5.0) to smoke during pregnancy as were
mothers of controls. The analysis of the relationship between maternal smoking after the child's
birth and frequency of SIDS showed clear evidence of a biological gradient of risk. Odds ratios
were as follows: 1.9 (95% C.I. = 1.0-3.5) for mothers who smoked 1 to 9 cigarettes per day; 2.6
(95% C.I. = 1.5-4.7) for mothers who smoked 10 to 19 cigarettes per day; and 5.1 (95% C.I. =
2.9-9.0) for mothers who smoked 20 or more cigarettes per day. The association between maternal
smoking and SIDS frequency was independent of antenatal care, maternal age, maternal education,
marital status, sex, neonatal problems, parity, socioeconomic status, birthweight, gestational age,
race, season of death, sleep position at death, and breastfeeding.
In summary, there is strong evidence that infants whose mothers smoke are at increased
risk of dying suddenly and unexpectedly during the first year of life. This relationship is
independent of all other known risk factors for SIDS, including low birthweight and low
gestational age. The finding that there is a biological gradient of risk extending from nonsmoking
mothers to those smoking more than 20 cigarettes per day adds to the evidence that exposure to
cigarette smoke products is involved in the sequence of events that result in SIDS. .
Available studies cannot differentiate the possible effects with respect to SIDS of exposure
to tobacco smoke products in utero from those related to passive smoking after birth. As
explained earlier (Section 7.2.2), both human and animal studies show that maternal smoking
during pregnancy may modify and potentiate the effects of postnatal ETS exposure. The
relationship between maternal smoking and SIDS is independent of low birthweight, which is the
most important known effect of maternal smoking during pregnancy. In addition, the incidence
of SIDS is apparently associated with days of higher air pollution levels (Hoppenbrouwers, 1981),
which could indicate a direct effect of airborne contaminants.
In view of the fact that the cause of SIDS is still unknown, it is not possible to assess the
biological plausibility of the increased incidence of SIDS related to exposure to ETS.
7-38
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Consequently, at this time this report is unable to assert wheth'er or not passive smoking is a risk
factor for SIDS.
7.8. PASSIVE SMOKING AND LUNG FUNCTION IN CHILDREN
The Surgeon General's report (U.S. DHHS, 1986) reviewed 18 cross-sectional and
longitudinal studies on the effects of ETS exposure on lung function in children (Table 7-9). The
report concluded that "the available data demonstrate that maternal smoking reduces lung function
in young children" (page 54). The hypothesis was proposed that passive smoking during
childhood, by affecting the maximal level of lung function attainable during early adult life, may
increase the subsequent rate of decline of lung function and, thus, increase the risk of chronic
obstructive lung disease.
The NRC report (1986) reached similar conclusions after reviewing 12 articles (Table 7-9).
The authors' summary asserted that "estimates of the magnitude of the effect of parental smoking
on FEV1 function in children range from 0 to 0.5% decrease per year. This small effect is
unlikely by itself to be clinically significant. However, it may reflect pathophysiologic effects of
exposure to ETS in the lungs of the growing child and, as such, may be a factor in the
development of chronic airflow obstruction in later life" (page 215).
7.8.1. Recent Studies on Passive Smoking and Lung Function in Children
Studies appearing since the 1986 reports are presented in Table 7-10.
Lung function measurements were included in the cross-sectional study by O'Connor and
collaborators (1987) described earlier (Section 7.6.1). When compared to 97 nonasthmatic children
of nonsmoking mothers (mean age ± SEM = 12.8 ± 0.3 years), 168 nonasthmatic children of
smoking mothers (mean age ± SEM = 12.9 ± 0.2 years) had significantly lower mean percentage of
predicted FEV1 (mean ± SEM = 108.0 ± 1.4 vs. 101.4 ± 1.1, respectively, p < 0.001) and
significantly lower FEF^.^ (103.0 ± 2.3 vs. 88.2 ± 1.5, respectively, p < 0.001). These effects
were independent of personal smoking by the child.
Lebowitz and coworkers (1987) reported on the results of a longitudinal study of
pulmonary function development in Tucson, Arizona. The authors analyzed 1,511 observations
over an average follow-up period of 8.8 years in 353 subjects aged 5.5 to 25 years. The last
available lung function value (as residuals after regressing the data with different power functions
of age and height) was used as outcome. Residuals for vital capacity were significantly higher
/
among subjects aged 14 years or less at entry whose mothers smoked cigarettes (mean = +3.3 vs.
-1.4 among nonexposed subjects, p < 0.001). Parental smoking had no direct effect on outcome
7-39
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
FEVj or Vmax50%, but showed significant interactions with personal smoking and parental history
of airway obstructive diseases in their effects on Vmax50%; subjects who had started smoking or
whose parents had airway obstructive diseases and were exposed to ETS had the lowest Vmax50%
residuals at the end of follow-up.
In subsequent reports, Lebowitz and Hblberg (1988) and Tager and coworkers (1987)
reanalyzed two sets of longitudinal pulmonary function data: the one on which the preceding
study from Tucson, Arizona, was based (Lebowitz, 1987) and data for children of similar age
from East Boston, Massachusetts (Tager, 1983). The objective was to determine if the different
answers in regard to the effect of maternal smoking (significant for the Boston study; no effect
for the Tucson study) were due to the use of different statistical tools. Applying the same
multivariable analysis of covariance for both data sets, Lebowitz and Holberg (1988) confirmed
the positive effect of maternal smoking of FEF25_75%.with the data from Boston (p < 0.05) and the
lack of a significant effect of maternal smoking on Vmax50% with the data from Tucson, Arizona.
A first-order autoregressive model applied by Tager and collaborators (1987) to both data sets
showed effects of maternal smoking on FEY^ with the Boston data but not with the Tucson data.
The authors concluded that the most likely factor responsible for the disparate results was the
exposure difference in the two populations.
Tsimoyianis and collaborators (1987) compared the prevalence of low levels of FEF25_75%
(< 70% of predicted) in athletes exposed and unexposed to ETS (for more information on this
study see Section 7.4.1). Eighteen out of 132 exposed athletes (13.6%) had low FEF^y^
compared with 2 out of 61 (3.3%) unexposed athletes (OR = 4.7; 95% C.I.= 1.1-20.8).
Kauffmann and collaborators (1989b) assessed familial factors related to lung function in a
cross-sectional study of 1,160 French children. Levels of lung function (FEVj and FEF^.^^
were significantly lower in children with mothers who smoked when compared to those whose
mothers were nonsmokers. The authors reported a loss of 10 mL of FEVj (p < 0.05) and of 15
mL/s of FEF25_75% (p < 0.01) for every gram of tobacco smoked per day by the mother. These
associations were independent of sex, town of origin, age, height, weight, and intrafamilial
aggregation of lung function. There was no effect of paternal smoking on lung function.
Chan and coworkers (1989) performed lung function tests in a cohort of 130 children of
low birthweight (under 2,000 grams) at 7 years. These authors had previously reported on the
respiratory outcome of these same children (see Section 7.5.1). Children of low birthweight whose
mothers smoked had significantly lower values of percentage of predicted Vmax75% than did low-
birthweight children whose mothers did not smoke (80.7% vs. 91.4%, p < 0.01). This association
was independent of sex, birthweight, neonatal respiratory illness, and treatment. As 92% and 79%
7-40
05/15/92
-------
DRAFT— DO NOT QUOTE OR CITE
of mothers who smoked when the child was 7 years old were smokers before and during their
pregnancy, respectively, it was not possible to determine whether the effect of maternal smoking
was fetal or postnatal.
The study by Dijkstra and collaborators (1990) has been described earlier (Section 7.5.1).
Together with respiratory symptoms, the authors studied lung function and its relationship with
indoor exposures to ETS and nitrogen dioxide in a population of 634 Dutch children 6 to 12 years
of age. When compared to unexposed children, children exposed to ETS had significantly lower
levels of FEV1 (—1.8%; 95% C.I. -- 0.2 to -3.3), FEF^.^ (—5.2%; 95% C.I. = —1.4 to -8.8) and
Peak Flow (—2.8%; 95% C.I. = —0.6 to —4.8). Adjustment for smoking by the mother when she
was pregnant with the investigated child removed little of the effect of current ETS exposure on
lung function. The authors suggested that this indicated that the associations seen at ages 6 to 12
years were not just mirroring harm that was caused when the children were exposed in utero to
tobacco smoke components inhaled by the mother. There was no association between exposure to
NO2 and lung function.
A previously mentioned study by Strachan and co workers (1990) (Section 7.5.1) included
lung function measurements in 757 children. Lung function variables were adjusted for sex,
height, and housing characteristics. The authors found a significant negative correlation between
salivary cotinine concentrations and levels of FEF25.75% (p < 0.05) and Vmax75% (p < 0.05). For
these indices, the difference between adjusted mean values for the top and bottom quintiles of
salivary cotinine was of the order of 7% of the mean value in the children with undetectable
levels.
The longitudinal study by Martinez and coworkers (1991b) has been reviewed earlier
(Section 7.6.1). In addition to their findings on incidence of childhood asthma, these authors
reported that, at the end of follow-up, children of mothers with <; 12 years of formal education
and who smoked z 10 cigarettes per day had 15% lower mean values for percentage of predicted
FEF25_75% than did children of mothers of the same level of education who were nonsmokers or
smoked < 10 cigarettes per day. Maternal smoking had no effect on percentage of predicted
FEF25_75% values in children of mothers who had at least some education beyond high school.
Female children of smoking mothers (^ 10 cigarettes) had 7% higher Vital Capacity than did
female children of mothers who were nonsmokers or light smokers (< 10 cigarettes/day), and this
was independent of maternal education. All differences were still significant after controlling for
parental history of respiratory disease.
7-41
05/15/92
-------
DRAFT— DO NOT QUOTE OR CITE
7.8.2. Summary and Discussion on Pulmonary Function in Children
This report concludes that there is a causal relationship between ETS exposure and
reductions in airflow parameters of lung function (FEVl5 FEF^..^, Vmax50%, or Vmax75%) in
children. For the population as a whole, these reductions are small relative to the intraindividual
variability of each lung function parameter; for FEF 25-75%' for example, reductions range from 3%
to 7% of the levels seen in unexposed children, depending on the study analyzed. Groups of
particularly susceptible or heavily exposed subjects have larger decrements: Exposed children of
low birthweight, for example, had 12% lower Vmax75% than did children of similar birthweight
who were not exposed to ETS (Chen, 1989). Likewise, children of less educated mothers who
smoked > 10 cigarettes per day were shown to have 15% lower mean FEF25_75% than children of
less educated mothers who did not smoke or smoked < 10 cigarettes per day. This stronger effect
may be explained by Strachan and coworkers' (1989) finding that children of lower socioeconomic
status have higher salivary cotinine levels, for any amount of parental smoking, than do children
of higher socioeconomic status.
The studies reviewed suggest that a continuum of exposures to tobacco products starting in
fetal life may contribute to the decrements in lung function found in older children. In fact,
exposure to tobacco smoke products inhaled by the mother during pregnancy may contribute
significantly to these changes, but there is strong evidence indicating that postnatal exposure to
ETS is an important part of the causal pathway.
New longitudinal studies have demonstrated that young adults who were exposed earlier in
life to ETS are also more susceptible to the effects of active smoking (Lebowitz et al., 1987). In
addition, Sherrill and collaborators (1990) showed, in a longitudinal study, that children who
entered a longitudinal study with lower levels of lung function still had significantly lower levels
later in life. The high degree of tracking shown by these spirometric parameters implies that the
decrements in lung function related to passive smoking may persist into adulthood. Although the
subsequent rates of decline in lung function of these subjects has yet to be studied in detail, the
findings by Sherrill and coworkers (1990) support the idea proposed by the Surgeon General's
report (U.S. DHHS, 1986) that, by the mechanisms described above, passive smoking may increase
the risk of chronic airflow limitation.
7.9. PASSIVE SMOKING AND RESPIRATORY SYMPTOMS AND LUNG FUNCTION IN
ADULTS
Both the NRC report (1986) and the Surgeon General's report extensively reviewed the
evidence then available on involuntary smoking and respiratory health in adults. The Surgeon
7-42
05/15/92
-------
DRAFT-DO NOT QUOTE OR CITE
General's report concluded that healthy adults exposed to ETS may have small changes on
pulmonary function testing but are unlikely to experience clinically significant deficits in
pulmonary function as a result of exposure to ETS alone. The report added that the small
magnitude of the effect implied that a previously healthy individual would not develop chronic
lung disease solely on the basis of ETS exposure in adult life. It was suggested that small changes
in lung function may be markers of an irritant response, possibly transient, to the irritants known
to be present in ETS.
The NRC report concluded that it. was difficult to document the extent to which a single
type of exposure like ETS affects lung function. The report attributed this difficulty to the large
number of factors, including other exposures, that affect lung function over a lifetime. The
report added that results in adults should be evaluated for possible misclassification of exsmokers
or occasional smokers as nonsmokers, as well as possible confounding by occupational exposures to
other pollutants. The authors of the report considered it "unlikely that exposure to ETS can cause
much emphysema" (page 212), but that, "as one of many pulmonary insults, ETS may add to the
total burden of environmental factors that become sufficient to cause chronic airway or
parenchymal disease" (page 212).
7.9.1. Recent Studies on Passive Smoking and Adult Respiratory Symptoms and Lung Function
Six recent studies of respiratory symptoms and lung function in adults are presented in
Table 7-11.
Svendsen and collaborators (1987) studied longitudinal data from 1,245 married American
men aged 35 to 57 years who reported that they had never smoked. Subjects who had smoking
wives had significantly higher mean levels of exhaled carbon monoxide (7.7 vs. 7.1 ppm,
p < 0.001) but not of serum thiocyanate. These men also had lower levels of age- and
height-adjusted FEVt (mean difference = 99 ml; 95% C.I. = 5 ml-192.4 ml). However, those with
wives who smoked 20 or more cigarettes per day had higher mean adjusted FEV, (3,549 ml) than
those with wives who smoked 1 to 19 cigarettes per day (3,412 ml), whereas nonexposed subjects
had mean adjusted FEV, of 3,592 ml.
Kalandidi and coworkers (1987) studied 103 Greek ever-married women aged 40 to 79
who were admitted in 1982 and 1983 to a hospital in Athens with obstructive or mixed type
reduction of pulmonary function, without improvement after bronchodilatation. The women
denied that they had ever been smokers, and their husband's smoking habits were compared with
those of 179 ever-married controls of the same age selected from visitors to the hospital. Patients
were 1.9 times more likely to have smoking spouses than were controls (95% C.I. = 1.0-4.0).
7-43
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
However, odds ratios were higher for women whose spouses smoked 20 or fewer cigarettes per
day (2.5) than for those whose spouses smoked more than 20 cigarettes per day. The unusually
high number of nonsmoking women hospitalized with chronic lung disease in a 2-year period
suggests that some could have severe asthma unresponsive to bronchodilators and that the results
could in part illustrate exacerbation of symptoms in asthmatic women exposed to ETS.
Masi and coworkers (1988) mailed questionnaires to 818 subjects aged 15 to 35 who had
previously performed detailed lung function testing and carboxyhemoglobin (COHb)
measurements. A total of 636 subjects responded to the questionnaire and 293 denied having
smoked regularly before the date of the lung function tests. All but five subjects had COHb
values below 5 grams %. Questionnaires assessed past and present ETS exposure, both at home
and at work. Indices of cumulative exposure to ETS at home and at work were calculated from
the number of reported smokers on each location, the smoking conditions reported for each area,
and the number of years of exposure. In men, there were significant inverse relationships
between cumulative exposure to ETS in the home and maximal expiratory flows at low lung
volumes. A more detailed analysis showed that in these subjects, exposure before 17 years of age
had the strongest effects on lung function, whereas exposure in the 5 years preceding the lung
function tests had no effect on lung function. Exposure at work significantly decreased the
diffusing characteristics of the lung in women.
Kauffmann and collaborators (1989a) compared the results obtained from a parallel
analysis of the association of passive smoking with respiratory symptoms and lung function in
2,220 American women aged 25 to 69 years and 3,855 French women aged 25 to 59 years. Women
were classified according to their personal and current spouse's smoking habits. After adjusting
for age, city of origin, educational level, and occupational exposure, ever-passive-smokers
(excluding active smokers) had significantly more wheeze than true never-smokers (i.e., never
active and with nonsmoking spouse) in the U.S. sample (OR of approximately 1.3; C.I. cannot be
calculated). There was a positive trend for French passive smokers to have more chronic cough
(OR « 1.4) and dyspnea (OR = 1.2), but both results could be due to chance (95% C.I. = 0.8-2.4
and 0.9-1.6, respectively). In both samples, no significant decrease of lung function was observed
for passive smokers compared to true never-smokers in the whole sample, although FEV,/FVC
values for ever-passive-smokers tended to be intermediate between those of true never-smokers
and ex-smokers or active smokers. French women aged 40 or older who were passive smokers had
significantly lower FVC (p < 0.01) and FEV, (p < 0.01) than did true never-smokers, but no such
effect was seen among American women of the same age.
7-44
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Hole and coworkers (1989) studied cardiorespiratory symptoms and mortality in a cohort
of 7,997 subjects aged 45 to 64 and followed for 11 years in urban west Scotland. A self-
administered questionnaire was used in 1972-76 to assess respiratory symptoms and active
smoking by each member of the household. When compared to true never-smokers (i.e., persons
who were not active smokers and did not live with an active smoker), passive smokers were
invariably at a higher risk of having each cardiorespiratory symptom examined (including infected
sputum, persistent sputum, and dyspnea), but all 95% confidence intervals for odds ratios included
1. F£V! (adjusted for sex, age, and height) was significantly higher in true never-smokers than in
passive smokers (p < 0.01), but this effect was mainly due to the low adjusted FEV, of passive
smokers with high exposure (i.e., exposed to a cohabitee who smoked > 15 cig./day; mean =
1.83 L) when compared to those with low exposure (mean = 1.89 L) or with no exposure (mean =
1.88 L). This study was initiated when there was little concern for the possible ill effects of
passive smoking and is based on self-reports of active smoking by cohabitees. It is thus probably
not affected by classification bias due to overreporting of symptoms by smokers.
Schwartz and Zeger (1990) studied data from a cohort of approximately 100 student nurses
in Los Angeles who kept diaries of acute respiratory symptoms (cough, phlegm, chest discomfort)
and for whom data on exposure to passive smoking and air pollution were available. After
controlling for personal smoking, a smoking roommate increased the risk of an episode of phlegm
(OR = 1.4; 95% C.I. = 1.1-1.9) but not of cough. The authors also excluded asthmatics (on the
assumption that medication could bias the results) and found that in this case, the odds ratio of
having phlegm increased to 1.8 (95% C.I. = 1.3-2.3). The greater sensitivity of diaries of acute
symptoms such as those used herein, compared with the indices of period prevalence of symptoms
used in other studies, may have increased the power of this study. However, overreporting by
exposed subjects is still a possible source of bias in a study that is solely based on self-report of .
symptoms.
7.9.2. Summary and Discussion on Respiratory Symptoms and Lung Function in Adults
Recent studies have confirmed the conclusion by the Surgeon General's report
(U.S. DHHS, 1986) that adult nonsmokers exposed to ETS may have small reductions in lung
function (approximately 2.5% lower mean FEV, in the studies by Svendsen et al. [1987] and Hole
et al. [1989]). Using modern statistical tools designed for longitudinal studies, new evidence has
also emerged suggesting that exposure to ETS may increase the frequency of respiratory symptoms
in adults. These latter effects are estimated to be 30% to 60% higher in ETS-exposed nonsmokers
compared to unexposed nonsmokers.
7-45
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Because active smoking causes significant reductions in lung function and significant
increases in prevalence of respiratory symptoms (U.S. DHHS, 1984), the reported effects of
passive smoking in adults are biologically plausible. From a quantitative point of view, effects of
passive smoking on lung function are approximately comparable to those reported for light (< 10
cig./day), male active smokers (Camilli et al., 1987). However, because of the self-selection of
smokers and other factors, it is difficult to make direct quantitative comparisons between the
effects of active and passive smoking. The process of self-selection is likely to occur among
smokers by which more susceptible individuals never start smoking or quit smoking early in life
(the "healthy smoker" effect). Therefore, lower lifetime doses may be required to elicit effects
among nonsmokers than among smokers. The different nature of ETS and MS has also been
discussed in previous chapters and has to be taken into account when comparing effects of active
and passive smoking.
Several sources of bias and confounding factors need to be considered in studies of the
effects of single exposures in adults. Classification bias due to underreporting of active smoking
or past smoking may significantly affect the results of these studies. Because there is marital
aggregation of smoking (i.e., smokers tend to marry smokers, and nonsmokers are more prone to
marry nonsmokers), this source of misclassification is more probable among spouses of smokers
and may introduce differential biases in some studies. The resulting small overestimation of
effect may be nevertheless substantial for effects that are particularly subtle, such as those
described for ETS exposure in adults. In addition, recent public concern with passive smoking
may increase the awareness of respiratory symptoms in exposed subjects, who may be thus more
prone to report symptoms than are unexposed subjects. Studies using objective measures of lung
function are obviously not affected by the latter type of bias.
Adults are exposed to multiple sources of potentially harmful substances during their
lifetimes, and it is not always possible to control for their effects, because often they are unknown
or are unmeasurable. In general, the majority of these exposures should introduce nondifferential
error to the studies, which would thus underestimate true effects. For example, a significant
nondifferential error may be introduced by ETS exposure during childhood, which is known to
cause decrements in lung function (see Section 7.7). Conversely, effects of ETS would be
overestimated if a certain noxious exposure were more likely to occur among ETS-exposed
subjects. In this sense, social factors need to be accurately controlled, because prevalence of
smoking is significantly higher among less educated than among higher educated subjects (Pierce
et al., 1989). Most reviewed studies have controlled for indices of socioeconomic level in a
satisfactory manner. Finally, lifestyles may differ between spouses of smokers and those of
7-46
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
nonsmokers, but it is not possible to determine a priori the effect of this eonfounder on the
relationship between passive smoking and respiratory health.
The influence of these factors and sources of bias, together with the subtlety of the
effects, may explain the inconsistent and sometimes contradictory results of the studies reviewed
in this report. In fact, such variability should be expected, particularly for studies with relatively
low power (i.e., low probability of finding a statistically significant difference when a difference
really exists). The lack of a dose-response relationship in some studies may also be explained by
the multiplicity of uncontrolled factors that may affect lung function.
In summary, recent evidence suggests that passive smoking has subtle but statistically
significant effects on the respiratory health of nonsmoking adults.
7-47
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
TABLE 7-1. Studies on respiratory illness referenced in the.
Surgeon General's and National Research Council's
reports of 1986
Study
No. of
subjects
Age of subjects
Surgeon
General
NEC
Cameron, 1969
Colley, 1971
Colley, 1974
Dutau et al., 1981
Fergusson et al., 1981
Leeder et al., 1976
Pedreira, 1985
Pullan and Hey, 1982
Rantakallio, 1978
Speizer et al., 1980
Ware et al., 1984
158 . Children (6-9)
2,205 Infants
1,598 Children (6-14)
892 Infants/children (0-6)
1,265 Infants
2,149 Infants
1,144 Infants
130 Children (10-11)
3,644 Infants/children (0-5)
8,120 Children (6-10)
8,528 Children (5-9)
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
7-48
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
1
1
o
I
I
O
I
O
I
B
8«
a
o
CM
O
oo
1
O
pa
I
g
1
TJ
a
I
s
J
it-*
I
00
r-a^
«> a •*•• P
•a g S o «
S2.3SbS-ioSl'oo
U S o rt S « «2 O.OQ (=
st-
•S'S 8
O Q 25
t« g)§
O ^ •**<
»5vo
Od
a?
-H A
§
i-s
^i
.§
ii
.1 i
OH 0*
2 S
en
1
.S >.
i*
o >>
2 tS
en
U
oo
7-49
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
O
o
es
t~>
0)
^Observations;;
"I
,. I:
." W'
' ' ' .Q
,g;
••: ;l:
•0}
••• -;.a
o
ii
2
S^
o a
ETSexpi
:;;"1i:|:assessine'
•o
'•o
s
V)
c
•s
<2
o.
o
Pk
o
.j3
; .-. ;3:'
1 CO >,
• ^ ^_^
;3 ro
*o ^ '^ <=>
C^^
o - o -
S° 6°°.
so m S rt
<-. II oo "
I8S8
S » i
.O co C .
CM o SO
§5 S S
°!£°f5
Q «*H O o **•* co
g Is 2 g ^ is «
.2 O, "S. ^2 " rt :2
§ 8 S g S 'S C
£ ,« 5J .5 & i o
5-8.S
2 S3 S
5 S g •
I1«
III
w cs cr
.S
a
o rt
^Q C
CO "5
1"
C3 "
1C on
^ §
^*t .*^
es co
•a
"5 /_s
CH °°
S 00
<-; O\
u o
O Q^
*rt ^K
g% rJ
^^ CO A O *
O c3 ^ ***
O *O VO '
Sg-'i:
Slfi-li
1
CU
1
1
CO
I •
a
>
o
•§
g
OJ
OQ
/_^
gs .
FgH Q^
OC
4)
1 U
•• d
il
• 1
S o +j
— J.H 1U1 UU
fed babies; i
intervals no
calculable
'
7-50
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
•o
I
1
u
ju
2
«
H
1
w>
6
I
i
1
i
'
s s.
*>• 53
ii
•o
, i
Sj
I
a
0
cu
2
*•»
»
^
o - «• oo Iti
(8 o J3 rt-gtn o § § .S
Bis-asfiaifi-i
GO l-l
"* • > 2
43 °° IH "S
8 "*' j§ o
8-H fl 00
"^ .S
£
s § «
C *3 c-
§ S I
OH CT S
.•s .»
,» .5 "3
111
g • a o
"** O NO
CO t-c O
cd NO
^ O, ^ "o, V 'S ^ .
O V3 CH H^ *2 'W »^H *O O ^*H
S S
•—i tso ***
c iA .S >A
•a o jtf (N
o - g -
g >-' B >~'
i-12
. 1 » « «
lo so
Tv>4 (^ ^jj,!
*O >»5 °
rt P oo 5
ill 5
O S 3 a
oo
• .s
l^j
. -3 il&i
"i is •« « §
« S *2 gj "^
I
c
"5
|
>f) rt
NO "3
-N
1
« oo
oo o\
OO
f g.
•O wT <-• ra
S 8 «> «
«_ ••-< C — .
- " 'S •-
s> s « g
CJ O. -C on
7-51
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
¥
9
a
8
" 1
s 3
;;l
\ '•',
•i
;l
i?
i 1
*• o
-.*
\ '*
i
s'£ ^
Rj -A 1..
If
, ,<& <«
S2 **
**"* 0§
B4 «o
-, "• x"1
/- -v
s
.1
j'T3
'1
1
O.
, <£
s
"• "•
vn
-. u<
0
^5
- ^S>
Selection bias
possible
"3 Jrt
^-* d>
g «
s °
a •- S
<« S«2
2 'S °° s
*S 'J4 S «2
list
3
O co
£_?
« B
"!"!
a M 1
£.5^.
«« o '3
CD Q n
CO CO «*H
•a
N • r
15
a v
o^-
* r?
(3 ^ eo
III
ill
i.aS
ci
0
CO
a>,oo
•O oo
C a\
ep^ >,i§ "S «
g S IM C J3 « §
s S ° s ° s -S -a
oh^ea'artc
ifltlfH
(jc,-(43i-irSootH
^. 4)
O ^ ^*^
O Q
^"c S fe « S?
^ -S -S, S e2 S
' j< .2? 3 fc a
"1 g £ 0 *s -3
C- S n 2 0 g-
_, M 0 73 5r co
°. rt «« 5 ^ —
11 C3 f d ni I*
p> -g | a g s
o^ I 8 § g,
oo a
.s ,° •* _
^"3 «H «^^
S S S .So
O S)
.a £
C3 "3
"™| 2
•5 §
"8-0.2
i a> <-»
^^ •*•* 5^
*o3 S ^
co S cr
OO
TJ
CD
oo
03
g
£.
2 ^
'•§0
in •£•
Si
*3
4-*
1
C3
•o 0s
o o\
o a\
CO
1
CO
7-52
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
a
e
.a
I
e
.2
1
i.
i
W • (JQ j,^ £*
•Js 8 a,
.-
X
Wg o o-S o
o o
00
1v-
* a 'S
a «'fi
•d -o _,
fli fl W
S ^ W
•ill
e
t/3 a* o
g
I
-S <
0
r- ^
oo .S
•a
«
(D
to
•«_*
d
I
a
_d
to
•a
u
o
o
7-53
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
TABLE 7-3. Studies on middle ear diseases referenced in the
Surgeon General's report of 1986
Study
Said et al., 1978
Iversen et al., 1985
Kraemer et al., 1983
Black, 1985
Pukander et al., 1985
"No. of subjects
3,290
337
76
450
264
Age of subjects (years)
10-20
0-7
Young children
(unspecified age)
4-9
2-3
7-54
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
S
S
I
a
o
s
o
•d
C3
S
i
d
o
00
g
'2
I
to
I
<*H
o
«•
g
O
4)
P4
•v
i
i~
—
5
^2
• ^
1
-i
f
\
2?
* *
V
«o
"C
<&
*
i
'I
£
a*-
ll
«t ®
& i
s>
53
i
1
•*-*
i
a
o
««
*<
WS
5
'o
TT BO
-1. -S
^^ ^s
s*^' S
^ M co
'-72*'
11 -is 3
Qd !•» O
0^5 a
1
2
Jl
1
.§ >>
1 "«
1 £ S
al 1
• CO
>, n o
^.o , is
^H ^^ di d
^ ed O
"H ,
"S &S "S
rt *^ 5? -B
gSll
2 • 5 5 w
O M CO _^
C C "
ON rt -4-» o3
** /^-s
*-2 oo
.^3 O\
rt S
J3 .5 «? o5'^ 0
S T3 w aT .S -2 °
5^ l.§ § 8 S
1
^ I e3
^J CO qj
^ t! d
o 2 S
**^ ^3
r~ o o d
Dd *d o ^*^
O ^ Z .S
S" d M
O ^^"-i -g
ca >— i to a>
v-> ^ c >
"^ C o ^
S"5-'5 ^
s |!s|
^j ^H rt CL
d
*^
1» s
O ?L
"^ S w
to ^ *rt
® CO
&§!
•
>>
IT)
V
§
2
U
i^j
10
'•a
•«-*
oo
•S p
£ 0\
d
No distinctio:
between
maternal and
paternal
smoking.
CO
ts
to
0 X) C
S ^ *3
E'ii'
" "S 2? '-3
« 2 .S «
W d 73 a
2
I*
'1
.2 *2
S «
al
to 0
d *^
IH J2
O t3
•S--
? -°
"S T3 >»
C3 fli *-
a>
o /— s
•-; oo
d oo
"<" 2
"a
O
13
7-55
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
d)
g
••g
O
O
CU
s
Observations
1
•S
JO
*1
1
o
a
o
d> : •• *
»•• :.-l
52 **?
ETSexpc
assessmei
*d
1
co
2
"5
1
D.
.:•&.
12
1
• *»
€
CO
1
o
i
Tf
ON
TT
II
g
•a
g
|
^
2
CO
I-.
•3
c
a
.2
*4-»
g
3
1
0
1
o
£
o
a
3
•d
ti
•d
cu
OJ
CD ^j
d 43 S
° *j O
s i-o
45 1 ~
13 c3 »
« U IH
having smokers at
home
^
o>
a
o
o.
*-
d
I
43
2
•+-*
§
u
in
CN
.a*
•d
a
oo
oo
ON
IH
No control fo
confounders;
selection bias
not ruled out
CM
O
OR = 2.1 (1.0-4.5)
having smoking
parents
lH ^
•d '5
« «
4_» CO
1*
w fl
oo 5
.S "
W oo
2
«
IH
•3
c
a
O co
'« 1
11
i^ *^
= fe.S 9S
.S S
KO
, ri
No distinctio]
between
paternal and
:lc2
0 "^
1 jl
o "* £i
& t§ & s
»H to ,CM
CM — < CM
4) C eu
a-S
W ON A
^ ' •—!
.2
CO
^
fl
^-»
CO
O
•S
*— 1
s
after controlling f(
O .IH Ctt CO
confounders.
^J M IH >H
•^ >.>>>»
«* — < co r~
r^- i_j i_i
r- ,0 o
OO CM CM
•a
"Jn
42 oT
HO
«.
O
£j
No distinctio:
between
maternal and
C«C,
ol°
JH O
OO 02 *T3
7 Ij
o a2
^
oo
™ "^ CO
^** pj ^>
2 § g
•S g- 3
U £ a
cu
C3
Parental
question
i
VO >>
•d ^g
!a"
S3 N
H 9
T) IH
W-i .S
t-H .
VO IH
•a
•*-»
.
%
JS
i
T5
.S*
'o
7-56
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
I
1
o
1
«J
's
.
1
1
•M
v is
P
"I
1
a
<&" ^
s
6* -H
, O Cj
xJ •Cj
JH «
1
s»
*»»
1
o
cu
e
' O
.c
3
v- 2
§3 8 «
3.2 ,.•§
O -O S3 -S
«« "S-S S
0 '§ 3 ••§
^.«« «s S
ii
00 ® >-^
Ili
|||
^4-(
O
H
>I1 £.S 8 .l-^a &
'S'S'StOOT3KI&J'^**
0\
1— (
(U
I
l-i
(1)
O
o
7-57
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
TABLE 7-5. Studies on chronic respiratory symptoms referenced in
the Surgeon General's and National Research Council's
reports of 1986
Study
Bland et al., 1978
Charlton, 1984
Colley et al., 1974
Dodge, 1982
Ekwo et al., 1983
Kasuga et al., 1979
Lebowitz and
Burrows, 1976
Schenker et al.,
1983
Schilling et al., 1977
Tager et al., 1979
Ware et al., 1984
Weiss et al., 1980
No.k6f "
subjects
3,105
15,000
2,426
628
1,355
1,937
1,525
4,071
816
444
10,106
650
Age of
subjects
Children/adol.
(12-13)
Children/adol.
(8-19)
Children (6-14)
Children (8-10)
Children (6- 12)
Children (6- 11)
Children (<15)
Children (5-14)
Children/adol.
(7-16)
Children/adol.
(5-19)
Children (6-13)
Children (5-9)
Respiratory
Symptoms
Cough
Cough
Cough
Wheeze, phlegm,
cough
Cough, wheeze
Wheeze, asthma
Cough, phlegm,
wheeze
Cough, phlegm,
wheeze
Cough, phlegm,
wheeze
Cough, wheeze
Cough, wheeze,
phlegm
Cough, phlegm,
wheeze
Surgeon
General
X
X
X
X
X
X
X
X
X
X
NRC
X
X
X
X
X
X
X
X
X
7-58
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
00
c
*N
CU
CU
1 .
•o
«s
1
o>
I
I
8
a
o
oo
a
'^ .
o
&
ca
.1 '
W
1.
%
£
O
a>
a^
c
CU
^.
o
• in
CU
.Si
•o
3
•»-»
CO
.Sn
oo
_©
o
•I-H
1
3
a
V
.4^
d
CU
J
{*
vo
a>
IS
O ^1 .
s
1
9
,.
•M
, -2
3
CO
^
.**
*W
4 ?
4>
lUttoli!
:-:•:•: •••:-:-:•:•:•:•:•:••""•>:-;-: :•:•
£
ll
g S
K i
W w
•o
•S
*o
s**
1
<^J
*§
I
45
)-<
o
^
,-4— •
<
||fi| Jililfl
^ o
«2 ^
^ g"S
T 'J S
0 0 «
^' w CX
rsl . ;_
^» *p3 ;^j BO
^ C C3
gl^l
O C cu oo
5,-
ti ^i
.5
'R o
8'5!
£ £
1? O,
(H
1
s .2
rt *-*^
c3 3
»H (T
o
*7 ><
^ z
*cu "
oo cu
c4 ^~*
05
1 o
fH ***
" .S
IN IH'
CN >.
•o
ci
^
•1
o
O S3
c3 5 ^^
u "SS"
S(§0
«« I = 1
^3" * 5^ CO rrt
^J W « fll
C^ A « "S O '3 O
O —J .S i2? S rt •*
tH Tf >,
O —i w e>J
^ 00 ^ '"1 ~00
^* CH 5 *O
^^" ^^ j ^^ *^^
• o K j^ —<
^-^ S O (S Al
- OT
^» ^^ ***** Qn
*N 4) rt ""S C
CM ^"* 00 ^« ri
O «4-l '3 ^ «
to
•S 2
d £ £
i'£f
ail
2
CU
'es
C! *C5
.2 j2 «
1 i"a
3 o 2
O*1 ji S
rr
1
0
•s «
00 g
o
£ I J
»
J.
*t3
•s ~
o-i O
*CJ 00
••5 §
member!
not have
chronic i
7-59
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
f
o
'«
o
o
*~^
•
1
t--
co
*«. co
W q>
(— ( 00
M S
"a
CO
a
o
*3
eti
*«<
w
D.
O
a,
w
1
S3
M ri
o> «4-. .53
M " ° TJ »
jmj **H Jzl . . ^^ ^^ *o
issi-lf 1«
^lifll&l
.2 en
S- AI
OO a "«
dl 1
•*-*
CO
to *w f 1 1
w > >.
«
•3
•3 §
a o
j5 ta
2 «
-^>
•o
^g
w ^^
S ^
c*
a a
•« .a
•/^ 'S
^t* .S-4
w-> ^>
.
"3
«
•o
DOOO
So
Ui
o
C<
No control
parental
symptoms
OR = 3.1 (1.1-8.9) of
having smoking
parents
N •— ' Q.
Persistent whee
as assessed by
physician after
yr. of follow-u]
CO
11
Cl1
VO
^-
00
1 §
.§«
3.9
° 0
o 2
oo e
(3
'1
1
CQ
^ — "I?
"3 * CT\
O aSd-
•4^ Q*
^ « ^
° o^
7S2
By 5 yr., 63% of
parents who smoked
had wheezing
children, compared
with 37% of
nonsmoking parents
(p<0.05)
CO
CD
««5 ts-S
I'll 11 8
<0
.53
ll
£ «
3 ^
Pk 0-
.y .53 M
'eo £> O
C«H *— I J>
S sl
1s!'
.0 -g 3
O H " .
O 03 <*^ I-"
— < o, r~- ;>•
^j:
•4-*
CO
"a3
aEo
o ^^
uo
7-60
*2
.2
Selection b
cannot be
ruled out
/— s
i|l
2 2 to
n "5 3
o
•a
Number of
children still
wheezing at en
follow-up
CD
°3
!•!
Q) CO
P"* O*
m •»
y cc*
C "cd
CD co
^^ ^J
r^H c3
«2 •^•*
o *o
>• £
N #
•g "3 (3
O.
oo IH cd
^f ;>> >-J
*
"3
«
S
°3
|P
o °*
H C-
(D
o
'
O *O
D, en CO
Q> P3 ^j
OH 'j3 S
No effect on
bronchitis, wheeze,
shortness of breath.
Increased frequency
of cough (p=0.08)
co" *3
"*3 S
i.* 5i
V-S S B
fli O ^T *£H
1^2 O ? co
2 „
£
'i ° TS ^
Ill-s
mi
2
i &
m 4)
S» ^^
ra
o|
O4 So
»*H
CO 00
ON *JH
-H ^ ;
•d
4-*
(D
CO
• 1-1
">>
It?
CO ^^ !
H C-
05
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
1
i
n,
57
CD
§
-.0
vo
I
4)
1
S
to
'3
S3
W
M a
21
?&
J; -E5 a T3
|t a!
DH C « «
.1
•§
OO O
• fH ^
a g-Tt
8 lia
"
-
O, 00
« a ^
V2 |
Ss § S
43
vd S?
S JO
||
43 W
"o •-«
«
O O
43
0
oo
VO
. co J«S
Q. 3 0
S .S 2
. • d
,-
^
.
—< r-
•a
(H OO
T3 00
(3 0\
tD
(D
tt> OO
S c°,
_Q —<
W
C3 ^^
If
t-c ON
7-61
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
0>
£
o
u
f>
Q>
x
i
Observations 1
Results*
3
*§'
^ •
G):
S
0
-B
**•*
'S3.
O
«
d
cu
S
assess:
Population studied
• 52
•»•* •
-.S3'
S3
1
o.
o
_d
CO
OO
e
t-;
li
g
d
.2
cd
3
o
o«
et>
VI
03
I
1
O
I
o
o
CO
o
j5
^
•2
d
cu
£
•o
O "00
rf^ffll-N
SSll&fill
t«— S^sSSfH
Occtco^uOicaO
2 ed 0) ^ 1 ^ OS
I +3 -O J« -H JQ O 5
^jo'o M ^"I^'S
*H «?* W £rt ^M C7\ -
euvo"* cu^ ®OQ^
SuT "O d -S1 «H S" "O
a) m (3 o D ^ (u
cSdcsvT'oobtd
• B d • ' ^ "H d
V>coo3f~W>dOco
&
o
T4
s
A
^•H
CO
0)
OO
d
1-4
•*-» ^ M
ctf ^»
fl) ^^
ui a> ^H
•j3 en
/rt Q)
rt ra 00
.2 ^ ^
111
(TJD rt
'I
•a
2
o
.S
a
oo
C7\
^
. E I *
iffifffill
lll'llil
symptoms
.. I
7-62
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
I
4-*
(4
8
j
1
O
I
1
1
1
1
O
s
8» ±J
a*
Sc £*
Q* 4^
il
*§
^*^
2
«>
1
Ck
s
o
s
d E
§ 3 |g
S g g 3 -J2 r
0 g 0. ,0 0 rf
4J JJ* fli %4^ QJ C
'«« « oi
0 S1 o
«0" *- „" «*H
"? « 'o g ^
CO N -H 5 ^O JH
M oo 3 .g C- oo
ilS-s^l
0 J £ 1 il J
1
o
i
8
1
2
• ^t
'3.1
C •ii*
If
4> "TS
c^ W 1
^ IH
o-s-^S
, M 4>
^* *S> 5 *"*
CH ^5 « ^_»
£ g^ ^I§
2 ti o § i
t> O O O JH
, ^__ ^j ^5 _— ^2
^T ^™* ** c^ TL
• •a"
ai
^*^
d ON
rt oo
/-j ON
1 « 1
Co *»-< pj
£
.tS . CO
iff l~
O O •* <*H —<
•^ o a o o
•3 S'-S ^ v'
gg.sga
111!!
>,.s
° i
II 2
a 9-"d
8 12
(^ i?o.
4)
d
is
8
C^ CO
;S "3
d
IH
•d
is
ts
O d eu
j- S >
~ S
«*H rt ^
0 -3 o«
-* *^
ti
Increase in prevalem
Respiratory
3
<3,
2
>r>
i
•d
1
d
ON
«N
2
CO
O
d
hJ
^»
.S
of "chronic
ai
bronchitis" in childri
•d
i"J
symptoms in
children repo:
V
o
•^2
CO
CO
O
1
2
a
X)
formaldehyde
Independent
of parental
2
« ™
i 1
<3 w
•d
respiratory
symptoms an
maternal
H
A
o ,H o
^f^ 3 ^-^.
*o o *o
«?2«li
III
^•^2
1,
oo
1?
I'S
lo -d
%
^fi
«5
II m
• -
O^
IT
S ON
Is
pregnancy
IH
OH
O
>,.S g.
° S d
2 o S 2
'S. 9"S g.
g g .-s
ed £"5.0
0
'3
- d
S o
^H *^
O ">"*
28
03 £3
PH O-
2
"3
C5
>«H
(N _d
vo S
£2
°.a
-H O
"3
«
rt
,i ^.^
« O
J^ON
CA
Netherland:
7-63
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
^8"
§
d
•c
C!
1
*
vo
Q)
s
05
H
' 1
J^,
*Tk
Q
^
,,
s;fe
'"!
<$
4i
3
, i
1
0
1 £
§1
*2
Si
•s
1
.: 1
1
^
Lj
s *»4
•*-*
§ X" «*H fi"
1 s 2 "S
M *H CO "tS
« .r eu oo f* >>
a S, rfi a .£J oo
a> 2? s2 •« -S ii
•£) « C — H ^3 gj
I'sstig
G»
VO co
rsi O
j_5 o<
I X
CTx »
-H' a
^ O ^ S3
C< Uc TT P ««
M a ? -e o
ll
CU CO
s s
OH CT
IH N
t>> CU
^ ^ J1
"^ * •£
l^H
^H Cj
« to «
^'•2 5 -a
NM
10 ? 43 PL,
13
4)
^
"3 .~.
£3^
IH O
tS °*
^H ^^^
•
S
(3
• ^H
CU
O
(3
CU
•o
G
G
8
?g
cs^
*«i
&
7-64
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
ta
fi
1
"s
I
1
1
1
O
I
a S
Illa^essmi
12 s ^ c3 F-H
HH O £.*3 Q. 0) CH A
o o
C*H t*H
x— s y—s
\f\ Q\
tN —
tN oo
d S
r-; tN
H £ "1 „
r^ r* rS. Tl
SJSe-a
CM
O
CO
J.
H
£ 3
cl
CO
co c3
.h >. o.
S -° «
2 T3 °
12 »
o £ S
••s S »
S SE •*
3 e -S
CX§ S
(S
eo"^
*M ^*
o a\ S
6 -r 2
CO ' "
CJ ° J3
O e« «>
S «5 OT
C>1 »rt *-j
^ w S
•* <^ Q*
m o H
•3
U
*3
?T^
J-H C7^
s «*
CQO
e ts
5 5 -o « ^
• PN QJ *M TM Tj
-^J >^ " C *T"i C
0 ? CS CO S . CO
.s ts-a- ob^ «- a-
f! S3
°' o"^ «
-H C 2 o
S3 '3 C3
00 °°o 73 "
rfl-sag
> O-B g> g.
2 11 1 J
-?1!1
CO vS 8 0 CO
i ill
^, +•> o
S-e-SIS
53 S "a ja "S
oo rt S *i S
IH CO 'S S •"
CO -^ O« CvJ *-.} .
1111 1
£
• v^
d
C3
1-2
C •*-*
(U CO
S §
PH CT
r-
'™H *><
•"* Z
T3 ^
CO ^!
iso v3
03 O
«** :
•O co
a z
u d
^, "*
o\ >-.'
•a
•*-»
^
"3
CO
o
S3 i^
o ^~
r i °°
X ON
O O
CO
o
tvlli
§ co ° 2
w o o3 o
^-^ so
-< d
o •«
O O
V ^
S CO
Higher scores
in children of
mothers
Asthma symptom
score for severity
of asthma
2
I"
^» C3
« 9
d *2
CO CO
£, co
o^ cr
r-
•-H
1
^H ci
•o.S "2
oo cj S
SJS
!,7 ^» *i
S CO IH
•S s ^
.•3 43 3
45 .tJ 0
0 > 0
^ ? d
^H ^H fcf^
•* >.
d
03 pj
>, o
3 0 ^
SSo
0>
1
CO
O
• 1-H
00
o
.2
1
V
7-65
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
I
o
r-
1
Observati
1
en
45
¥&
,S
§
i
i
8
*>.
IB
x a
0> «
„_ en
S «
w 3
•o
•3
to
0
40
1
D.
£
52
Q
*3
0
,2
"H.
to
1
T»
0 %
S'| *
I ^
4- 2 •§
G-S-s £*g
S^all
O\ 1*3 c/j ,23 O.
„ 43 K w X
iSEgf
5<22<2 §
o
£2 u
S'S
1*
n (3
•— • — 1 CS
S * S
gll
C3 to 43
PL, d o
IM
.S
|
2 o
0 '-a
CD >
0)
en
^
Crt
g. S"
f? <^
l-i . O\
So >> o n
||||||i]
CD CO
•a SI
ill
TO 1> C|^
Q.
Ill
*c3
(3 en
|||
£•! i
-4 ?>
T3 *-i
S 'S
08 0
CD -*n g
rt "9 !3
CU co CT
o7
>A ""
"O ^,
CD
9 =
rt s ^
fl> O ^4
1^ ^*H ^*
•O T3 **
Hi
C5 ^Z< "^
P:c2n
•a
0
a o*
*-4 Q^
432
00 v-'
•4^ A
0 «'"
« S 8 >•
CX s *-^
CO H-l O C+H
<2 ^
S 5
**! o o
^•S a
. ^ en
«• S e
II -O 4?
ftS-a g
o-S a
1 £
' |^1
|JJ
^JJ'S
£
M
||
11
^; u
0 "g
"O ffi
CD —
e3 53
(3.9
2 t3
*«H X-l 't^'
.1^ ^f
r* * "^
*^ r/\ ^
o "^ .s
co >-^ S;
TT" >.£
.
4^
CD
0
CJ
|x-s
• ^H ON
t2 ^
fS
'S <3 ««
o rt *^
S Q- *^
CO -M *2 fl
sill
0. 43 0 IH
00
• ^H
O
c
«««
a o
« «
gg.a
a%|
5^43
Z o o
4»*
O
CD
CD
0°
• ^
_0
en 0
l-s
|a
u<2
o
•£<
£ ^
S §
OfH
l*^
&-
4) Ui
IP
Q W
0
2
2
43
O
.9
•*_•
ca
Q
i
""*
•a
CD
en /-^
oo — i
•••4 ^^
*Q «v
-i oj\
exposure
7-66
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
§
u
r-
t-
-
1
^
1
i
Tl
^
•o
12
|
w O
8
g , » E
*.
.S
o
"*• ,2 3 S o
; *^ JZ f« *-H
11.
m
£••33
8 2
e *s
gsl
0 5 "
t~- "3 «J
r- Cl >.
N
-------
DRAFT—DO NOT QUOTE OR CITE
i
Q
I
a
.9
rt
cu
•a
1
CM
O
§
a
IS
O
**««
d
o
oo
a
*»^
1
CO
CD
o.
CM
O
a>
CM
O
^
••a
3
00
JO
"3
• 1^
a
§
*
oo
I
f-
.2
3
Observations ||
rf
t/t
3
• IS
ETS exposure
assessment
Population studied
',
2
0
*-•
9
No control for socio-
economic status or
o
CO
.13
f 2\
ri~
1 *™J
^H j*j
O °
— < S
O4 VH
ll 2
O/ O
o a
a
0
ti«
1; Maternal repo
hospital recort
S
CM -S
*•* CO
•o o
CD U
•M H-*
•0 o
II
S o
G +••
a a
•s*
0 vr>
oo —<
£
o
oo
g
>J
g
... /—N
JS *o
: 2.2?
a «j3 o
•>s^ ^
00 ST^
•3t-- Al
1
Ontario, Canada
o
g
°. CM
fc 0
CM «
£?•§
•*"* «
— 43
O 43
IS
3*
^ OQ
CM .8
^"i
«N a
1 CO
O >— 4
C- a
vo »
— < td
ii a.
OH >>
O §
a
O
>r^ ^«
t-2
Maternal repo
hospital recor<
_j
•i^
ll
c2 ^
•3
•4-*
•S.vS'
CD f"^
rt ON
delivery, gestational
age, sex, race, socio-
|^ irt
<2
f.i>
r^. «^
-H_ §
• £/2
"
^.§|
N 4-« O
ii ° a
OH CM* "-1
O Wc2
i
•o
l_l
Maternal
questionnaire
answered aftei
.S
CO
100 cases of SIDS;
100 matched contro]
King County, WA
IH
CD
CO
g
•o
3
a
c
^4-»
a
s °^
education, parity, and
birthweight
Controlling for age,
birthweight, and race
o
o
CO
•S
^'
ro
'll
OH
o
Maternal
questionnaire
Q
n-!
800 SIDS cases;
1,600 controls (NIC!
cooperative study)
;
"3
0>
g
tS oo
£E
7-68
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
§
u
oo
i
i
09
&
X
£
w
•.
i
o
52
"»
1
04
•*«*
1
1
TJ
•13
MHf
»
CS
,9
sjj[
0,
o
Cw
£
i
<
Independent of
/^>
^
I*
tN
^
OO
^*
II
o
•3
1
s
V
•o
(D
^
00
_0
i
£
0
o
Os
f-
tN
1
1
1
birthweight, matei
1
o
g
eo
ti
'2
i
VN
CO
1
>"
K
CU
_v<
'3
c
1
er
CO
1
1—1 •+••*
oo ed
rt §
a
o
CM
II '
03 o>
•5 to
oo « -a
rt O.43
^^
Os
CO
OS
*— 1
f^
CN
II
§
o1
Os
OS
C'
Independent of
04
O
3
1
o
%
55
(N
^D
-3
"3
u
•i^
s
prenatal care,
maternal age,
0
^— '
Os
"H
CO
\
.go
'o
1 "1
.
•o
^>
OQ
.«
O
tN
Al
birthweight, race,
season of death,
o*
Os
breastfeeding
CU
C3
O,
d
• *H
V
o
o
7-69
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
TABLE 7-9. Studies on pulmonary function referenced in the
Surgeon General's and National Research Council's
reports of 1986
Study
No. of,
subjects
Age of subjects,
Surgeon
General
NR.C
Berkey et al., 1986 7,834
Brunekreef et al., 1985 173
Burchfiel et al., 1986 3,482
Chen and Li, 1986 571
Comstock et al., 1981 1,724
Dodge, 1982 558
Ekwo et al., 1983 1,355
Ferris et al., 1985 10,000
Hasselblad et al., 1981 16,689
Kauffmann et al., 1983 7,818
Kentner et al., 1984 1,851
Lebowitz, 1984 117
Lebowitz and Burrows, 271
1976
Schilling et al., 1977 816
Tager et al., 1979 444
Tager, 1983 1,156
Tashkin et al., 1984 1,080
Vedal et al., 1984 4,000
Ware et al., 1984 10,106
Weiss et al., 1980 650
White and Froeb, 1980 2,100
Children (6-10)
Adult women
Infants/children (0-10)
Children/adol. (8-16)
Adults
Children (8-10)
Children (6-12)
Children/adol. (6-13)
Children (5-17)
Adults
Adults
Families
Children/adol. (<16)
Children/adol. (<18)
Children (5-19)
Children (5-9)
Children (7-17)
Children (6-13)
Children (6-13)
Children (5-9)
Adults
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
7-70
05\15\92
-------
DRAFT-DO NOT QUOTE OR CITE
I
0>
{
a
"3
£
Id
n
eM\
«
3
f
•a
*M:
to
'1
a.
£
t2
5
*jr2
O
^
Independent of
personal smoking
habits
•<* • >r>
. . ^^ •
Nonsmoking mothers vs
smoking mothers: FEV,
(% predicted) 108.0 ± 1
vs. 101.4 ± 1.1 (p<0.001
FEF2J.7J (% predicted)
103.0 ± 2.3 vs. 88.2 ± 1
(p<0.001)
.§
ll
>
§
gp
V^ Q^
be-
Interaction between
family history of
respiratory illnesses
and passive smoking
for V.^50% residuals
i
d
Smoking mothers vs. no
smoking mothers FVC
(residuals)
+3.3 vs. -1.4(p<0.001)
.§
03
a!
d +5
CO CO
cd 3
&< cr
.
>,
>o
CM
1
•S
CO
c?N
a <
O _T
_« d
3* CO
L^
CO »^N
"cS
•g
.•S
o t^
o OO
*~ ON
jc
oo
ij
J5
00
H
•0 W
S 0
b *-
£-S
V g
"3 x
C/3 CO
t/f
H 0
w 2
O ^3
•*-• CO
•g I
Q. CO
athletes exi
thletes not
CM w p
2 vo W
"cJ
"co
CO
1
O {^
d oo
H C-
Independent of sex,
town of origin, age,
height, weight, and
family aggregation of
lung function
£ •s
^ 0
CM CO /-N
o -i0,-^
ul
o o o g
CO O OT fT .
CO H CO ^
O ^ 0 W
.§
ll
CO CO
V-l .CO
PL, CT
'd
CO
2
2
u
1
CO
o
VO
^
•a
w
d
1 —
C«H ON
9 oo
$2E.
Independent of sex,"
birthweight, neonatal
respiratory illness,
treatment
t/i rf
Mean V,^ 75% (%
predicted) in exposed vi
nonexposed 80.7 vs. 91.
(p<0.01)
° oo
I!
&1
t-c E5
is
t-H CO
CO t-4
^ .2?
S
.S
c
& f-
Q CO
" ' .^
° S
« 2
2 •£
•O op
.1'S-s
43 £ C
0 43 —
o t §>
rn .2 C3
22 w
s
"ca
4^
d ON"
crt oo
_i-; ON
uo
7-71
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
§
o
i
t-
^
s
',
-i
1
s
^
.1
,
5/5
«» Is
p" «
w. «
T>
.S
1
(S
o
a
1
O*
P-
2
o
•*-«
<
00 >.
C o
*4H "^ 0
« a 50
CU CO U
"0 — Q.
« 2
CU Q 00
' "* CM5
27^'
3*8
. l . t!^ vo
* "* «-*^e»
a> U fl PJ ' IL
Q E en E 2 S
CU
1_
1
»-H £
CO Q
c3 *EJ
II
PH O1
_g
^
tN
1
VO
1-8
SI
||
^H "<•*
"J3 »7
o ^
^f 2*
2 H
•a*
4^
O)
CJ
+•» /'— N
.•S S
Sc-
8
a
ID •-} .
>+^ » /"^
s s i
Approx. 7% d
between maxi
exposure and
exposure
s
+-» K^T
'> ^
Negative correlation
FEF25.75ij (p<0.05) an
75% (p<0.05)
(D
.S
"S
• 1H
8
^
tH
£3 CO
'•i^
oo JJ
13
,2
o
o
00
.S
a
2
"O
u
t-
r-
13
CU
g
43 o"
^§
00
8 0
*§£*
15% lower levels of (
predicted FEF^,; a
children of mothers
smoked and had a lo
level of education
CU
tj
-1
Cv O
pH *pi
CU CO
PH §•
a? £3
ta "§ S .
•o PS ***
"2< £
a „ £
2 S «2
•^ i
t— < f-H ^3
i.si
^* *o ^2
r- i o
r-* o *n
•a
4-*
CU
N
0) x-^
•S *^
t3 S
IS
CO
d>
I
C3
8
a
I'
a
o
o
7-72
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
1
%
O
1
•a
§
1
en
&
o
,0
•
i
s
O
CO
.a
"S
oo
o
.2
1
.•s
a
CO
•1^
8
w
1
H
<«
.2
1
a
No dose
response
effect
o
4> S~
£2 S
--
ON
ON
ON
03
13
i
« eo
1-3
&1
<-> a
s
C/3 w
« S5
ol
a o
C
S*S8
d>
CO
§*
o« ^
co J3
^OT
B
C/3
iS
S3 M
|S
< g
•o S
oo
S3 -3
ro ^
58
"?
§•«
C co
-3
«
CO
. „ co ON
fl5 ^ ^"
S'o <=>
r>) S ";t
u i»-d
CO CO CO
O ,0 -
—< O
^_
gp
43 oo
CJ O\
CO
' ll-
+^ W 00
8-S g. rt
OQ ILJ £7
co a>
V- >-i
O 3
co co
•—> O
*^ Q«
— CO
CO
CO
CO 1-c
VO >.
•a
I
7-73
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
,
o>
8
i
i
s
X
£
ja
1
S
Is
5, g
8 S
a> w
«_ w
S s
S 8
"S
•o'
B
«
a
.2
£4
a
p
a,
$2
o
n
<
TS
'•a g
l^ffllj
gaa&r-al
g 3 S 1 § 2 .S?
>—4 tH IH co G co co
•o
§ -3 * ps^ £
1 Is! 5s?
•s if 8 <~ § •§ -g «
S "S b ^ £ 3 £ A>
<2 S «2 ~* a O a &
CO §<* " •» .^X"
H^l uO§'§.»-icJ2
*— ^ ex y* oi o H
cd ^ od "^ ^S £H o
O.SoS-oS5a?
00
13 s
M >> !^
Is I «
W ctf O O
I'lll
(U 0> >•> 3
CO l-c eo «*H
(M
o M
S.S
o5
O. O
C3
« M
^ en
" "b en
•r> P •-;
•S o -9
9 Q. Cv
CO en J3
ti
a>
s s
1 I
s .. 1 .
** i-T in
•»H ^* *fl ^*
CU O\ rt ^
S *? £ *?
J5
^^ JG '^ JG
O _^ 0 _^
es ^5 m ^3
Ol **^ OO ^
A oo i bo
c*l cl co e$
•a
Oi
p*
g
Cj
^*-4 "
C ON
0 00
03 ON
Kyi ^"^
S •«
No significant increa
in risk of symptoms;
decrease in FEVi
(60 ml) when a
cohabitee smoked > '.
cig./day
>,
o «
13
Q. ^
Hi
s Eg
r \ p^* F»
^^ CO ^^
CO
V
.ts >.
§5-0
§lie
•+3 w "53 ^
« ^ en g
,
\fi
"*f
W)*?
rt rt
•J2 73
o o
_> 5?
•° I a
oo B "3
o -a s S
sills
Sltl!
ii
Increased risk of
having phlegm (OR =
1.4 [1.1-1.9])
1
en
Respiratory
symptoms asse
by self-
administered
questionnaire
1
S
Questionnaire
answered by
subject on
presence of a
smoking room
en
0
^^
£
en
Ui
3
^C
C U
4)
*O en
^ O
to "3
C3 5
•o
c
N
•4^
at *~*^
P o o
1 §?£
co N O
(U
s
a
.S
43
1
8
7-74
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
8. ASSESSMENT OF INCREASED RISK FOR RESPIRATORY ILLNESSES
IN CHILDREN FROM ENVIRONMENTAL TOBACCO SMOKE
In the preceding chapter, a review was presented of recently published studies regarding
the association between respiratory illnesses in children and environmental tobacco smoke (ETS)
exposure. The biological plausibility and the possible pathogenetic mechanisms involved in each
group of illnesses included in the chapter were also discussed. The purpose of this chapter is to
consider the weight of the evidence as a whole, to analyze in detail possible sources of systematic
bias or confounding that may explain the observed associations, and to estimate the population
impact of ETS-associated respiratory illnesses.
8.1. POSSIBLE ROLE OF CONFOUNDING
In the review of the available evidence indicating an association (or lack thereof) between
ETS exposure and the different outcomes considered in this report, the possible role of several
confounding factors was analyzed in detail (see Chapter 7). Such analysis will only be summarized
here.
• Other indoor air pollutants (wood smoke, NO2, formaldehyde, etc.) have not been
found to explain the effects of ETS, but may interact with it to increase the risk of
both respiratory illnesses and of decreased lung function in children.
• Many of the studies reviewed in this report and in those of the National Research
Council (NRC) (1986) and the Surgeon General (U.S. DHHS, 1986) used either
multivariate statistical methods of analysis or poststratification of. the sample to
control for the possible confounding effects of socioeconomic status. Others
controlled for this effect by study design. It can be concluded that socioeconomic
status does not explain the reported effects of ETS on children's health, although
children belonging to some social groups may be at increased risk of suffering the
effects of passive smoking (see also Section 8.3).
• The effect of parental symptoms on the association between ETS and child health
has also been extensively analyzed. It can be concluded that, although parents with
symptoms may be more aware of their children's symptoms than are parents
without symptoms, it is unlikely that this fact by itself explains the association. In
fact, objective parameters of lung function, bronchial responsiveness, and atopy,
8-1
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
which are not subject to such sources of bias, have been found to be altered in
children exposed to ETS.
• The effects of passive smoking may be modified by several characteristics of the
exposed child. Increased risk has been reported in premature infants and infants
of low birthweight, infants who are not breast-fed, infants who are kept at home
with smoking mothers and not sent to day-care centers, asthmatic children, and
children who are active smokers.
• Maternal smoking during pregnancy has significant effects on fetal growth and
development and may affect lung growth as well as the immunologic system.
However, reports of important effects of paternal smoking on the child's health
and studies in which ETS exposure was found to have effects that were
independent of in utero exposure indicate that maternal smoking during pregnancy
does not explain the relation between passive smoking and child health, but
modifies the effects of ETS.
In summary, there is no single or combined confounding factors that can explain the
observed respiratory effects of passive smoking in children.
8.2. MISCLASSIFICATION OF EXPOSED AND UNEXPOSED SUBJECTS
The importance of misclassification of exposed and unexposed children has not been
addressed and will be analyzed in detail below.
Two possible sources of systematic bias related to subject misclassification are considered.
The first is upward bias from the effect of active smoking in children; the second is downward
bias due to misreporting and background exposure. Both have also been considered in the
assessment of ETS and lung cancer in adults. Adjustment for background exposure will be similar
to that presented in Chapter 6, except that data for increased incidence of some ETS-associated
respiratory diseases show some evidence of thresholds that must also be taken into account.
8.2.1. Effect of Active Smoking in Children
The possibility needs to be considered that some children may be smokers themselves and
that this may happen more often among children of smoking parents than among those of
nonsmoking parents. This would bias the results upwards or against the null effect. This source
of bias is only applicable to studies of older children; regular active smoking may occur but is rare
before early adolescence. A study of third graders in Edinburgh, Scotland, by Strachan and
8-2
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
coworkers (Strachan et al., 1989, see Section 7.4.1), for example, showed that salivary cotinine
levels compatible with active smoking were found in 6 of 770 children ages 6-1/2 to 7-1/2 years,
suggesting only a small potential for bias. Consideration should also be given to the fact that some
of the effects described in Chapter 7 (for example, the increased risks for acute respiratory
illnesses [Section 7.3.1] and for cough, phlegm, and wheezing [Section 7.5.1]) have been found to
be stronger in younger children (i.e., those less likely to be active smokers) than in older children.
This observed reduced effect with increasing age may be in part due to an age-related increase in
misclassification of exposed subjects as "unexposed" (see below), but it is clear that these specific
effects of ETS do not increase with age, as would be expected if active smoking biased the results
of studies of ETS effects in older children. It can thus be concluded that the association between
respiratory health in children and ETS is not attributable to active smoking by some children. It
has been suggested that active and passive smoking may interact to increase the effects of either
exposure separately (Lebowitz, 1988). This interaction is biologically plausible, because it is likely
that active smoking may be more harmful in children whose lungs have been previously affected
by ETS (see Section 7.1).
8.2.2. Misreporting and Background Exposure
Various investigators have measured cotinine levels in body fluids in infants and children
and correlated the results with parental reports of ETS exposure. Coultas and coworkers (1987)
reported that 37% of children under 5 years of age whose parents were both nonsmokers had a
salivary cotinine level greater than 0, compared with 32% of children ages 6 to 12 and with 35% of
children ages 13 to 17. These authors did not ask parents to report possible sources of ETS
exposure for their children other than their own tobacco consumption. Strachan and coworkers'
study in 6-1/2- to 7-1/2-year-old children in Scotland (Strachan et al., 1989) showed that 73% of
children from households with no smokers had detectable concentrations of cotinine in saliva,
whereas only 1 in 365 children from households with one or more smokers had no detectable
salivary cotinine. The assay used by Strachan and coworkers was 10 times more sensitive than that
used by Coultas and coworkers, and this may explain the larger number of subjects with
detectable levels in the former study when compared to the latter.
Greenberg and coworkers (1984) studied cotinine levels in 32 infants in North Carolina
with reported exposure to tobacco smoke within the previous 24 hours and in 19 unexposed
infants. All subjects were under 10 months old. Urine samples of all exposed infants contained
8-3
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
cotinine, whereas all unexposed infants except 2 (11%) had undetectable urine cotinine or levels
below those of exposed infants with the lowest levels of urine cotinine. This same group of
researchers reported results for a larger sample (433 infants at a mean age of 18 days) of the same
population (Greenberg et al., 1989). They found that, of 157 infants who reportedly lived in
nonsmoking households and were also not in contact with smokers the previous week, 37 infants
(24%) had cotinine in their urine. They concluded that these infants had contact with tobacco
smoke during the previous week and that this contact was unknown to or was not reported by
their mothers.
Greenberg and coworkers (1991) followed 152 of the 433 infants originally enrolled and
reassessed exposure to ETS (through maternal interviews) and urine cotinine levels when the child
was 12.3 ± 0.6 months old. They found a significant increase in the prevalence of tobacco smoke
absorption, indicated by excretion of cotinine, during the first year of life (from 53% at a mean
age of 3 weeks to 77%). The interviews showed that this was mainly due to an increased exposure
to nonhousehold sources of smoke (from 14% to 36%). The proportion of infants who reportedly
had no contact with smokers but had cotinine in their urine increased from 24% at 3 weeks to 49%
at 1 year of age.
These results indicate that studies that rely exclusively on parental questionnaires to
ascertain ETS exposure in children may misclassify many exposed subjects as nonexposed.
Moreover, the degree of misclassification may increase with the child's age.
The possible consequences of this misclassification of exposure need to be discussed in
detail. Nondifferential misclassification (i.e., exposure classification that is incorrect in equal
proportions of diseased and nondiseased subjects) biases the observed results towards a conclusion
of no effect (Rothman, 1988; see below). The effect of differential misclassification depends on
the direction in which misclassification occurs. If true ETS exposure is preferentially reported by
parents of diseased subjects (i.e., there is reporting bias), an excess of disease prevalence would be
found among exposed subjects when compared to unexposed subjects that is unrelated to any
biological effect of ETS. The evidence available clearly indicates that this is a very unlikely
explanation for the reported misclassification of ETS exposure in infants and children. In fact,
reporting bias cannot explain the substantial increase in "underreporting" of exposure with age.
The logical explanation is provided by the finding that exposure to nonhousehold smokers
increases significantly with age and parallels the increase in the proportion of subjects who have
cotinine in their urine (Greenberg et al., 1991). There is no reason to believe that exposure to
smokers may occur preferentially among diseased children, and the contrary may be more
8-4
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
reasonable; the increased awareness of the ill effects of ETS inhalation may induce parents to limit
contact between their diseased children and nonhousehold smokers. Thus, the net effect of
misclassification of exposure, both nondifferential and differential, should be a systematic
downward bias or bias toward observing no effect. A correction for the nondifferential
misclassification bias of background exposure is made below.
8.3. ADJUSTMENT FOR BACKGROUND EXPOSURE
An important conclusion of the previous discussion is that studies based on parental
questionnaires may underestimate the health risk from ETS in children due to underreporting of
ETS exposure. The NRC report on passive smoking (NRC, 1986) adopted the use of cotinine
measures to correct for misreporting of ETS exposure for lung cancer effects, and this approach
was adapted for use in Chapter 6 of this report. It will also be employed here, with the cotinine
ratios, however, based on exposure data in children rather than in adults. The method is based on
several assumptions: (1) cotinine concentrations in body fluids of nonsmokers are linearly related
to ETS exposure, (2) the excess risk of respiratory illness in subjects exposed to ETS is linearly
related to the dose of ETS absorbed, (3) the relationship between ambient and absorbed ETS is
linear, and (4) one cotinine determination may adequately represent average childhood exposure to
ETS. While considerable evidence exists for assumptions 1 through 3, there is now some evidence
that assumption 4 may not be entirely warranted. Coultas and co workers (1990b) in a small study
of 9 children from 10 homes with at least 1 smoker reported that there is considerable variability
in cotinine levels in body fluids within individuals exposed to ETS when such levels are
repeatedly measured in different days. Thus, while the method of adjustment is based on group
mean body cotinine levels, which apparently reflect well household ETS levels (see below), the
intraindividual variability may subject these means to some error.
Application of the algorithms proposed by the NRC requires some knowledge of Z, the
ratio between the operative mean dose level in the "exposed" group, dE, and the mean dose level in
the "unexposed" group, dN. RRCdjj), the relative risk for the group identified as "exposed"
compared to the group identified as "unexposed", is thus given by
= (l+Z*/JdN)/(l+/?dN) (8-1)
where /3 is the amount of increase per unit dose and Z > RR(dE) > 1. (The "unexposed" group
actually contains those with background exposure plus those truly unexposed.)
8-5
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Several studies are available that could be used for the purpose of estimating Z. Jarvis and
coworkers (1985) studied 569 nonsmoking schoolchildren ages 11 to 16 in Great Britain. The
investigators reported that, when compared to salivary cotinine levels in children of nonsmoking
parents (N = 269), mean levels of salivary cotinine were 3.0 times as high in children whose father
smoked (N - 96), 4.4 times as high in children whose mother smoked, and 7.7 times as high in
children whose parents were both smokers. Pattishall and coworkers (1985) reported that children
from homes with smokers (N = 20) had 4.1 times as high mean levels of serum cotinine as children
from nonsmoking families. Black children, however, in the same study had lower values of Z
(2.8) than did white children. Coultas and coworkers (1987) found that, among 600 U.S. children
up to age 17 years, mean salivary cotinine levels were between 1.3 and 2.6 times as high among
subjects exposed to one cigarette smoker at home as among unexposed subjects, and between 2.9
and 3.5 times as high among subjects exposed to two or more smokers at home as among subjects
not exposed to cigarette smokers at home. Strachan and coworkers (1989) reported separate results
for 6-1/2- to 7-1/2-year-old Scottish children belonging to families living in their own homes
and for those belonging to families living in rented homes. In the former, geometric mean
salivary cotinine was 6 times as high among subjects exposed to one cigarette smoker at home as
among unexposed subjects and 16 to 17 times as high among subjects exposed to two or more
smokers at home as among unexposed subjects. For children belonging to families living in rented
homes, the same ratios were 3 to 5.5 times and 4 to 7 times, respectively.
While these studies show consistent relationships between mean body cotinine levels in
children and home smoker occupancy, there is also a wide variability in the estimated Z ratios,
ranging from 1+ to 17. These different estimates may have very important effects on the
background exposure adjustment and, thus, on the calculation of adjusted relative risks for
different studies (see also Chapter 6). For example, for a study in which the observed relative risk
(RR) is 2.0 but for which the Z ratio is 3, equation 8-1 can be solved for /?dN, which is the
estimated increase in relative risk for the group called "unexposed" but who in fact have been
exposed to some recent ETS. Solving,
/3dN=l.
Thus, the adjusted RR for the group identified as "unexposed" would be 2, and the adjusted RR
for an "exposed" group compared to a truly unexposed group would be 1 + (3*1) = 4, i.e., twice the
observed risk. For a similar example (observed RR = 2) but with Z = 5, /3dN = 0.3, the RR for a
8-6
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
group identified as "unexposed" in this case would be 1.3, and the adjusted RR for an "exposed" to
a truly unexposed group would be 2.67. Finally, if the observed RR is still 2 but Z = 17, /JdN =
0.07, RR for "unexposed" would be 1.07 and the adjusted RR for exposed children would be 2.13.
These results are shown in Table 8-1.
These calculations show that when use of parental questionnaires significantly
underestimates their children's exposures to other sources of ETS (other than via the parental ETS)
and values of Z are lower (as found in black children by Pattishall and coworkers [1985], and in
children of lower socioeconomic status by Strachan and coworkers [1989]), the "true" RR of
children exposed to ETS may be considerably underestimated. But perhaps the most important
conclusion that may be derived from the above analysis is that exposure to ETS from sources other
than smoking parents may be high enough to constitute a significant risk for their health. This
may be particularly consequential for children of lower socioeconomic levels, whose nutritional
status, crowding conditions at home, and opportunity for contact with biological agents of disease
make them a part of the population that is particularly susceptible to respiratory illnesses during
infancy and childhood. Available data show that ETS exposure via honhousehold members in
these children, as measured by cotinine levels in body fluids, may be as much as one-third that of
children exposed to one smoking parent (Z = 3). In the example presented above (observed
RR = 2), the estimate of the adjusted relative risk is 4 for children of smoking parents to the truly
unexposed children. However, using the same assumptions, children of nonsmoking parents who
are exposed to ETS (at background levels found in some of the studies) would have twice as high a
risk of developing the illness under study as children truly unexposed to ETS.
A cautionary note about the model is appropriate. Table 8-1 shows that, for observed
RR = 2 and Z = 3, the adjusted relative risk is 4. However, as the observed RR and Z get closer
together, the behavior of the model becomes erratic. This is shown in Table 8-2. In fact, the
model (equation 8-1) becomes undefined if Z is less than or equal to the observed RR, and it
reaches some stability only as Z becomes at least 30% to 50% greater than RR.
Fortunately, the estimates of Z presented above are appreciably greater than the observed
relative risk estimates seen in Chapter 7, and in the observed range of both RR and Z, the model
yields relatively stable estimates of the adjusted RR. Furthermore, as discussed in Chapter 6, the
values of RR and Z are expected to be correlated for each study, i.e., the greater the Z ratio
between exposed and unexposed groups in each study, the greater should be the observed RR and
the less the effect of the (equation 8-1) adjustment.
8-7
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
If the above model is correct, then exposure of children to ETS, other than at home
(parental smoking) may be an important risk factor for respiratory illness in childhood. On the
other hand, it is also possible that for at least some respiratory illnesses, outside exposure to ETS
has relatively little effect, either because outside exposures in younger children tend to be less
than those of older children or because there may be a threshold of exposure below which certain
respiratory effects may not be expected to occur. For this latter case, equation 8-1 is not an
appropriate model, and the observed relative risk would be taken to be the true risk. Both models
are addressed in the sections that follow.
8.4. ASSESSMENT OF RISK
Neither the NRC report (1986) nor the Surgeon General's report (U.S. DHHS, 1986)
attempted to assess the population or public health impact of the increased risk of respiratory
disorders in children attributable to ETS exposure. In this section, estimates will be derived for
the number of ETS-attributable lower respiratory tract infections in infants and for the induction
and exacerbation of childhood asthma. Quantifying the public health impact of other conditions,
such as reduced lung function, coughing, wheezing, and middle ear effusion, is difficult, either
because of the lack of overt symptoms or because some necessary U.S. population health statistics
are not available. Estimates of sudden infant death syndrome (SIDS) deaths attributable to ETS
will not be made but will be discussed in Section 8.4.3.
For the following quantitative analyses, estimates will be developed in terms of ranges.
The ranges are derived by the use of both threshold and nonthreshold (equation 8-1) models,
different estimates for population incidence and prevalence, and estimated values of Z and RR
from studies reviewed above. Various differences in design, disease definition, and conduct
among these studies make them less adaptable to meta-analysis techniques than were the lung
cancer studies. To the extent that a less rigorous statistical analysis is attempted here, the ranges
should reflect that uncertainty. •
8.4.1. Asthma
From the analysis of studies regarding risk for asthma and ETS exposure, it was concluded
that passive smoking increases both the number and severity of episodes in asthmatic children. It
was further concluded that ETS is a risk factor for new cases among previously asymptomatic
children, since the evidence is suggestive, but not conclusive, of a causal association (see Section
7.6). Relative risks for asthma ranged from 1.0 to 2.5 in the studies analyzed, but methodologies
8-8
05/15/92
-------
DRAFT—DO NQT QUOTE OR CITE
differed considerably among studies, and effects were often found only in children of mothers
who smoke heavily. Of the four large studies, totaling over 9,000 children (Burchfiel et al., 1986;
Sherman et al., 1990; Weitzman et al., 1990; Martinez et al., 1991b), three showed statistically
significant risk estimates ranging from 1.7 to 2.5, with the two largest ratios, 2.5 (Martinez et al.,
1991b) and 2.1 (Weitzman et al., 1990), coming from comparisons using children of heavily
smoking mothers (;»10 cig./day) as the exposed group. The third study (Burchfiel et al., 1986)
had OR = 1.7 for males with two smoking parents, but results were not significant either for girls
or for children with one parental smoker. The fourth study (Sherman et al., 1990) (770 children)
did not find an effect, but made no effort to assess the effect of heavy smoking by parents, nor
was there control for socioeconomic status. Thus, assigning a range of 1.75 to 2.25 for the
estimated relative risk of developing asthma for children of mothers who smoke 10 or more
cigarettes per day appears reasonable and is within the ranges of observed risk.
The above results suggest two possible scenarios. One scenario is that relatively heavy
exposure to ETS is needed to bring on asthma, i.e., there is a threshold of exposure below which
effects will not occur. Alternatively, lesser exposures may merely induce fewer effects, not
detectable statistically with these study designs. The choice of scenario does not affect the
observed relative risk but will affect whether or not an adjustment for background exposure (Z
ratio) is appropriate. Under the first (threshold) scenario, the estimates of RR = 1.75 to 2.25 need
no adjustment; under the alternative (nonthreshold) scenario, equation 8-1 applies.
Considering the nonthreshold model first, from the discussion in Section 8.3, it can be
assumed that values of 3 to 10 may be a reasonable range for estimates of Z (i.e., the ratio of body
cotinine levels in children whose mothers smoke heavily to those of children whose mothers do not
smoke). Lower values of Z would yield significantly larger estimates of asthma cases attributable
to ETS. Based on the above estimates for a range of Z and RR and use of the nonthreshold
model, the estimated range of adjusted relative risks for children of mothers who smoke 10 or
more cigarettes per day would be approximately 1.91 to 6.00 (see Table 8-3). Transforming
relative risks to attributable risks (Rothman, 1986), 48% to 83% of all cases of asthma among
children of mothers who smoke 10 or more cigarettes per day may be attributable to passive
smoking based on
ARE= 100
where ARE is the attributable risk (%) for the exposed population.
8-9
(8-2)
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Under the assumptions of the threshold model, RR = 1.75 to 2.25 for children of heavily
smoking mothers, and the ARE = 43% to 56% (see Table 8-3); for children of light smoking
mothers, RR = 1, and the ARE = 0.
To calculate the percentage of all cases occurring in a mixed population of exposed and
unexposed individuals that is attributable to exposure (ART), knowledge of the prevalence of
mothers smoking 10 or more cigarettes per day is needed because
ART = ARE
(8-3)
where Pj is the proportion of cases that is exposed (Rothman, 1986). It has been reported that
approximately 26% of the population of women of childbearing age smoked in the United States
in 1988 (CDC, 1991b) and in 1990 (CDC, unpublished). For the number of cigarettes smoked,
Weitzman and coworkers (1990), using the 1981 National Health Information Survey (NHIS),
found that approximately 50% of smoking mothers of children ages 0 to 5 years smoke 10 or more
cigarettes per day. The 1990 NHIS reports that 78% of smoking women ages 18 to 44 smoke at
least 10 cigarettes per day (data courtesy of Dr. Gary Giovino, CDC, unpublished). We have used
an average of 65% to derive the estimates in Table 8-3. Based on these figures and the threshold
model, it can thus be estimated that approximately 7% to 9% of all cases of asthma may be
attributable to exposure to ETS from mothers who smoke 10 or more cigarettes per day. Estimates
of the prevalence of asthma among U.S. children less than age 18 vary from 5% to 10% (Clark and
Godfrey, 1983) to 3% to 8% (R. Evans et al., 1987), depending on disease definition. This latter-
paper uses the data from the 1979-1981 NHIS and derives a population asthma prevalence of 2
million to 5 million. A more recent estimate from the 1989 NHIS is 3.9 million (U.S. DHHS,
1990b). Use of these population prevalence figures and the threshold model provides a range of
8,000 to 26,000 as the annual number of new cases of childhood asthma attributable to mothers
who smoke 10 or more cigarettes per day. The confidence in this estimate is medium and is
dependent on the conclusion that ETS is a risk factor for asthma induction.
If the nonthreshold model applies, use of the same prevalence figures leads to a range of
13,000 to 60,000 new cases per year attributable to all ETS exposures (Table 8-3).
While the range of 8,000 to 60,000 is plausible, the existing data are more supportive of the
threshold model, which assumes that rather heavy exposures to ETS are required to induce asthma
in previously asymptotic children (Section 7.6.2). Thus, the range of 8,000 to 26,000 will be
8-10
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
adopted as the more probable range of new cases among children per year attributable to ETS
exposure.
In view of the increased number and severity of asthmatic episodes also caused by ETS,
the public health impact of ETS on asthmatic children is considerably greater than the range of
estimates for new cases presented above. Shephard (1992), after reviewing several studies,
concludes that ETS exposure (from any source) exacerbates preexisting asthma in approximately
20% of patients. If this figure is correct, up to 1 million asthmatic children could be affected.
Also, in an earlier study, O'Connell and Logan (1974) found that parental smoking aggravated
clinical symptoms of 67% of 265 asthmatic children in the Midwest versus 16% of 137 controls
(p < 0.0001) and that 10% of 400 asthmatic patients (of both smoking and nonsmoking parents)
considered tobacco smoke a major aggravating factor. D. Evans and coworkers (1987) found that
passive smoking by asthmatic children in New York City (via presence of smokers in the
household) was associated with a mean annual increase of 1.34 emergency room visits per year for
asthmatic symptoms, an increase of 63% over asthmatic children from nonsmoking households.
Thus, exposure to ETS in general and especially to parental ETS adversely affects hundreds of
thousands of asthmatic children.
8.4.2. Lower Respiratory Illness
From the assessment of available data (see Section 7.3), it was concluded that exposure of
infants and young children to ETS causes an increased incidence of lower respiratory illness
(LRI). An examination of the data in the referenced studies of both Tables 7-1 and 7-2 leads to
the conclusion that the observed risk of having LRIs is approximately 1.5 to 2.0 times as high in
young children whose mothers smoke as in those whose mothers do not smoke and that the risk is
probably higher in infants than in toddlers.
This estimate is also consistent with that of the NRC (1986), which estimated a relative risk
of up to 2 for infants who have one or more parents who smoke. The more recent evidence
reviewed here 'strongly suggests that the increased risk due to ETS exposure lasts for at least the
first 18 months and decreases after that. Based on this evidence, this chapter estimates a relative
risk range of 1.5 to 2.0 for infants and children up to 18 months old who have smoking mothers.
It will assume that the increased risk is zero after 18 months.
Based on these findings, and following equation.8-1 with a range of Z = 3 to 10 and
RR = 1.5 to 2.0, the adjusted relative risk range becomes 1.6 to 4.0, and ARE takes the range 38%
to 75%. As in the previous section, for equation 8-3, the mixed population attributable risk ART
8-11
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
takes the range 10% to 20%, again based on 1988 and 1990 estimates of approximately 26% women
of childbearing age who smoked (CDC, 1991b; CDC, unpublished). Because the estimated mean
number of cigarettes smoked by these women is approximately 17 to 20 per day (CDC 1991b;
CDC, unpublished), it is reasonable to assume that most children of smoking mothers will be
exposed. Therefore, the proportion of cases exposed, Pz, is estimated to be 0.26.
It has recently been shown that the incidence of LRIs early in life is approximately 30%
(Wright, 1991). When the analysis is limited to the first 18 months of life, the population at risk is
approximately 5.5 million children. Application of the same algorithms described above yields
150,000 to 300,000 cases of LRIs annually in children under 18 months old attributable to
exposure to ETS generated mostly by smoking mothers. Approximately 5% of these LRIs require
admission to a hospital (Wright, 1989), and therefore, it is estimated that 7,500 to 15,000
hospitalizations yearly for LRIs may be attributable to ETS exposure.
While these estimates may appear large, three factors suggest that they are on the low side.
First, although these estimates are calculated only for children less than 18 months old, Section 7.2
presents evidence that these ETS-attributed increased risks extend at a decreasing rate up to 3
years of age. Second, no estimates have been calculated for exposure in a smoking father-
nonsmoking mother household. Third, these numbers do not take into account the fact that many
infants and young children have recurrent LRIs, and therefore, more than one episode of such
illnesses may be attributable to ETS in each exposed child.
8.4.3. Sudden Infant Death Syndrome
Because this report concludes that there is an association between maternal smoking and
SIDS but is unable to determine the contribution that ETS makes to that association (see Section
7.7), no estimate of ETS-attributable SIDS deaths will be calculated. The Centers for Disease
Control (CDC) (1991a) provides an estimate of 702 SIDS deaths attributable to maternal smoking,
based on a relative risk of 1.5 for infants of actively smoking mothers. While this report concurs
with the numbers and the methodology used to determine that estimate, it is unable to apportion
the in utero, lactation, and ETS exposure components of the risk.
8.5. CONCLUSIONS
This chapter has attempted to estimate the U.S. population impact of ETS exposure on
childhood asthma and lower respiratory tract infections in young children. For new cases of
asthma in previously symptomatic children under 18 years of age, we estimate that 8,000 to 26,000
8-12
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
is a probable range of new cases per year that are attributable to ETS exposure from mothers who
smoke at least 10 cigarettes per day. The confidence in this range is medium and is dependent on
the conclusion that ETS is a risk factor for asthma induction.
While the data are most supportive of a situation in which heavy exposures to ETS are
required to induce new cases of asthma, two other scenarios would lead to larger estimates. The
first is that even in the absence of smoking mothers, a child could receive heavy ETS exposure
from other sources. The second is that lesser ETS exposures induce lesser numbers of new cases,
and the increase is not statistically detectable. Under this latter (nonthreshold) scenario, the range
of new cases of asthma annually attributable to ETS exposure is 13,000 to 60,000.
This report concludes that, in addition to inducing new cases of asthma, ETS exposure
increases the number and severity of episodes among this country's 2 million to 5 million
asthmatic children. This chapter considers exposure to parental smoking to be a major
aggravating factor to approximately 10%, or 200,000, asthmatic children. Estimates of the number
of asthmatics whose condition is aggravated to some degree by ETS exposure are very approximate
but could run well over 1 million.
This chapter also estimates that 150,000 to 300,000 cases annually of lower respiratory tract
infections in children up to 18 months old are attributable to ETS exposure, most of which comes
from smoking parents (mostly mothers). These ETS-attributable cases are estimated to result in
7,500 to 15,000 hospitalizations annually. Confidence in these estimates is high based on the
conclusion of a causal association and the strong validity of parental smoking as a surrogate of
temporally-relevant ETS exposure in infants and young children. Additional cases and
hospitalizations are expected to occur in children up to 3 years old in decreasing numbers, but this
report makes no further quantitative estimates.
Infants' exposure to ETS may also be responsible for a portion of the more than 700 deaths
from SIDS attributable to maternal smoking by the CDC (199la), but this report is unable to
determine whether and to what extent these deaths can be attributed specifically to ETS exposure.
The estimates of population impact presented above are given in ranges and approximate
values to reflect the uncertainty of extrapolating from individual studies to the population. As
with the lung cancer population impact assessment (Chapter 6), these extrapolations are all based
on human studies conducted at true environmental levels. Therefore, they suffer from none of
the uncertainties associated with either animal-to-human or high-to-low exposure extrapolations.
In addition to the estimates presented above, ETS exposure in children also leads to
reduced lung function, increased symptoms of respiratory irritation, and increased prevalence of
8-13
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
middle ear effusion, but this report does not provide estimates of the population impact of ETS
exposure for these conditions.
8-14
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table 8-1. Adjusted relative risks for "exposed children." Adjusted or background exposure based
on body cotinine ratios between "exposed" and "unexposed" and equation 8-1
Z RATIO OF BODY COTININE LEVELS ("EXPOSED"/"UNEXPOSED")
1.50 2.00 3.00 5.00 7.00 10.00 13.00 17.00
OBSERVED
RELATIVE
RISKS
(RR)
1.0
1.50
1.75
2.00
2.50
3.00
1
-
-
-
-
-
1
3.00
7.00
' -
-
-
1
2.00
2.80
4.00
10.00
-
1
1.71
2.15
2.67
4.00
6.00
1
1.64
2.00
2.40
3.33
4.50
1
1.59
1.91
2.25
3.00
3.86
1
1.57
1.87
.2.18
2.86
3.60
1
1.55
1.84
2.13
2.76
3.43
Table 8-2. Behavior variations in adjusted relative risks from equation 8-1 when the observed
relative risks and Z ratios are close together
OBSERVED
RELATIVE
RISKS
(RR)
1.50
1.75
2.00
2.25
2.50
1.50
1.75
Z RATIO
2.00 2.25
2.50 2.75
3.00 10.00
'
-3.5
-2.0
-1.5
-1.25
4.50
-
-6.00
-3.38
-2.50
3.00
7.00
-9.00
-5.00
2.50
4.38
10.00
-
-12.50
2.25
3.50
6.00
13.50
-
2.10
3.06
4.67
7.88
17.50
2.00
2.80
4.00
6.00
10.00
1.59
1.91
2.25
2.62
3.00
8-15
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
3^
"S •«
v3 u)
t; «
JJS
•aP
*1
"d ^
H w
« a
^J °
W T3
5T3 a>
5»
> ej
'Si
•32
«2
•CJ 0
•S.2
.i2, esw
"5 o w
•<•£.§
«« 3 O
O t3 5
u> O *
« >-" T3
cs ^ *S
III
•*^
« 1« £
w < *
^< ^^
"S *- a
o o S
«> «« 5r
S5^?
as a
n
f;
±A
Relative Ris
t3
CD
>
IH
CD
CO
.0
O
O
O
VO
o
Tj"
0
oo
CM*
CM
VO
CM*
»-H
CT\
*
1
1
.
Relative Ris:
•o
-4_l
CO
•^"»
T>
*•
OO
o
t-;
O
vq
o
CM
VO
0
00
"^
o
VO
in
o
to
o
•*
PH
"*
1
I
'
I
I
C7N
O
C>
O
o
o
II
u>
**~s
tn
PH
**
CM
CM
O
CM
O
t-~
O
S
O
CM
»— t
c^
i
•i
^
fM
fe
K
<
O O
o o
C5 O ^^
T^* ^"5
CM vo
o o
o o
°.,2 °.
CM TT
CM in
o o
o o
o\ *^ vb"
t— 1 ^f
I si
oo !Q
o o
o o o
C^ •*"* O^
CO ^f
—i to
0 0
0 0
o^ o o^
d vo"
—< CM
14
cN
"%
CD 0
tt
£j 0
-^-> *j3
i *3
E™1 o
W An
•s
2
.2?
*o
o
Al
00
o
CO
CD
1
3
CO
O
U
X
CD
w
t"
CJ
&
1^
cd
CO
CD
s
Threshold model ass
to induce new cases.
TS exposure can produce some new cases of asthma.
w
"o3
i *
A
+^
CO
Q
%
&
Nonthreshold model
Equation 8-1.
sed population.
o
CD
H
0
p.
_o
0
Attributable risk fra
(mixed) population.
1— (
GJ
•4-*
o
CD
•S
IH
pj
.2
o
Attributable risk fra
vq
C5
X
vo
CM
o,
•o
IH
CD
a
CO
a>
5
.2?
*o
o
CO
CS
1
CO
CD
oo
CD
4—»
73
O
IH
a
CD
IH
O
C3
Proportion of womei
age who smoke cigarettes.
CD
T!
O
-s
o
2
o
d
Proportion of womei
§
co
CD
• i-i
CD
•S3
2
"o
CO
CD
OO
CD
o
|
o
o
I
CO
OS
d
.2
3
S
l~i
o
e!
.2
n3
tN
d
o
CD
CO
CD
OO
d
0
n
N
CO
(11
>
CD
»— c
CD
'd
.1-1
o
o
o
1
CD
a
'o
.s
•4— '
(3
to
II
N
CO
*7Ti
>
*-H
CD
d
d
o
o
ctt
CD
^a
'o
o
1
+ *
8-16
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
SELECTED BIBLIOGRAPHY
Note: This section includes all references cited in the document as well as additional
references that were reviewed during the preparation of this document. This is not intended to be
a comprehensive list of all references available on the topic.
Adams, J.D.; O'Mara-Adams, K.J.; Hoffmann, D. (1987) Toxic and carcinogenic agents in
undiluted mainstream smoke and sidestream smoke of different types of cigarettes.
Carcinogenesis 8(5):729-731.
Aguayo, S.M.; Kane, M.A.; King, T.E.; Schwarz, M.I.; Grauer L.; Miller, Y.E. (1989) Increased
levels of bombesin-like peptides in the lower respiratory tract of asymptomatic cigarette
smokers. J. Clin. Invest. 84:1105-1113.
Akiba, S.; Kato, H.; Blot, WJ. (1986) Passive smoking and lung cancer among Japanese women.
Cancer Res. 46:4804-4807.
Alderson, M.R.; Lee, P.N.; Wang, R. (1985) Risks of lung cancer, chronic bronchitis, ischaemic
heart disease and stroke in relation to type of cigarette smoked. J. Epidemiol. Community
Health 39:286, ,
Alfheim, I.; Ramdahl, T. (1984) Contribution of wood combustion to indoor air pollution as
measured by mutagenicity in Salmonella and polycyclic aromatic hydrocarbon
concentration. Environ. Mol. Mutagen. 6:121-120.
Almanac of the 50 States. (1989) Hornon, E.R., ed. Palo Alto, CA: Information Publications.
American Academy of Pediatrics (1986) (p. 1-7).
Anderson, L.J.; Parker, R.A.; Strikas, R.A.; et al. (1988) Day-care center attendance and
hospitalization for lower respiratory tract illness. Pediatrics 82:300-308.
R-l
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Andrae, S.; Axelson, O.; Bjorksten, B.; Fredriksson, M.; Ljellman, N-IM. (1988) Symptoms of
bronchial hyperreactivity and asthma in relation to environmental factors. Arch. Dis.
Child. 63:473-478.
Aoki,_j et al. (1987) Smoking and Health. _. Elsevier Science Publishers B.V. (Biomedical
Division), pp. 483-486.
Arundel, A.; Sterling, T.; Weinkam, J. (1987) Never smoker lung cancer risks from exposure to
particulate tobacco smoke. Environ. Intl. 13:409-426.
Badre, R.; Guillerm, R.; Abran, N.; Bourdin, M.; Dumas, C. (1978) Atmospheric pollution by
cigarette smoking. Ann. Pharm. Fr. 36:443-452.
Basrur, P.K.; McClure, S.; Zilkey, B. (1978) A comparison of short term bioassay results with
carcinogenicity of experimental cigarettes. In: Nieburgs, H.E., ed. Prevention and
detection of cancer, Part 1, Vol. 2, New York: Marcel Dekker, pp. 2041-2048.
Bellofiore, S.; Eidelman, D.H.; Macklem, P.T.; Martin, J.G. (1989) Effects of elastase-induced
emphysema on airway responsiveness to methacholine in rats. J. Appl. Physiol. 66:606-612.
Benner, C.L.; Bayona, J.M.; Caka, F.M.; Tang, H.; Lewis, L.D.; Eatough, D.J. (1989) Chemical
composition of environmental tobacco smoke. 2. Particle-phase compounds. Environ. Sci.
Technol. 23:688-699.
Bergman, A.B.; Wiesner, B.A. (1976) Relationship of passive cigarette-smoking to sudden infant
death syndrome. Pediatrics 58:665-668.
Berkey, C.S.; Ware, J.H.; Dockery, D.W.; Ferris, E.G., Jr.; Speizer, F.E. (1986) Indoor air pollution
and pulmonary function in preadolescent children. Am. J. Epidemiol. 123:250-260.
Bernfeld, P.; Homburger, F.; Soto, E.; Pai, KJ. (1979) Cigarette smoke inhalation studies in
inbred Syrian golden hamsters. J. Natl. Cancer Inst. 63:675-689.
R-2
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Best, E.W.R.; Josie, G.H.; Walker, C.B. (1961) A Canadian study of mortality in relation to
smoking habits. A preliminary report. Can. J. Public Health 52:99-106.
Bewley, B.R.; Halil, T.; Snaith, A.H. (1973) Smoking by primary schoolchildren: prevalence and
associated respiratory symptoms. Br. J. Prev. Soc. Med. 27:150-153.
Bisgaard, H.; Dalgaard, P.; Nyboe, J. (1987) Risk factors for wheezing during infancy: a study of
5953 infants. Acta Paediatr. Scand. 76:719-726.
Black, N. (1984) Surgery for glue ear—a modern epidemic? Lancet 1:835-837,
Black, N. (1985) The etiology of glue ear—a case-control study. Int. J. Pediatr. Otolaryngol.
(Stockholm) 9:121-133.
Bland, M.; Bewley, B.R.; Pollard, V.; Banks, N.M. (1978) Effect of children's and parents'
smoking on respiratory symptoms. Arch. Dis. Child. 53:100-105.
Blot, W.J.; Fraumeni, J.F. (1986) Passive smoking and lung cancer. J. Natl. Cancer Inst.
77(5):993-999.
Breese-Hall, C.; Hall, J.H.; Gala, C.L.; MaGill, F.B.; Leddy, J.P. (1984) Long-term prospective
study in children after respiratory syncytial virus infection. J. Pediatr. 105:358-364.
Breslow, N.E.; Day, N.E. (1980) Statistical methods in cancer research. Lyon: IARC Sci. Publ.
No. 32.
Britten, N. (1988) Validity of claims to lifelong non-smoking at age 36 in a longitudinal study. Int.
J. Epidemiol. 17(3):525-529.
Bross, I.D.J.; Gibson, R. (1968) Risk of lung cancer in smokers who switch to filter cigarettes.
Am. J. Public Health 58:1396-1403.
R-3
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Brown, C.C.; Chu, K. (1987) Use of multistage models to infer stage affected by carcinogenic
exposure: example of lung cancer and cigarette smoking. J. Chronic Dis. 40(2):171A-179A.
Brownson, R.C.; Reif, J.S.; Keefe, T.J.; Ferguson, S.W.; Pritzl, J.A. (1987) Risk factors for
adenocarcinoma of the lung. Am. J. Epidemiol. 125:25-34.
Brunekreef, B.; Boleij, J.S.M. (1982) Long-term average suspended particulate concentrations in
smokers' homes. Int. Arch. Occup. Environ. Health 50:299-302.
Brunekreef, B.; Fischer, P.; Remijn, B.; Van der Lende, R.; Schouten, J.; Quanjer, P.H. (1985)
Indoor air pollution and its effects on pulmonary function of adult non-smoking women.
III. Passive smoking and pulmonary function. Int. J. Epidemiol. 14:227-230.
Brunnemann, K.D.; Hoffmann, D.; Wynder, EX.; Gori, G.B. (1976) Chemical studies on tobacco
smoke: 37. Determination of tar, nicotine, and carbon monoxide in cigarette smoke. A
comparison of international smoking conditions. In: Wynder, E.L.; Hoffmann, D; Gori,
G.B., eds. Modifying the risk for the smoker: proceedings of the third world conference on
smoking and health, DHEW Pub. No. (NIH) 76-1221, New York: June 1975; Vol 1. U.S.
Department of Health, Education, and Welfare, National Cancer Institute, pp. 441-449.
Brunnemann, K.D.; Adams, J.D.; Ho, D.P.; Hoffmann D. (1978) The influence of tobacco smoke
on indoor atmospheres. II. Volatile and tobacco-specific nitrosamines in main and
sidestream smoke and their contribution to indoor pollution: proceedings of the 4th joint
conference on sensing of environmental pollutants. Washington, DC: American Chemical
Society, pp. 876-880.
Buffler, P.A.; Pickle, L.W.; Mason, T.J.; Contant, C. (1984) The causes of lung cancer in Texas. In:
Mizell, M.; Correa, P., eds. Lung cancer: causes and prevention. New York: Verlag Chemie
International, pp. 83-99.
Burchfield, C.M.; Higgins, M.W.; Keller, J.B.; Howatt, W.F.; Butler, W.J.; Higgins, I.T.T. (1986)
Passive smoking in childhood: respiratory conditions and pulmonary function in
Tecumseh, Michigan. Am. Rev. Respir. Dis. 133:966-973.
R-4
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Burney, P.O.; Britton, J.R.; Chinn, S.; et al. (1987) Descriptive epidemiology of bronchial
reactivity in an adult population: results from a community study. Thorax 42:38-44.
Burrows, B.; Halonen, M.; Barbee, R.A.; Lebowitz, M.D. (1981) The relationship of serum
immunoglobulin E to cigarette smoking. Am. Rev. Respir. Dis. 124:523-525.
Burrows, B.; Knudson, R.J.; Cline, M.G.; Lebowitz, M.D. (1988) A reexamination of risk factors
for ventilatory impairment. Am. Rev. Respir. Dis. 138:829-836.
Burrows, B.; Martinez, F.D.; Halonen, M.; Barbee, R.A.; Cline, M.G. (1989) Association of asthma
with serum IgE levels and skin-test reactivity to allergens. N. Engl. J. Med. 320:271-277.
Butler, T.L. (1988) The relationship of passive smoking to various health outcomes among
Seventh-Day Adventists in California [dissertation]. Los Angeles: University of California
at Los Angeles.
Butler, W.J. (1990) Unpublished comments submitted to EPA on its draft report entitled "Health
Effects of Passive Smoking: Assessment of Lung Cancer in Adults and Respiratory
Disorders in Children," EPA/600/6-90/006A, May, 1990.
Butler, W.J. (1991) Supplementary information on confounding and the reported epidemiologic
association between lung cancer and spousal smoking status (unpublished).
Caddeback, J.E.; Donovan, J.R.; Burg, W.R. (1976) Occupational aspects of passive smoking. Am.
Ind. Hyg. Assoc. J. 37:263-267.
Cameron, P.; Kostin, J.S.; Oselett, B.; Stocker, R.; Tighe, G.; Winton, J.; Wolfe, J.H.; Zaks, J.M.
(1969) The health of smokers' and nonsmokers' children. Journal of Allergy 43(6):336-
341.
Camilli, A.E.; Burrows, B.; Knudson, R.J.; Lyle, S.K.; Lebowitz, M.D. (1987) Longitudinal
changes in FEV1 in adults: effects of smoking and smoking cessation. Am. Rev. Respir.
Dis. 135:794-799.
R-5
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Cano, J.P.; Catalin, J.; Badre, R.; Dumas, C.; Viala, A.; Guillerme, R. (1970) Determination de la
nicotine par chromatographie en phase gazeuse. II. Applications. Ann. Pharm. Fr.
28:633-640.
CDC. See Centers for Disease Control.
Cederlof, R.; Friberg, L.; Hrubec, Z.; Lorich, U. (1975) The relationship of smoking and some
social covariables in mortality and cancer morbidity: a ten year follow-up on a probability
sample of 55,000 Swedish subjects age 18-69, Parts 1 and 2. Stockholm, Sweden:
Department of Environmental Hygiene, Karolinska Institute.
Centers for Disease Control. (1989) Chronic disease reports: mortality trends—United States, 1979-
1986. MMWR 38:189-191.
Centers for Disease Control. (1989) Years of potential life lost before age 65—United States, 1987.
MMWR 38:27-28.
Centers for Disease Control. (1991a) Smoking-attributable mortality and years of potential life
lost-United States, 1988. MMWR 40:62-71.
Centers for Disease Control. (1991b) Cigarette smoking among adults—United States, 1988.
MMWR 40:757-765.
Chan, W.C.; Fung, S.C. (1982) Lung cancer in non-smokers in Hong Kong. In: Grundmann, E.,
ed. Cancer campaign, Vol. 6, cancer epidemiology. Stuttgart: Gustav Fischer Verlag, pp.
199-202.
Chan, W.C.; Colbourne, M.J.; Fung, S.C.; Ho, H.C. (1979) Bronchial cancer in Hong Kong, 1976-
1977. Br.J. Cancer 39:182-192.
Chan, K.M.; Noble-Jamieson, C.M.; Elliman, A.; Bryan, E.M.; Silverman, M. (1989) Lung
function in children of low birth weight. Arch. Dis. Child. 64:1284-1293.
R-6
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Chan, K.N.; Elliman, A.; Bryan, E.; Silverman, M. (1989) Respiratory symptoms in children of
low birth weight. Arch. Dis. Child. 64:1294-1304.
Charlton, A. (1984) Children's coughs related to parental smoking. Brit. Med. J. 288:1647-1649.
Chen, M.F.; Kimisuka, G.; Wang, N.S. (1987) Human fetal lung changes associated with maternal
smoking during pregnancy. Pediatr. Pulmonol. 3:51-58.
Chen, Y. (1989) Synergistic effect of passive smoking and artificial feeding on hospitalization for
respiratory illness in early childhood. Chest 95:1004-1007.
Chen, Y.; Li, W. (1986) The effect of passive smoking on children's pulmonary function in
Shanghai. Am. J. Public Health 76:515-518.
Chen, Y.; Li, W.; Yu, S. (1986) Influence of passive smoking on admissions for respiratory illness
in early childhood. Br. Med. J. 293:303-306.
Chen, Y.; Li, W.; Yu, S.; Qian, W. (1988) Chang-Ning epidemiological study of children's health: I
Passive smoking and children's respiratory diseases. Int. J. Epidemiol. 17:348-355.
Chilmonczyk, B.A.; Knight, G.J.; Palomaki, G.E.; Pulkkinen, A.J.; Williams, J.; Haddow, J.E.
(1990) Environmental tobacco smoke exposure during infancy. Am. J. Public Health
80:1205-1208.
China Map Press. (1979) Atlas of Cancer Mortality in the People's Republic of China.
Clark, T.J.H.; Godfrey, S. (1983) Asthma. 2nd edition. Cambridge, _: Chapman and Hall Medical.
Claxton, L.D.; Morin, R.S.; Huges, T.J.; Lewtas, J. (1989) A genotoxic assessment of
environmental tobacco smoke using bacterial bioassays. Mutat. Res. 222:81-99.
Coghlin, J.; Hammond, S.K.; Gann, P. (1989) Development of epidemiologic tools for measuring
environmental tobacco smoke exposure. Am. J. Epidemiol. 130:696-704.
R-7
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Cogswell, J.J.; Mitchel, E.B.; Alexander, J. (1987) Parental smoking, breast feeding, and
respiratory infection in development of allergic diseases. Arch. Dis. Child. 62:336-344.
Colley, J.R.T. (1971) Respiratory disease in childhood. Br. Med. Bull. 27:9-14.
Colley, J.R.T. (1974) Respiratory symptoms in children and parental smoking and phlegm
production. Br. Med. J. 2:201-204.
Colley, J.R.T.; Holland, W.W.; Corkhill, R.T. (1974) Influence of passive smoking and parental
phlegm on pneumonia and bronchitis in early childhood. Lancet 2:1031-1034.
Collier, A.M.; Goldstein, G.M.; Shrewsbury, R.P.; Zhang, C.A.; Williams, R.W. (1990) Urine
cotinine elimination half-life in young children exposed to sidestream cigarette smoke. In:
Indoor air '90: proceedings of the 5th international conference on indoor air quality and
climate, July 29-August 3. Vol. 2: Characteristics of indoor air. Ottawa, Ontario, Canada:
Canada Mortgage and Housing Corporation, pp. 195-200.
Collins, M.H.; Moessinger, A.C.; Kleinerman, J.; Bassi, J.; Rosso, P.; Collins, A.M.; James, L.S.;
Blanc, W.A. (1985) Fetal lung hypoplasia associated with maternal smoking: a
morphometric analysis. Pediatr. Res. 19:408-412.
Comstock, G.W.; Meyer, M.B.; Helsing, K.J.; Tockman, M.S. (1981) The respiratory effects of
household exposure to tobacco smoke and gas cooking. Am. Rev. Respir. Dis. 124:143-148.
Corbo, G.M.; Fuciarelli, F.; Foresi, A.; De Benedetto, F. (1989) Snoring in children: association
with respiratory symptoms and passive smoking. Br. Med. J. 299:1491-1494.
Correa, P.; Fontham, E.; Pickle, L.; Lin, Y.; Haenszel, W. (1983) Passive smoking and lung cancer.
Lancet 2:595-597.
Coultas, D.B.; Howard, C.A.; Peake, G.T.; Skipper, B.J.; Samet, J.M. (1987) Salivary cotinine
levels and involuntary tobacco smoke exposure in children and adults in New Mexico. Am.
Rev. Respir. Dis. 136:305-309.
R-8
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Coultas, D.B.; Howard, C.A.; et al. (1988) Discrepancies between self-reported and validated
cigarette smoking in a community survey of New Mexico Hispanics. Am. Rev. Respir. Dis.
137:810-814.
Coultas, D.B.; Peake, G.T.; Samet, J.M. (1989) Questionaire assessment of lifetime and recent
exposure to environmental tobacco smoke. Am. J. Epidemiol. 130:338-347.
Coultas, D.B.; Samet, J.M.; McCarthy, J.F.; Spengler, J.D. (1990a) A personal monitoring study to
assess workplace exposure to environmental tobacco smoke. Am. J. Public Health
80:988-990.
Coultas, D.B.; Samet, J.M.; McCarthy, J.F.; Spengler, J.D. (1990b) Variability of measures of
exposure to environmental tobacco smoke in the home. Am. Rev. Respir. Dis. 142:602-606.
Cummings, K.M. (1990) Statement to the U.S. Environmental Protection Agency Science Advisory
Board Indoor Air Quality and Total Human Exposure Committee. Environmental Tobacco
Smoke Review, December 4, 1990.
Cummings, K.M.; Markello, S.J.; Mahoney, M.C.; Marshall, J.R. (1989) Measurement of lifetime
exposure to passive smoke. Am. J. Epidemiol. 130:122-132.
Cummings, K.M.; Markello, S.J.; et al. (1990) Measurement of current exposure to environmental
tobacco smoke. Arch. Environ. Health 45:74-79.
Cutz, E.; Chan, W.; Track, N.S. (1981) Bombesin, calcitonin and leuehkephalin immunoreactivity
in endocrine cells of the human lung. Experientia (Basel) 37:765-767.
Dagle, G.E.; McDonald, K.E.; Smith, L.G.; Stevens, D.L., Jr. (1978) Pulmonary carcinogenesis in
rats given implants of cigarette smoke condensate in beeswax pellets. J. Natl. Cancer Inst.
61:905-910.
Dalager, N.A.; Pickle, L.W.; Mason, T.J.; Correa, P.; Fontham, E.; Stemhagen, A.; Buffler, P.A.;
Ziegler, R.G.; Fraumeni, J.F. (1986) The relation of passive smoking to lung cancer.
Cancer Res. 46:4808-4811.
R-9
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Davies, R.F.; Day, T.D. (1969) A study of the comparative carcinogenicity of cigarette and cigar
smoke condensate on mouse skin. Br. J. Cancer 23:363-368.
Day, N.E.; Brown, C.C. (1980) Multistage models and primary prevention of cancer. J. Natl.
Cancer Inst. 64:977-989.
Dean, G.; Lee, P.N.; Todd, G.F.; Wicken, AJ. (1977) Report on a second retrospective study in
North-east England. Part I. Factors related to mortality from lung cancer, bronchitis, heart
disease and stroke in Cleveland County, with particular emphasis on the relative risks
associated with smoking filter and plain cigarettes. Research Paper 14. London: Tobacco
Research Council.
Deaner, R.M.; Trummer, MJ. (1970) Carcinoma of the lung in women. J. Thorac. Cardiovasc.
Surg. 59:551-554.
DeMarini, D.M. (1983) Genotoxicity of tobacco smoke and tobacco smoke condensate. Mutat. Res.
114:59-89.
Dijkstra, L.; Houthuijs, D.; Brunekreef, B.; Akkerman, I.; Boleij, J.S.M. (1990) Respiratory health
effects of the indoor environment in a population of Dutch children. Am. Rev. Respir.
Dis. 142:1172-1178.
Dodge, R. (1982) The effects of indoor pollution on Arizona children. Arch; Environ. Health.
37:151-155.
Doll, R. (1971) The age distribution of cancer. Implication for models of carcinogenesis. J. R. Stat.
Soc. A. 134:133-155.
Doll, R.; Hill, A.B. (1964a) Mortality in relation to smoking: ten years' observations of British
doctors. Br. Med. J. i: 1399-1410.
Doll, R.; Hill, A.B. (1964b) Mortality in relation to smoking: ten years' observations of British
doctors. Br. Med. J. i: 1460-1467.
R-10
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Doll, R.; Peto, R. (1976) Mortality in relation to smoking: 20 years' observations on male British
doctors. Br. Med. J. ii: 1525-1536.
Doll, R.; Peto, R. (1978) Cigarette smoking and bronchial carcinoma: dose and time relationships
among regular smokers and lifelong non-smokers. J. Epidemiol. Community Health
32:303-313.
Doll, R.; Hill, A.B.; Kreyberg, L. (1957) The significance of cell type in relation to the aetiology
of lung cancer. Br. J. Cancer 11:43-48.
Doll, R.; Gray, R.; Hafner, B.; Peto, R. (1980) Mortality in relation to smoking: 22 years'
observations on female British doctors. Br. Med. J. i:967-971.
Dontenwill, W.; Chevalier, H.J.; Harke, H.P.; Lafrenz, U.; Reckzeh, G.; Schneider, B. (1973)
Investigations on the effects of chronic cigarette smoke inhalation in Syrian golden
hamsters. J.Natl. Cancer Inst. 51:1781-1832.
Dube, M.F.; Green, C.R. (1982) Methods of collection of smoke for analytical purposes. Recent
Adv. Tobacco Sci. 8:42-102.
Dupont, W.D.; Plummer, W.D. (1990) Power and sample size calculations: a review and computer
program. In: Controlled Clinical Trials, Vol. 11, Elsevier Science Publishing Co., New
York, pp. 116-128.
Dutau, G.; Enjaume, C.; Petrus, M.; Darcos, P.; Demeurisse, P.; Rochiccioli, P. (1981) Enquete
epidemiologue sur le tabagisme passif des enfants de 0 a 6 ans. Arch. Fr. Pediatr.
38:721-725.
Eatough, D.J.; Benner, C.; Mooney, R.L.; Batholomew, D.; Steiner, D.S.; Hansen, L.D.; Lamb,
J.D.; Lewis, E.A. (1986) Gas and particle phase nicotine in environmental tobacco smoke:
proceedings of 79th annual meeting of the Air Pollution Control Association, Paper 86-
68.5. June 22-27, 1986; Minneapolis, MN. . '.
R-ll
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Eatough, D.J.; Benner, C.L.; Bayona, J.M.; Caka, H.; Tang, H.; Lewis, L.; Lamb, J.D.; Lee, M.L.;
Lewis, E.A.; Hansen, L.D. (1989a) Measurement of toxic and related air pollutants.
Pittsburgh, PA: Air Pollution Control Association, pp. 132-139.
Eatough, D.J.; Benner, C.L.; Tang, H.; Landon, V.; Richards, G.; Caka, P.M.; Crawford, F.;
Lewis, E.A.; Hansen, L.D.; Eatough, N.L. (1989b) The chemical composition of
environmental tobacco smoke. III. Identification of conservative tracers of environmental
tobacco smoke. Environ. Int. 15:19-28.
Eatough, D.J.; Caka, P.M.; Crawford, J.; Braithwaite, S.; Hansen, L.D.; Lewis, E.A. (1990)
Environmental tobacco smoke in commercial aircraft. In: Indoor air '90: proceedings of the
5th international conference on indoor air quality and climate, July 29-August 3. Vol. 2.
Characteristics of indoor air. Ottawa, Ontario, Canada: Canada Mortgage and Housing
Corporation, pp. 311-316.
Ekwo, E.; Weinberger, W.M.; Lachenbruch, P.A.; Humtley, W.H. (1983) Relationship of parental
smoking and gas cooking to respiratory disease in children. Chest 84:662-668.
Elliot, L.P.; Rowe, D.R. (1975) Air quality during public gatherings. J. Air Pollut. Control Assoc.
25:635-636.
Etzel, R.A.; Greenberg, R.A.; Haley, N.J.; Loda, F.A. (1985) Urinary cotinine excretion in
neonates exposed to tobacco smoke products in utero. J. Pediatr. 107:146-148.
Eudy, L.W.; Thorne, F.A.; Heavor, D.L.; Green, C.R.; Ingebrethsen, B.J. (1985) Studies on the
vapour-phase distribution of environmental nicotine by selected trapping and detection
methods. Presented at: 39th Tobacco'Chemists Research Conference; October; Montreal.
Eudy, L.W.; Thomas, F.A.; Heavner, D.L.; Green, C.R.; Ingebrethsen, B.J. (1986) Studies on the
vapor-particulate phase distribution of environmental nicotine by selective trapping and
detection methods: proceedings of the Air Pollution Control Association 79th annual
meeting, June 22-27, pp. 2-14.
R-12
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Evans, D.; Levison, J.; Feldman, C.H.; et al. (1987) The impact of passive smoking on emergency
room visits of urban children with asthma. Am. Rev. Respir. Dis. 135:567-572.
Evans, R.; Mullolly, D.I.; Wilson, R.W.; Gergen, P.J., Rosenberg, H.M.; Grauman, T.S.; Chevarley,
F.M.;
Feinlein, M. (1987) National trends in the morbilidity and mortality of asthma in the U.S. Chest
91(6):655-745.
Fergusson, D.M.; Horwood L.J.; Shannon, F.T.; Taylor, B. (1981) Paternal smoking and lower
respiratory illness in the first three years of life. J. Epidemiol. Community Health
35:180-184.
Ferris, E.G.; Ware, J.H.; Berkey, C.S.; Dockery, D.W.; Spiro, A.; Speizer, F.E. (1985) Effects of
passive smoking on health of children. Environ. Health Perspect. 62:289-295.
Feyerabend, C.; Higenbottam, T.; Russell, M.A.H. (1982) Nicotinic concentrations in urine and
saliva of smokers and nonsmokers. Brit Med J 284:1002-04.
Fielding, I.E. (1985) Smoking: health effects and control (Part I). N. Engl. J. Med. 313:491.
Fiore, M.C.; Novotny, T.E.; Pierce, J.P.; Hatziandreu, E.J.; Patel, K.M.; Davis, R.M. (1989) Trend
in cigarette smoking in the United States. The changing influence of gender and race.
JAMA 261:49-55.
First, M.W. (1984) Environmental tobacco smoke measurements: retrospect and prospect. Eur. J.
Respir. Dis. 65 (Suppl. 133):9-16.
Fleming, D.W.; Cochi, S.L.; Hightower, A.W.; Broome, C.V. (1987) Childhood upper respiratory
tract infections: to what degree is incidence affected by day-care attendance? Pediatrics
79:55-60.
Floderus, B.; Cederlof, R.; Friberg, L. (1988) Smoking and mortality: a 21-year follow-up based
on the Swedish Twin Registry. Int. J. Epidemiol. 17(2):322.
R-13
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Fontham, E.T.H.; Correa, P.; Wu-Williams, A.; Reynolds, P.; Greenberg, R.S.; Buffler, P.A.;
Chen, V.W.; Boyd, P.; Alterman, T.; Austin, D.F.; Liff, J.; Greenberg, S.D. (November
1991) Lung cancer in nonsmoking women: a multicenter case-control study. Cancer
Epidemiol. Biomarkers and Prev., Vol. 1:1, pp. 35-44.
Fraser, G.E.; Beeson, W.L., and Phillips, R.L. (1991) Diet and lung cancer in California Seventh
Day Adventists. Am J Epidemiol 133:683-93.
Freedman, D.A.; Navidi, W. (1987a) On the multistage model for carcinogenesis. Tech. Report.
No. 47, University of California.
Freedman, D.A.; Navidi, W. (1987b) On the risk of lung cancer for ex-smokers. Tech. Report. No.
135, University of California.
Friedman, G.D.; Petitti, D.B.; Bawol, R.D. (1983) Prevalence and correlates of passive smoking.
Am. J. Public Health 73:401- .
Gaffney, M.; Altshuler, B. (1988) Examination of the role of cigarette smoke in lung
carcinogenesis using multistage models. J. Natl. Cancer Inst. 80:925-931.
Gairola, C. (1982) Genetic effects of fresh cigarette smoke in Saccharomyces cerevisiae. Mutat.
Res. 102:123-136.
Gao, Y.; Blot, W.J.; Zheng, W.; Ershow, A.G.; Hsu, C.W.; Levin, L.I.; Zhang, R.; Fraumeni, J.F.
(1987) Lung cancer among Chinese women. Intl. J. Cancer 40:604-609.
Garfinkel, L. (1981) Time trends in lung cancer mortality among nonsmokers and a note on
passive smoking. J. Natl. Cancer Inst. 6:1061-1066.
Garfinkel, L. (1984) Passive smoking and cancer: American experience. Prev. Med. 13:691-697.
Garfinkel, L.; Silverberg, E. (1991) Lung cancer and smoking trends in the United States over the
past 25 years. CA 41:137-145.
R-14
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Garfinkel, L.; Stellman, S.D. (1988) Smoking and lung cancer in women: findings in a prospective
study. Cancer Res. 48:6951-6955.
Garfinkel, L. (1985) Selection, follow-up, and analysis in the American Cancer Society
Prospective Studies. Presented at a workshop on the selection, follow-up, and analysis in
prospective studies held at the Waldorf-Astoria Hotel* October 3-5, 1983; New York.
Garfinkel, L.; Auerbach, O.; Joubert, L. (1985) Involuntary smoking and lung cancer: a
case-control study. J. Natl. Cancer Inst. 75:463-469.
Geller-Bernstein, B.; Kenett, R.; Weisglass, L.; Tsur, S.; Lahav, M.; Levin, S. (1987) Atopic babies
with wheezy bronchitis. Allergy 42:85-91.
Geng, G.; Liang, Z.H.; Zhang, G.L. (1988) On the relationship between smoking and female lung
cancer. Smoking and Health, Elsevier Science Publishers, pp. 483-486.
Geng, G.; Liang, Z-H; et al. Effects of smoking and passive smoking on female lung cancer in
Tianjin area. Tianjin, People's Republic of China: Dept. of Epidemiol., Tianjin Med. Col.
Gerrard, J.W.; Helner, D.C.; Ko, C.G.; Mink, J.; Meyers, A.; Dosman, J.A. (1980) Immunoglobulin
levels in smokers and non-smokers. Ann. Allergy 44:261-262.
Gillis, C.R.; Hole, D.J.; Hawthorne, V.M.; Boyle, P. (1984) The effect of environmental tobacco
smoke in two urban communities in the west of Scotland. Eur. J. Respir. Dis.
133(Suppl.):121-126.
Glantz, S.A.; Parmley, W.W. (1991) Passive smoking and heart disease. Circulation 83:1-12.
Goldstein, G.M.; Collier, A.; Etzel, R.; Lewtas, t.; Haley, N. (1987) Elimination of urinary
cotinine in children exposed to known levels of side-stream cigarette smoke. In: Seifert,
H.; Esdorn, M.; Fischer, H.; Ruden, J.; Wegner, eds. Indoor air '87, Vol. 2, B. Berlin:
Oraniendruck GmbH, pp. 61-67,
R-15
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Gortmaker, S.L.; Walker, D.K.; Jacobs, F.H.; Ruch-Ross, H. (1982) Parental smoking and the risk
of childhood asthma. Am. J. Pub Health 72:572-579.
Graham, N.M.H.; Woodward, A.J.; Ryan, P.; Douglas, R.M. (1990) Acute respiratory illness in
Adelaide children. It The relationship of maternal stress, social supports and family
functioning. Int. J. Epidemiol. 19:937-944.
Greenberg, R.A.; Haley, N.J.; Etzel, R.A.; Loda, F.A. (1984) Measuring the exposure of infants to
tobacco smoke. N. Engl. J. Med. 310:1075-1078.
Greenberg, R.A.; Bauman, K.E.; Glover, L.H.; Strechar, V.J.; Kleinbaum, D.G.; Haley, N.J.;
Stedman, H.C.; Fowler, M.G.; Loda, F.A. (1989) Ecology of passive smoking by young
infants. J. Pediatr. 114:774-780.
Greenberg, R.A.; Bauman, K.E.; Strecher, V.J.; Keyes, L.L.; Glover, L.H.; Haley, N.J.; Stedman,
H.C.; Loda, F.A. (1991) Passive smoking during the first year of life. Am. J. Public Health
81:850-853.
Greenland, S. (1987) Quantitative methods in the review of epidemiologic literature. Epidem Rev
9:1-30.
•>
Grimmer, G.; Brune, H.; Dettbarn, G.; Naujack, K.-W.; Mohr, U.; Wenzel-Hartung, R. (1988)
Contribution of polycyclic aromatic compounds to the carcinogenicity of sidestream smoke
of cigarettes evaluated by implantation into the lungs of rats. Cancer Lett. 43:173-177.
Guerin, M. (1987) Formation and physicochemical nature of sidestream smoke. In: O'Neill, I.K.;
Brunnemann, K.D.; Dodet, B.; Hoffmann, D., eds. Environmental carcinogens—selected
methods of analysis, v. 9, passive smoking. IARC Monographs No. 81. Lyon, France:
International Agency for Research on Cancer.
Guerin, M.R.; Jenkins, R.A.; Tomkins, B.A. (1992) The chemistry of environmental tobacco
smoke: Composition and measurement. Lewis Publishers, Chelsea, MI.
R-16
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Gwinn, J.V.; Martinez, F.D.; Wright, A.L.; Ray, C.G.; Taussig, L.M. (1991) Etiology and
symptomatology of lower respiratory illnesses (LRIs) in the first three years of life. Am.
Rev. Respir. Dis. 143:A509.
Haddow, J.E.; Palomaki, G.E.; Knight, GJ. (1986) Use of serum cotinine to assess the accuracy of
self-reported non-smoking. Br. Med. J. 293:1306.
Haddow, J.E.; Knight, G.J.; et al. (1987) Cigarette consumption and serum cotinine in relation to
birthweight. Br. J. Obstet. Gynaecol. 94:678-681.
Haddow, J.E.; Knight, G.J.; Palomaki, G.E.; et al. (1988) Second-trimester serum cotinine levels
in nonsmokers in relation to birth weight. Am. J. Obstet. Gynecol. 159:481-484.
Haenszel, W.; Taeuber, K.E. (1964) Lung-cancer mortality as related to residence and smoking
histories. II. White females. J. Natl. Cancer Inst. 32:803-838.
Haglund, B.; Cnattingius, S. (1990) Cigarette smoking as a risk factor for sudden infant death
syndrome: a population-based study. Am. J. Public Health 80:29-32.
Haley, N.J.; Colosimo, S.G.; Axelrad, C.M.; Harris, R.; Sepkovic, D.W. (1989) Biochemical
validation of self-reported exposure to environmental tobacco smoke. Environ. Res.
49:127-135.
Halonen, M.; Barbee, R.A.; Lebowitz, M.D.; Burrows, B. (1982) An epidemiologic study of the
interrelationships of total serum immunoglobulin E, allergy skin-test reactivity, and
eosinophilia. J. Allergy Clin. Immunol. 69:221-228.
Halonen, M.; Stern, D.; Lyle, S.; Wright, A.; Taussig, L.; Martinez, F.D. (1991) Relationship of
total serum IgE levels in cord and 9-month sera of infants. Clin. Exp. Allergy 21:235-241.
Hammond, E.G. (1966) Smoking in relation to the death rates of one million men and women. In:
Haenszel, W., ed. Epidemiological approaches to the study of cancer and other chronic
diseases. National Cancer Institute Monograph No. 19, Washington, DC: pp. 127-204.
R-17
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Hammond, E.G.; Horn, D. (1958a) Smoking and death rates: report on forty-four months of
follow-up of 187,783 men. I. Total mortality. JAMA 166:1159-1172.
Hammond, E.G.; Horn, D. (1958b) Smoking and death rates: report on forty-four months of
follow-up of 187,783 men. II. Death rates by cause. JAMA 166:1294-1308.
Hammond, S.K.; Leaderer, B.P. (1987) A diffusion monitor to measure exposure to passive
smoking. Environ. Sci. and Technol. 21:494-497.
Hammond, E.G.; Seidman, H. (1980) Smoking and cancer in the United States. Prev. Med. 9:169-
173.
Hammond, E.G.; Selikoff, I.J. (1981) Commentary: passive smoking and lung cancer with
comments on two new papers. Environ. Res. 24:444-452.
Hammond, S.K.; Leaderer, B.P.; Roche, A.C.; Schenker, M. (1987) Collection and analysis of
nicotine as a marker for environmental tobacco smoke. Atmos. Environ. 21:457-462.
Hammond, S.K.; Smith, T.J.; Woskie, S.R.; Leaderer, B.P.; Bettinger, N. (1988) Markers of
exposure to diesel exhaust and cigarette smoke in railroad workers. Am. Ind. Hyg. Assoc.
J. 49:516-522.
Hammond, S.K.; Lewtas, J.; Mumford, J.; Henderson, J.; Henderson, F.W. (1989) Exposures to
environmental tobacco smoke in homes. In: Measurement of toxic and related air
pollutants, Environmental Protection Agency/Air and Waste Management Association
international symposium, Pittsburgh, PA: Air and Waste Management Association,
pp. 590-595.
Hammond, S.K.; Gann, P.H.; Coughlin, J.; Tannenbaum, S.R.; Skipper, P.L. (1990) Tobacco
smoke exposure and carcinogen-hemoglobin adducts. In: Indoor air '90: proceedings of the
5th international conference on indoor air quality and climate, July 29-August 3. Vol. 2:
Characteristics of indoor air, Ottawa, Ontario, Canada: Canada Mortgage and Housing
Corporation, pp. 157-161.
R-18
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Hanbury, W.J. (1964) Bronchogenic carcinoma in women. Thorax 19:338-342, :
Hanrahan, J.P.; Tager, I.B.; Segal, M.R.; et al. (1990) Effect of prenatal smoking on infant lung
function. Am. Rev. Respir. Dis. 141:A282.
Hanson, B.; McGue, M.; Roitman-Johnson, B.; Segal, N.L.; Bouchard, T.J.; Blumenthal, M.N.
(1991) Atopic disease and immunoglobulin E in twins reared apart and together. Am. J.
Hum. Genet. 48:873-879.
Harmsen, H.; Effenberger, E. (1957) Tobacco smoke in public transportation, dwellings and work
rooms. Arch. Hug. 141:383-400.
Hasselblad, V.; Humble, C.G.; Graham, M.G.; Anderson, H. (1981) Indoor environmental
determinants of lung function in children. Am. Rev. Respir. Dis. 123:479-485.
Hawthorne, A.R.; Gammage, D.; Dudney, C.S.; Hingerty, B.E.; Schuresko, D.D.; Parzyek, D.C.;
Womak, D.R.; Morris, S.A.; Westeley, R.R.; White, D.A.; Schrimscher, J.M. (1984) Air
indoor quality study of forty east Tennessee homes. ORNL 5965. Oak Ridge, TN: Oak
Ridge National Laboratory, pp. 134.
Hawthorne, V.M.; Fry, J.S. (1978) Smoking and health: the association between smoking
behaviour, total mortality and cardiorespiratory disease in West Central Scotland. J.
Epidemiol. Community Health 32:260-266.
Higgins, C.E. (1987) Organic vapor phase composition of sidestream and environmental tobacco
smoke from cigarettes: Proceedings of the Environmental Protection Agency/Air Pollution
Control Association symposium on measurement of toxic and related air pollutants,
pp; 140-151.
Hill, A.B. (1953) Observation and experiment. N. Engl. J. Med. 248:995-1001.
Hill, A.B. (1965) The environment and disease: association or causation? proceedings of the Royal
Society of Medicine 58:295-300.
R-19
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Hinds, W.C.; First, M.W. (1975) Concentrations of nicotine and tobacco smoke in public places. N.
Engl. J. Med. 292:844-845.
Hinton, A.E. (1989) Surgery for otitis media with effusion in children and its relationship to
parental smoking. J. Laryngol. Otol. 103:559-561.
Hirayama, T. (1967) Smoking in relation to the death rates of 265,118 men and women in Japan.
Tokyo: National Cancer Center Research Institute.
Hirayama, T. (1975a) Smoking and cancer: a prospective study on cancer epidemiology based on a
census population in Japan. In: Steinfeld, J.; Griffiths, W.; Ball, K.; et al., eds. Proceedings
of the 3rd world conference on smoking and health, Vol. II, Washington, DC: U.S.
Department of Health, Education, and Welfare, pp. 65-72.
Hirayama, T. (1975b) Prospective studies on cancer epidemiology based on a census population in
Japan. In: Bucalossi, P.; Veronesi, U.; Cascinelli, N., eds. Proceedings of the Xlth
international cancer congress, Florence, Italy, 1974, Vol. 3, Cancer epidemiology,
environmental factors. Amsterdam: Excerpta Medica, pp. 26-35.
Hirayama, T. (1977) Epidemiology of lung cancer based on population studies. In: National Cancer
Center Library, ed. Collected papers from the National Cancer Center Research Institute,
Vol. 12. Tokyo: National Cancer Center, pp. 452-461.
Hirayama, T. (1978) Prospective studies on cancer epidemiology based on census population in
Japan. In: Nieburgs, H.E., ed. Prevention and detection of cancer, Vol. 1, Etiology. New
York: Marcel Dekker, pp. 1139-1147.
Hirayama, T. (198la) Non-smoking wives of heavy smokers have a higher risk of lung cancer: a
study from Japan. Br. Med. J. 282:183-185.
Hirayama, T. (1981b) [Letter] Br. Med. J. 283:916-917.
Hirayama, T. (1982) Smoking and cancer in Japan. A prospective study on cancer epidemiology
based on census population in Japan. Results of 13 years follow up. In: Tominaga, S.; Aoki,
R-20
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
K., eds. The UICC smoking control workshop, Nagoya, Japan,, August 24-25, 1981.
Nagoya: University of Nagoya Press, pp. 2-8.
Hirayama, T. (1983a) Passive smoking and lung cancer: consistency of association. Lancet
2:1425-1426.
Hirayama, T. (1983b) Passive smoking and lung cancer. Presented at the 5th world conference on
smoking and health, Winnipeg, Canada.
f
Hirayama, T. (1984) Cancer mortality in nonsmoking women with smoking husbands based on a
large-scale cohort study in Japan. Prev. Med. 13:680-690.
Hirayama, T. (1985) A cohort study on cancer in Japan. In: Blot, W.J.; Hirayama, T.; Hoel, D.G.,
eds. Statistical methods in cancer epidemiology. Hiroshima: Radiation Effects Research
Foundation, pp. 73-91.
Hirayama, T. (1988) Duration of exposure as a determinant of lung cancer risk in passive smokers.
Environ. Tech. L. 9:731-732.
Hirayama, T. (1989) Dietary habits are of limited importance in influencing the lung cancer risk
among Japanese females who never smoked. In: Bieva, D.J.; Courtois, Y.; Govaerts, M.,
eds. Present and future of indoor air quality. New York: Elsevier.
Hirayama' T. (1990) Passive smoking and cancer: the association between husbands' smoking and
cancer in the lung of non-smoking wives. In: Kaskuga, H., ed. ___^_
Berlin: Springer-Verlag.
Hoegg, U.R. (1972) Cigarette smoke in closed spaces. Environ. Health Perspect. 2:117-128.
Hoffman, H.J.; Damus, K.; Hillman, L.; Krongrad, E. (1988) Risk factors for SIDS. Results of the
National Institute of Child Health and Human Development SIDS Cooperative
Epidemiological Study. Ann. N.Y. Acad. Sci. 533:13-30.
R-21
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Hoffmann, D.; Hecht, S.S. (1989) Advances in tobacco carcinogenesis. In: Springer handbook of
experimental pharmacology: chemical carcinogenesis and mutagenesis (in press).
Hoffmann, D.; Wynder, E.L. (1971) A study of tobacco carcinogenesis. XI. Tumor initiators,
tumor accelerators, and tumor promoting activity of condensate fractions. Cancer 27:848-
864.
Hoffmann, D.; Haley, N.J.; Adams, J.D.; Brunnemann, K.D. (1984) Tobacco sidestream smoke.
Uptake by nonsmokers. Prev. Med. 13:608-617.
Hoffmann, D.; Brunnemann, K.D.; Haley, NJ. (1989) Absorption of smoke constituents by
nonsmokers. In: Reducing workplace exposures to environmental tobacco smoke.
EPA/HHS Manual.
Hole, D.J; Gillis, C.R.; Chopra, C.; Hawthorne, V.M. (1989) Passive smoking and
cardiorespiratory health in a general population in the west of Scotland. Br. Med. J.
299:423-427.
Hopp, R.J.; Bewtra, A.; Nair, N.M.; Townley, R.G. (1985) The effect of age on methacholine
response. J. Allergy Clin. Immunol. 76:609-613. :
Hopp, R.J.; Townlwy, R.G.; Biven, R.E.; Bewtra, A.K.; Nair, N.M. (1990) The presence of airway
reactivity before the development of asthma. Am. Rev. Respir. Dis. 141:2-8.
o
Hoppenbrouwers, T.; Ceeb, M.; Arakawa, K.; Hodgman, J.E. (1981) Seasonal relationship of
sudden infant death syndrome and environmental pollutants. Am. J. Epidemic!.
113:623-635.
Hornon, E.R. (ed) Almanac of the 50 States. Information Publications, Palo Alto, California.
Horwood, L.J.; Fergusson, D.M.; Shannon, F.T. (1985) Social and familial factors in the
development of early chilhood asthma. Pediatrics 75:859-868.
R-22
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Humble, C.G.; Samat, J.M.; Pathak, D.R.; Skipper, BJ. (1985) Cigarette smoking and lung cancer
in 'Hispanic' whites and other whites in New Mexico. Am J Public Health 75:145-8.
Humble, C.G.; Samet, J.M.; Pathak, D.R. (1987) Marriage to a smoker and lung cancer risk. Am.
J. Public Health 77:598-602.
Husgafvel-Pursiainen, K.; Sorsa, M.; Mpller, M.; Benestad, C. (1986) Genotoxicity and
polynuclear aromatic hydrocarbon analysis of environmental tobacco smoke samples from
restaurants. Mutagenesis 1:287-291.
IARC. See International Agency for Research on Cancer.
Idle, J.R. (1990) Titrating exposure to tobacco smoke using cotinine—a minefield of
misunderstandings. J. Clin. Epidemiol. 43:313-317.
Inoue, R.; Hirayama, T. (1988) Passive smoking and lung cancer in women. In: Smoking and
health. Elsevier Science Publishers, pp. 283-285.
International Agency for Research on Cancer. (1986) IARC monographs on the evaluation of the
carcinogenic risk of chemicals to man. Vol. 38. Tobacco smoking; Lyon, France: World
Health Organization.
International Agency for Research on Cancer. (1987) Environmental carcinogens—methods of
analysis and exposure measurement. Vol. 9: Passive smoking. O'Neill, I.K.; Brunnemann,
K.D.; Dodet, B.; Hoffmann, D. eds. Lyon, France: IARC Scientific Publications No. 81.
Ishizu, Y. (1980) General equation or the estimation of indoor pollution. Environ. Sci. Technol.
14:1254-1257.
Iversen, M.; Birch, L.; Lundqvist, G.R.; Elbrond, O. (1985) Middle ear effusion and the indoor
environment. Arch. Environ. Health 40:74-79.
Janerich, D.T.; Thompson, W.D.; Varela, L.R.; et al. (1990) Lung cancer and exposure to tobacco
smoke in the household. N. Engl. J. Med. 323:632-636.
R-23
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Jarvis, MJ. (1987) Uptake of environmental tobacco smoke. IARC Sci. Publ. 81:43-58.
Jarvis, MJ. (1989) Application of biochemical intake markers to passive smoking measurement
and risk estimation. Mutat. Res. 222:101-110.
Jarvis, MJ.; Russell, M.H.; Feyerabend, C. (1983) Absorption of nicotine and carbon dioxide
from passive smoking under natural conditions of exposure. Thorax 38:829-833.
Jarvis, M.; Tunstall-Pedoe, H.; Feyerabend, C.; Vesey, C.; Kabat, G.C.; Wynder, E.L. (1984) Lung
cancer in nonsmokers. Cancer 53:1214-1221.
Jarvis, MJ.; Russell, M.A.H.; Feyerabend, C.; Eiser, J.R.; Morgan, M.; Gammage, P.; Gray, E.M.
(1985) Passive exposure to tobacco smoke: saliva cotinine concentrations in a representative
sample of non-smoking children. Br. Med. J. 291:927-929.
Joly, O.G.; Lubin, J.H.; Caraballoso, M. (1983) Dark tobacco and lung cancer in Cuba. J. Natl.
Cancer Inst. 70:1033-1039.
Kabat, G.C. (1990) Epidemiologic studies of the relationship between passive smoking and lung
cancer. Presented at the 1990 annual winter meeting of the Toxicology Forum in
Washington, DC.
Kabat, G.C.; Wynder, E.L. (1984) Lung cancer in nonsmokers. Cancer 53:1214-1221.
Kabat, G.C.; Wynder, E.L. (1984) Lung cancer in nonsmokers. Cancer 53:1214-1221.
Kahn, H.A. (1966) The Dorn study of smoking and mortality among U.S. veterans: report on eight
and one-half years of observation. In: Haenszel, W., ed. Epidemiological approaches to the
study of cancer and other chronic diseases. National Cancer Institute Monograph No. 19,
Washington, DC, pp. 1-125.
Kalandidi, A.; Trichopoulos, D.; Hatzakis, A.; Tzannes, S.; Saracci, R. (1987) Passive smoking and
chronic obstructive lung disease. Lancet ii: 1325-1326.
R-24
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Kalandidi, A; Katsouyanni, K.; Voropoulou, N.; et al. (1990) Passive smoking and diet in the
etiology of lung cancer among non-smokers. Cancer Causes and Control 1:15-21.
Kaplan, H.S.; Tsuchitani, P,J. (eds) (1978) Cancer in China, Alan R. Liss, Inc., New York, 1978.
Kasuga, H.; Hasebe, A.; Osaka, F.; Matuski, H. (1979) Respiratory symptoms in school children
and the role of passive smoking. Tokai J. Exp. Clin. Med. 4:101-104.
Katada, H.; Mikami, R.; Konishi, M.; Koyama, Y.; Narita, N. (1988) Effect of passive smoking in
lung cancer development in women in the Nara region. Gan No Rinshp 34(l):21-27.
Katsouyanni, K.; Kogevinas, M.; Dontas, N.; Maisonneuve, P.; Boyle, P.; Trichopoulos, D. (1990)
Mortality from malignant neoplasms in Greece 1960-1985. Greek Cancer Society, Athens,
pp. 1-125 [in Greek].
Katz, D.; Baptista, J.; Azen, S.P.; Pike, M.C. Obtaining confidence intervals for the risk ratio in
cohort studies. Biometrics 78:469-474.
Kauffmann, F.; Tessier, J.F.; Oriol, P. (1983) Adult passive smoking in the home environment: a
risk factor for chronic airflow limitation. Am. J. Epidemiol. 117:269-280.
Kauffmann, F.; Neukirch, F.; Orobaeff, M.; Marne, M.J.; Claud, J.R.; Lellouch, J. (1986)
Eosinophils, smoking, and lung function. An epidemiologic survey among 912 working
men. Am. Rev. Respir. Dis. 134:1172-1175.
Kauffmann, F.; Dockery, D.W.; Speizer, F.E.; Ferris, B.G. (1989a) Respiratory symptoms and lung
function in relation to passive smoking: a comparative study of American and French
Women. Int. J. Epidemiol. 18:334-344.
Kauffmann, F.; Tager, I.B.; Munoz, A.; Speizer, F.E. (1989b) Familial factors related to lung
function in children aged 6-10 years: results from the PA ARC epidemiologic study. Am. J.
Epidemiol. 129:1289-1299.
R-25
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Kentner, M.; Triebig, G.; Weltle, D. (1984) The influence of passive smoking on pulmonary
function: a study of 1351 office workers. Prev. Med. 13:656-69.
Kleinbaum, D.G.; Kupper, L.L.; Morgenstern, H. Epidemiologic Research, Lifetime Learning
Publications, Belmont, California, 1982.
Knoth, A.; Bohn, H.; Schmidt, F. (1983) Passive smoking as a causal factor of bronchial carcinoma
in female nonsmokers. Med. Klin. 78:66-69.
Koo, L.C. (1988) Dietary habits and lung cancer risk among Chinese females in Hong Kong who
never smoked. Nutr. Cancer 11:155-172.
Koo, L.C. (1989) Environmental tobacco smoke and lung cancer: is it the smoke or the. diet?
Elsevier Science Publishers.
Koo, L.C.; Ho, J.H-C.; Saw, D. (1983) Active and passive smoking among female lung cancer
patients in Hong Kong. J. Exp. Clin. Cancer Res. 4:367-375.
Koo, L.C.; Ho, J.H-C.; Saw, D. (1984) Is passive smoking an added risk factor for lung cancer in
Chinese women? J. Exp. Clin. Cancer Res. 3:277-283.
Koo, L.C.; Ho, J.H-C.; Lee, N. (1985) An analysis of some risk factors for lung cancer in Hong
Kong. Int. J. Cancer 35:149-155.
Koo, L.C., Ho, J.H.; Saw, D.; Ho, C.Y. (1987) Measurements of passive smoking and estimates of
lung cancer risk among non-smoking Chinese females. Int J Cancer 39:162-169.
Koo, L.C.; Ho, J. H-C.; Rylander, R. (1988b) Life-history correlates of environmental tobacco
smoke: a study on nonsmoking Hong Kong Chinese wives with smoking versus
nonsmoking husbands. Soc. Sci. Med. 26(7):751-760.
Koutrakis, P.; Fasano, A.M.; Slater, J.L.; Spengler, J.D.; McCarthy, J.F.; Leaderer, B.P. (1989)
Atmos. Environ. 23:2767-2773.
R-26
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Kraemer, M.J.; Richardson, M.A.; Weiss, N.S.; et al. (1983) Risk factors for persistent middle-ear
effusions: otitis media, cattarah, cigarette smoke exposure, and atopy. JAMA
249:1022-1025.
Krall, E.A.; Valadian, I.; Dwyer, J.T.; Gardner, J. (1989) Accuracy of recalled smoking data. Am
J Publ Health 79:200.
Krzyzanowski, M.; Quackenboss, J.J.; Lebowitz, M.D. (1990) Chronic respiratory effects of indoor
formaldehyde exposure. Environ. Res. 52:117-125.
Kuller, L.H.; Garfinkel, L.; Correa, P.; Haley, N.; Hoffmann, D.; Preston-Martin, S.; Sandier, D.
(1986) Contribution of passive smoking to respiratory cancer. Environ. Health Perspect.
70:57-69.
Kurihara, M.; Aoki, K.; Miller, R.W.; Muir, C.S. (eds) Changing Cancer Patterns and Topics in
Cancer Epidemiology, Plenum Press, New York, 1989.
Lam, W.K. (1985) A clinical and epidemiological study of carcinoma of lung in Hong Kong
[doctoral thesis]. Hong Kong: University of Hong Kong.
Lam, T. H.; Kung, I.T.M.; Wong, C.M.; Lam, W.K.; Kleevens, J.W.L.; Saw, D.; Hsu, C.;
Seneviratne, S.; Lam, S.Y.; Lo, K.K.; Chan, W.C. (1987) Smoking, passive smoking and
histological types in lung cancer in Hong Kong Chinese women. Br. J. Cancer 6:673-678.
Layard, M.W.; Viren, J.R. (1989) Assessing the validity of a Japanese cohort study. Elsevier
Science Publishers B.V. (Biomedical Division).
Leaderer, B.P. (1988) Measuring exposure to environmental tobacco smoke. Report prepared for
the U. S. Environmental Protection Agency.
Leaderer, B.P. (1989) Unpublished data.
Leaderer, B. (1990) Assessing exposure to environmental tobacco smoke. Risk Anal. 10(l):19-26.
R-27
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Leaderer, B.P.; Hammond, S.K. (1991) Evaluation of vapor-phase nicotine and respirable
suspended particle mass as markers for environmental tobacco smoke. Environ. Sci.
Technol. 25(4):770-777.
Leaderer, B.P.; Cain, W.S.; Isseroff, R.; Berglund, L.G. (1984) Ventilation requirements in
buildings II. Particulate matter and carbon monoxide from cigarette smoking. Atmos.
Environ. 18:99-106.
Leaderer, B.P.; Koutrakis, P.; Briggs, S.; Rizzuto, J. (1990) Impact of indoor sources on residential
and concentrations. In: Indoor air '90: proceedings of the 5th international conference on
indoor air quality and climate, July 29-August 3. Vol. 2: Characteristics of indoor air.
Ottawa, Ontario, Canada: Canada Mortgage and Housing Corporation, pp. 269-274.
Lebowitz, M.D. (1984) The effects of environmental tobacco smoke exposure and gas stoves on
daily peak flow rates in asthmatic and non-asthmatic families. Eur. J. Respir. Dig.
133:190-195.
Lebowitz, M.D.; Burrows, B. (1976) Respiratory symptoms related to smoking habits of family
adults. Chest 69:49-50. -
Lebowitr, M.D.; Holberg, CJ. (1988) Effects of parental smoking and other risk factors on the
development of pulmonary function in children and adolescents. Analysis of two
longitudinal population studies. Am. J. Epidemiol. 128:589-597.
Lebowitz, M.D.; Holberg, C.J.; Knudson, R.J.; Burrows, B. (.1987) Longitudinal study of
pulmonary function development in childhood, adolescence, and early adulthood. Am.
Rev. Respir. Dis. 136:69-75.
Lebowitz, M.D.; Quackenboss, J.J. (1990) The effect of environmental tobacco smoke on
pulmonary function. Int. Arch. Occup. Environ. Health (Suppl): 147-152.
Lebret, E.; Boley, J.; Brumekreef, B. (1990) Environmental tobacco smoke in Dutch homes. In:
Indoor air '90: proceedings of the 5th international conference on indoor air quality and
R-28
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
climate, July 29-August 3. Vol. 2: Characteristics of indoor air, Ottawa, Ontario, Canada:
Canada Mortgage and Housing Corporation, pp. 263-268.
Lee, P.N. (1986) Does breathing other people's tobacco smoke cause lung cancer? [Letter] Br. Med.
J. 293:1503-1504.
Lee, P.N. (1987a) Lung cancer and passive smoking: association an artifact due to misclassification
of smoking habits. Toxicol. Lett. 35:157-162.
Lee, P.N. (1987b) Passive smoking and lung cancer association: a result of bias? Hum. Toxicol.
6:517-524.
Lee, P.N. (1988) Misclassification of smoking habits and passive smoking. Berlin: Springer.
Lee, P.N. (1989) Passive smoking and lung cancer: fact or fiction. In: Bieva, C.J.; Courtois, Y.;
Govaerts, M., eds. Present and Future of Indoor Air Quality, Excerpta Medica,
Amsterdam, pp. 119-128.
Lee, P.N. (1990) A detailed review of epidemiological evidence relating environmental tobacco
smoke (ETS) to the risk of cancer, heart disease and other causes of death in adults who
have never smoked. Draft 3 of Annex A.
Lee, P.N. (1991) Correcting meta-analyses of the association of lung cancer in females with spouse
(or household) exposure for bias due to misclassification of active smoking status.
Submitted to U>S. EPA, dated November 29, 1991.
Lee, P.N.; Chamberlain, J.; Alderson, M.R. (1986) Relationship of passive smoking to risk of lung
cancer and other smoking-associated diseases. Br. J. Cancer 54:97-105.
Lee, B.L.; Bendwitz, N.L.; Jacob P. (1987) Influence of tobacco abstinence of the disposition
kinetics and effects of nicotine. Clin. Pharmacol. Ther. 41:474-479.
R-29
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Leader, S.R.; Corkhill, R.T.; Irwig, L.M.; Holland, W.W.; Colley, J.R. (1976) Influence of family
factors on the incidence of lower respiratory illness during the first year of life. Br. J.
Prev. Soc. Med. 30:203-212.
Lehnert, G. (1984) Roundtable discussion from symposium on medical perspectives on passive
smoking. Prev. Med. 13:732-733.
Levin (1953) (p. 6-13).
Lewak, N.; van den Berg, B.J.; Beckwith, J.B. (1979) Sudden infant death syndrome risk factors.
Clin. Pediatr. (Phila) 18:404-411.
Lewis, H.M.; Haeney, M.; Jeacock, J.; Thomas, H. (1989) Chronic cough in a hospital population:
its relationship to atopy and defect in host defence. Arch. Dis. Child. 64:1593-1598.
Lewtas, J.; Claxton, L.; Mumford, J.L. (1987) Human exposure to mutagens from indoor
combustion sources. In: Seifert, B.; Esdorn, M.; Fischer, M.; Ruden, H.; Wegner, J., eds.
Indoor air '87, Vol. 1, Oraniendruck GmbH, Berlin, pp. 473-477.
Ling, P.I.; Lofroth, G.; Lewtas, J. (1987) Mutagenic determination of passive smoking. Toxicol.
Lett. 35:147-151.
Liu, Z.; He, X.; Chapman, R.S. (1991) Smoking and other risk factors for lung cancer in Xuanwei,
China. Int. J. Epidemiol. 20:26-31.
Lofroth, G.; Nilsson, L.; Alfeim, I. (1983) Passive smoking and urban air pollution:
Salmonella/microsome mutagenicity assay of simultaneously collected indoor and outdoor
particulate matter. In: Waters, M.D.; Sandhu, S.S.; Lewtas, J.; Claxton, L.; Chernoff, N.;
Nesnow, S., eds. Short-term bioassays in the analysis of complex environmental mixtures,
Vol. m. New York: Plenum, pp. 515-525.
Lofroth, G.; Ling, P.I.; Agurell, E. (1988) Public exposure to environmental tobacco smoke.
Mutat Res. 202:103-110.
R-30
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Lofroth, G.; Burtin, R.; Forehand, L.; Hammond, K.; Selia, R.; Zwiedinger, E.; Lewtas, J, (1989)
Characterization of genotoxic components of environmental tobacco smoke. Environ. Sci.
Technol. 23:610-614.
Lossing, E.H.; Best, E.W.R.; McGregor, J.T.; Josie, G.H.; Walker, C.B.; Delaquis, P.M.; Baker,
P.M.; McKenzie, A.C. (1966) A Canadian study of smoking and health. Ottawa, Canada:
Department of National Health and Welfare.
Lubin, J.H.; Blot, W.J.; Berrino, F.; Flamant, R.; Gillis, C.R.; Kunze, M.; Schmahl, D.; Visco, G.
(1984) Modifying risk of developing lung cancer by changing habits of cigarette smoking.
Br. Med. J. 288:1953-1956.
MacDonald, EJ. (1981) [Letter] Br. Med. J. 283:915-916.
Machlin, S.R.; Kleinman, J.C.; Madans, J.H. (1989) Validity of mortality analysis based on
retrospective smoking information. Stat. Med. 8:997-1009.
Maclure, M.; Katz, R.B.-A.; Bryant, M.S.; Skipper, P.L.; Tannenbaum, S.R. (1989) Elevated blood
levels of carcinogens in passive smokers. Am. J. Public Health 79:1381-1384.
Magnusson, C.G.M. (1986) Maternal smoking influences cord serum IgE and IgD levels and
increases the risk for subsequent infant allergy. J. Allergy Clin. Immunol. 78:898-904.
Makin, J.; Fried, P.A.; Watkinson, B. (1991) A comparison of active and passive.smoking during
pregnancy: long-term effects. Neurotoxicol. Teratol. 13:5-12.
Malloy, M.H.; Kleinman, J.C.; Land, G.H.; Schramm, W.F. (1988) The association of maternal
smoking with age and cause of infant death. Am. J. Epidemiol. 128:46-55.
Mantel, N. (1963) Chi-square tests with one degree of freedom: extensions of the Mantel-Haenszel
procedure. J. Am. Stat. Assoc. 58:690-700.
Mantel, N.; Haenszel, W. (1959) Statistical aspects of the analysis of data from retrospective
studies of disease. J. Natl. Cancer Inst. 22:719-748.
R-31
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Maran, A.G.D.; Wilson, J.A. (1986) Glue ear and speech development. Br. Med. J. 293:713-714.
Marbury, M.C.; Hammond, S.K.; Haley, N.J. (1990) Assessing exposure to environmental tobacco
smoke in epidemiological studies of acute health effects. In: Indoor air '90: proceedings of
the 5th international conference on indoor air quality and climate, July 29-August 3. Vol.
2. Characteristics of indoor air. Ottawa, Ontario, Canada: Canada Mortgage and Housing
Corporation, pp. 189-194.
Marchand, L.L.; Wilkens, L.R.; Hankin, J.H.; Haley, N.J. (1991) Dietary patterns of female
nonsmokers with and without exposure to environmental tobacco smoke. Cancer Causes
and Control 2:11 -16.
Martin, T.R.; Bracken, M.B. (1986) Association of low birth weight with passive smoke exposure
in pregnancy. Am. J. Epidemiol. 124:633-642.
Martinez, F.D.; Morgan, W.J.; Wright, A.L.; Holberg, C.J.; Taussig, L.M. (1988a) Diminished lung
function as a predisposing factor for wheezing lower respiratory tract illness in infants. N.
Engl. J. Med. 319:1112-1117.
Martinez, F.D.; Antognoni, G.; Macri, F.; Bonci, E.; Midulla, F.; De Castro, G.; Ronchetti, R.
(1988b) Parental smoking enhances bronchial responsiveness in nine-year-old children.
Am. Rev. Respir. Dis. 138:518-523.
Martinez, F.D.; Morgan, W.J.; Wright, A.L.; Holberg, C.; Taussig L.M. (1991a) Initial airway
function is a risk factor for recurrent wheezing respiratory illnesses during the first three
years of life. Am. Rev. Respir. Dis. 143:312-316. ;
Martinez, F.D.; Cline, M.; Burrows, B. (1991b) Increased incidence of asthma in children of
smoking mothers. Pediatrics: in press.
Masi, M.A.; Hanley, J.A.; Ernst, P.; Becklake, M.R. (1988) Environmental exposure to tobacco
smoke and lung function in young adults. Am. Rev. Respir. Dis. 138:296-299.
R-32
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Mattson, ME.; Boyd, G.; Byar, D.; Brown, C.; Callahan, J.F.; Cullen, J.W.; Grenblatt, J.; Haley;
Hammond, N.; J.F.; Lewtas, J.; Reeves, W. (1989) Passive smoking on commercial airline
flights. JAMA 261:867-872.
McConnochie, K.M.; Roghmann, KJ. (1986a) Parental smoking, presence of older sibling, and
family history of asthma increase risk of bronchiolitis. Am. J. Dis. Child. 140:806-812.
McConnochie, K.M.; Roghmann, KJ. (1986b) Breast feeding and maternal smoking as predictors
of wheezing in children age 6 to 10 years. Pediatr. Pulmonol. 2:260-268.
Mertsola, J.; Ruuskanen, O.; Vanto, T.; Koivikko, A.; Halonen, P. (1991) Recurrent wheezy
bronchitis and viral respiratory infections. Arch. Dis. Child. 66:124-129.
Meyers, D.A.; Beaty, T.H.; Freidhoff, L.R.; Marsh, D.G. (1987) Inheritance of total serum IgE
(basal level) in man. Am. J. Hum. Genet. 41:51-62.
Miesner, E.A.; Rudnick, S.N.; Hu, F.; Spengler, J.D.; Preller, L.; Ozkaynak, H.; Nelson, W.,(1989)
Particule and nicotine sampling in public facilities and offices. J. Air Pollut. Control
Assoc. 39:1577-1582.
Miller, G.H. (1984) Cancer, passive smoking and nonemployed and employed wives. West. J. Med.
140(4):632-635.
Mitchell, E.A.; Scragg, R.; Stewart, A.W.; et al. (1991) Results from the first year of the New
Zealand cot death study. N.Z. Med. J. 104:71-6.
Moolgavkar, S.H.; Dewanji, A.; Luebeck, G. (1989) Cigarette smoking and lung cancer: reanalysis
of the British doctors'data. J. Am. Cancer Inst. 81 (6):415-420.
Moschandreas, D.J. (1981) Exposure to pollutants and daily time budgets of people. Bull. N.Y.
Acad. Med. 57:845-859.
Mumford, J.L.; Lewtas, J.; Burton, R.M.; Henderson, F.W.; Forehand, L.; Allison, J.C.;
Hammond, S.K. (1989) Assessing environmental tobacco smoke exposure of preschool
R-33
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
children in homes by monitoring air particles, mutagenicity, and nicotine. In: Measurement
of toxic and related air pollutants, Environmental Protection Agency/Air and Waste
Management Association international symposium. Pittsburgh, PA: Air and Waste
Management Association, pp. 606-610.
Muramatsu, M.; Umemura, S.; Okada, T.; Tomita, H. (1984) Estimation of personal exposure to
tobacco smoke with a newly developed nicotine personal monitor. Environ. Res.
35:218-227.
Murray, A.B.; Morrison, BJ. (1986) The effect of cigarette smoke from the mother on bronchial
responsiveness and severity of symptoms in children with asthma. J. Allergy Clin.
Immunol. 77:575-581.
Murray, A.B.; Morrison, B.J. (1988) Passive smoking and the seasonal difference of severity of
asthma in children. Chest 94:701-708.
Murray, A.B.; Morrison, BJ. (1989) Passive smoking by asthmatics: its greater effect on boys than
on girls and on older than on younger children. Pediatrics 84:451-459.
Naeye, R.L.; Ladis, B.; Drage, J.S. (1976) Sudden infant death syndrome. A prospective study.
Am. J. Dis. Child. 130:1207-1210.
Nagda, N.; Fortmann, R.; Koontz, M.; Konheim, A. (1990) Investigation of cabin air-quality
aboard commercial airlines. In: indoor air '90: proceedings of the 5th international
conference on indoor air quality and climate, July 29-August 3.. Vol. 2: Characteristics of
indoor air. Ottawa, Ontario, Canada: Canada Mortgage and Housing Corporation, pp. 245-
250.
National Institute for Occupational Safety and Health. (1991) Environmental tobacco smoke in the
workplace: lung cancer and other health effects. Current Intelligence Bulletin 54. U.S.
Department of Health and Human Services, National Institute for Occupational Safety and
Health.
R-34
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
National Research Council. (1986) Environmental tobacco smoke: measuring exposures and
assessing health effects. Washington, DC: National Academy Press.
National Research Council. (1987) Indoor pollutants. Washington, DC: National Academy Press,
p. 537.
Neal, A.D.; Wadden, R.A.; Rosenberg, S. (1978) Evaluation of indoor particulate concentrations
for an urban hospital. Am. Ind. Hyg. Assoc. J. 29:578-582.
Neddenriep, D.; Martinez, F.D.; Morgan, WJ. (1990) Increased specific lung compliance in
newborns whose mothers smoked during pregnancy. Am. Rev. Respir. Dis. 141:A282.
Nelson, P.A.; Qvant, F.R.; Sem, G.F. (1982) Experimental measurements of aerosol concentrations
in offices. Environ. Int. 8:223-227.
Neuspiel, D.R.; Rush, D.; Butler, R; Golding, J.; Bijur, P.E.; Kurzon, M. (1989) Parental smoking
and post-infancy wheezing in children: a prospective cohort study. Am. J. Public Health
79:168-171.
NIOSH. See National Institute for Occupational Safety and Health.
Nitschke, I.A.; Clarke, W.A.; Clarkin, M.E.; Traynmor, G.W.; Wadach, J.B. (1985) Indoor air
quality infiltration and ventilation in residential buildings. NYSERDA #85-10. Albany,
NY: New York State Energy Research and Development Authority.
NRC. See National Research Council.
O'Connell, E.J.; Logan, G.B. (1974) Parental smoking in childhood asthma. Annals of Allergy
32:142-145.
O'Connor, G.T.; Weiss, S.T.; Tager, I.E.; Speizer, F.E. (1987) The effect of passive smoking on
pulmonary function and non-specific bronchial responsiveness in a population based
sample of children and young adults. Am. Rev. Respir. Dis. 135:800-804.
R-35
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
O'Connor, G.T.; Sparrow, D.; Weiss, S.T. (1989) The role of allergy and nonspecific airway
hyperresponsiveness in the pathogenesis of chronic obstructive pulmonary disease. Am.
Rev. Respir. Dis. 140:225-252.
Ogden, M.W.; Nystrom, C.W.; Oldaker, G.B., III; Conrad, F.W., Jr. (1989) Evaluation of a personal
passive sampling device for determining exposure to nicotine in environmental tobacco
smoke. In: Measurement of toxic and related air pollutants. Pittsburgh, PA: Air Pollution
Control Association, pp. 552-558.
Ogston, S.A.; Florey, C. du V.; Walker, C.M. (1987) Association of infant alimentary and
respiratory illness with parental smoking and other environmental factors. J. Epidemiol.
Community Health 41:21-25.
Oldaker, G.B., III; Conrad, F.W., Jr. (1987) Estimation of the effect of environmental tobacco
smoke (ETS) on air quality within aircraft cabins of commercial aircraft. Environ. Sci.
Technol. 21:994-999.
Oldaker, G.B.; Ogden, M.W.; Maiolo, K.C.; Conner, J.M.; Conrad, F.W.; DeLuca, P.O. (1990)
Results from surveys of environmental tobacco smoke in restaurants in Winston-Salem,
North Carolina. In: Indoor air '90: proceedings of the 5th international conference on
indoor air quality and climate, July 29-August 3. Vol. 2: Characteristics of indoor air,
Ottawa, Ontario, Canada: Canada Mortgage and Housing Corporation, pp. 281-285.
Oldigs, M.; Jrres, R.; Magnussen, H. (1991) Acute effects of passive smoking on lung function and
airway responsiveness in asthmatic children. Pediatr. Pulmonol. 10:123-131.
Ong, T.M.; Stewart, J.; Whong, W.Z. (1984) A simple in situ mutagenicity test for detection of
mutagenic air pollutants. Mutat. Res. 139:177-181,
Park, J.K.; Kim, I.S. (1986) Effect of family smoking on acute respiratory disease in children.
Yonsei Med. J. 27:261 -270.
Parker, G.B.; Wilfert, G.L.; Dennis, G.W. (1984) Indoor air quality and infiltration in multifamily
naval housing. Annual PNWIS/APCA meeting, Portland, OR, Nov. 12-14, pp. 1-14.
R-36 05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Pattishall, E.N.; Strope, G.L.; Etzel, R.A.; Helms, R.W.; Haley, N.J.; Denny, F.W. (1985) Serum
cotinine as a measure of tobacco smoke exposure in children. Am. J. Dis. Child.
139:1101-1104.
Pedreira, F.A.; Guandolo, V.L.; Feroli, E.J.; Mella, G.W.; Weiss, I.P. (1985) Involuntary smoking
and incidence of respiratory illness during the first year of life. Pediatrics 75:594-597.
Perera, F.R.; Santella, R.M.; Brenner, D.; Poirier, M.C.; Munshi, A.A.; Fischman, H.K.; Van
Ryzin, J. (1987) DNA adducts, protein adducts, and sister chromatid exchange in cigarette
smokers and nbrismokers. J. Natl. Cancer Inst. 79:449-456.
Pershagen, G.; Hrubec, Z.; Svensson, C. (1987) Passive smoking and lung cancer in Swedish
women. Am. J. Epidemiol. 125(l):17-24.
Peto, R.; Doll, R. (1984) Keynote address: the control of lung cancer. In: Mizell, M.; Correa, P.,
eds. Lung cancer: causes and prevention. New York: Verlag Chemie International,
pp. 1-19.
Phillips, R.L.; Garfinkel, L.; Kuzma, J.W.; et al. (1980a) Mortality among California Seventh-Day
Adventists for selected cancer sites. J. Natl. Cancer Inst. 65:1097-1107.
Phillips, R.L.; Kuzma, J.W.; Beeson, W.L.; et al. (1980b) Influence of selection versus lifestyle on
risk of fatal cancer and cardiovascular disease among Seventh-Day Adventists. Am. J.
Epidemiol. 112:296-314.
Pierce, J.P.; Dwyer, T.; et al. (1987) Cotinine validation of self-reported smoking in commercially
run community surveys. J. Chronic. Dis. 40:689-695.
Pierce, J.P.; Fiore, M.C.; Novotny, T.E.; Hatziandreu, E.J.; Davis, R.M. (1989) Trends in cigarette
smoking in the United States. Projections to the year 2000. JAMA 261:61-65.
Pimm, P.E.; Silverman, F.; Shepard, RJ. (1978) Physiological effects of acute passive exposure to
cigarette smoke. Arch. Environ. Health 33:201-213.
R-37
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Pojer(1984).
Pukander, J.; Luotonen, J.; Timonen, J.; Karma, P. (1985) Risk factors affecting the occurrence of
acute otitis media among 2-3-year old urban children. Acta Otolaryngol. 100:260-265.
Pullan, C.R.; Hey, E.N. (1982) Wheezing, asthma, and pulmonary dysfunction 10 years after
infection with respiratory sycytial virus in infancy. Br. Med. J. 284:1665-1669.
Putnam, D.L.; David, R.M.; Melhorn, J.M.; Dansie, D.R.; Stone, C.J.; Henry, C.J. (1985) Dose-
responsive increase in sister-chromatid exchanges in bone-marrow cells of mice exposed
nose-only to whole cigarette smoke. Mutat. Res. 156:181-186.
Quant, F.R.; Nelson, P.A.; Sem, GJ. (1982) Experimental measurements Of aerosol concentrations
in offices. Environ. Int. 8(l-6):249-258.
Rantakallio, P. (1978) Relationship of maternal smoking to morbidity and mortality of the child
up to the age of five. Acta Paediatr. Scand. 67:621-631.
Reed, B.D.; Lutz, LJ. (1988) Household smoking exposure-association with middle ear effusions.
Fam. Med. 20:426-430.
Remmer, H. (1987) Passively inhaled tobacco smoke: a challenge to toxicology and preventive
medicine. Arch. Toxicol. 61:89-104.
Repace, J.L. (1987) Indoor concentrations of environmental tobacco smoke: models dealing with
effects of ventilation and room size. In: O'Neill, I.K.; Brunnemann, K.D.; Dodet, B.;
Hoffmann, D., eds. Environmental carcinogens-selected methods of analysis, v. 9, passive
smoking. Chapter 3, IARC Monographs No. 81. Lyon, France: International Agency for
Research on Cancer.
Repace, J.L. (1989) Workplace restrictions on passive smoking: justification on the basis of cancer
risk (in press).
R-38
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Repace, J.L.; Lowrey, A.H. (1980) Indoor air pollution, tobacco smoke and public health. Science
208:464-472.
Repace, J.L.; Lowrey, A.H. (1982) Tobacco smoke, ventilation and indoor air quality. Am. Soc.
Heat Refrig. Air Cond. Eng. Trans. 88:894-914.
Repace, J.L.; Lowrey, A.H. (1985) A quantitative estimate of nonsmokers' lung cancer risk from
passive smoking. Environ. Intl. 11:3-22.
Repace, J.L.; Lowrey, A.H. (1990) Risk assessment methodologies in passive smoking. Risk Anal.
10(l):27-37.
Riboli, E.; Preston-Martin, S.; Saracci, R.; Haley, N.J.; et al. (1990) Exposure of nonsmoking
women to environmental tobacco smoke: a 10-country collaborative study. Cancer Causes
and Control 1:243-252.
Rickert, W.S.; Robinson, J.C.; Collinshaw, N.E. (1984) Yields of tar, nicotine and carbon
monoxide in the sidestream smoke from 15 brands of Canadian cigarettes. Am. J. Public
Health 74:228-231.
Rijcken, B.; Schouten, J.P.; Weiss, S.T.; Speizer, F.E.; Van der Lende, R. (1987) The relationship
of nonspecific bronchial responsiveness to respiratory symptoms in a random population
sample. Am. Rev. Respir. Dis. 136:62-68.
Rimington, J. (1981) The effect of filters on the incidence of lung cancer in cigarette smokers.
Environ. Res. 24:162-166.
Robins, J.M.; Blebins, D.; Schneiderman, M (1989) The effective number of cigarettes inhaled
daily by passive smokers: are epidemiologic and dosimetric estimates consistent? J. Haz.
Mat. 21:215-238.
Rogot, E.; Murray, J.L. (1980) Smoking and causes of death among U.S. veterans: 16 years of
observation. Public Health Rep. 95:213-222.
R-39
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Ronchetti, R.; Macri, F.; Ciofetta, G.; et al. (1990) Increased serum immunoglobulin E and
increased prevalence of eosinophilia in 9-year-old children of smoking parents. J. Allergy
Clin. Immunol. 86:400-407.
Rothman, K.J. (1986) Modern epidemiology. Boston: Little, Brown and Co.
Russell, M.A.H.; Feyerabend, C. (1975) Blood and urinary nicotine in nonsmokers. Lancet
1:179-181.
Russell, M.A.H.; Jarvis, M.J.; West, RJ. (1986) Use of urinary nicotine concentrations to estimate
exposure and mortality from passive smoking in non-smokers. Br. J. Addict. 81:275-281.
Rylander, E.; Eriksson, M.; Freyschuss, U. (1988) Risk factors for occasional and recurrent
wheezing after RSV infection in infancy. Acta Pediatr. Scand. 77:711 -715.
Said, G.; Zalokar, J.; Lellouch, J.; Patois, E. (1978) Parental smoking related to adenoidectomy and
tonsillectomy in children. J. Epidemiol. Community Health 32:97-101.
Samet, J.M.; Tager, I.E.; Speizer, F.E. (1983) The relationship between respiratory illness in
childhood and chronic air-flow obstruction in adulthood. Am. Rev. Respir. Dis.
127:508-523.
Sandier, D.P.; Everson, R.B.; Wilcox, A.J. (1985) Passive smoking in adulthood and cancer risk.
Am. J. Epidemiol. 121(l):37-48.
Saracci, R. (1989) Passive smoking and cancer risk. IARC report of panel of experts. Prepared at
the request of the European School of Oncology through the Europe Against Cancer
program of the European Economic Community.
Saracci, R.; Riboli, E. (1989) Passive smoking and lung cancer: current evidence and ongoing
studies at the International Agency for Research on Cancer. Mutat. Res. 222:117-127.
Schelesselman, JJ. (1978) Assessing effects of confounding variables. Am. J. Epidemiol. 108:3-8.
R-40
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Schelesselman, J.J. (1982) Case-control studies: design, conduct, analysis. New York: Oxford
University Press.
Schenker, M.B.; Samet, J.M.; Speizer, E.F. (1983) Risk factors for childhood respiratory disease.
The effect of host factors and home environment exposures. Am. Rev. Respir. Dis.
128:1038-1043.
Schenker, M.B.; Samuels, S.J.; Kado, N.Y.; Hammond, S.K.; Smith, T.J.; Woskie, S.R. (1990)
Markers of exposure to diesel exhaust. Research Publication Number 33. Cambridge, MA:
Health Effects Institute.
Schilling, R.S.F.; Letai, A.D.; Hui, S.L.; Beck, G.J.; Schoenberg, J.B.; Bouhuys, A. (1977) Lung
function, respiratory disease, and smoking in families. Am. J. Epidemiol. 106:274-283.
Schwartz, J.; Zeger, S. (1990) Passive smoking, air pollution, and acute respiratory symptoms in a
diary of student nurses. Am. Rev. Respir. Dis. 141:62-67.
Seely, J.E.; Zuskin, E.; Bouhuys, A. (1971) Cigarette smoking: objective evidence for lung damage
in teen-agers. Science 172:741-743.
Seigel, D.G.; Greenhouse, S.W. (1973) Validity in estimating relative risk in case-control studies. J.
Chronic Dis. 26:219-226.
Sexton, K.; Spengler, J.D.; Treitman, R.D. (1984) Personal exposure to respirable particulates: a
case-study in Waterbury, Vermont. Atmos. Environ. 18:1385-1398.
Shephard, R.J. (1992) Respiratory irritation from ETS. Arch. Env. Health 47:(2): 123-130.
Shephard, R.J.; Collis, R.; Silverman, F. (1979) "Passive" exposure of asthmatic subjects to
cigarette smoke. Environ. Res. 20:392-402.
Sherman, C.B.; Tosteson, T.D.; Tager, I.E.; Speizer, F.E.; Weiss, S.T. (1990) Early childhood
predictors of asthma. Am. J. Epidemiol. 132:83-95.
R-41
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Sherrill, D.; Holberg, C.J.; Lebowitz, M.D. (1990) Differential rates of lung growth as measured
longitudinally by pulmonary function in children and adolescents. Pediatr. Pulmonol.
8:145-154.
Shimizu, H.; Morishita, M.; Mizuno, K.; Masuda, T.; Ogura, Y.; Santo, M.; Nishimura, M.;
Kunishima, K.; Karasawa, K.; Nishiwaki, K.; Yamamoto, M.; Hisamichi, S.; Tominaga, S.
(1988) A case-control study of lung cancer in nonsmoking women. Tohoku J. Exp. Med.
154:389-397.
Shultz, T.M., Novotny, T.E.; Rice, D.P. (1991) Quantifying the disease impact of cigarette
smoking with SAMMEC II Software. Public Health Reports 106(3)326-333.
Sidney, S.; Caan, B.; Friedman, G. (1989) Dietary intake of carotene in nonsmokers with and
without passive smoke at home. Am J Epidimiol 129:1305-9.
Sims, D.G.; Downham, M.A.P.S.; Gardner, P.S.; Webb, J.K.G.; Weightman, D. (1978) Study of
8-year-old children with a history of respiratory syncytial virus bronchiolitis in infancy.
Br. Med. J. 1:11-14.
Sobue, T.; Suzuki, R.; Nakayama, N.; Inubuse, C.; Matsuda, M.; Doi, O.; Mori, T.; Furuse, K.;
Fukuoka, M.; Yasumitsu, T.; Kuwabara, O.; Ichigaya, M.; Kurata, M.; Nakahara, K.;
Endo, S.; Hattori, S. (1990) Passive smoking among nonsmoking women and the
relationship between indoor air pollution and lung cancer incidence—results of a
multicenter case controlled study. Gan to Rinsho 36(3):329-333.
Somerville, S.M.; Rona, R.J.; Chinn, S. (1988) Passive smoking and respiratory conditions in
primary school children. J. Epidemiol. Community Health 42:105-110.
Sparrow, D.; O'Connor, G.; Colton, T.; Barry, C.L.; Weiss, S.T. (1987) The relationship of
nonspecific bronchial responsiveness to the occurrence of respiratory symptoms and
decreased levels of pulmonary function. The normative aging study. Am. Rev. Respir. Dis.
135:1255-1260.
R-42
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Speizer, F.E.; Ferris, B., Jr.; Bishop, Y.M.; Spengler, J. (1980) Respiratory disease rates and
pulmonary function in children associated with NO exposure. Am. Rev. Respir. Dis.
121:3-10.
Spengler, J;D.; Dockery, D.W.; Turner, W.A.; Wolf son, J.M.; Ferris, E.G. (1981) Long-term
measurements of respirable sulphates and particles inside and outside homes. Atmos.
Environ. 15:23-30.
Spengler, J.D.; Treitman, R.D.; Tosteson, T.D.; Mage, D.T.; Soczek, M.L. (1985) Personal
exposures to respirable particulates and implications for air pollution epidemiology.
Environ. Sci. Technol. 19:700-707.
Spitzer, W.O.; Lawrence, V.; Dales, R.; Hill, G.; Archer, M.C.; Clark, P.; Abenhaim, L.; Hardy, J.;
Sampalis, J.; Pinfold, S.P.; Morgan, P.P. (1990) Links between passive smoking and disease:
a best evidence synthesis. A report of the working group on passive smoking. Clin. Invest.
Med. 13:17-42.
Stahlman, M.T.; Gray, M.E. (1984) Ontogeny of neuroendocrine cells in human fetal lung. I. An
electron microscopic study. Lab. Invest. 51:449-463.
Stanton, M.F.; Miller, E.; Wrench, C.; Blackwell, R. (1972) Experimental induction of epidermoid
carcinoma in the lungs of rats by cigarette smoke condensate. J. Natl. Cancer Inst.
49:867-877.
Steele, R.; Langworth, J.T. (1966) The relationship of antenatal and postnatal factors to sudden
unexpected death in infancy. Can. Med. Assoc. J. 94:1165-1171.
Steenland, K. (1992) Passive smoking and the risk of heart disease. JAMA 267:94-99.
Stellman, S. D.; Garfinkel, L. (1986) Smoking habits and tar levels in a new American Cancer
Society prospective study of 1.2 million men and women. J. Natl. Cancer Ins.t.
76(6):1057-1063.
R-43
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Sterling, T.D.; Sterling, E.M. (1983) Investigations on the effect of regulating smoking on levels of
indoor pollution and on the preception of health on comfort of office workers. Eur. J.
Respir. Dis. 65 (Suppl. 133):17-32. •
Stockwell (1990) (p. B-7).
Stockwell, H.G.; Candelora, E.G.; Armstrong, A.W.; Pinkham, P.A. (1991) Environmental tobacco
smoke and lung cancer in never smoking women. Presented at the annual meeting abstract
forum: Society for Epidemiologic Research, June 12-14, 1991, Buffalo, Ne\V York.
Strachan, D.P. (1988) Damp housing and childhood asthma: validation of reporting of symptoms.
Br. Med. J. 297:1223-6.
Strachan, D.P.; Jarvis, M.J.; Feyerabend, C. (1990) The relationship of salivary cotinine to
respiratory symptoms, spirometry, and exercise-induced bronchospasm in seven-year-old
children. Am. Rev. Respir. Dis. 142:147-151.
Strachan, D.P.; Jarvis, M.J.; Feyerabend, C. (1989) Passive smoking, salivary cotinine
concentrations, and middle ear effusion in 7 year old children. Br. Med. J. 298:1549-1552.
Sutton (1980) (p. B-14).
Svendsen, K.H.; Kuller, L.H.; Martin, M.J.; Ockene, J.K. (1987) Effects of passive smoking in the
multiple risk intervention trial. Am. J. Epidemiol. 126:783-795.
Svensson, D. (1988) Lung cancer etiology in women. [Dissertation] Karolinska Institute,
Stockholm, Sweden.
Svensson, C.; Pershagen, G.; Klominek, J. (1988) Smoking and passive smoking in relation to lung
cancer in women. Department of Epidemiology, National Institute of Environmental
Medicine, Stockholm, Sweden.
Svensson, C.; Pershagen, G.; Klominek, J. (1989) Smoking and passive smoking in relation to lung
cancer in women. Acta Oncol. 28:623-629.
R-44
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Tager, I.E.; Segal, M.R.; Munoz, A.; Weiss, S.T.; Speizer, F.E. (1987) The effect of maternal
smoking on the pulmonary function of children and adolescents. Analysis of data from two
populations. Am. Rev. Respir. Dis. 136:1366-1370.
Tager, I.E.; Weiss, S.T.; Munoz, A.; Rosner, B.; Speizer, F.E. (1983).Longitudinal study of the
effects of maternal smoking on pulmonary function in children. N. Engl. J. Med.
309:699-703.
Tager, I.E.; Weiss, S.T.; Rosner, B.; Speizer, F.E. (1979) Effect of parental cigarette smoking on
the pulmonary function of children. Am. J. Epidemiol. 110:15-26.
Tainio, V.M.; Savilahti, E.; Salmenpera, L.; Arjomaa, P.; Siimes, M.A.; Perheentupa, J. (1988)
Risk factors for infantile recurrent otitis media: atopy but not type of feeding. Pediatr.
Res. 23:509-12.
Takasaka, T. (1990) Incidence, prevalence, and natural history of otitis media in different
geographic areas and populations. Ann. Otol. Rhinol. Laryngol. 99:13-14.
Tashkin, D.P.; Clark, V.A.; Simmons, M.; Reems, C.; Coulson, A.M.; Bourque, L.B.; Sayre, J.W.;
Detels, R.; Rokaw, S. (1984) The UCLA population studies of chronic obstructive
respiratory disease. VII. Relationship between parental smoking and children's lung
function. Am. Rev. Respir. Dis. 129:891-897.
Taylor, R.G.; Gross, E.; Joyce, H.; Holland, F.; Pride, N.B. (1985) Smoking, allergy, and the
differential white blood cell count. Thorax 40:17-22.
Teele, D.W.; Klein, J.O.; Rosner, B. (1989) Epidemiology of otitis media during the first seven
years of life in children in greater Boston: a prospective, cohort study. J. Infect. Dis. 160:
83-94.
Toyoshima, K.; Hayashida, M.; Yasunami, J.; Takamatsu, I.; Niwa, H.; Muraoka, T. (1987) Factors
influencing the prognosis of wheezy infants. J. Asthma. 24:267-270.
R-45
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Trichopoulos, D.; Kalandidi, A.; Sparros, L. (1983) Lung cancer and passive smoking: conclusion
of Greek study. [Letter] Lancet 667-668.
Trichopoulos, D.; Kalandidi, A.; Sparros, L.; MacMahon, B. (1981) Lung cancer and passive
smoking. Int. J. Cancer 27:1-4.
Trichopoulos, D. (1988) Passive smoking and lung cancer. Scand. J. Soc. Med. 16:75-79.
Tsimoyianis, G.V.; Jacobson, M.S.; Feldman, J.G.; Antonio-Santiago, M.T.; Clutario, B.C.;
Nussbaum, M.; Shenker, I.R. (1987) Reduction in pulmonary function and increased
frequency of cough associated with passive smoking in teenage athletes. Pediatrics
80:32-36.
Uberla, K.; Ahlborn, W. (1987) Passive smoking and lung cancer; a reanalysis of Hirayama's data.
In: Proceedings of international conference on indoor air quality. Tokyo: Council for
Environment and Health, p. 41.
U.S. Bureau of the Census (1990). Statistical abstract of the United States, 1990.
U.S. Department of Health and Human Services. (1982) The health consequences of smoking:
cancer. A report of the Surgeon General. U.S. Department of Health and Human Services,
Public Health Service.
U.S. Department of Health and Human Services. (1983) The health consequences of smoking:
cardiovascular disease. A report of the Surgeon General. DHHS Pub. No. (PHS) 84-50204.
U.S. Department of Health and Human Services, Public Health Service, Office of the
Assistant Secretary for Health, Office of Smoking and Health.
U.S. Department of Health and Human Services.. (1984) The health consequences of smoking:
chronic obstructive lung disease. A report of the Surgeon General. DHHS Pub. No. (PHS)
84-50205. U.S. Department of Health and Human Services, Public Health Service, Office
of the Assistant Secretary for Health, Office of Smoking and Health.
R-46
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
U.S. Department of Health and Human Services. (1986) The health consequences of involuntary
smoking. A report of the Surgeon General. DHHS Pub. No. (PHS) 87-8398. U.S.
Department of Health and Human Services, Public Health Service, Office of the Assistant
Secretary for Health, Office of Smoking and Health.
U.S. Department of Health and Human Services. (1989) Reducing the health consequences of
smoking: 25 years of progress. A report of the Surgeon General. U.S. Department of
Health and Human Services, Public Health Service.
U.S. Department of Health and Human Services. (1990) The health benefits of smoking cessation.
A report of the Surgeon General. U.S. Department of Health and Human Services, Public
Health Service.
U.S. Department of Health and Human Services. (1990b) Vital and health statistics: current
estimates from the national health interview survey, 1989. p. 129.
U.S. Department of Health, Education, and Welfare. (1964) Smoking and health. Report of the
Advisory Committee to the Surgeon General of the Public Health Service. PHS Pub. No.
1103.
U.S. Department of Health, Education, and Welfare. (1977) Nicotine. NIOSH manual of analytical
methods, Vol. 3, 2nd Ed., Publication No. 77-157-C.
U.S. DHEW. See U.S. Department of Health, Education, and Welfare.
U.S. DHHS. See U.S. Department of Health and Human Services.
U.S. Department of Transportation. (1971) Health aspects of smoking in transport aircraft.
Washington, DC: U.S. Department of Transportation, National Technical Information
Service.
U.S. Environmental Protection Agency. (1986a) Guidelines for carcinogen risk assessment. Federal
Register 51:33992-34003.
R-47
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
U.S. Environmental Protection Agency. (1986b) Guidelines for the health risk assessment of
chemical mixtures. Federal Register 51:34014-34025.
U.S. Environmental Protection Agency. (1989) Workshop report on EPA guidelines for carcinogen
risk assessment: Use of human evidence, September 1989.
U.S. Environmental Protection Agency (EPA). (1990) Health effects of passive smoking:
assessment of lung cancer in adults and respiratory disorders in children. External Review
Draft. EPA/600/6-90/006A.
U.S. EPA. See U.S. Environmental Protection Agency.
Varela, L.R. (1987) Assessment of the association between passive smoking and lung cancer.
[doctoral dissertation]. New Haven, CT: Yale University.
Vaughan, W.M.; Hammond, S.K. (1990) Impact of designated smoking area. Policy on nicotine
vapor and particle concentrations in a modern office building. J. Air Waste Management
Assoc. 40:1012-1017.
Vedal, R.E.; Schenker, M.B.; Samet, J.M.; Speizer, F.E. (1984) Risk factors for childhood
respiratory disease. Am. Rev. Respir. Dis. 130:187-192.
Vidic, B.; Ujevic, N.; Shabahang, M.M.; Van de Zande, F. (1989) Differentiation of interstitial
cells and stromal proteins in the secondary septum of early postnatal rat: effect of maternal
chronic exposure to whole cigarette smoke. Anat. Rec. 165:165-173.
Vincent, T.N.; Satterfield, J.V.; Ackerman, L.V. (1965) Carcinoma of the lung in women. Cancer
18:559-570.
Vincent, R.G.; Pickren, J.W.; Lane, W.W.; Bross, I.; Takita, H.; Houten, L.; Gutierrez, A.C.;
Rzepka, T. (1977) The changing histopathology of lung cancer: a review of 1682 cases.
Cancer 39:1647-1655.
Vutuc, C. (1984) Quantitative aspects of passive smoking and lung cancer. Prev. Med. 13:698-704.
R-48
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Wald, N.J.; Boreham, J.; Bailey, A.; et al. (1984) Urinary cotinine as a marker of breathing other
people's tobacco smoke. Lancet 1:230-231.
Wald, N.J.; Ritchie, C. (1984) Validation of studies on lung cancer in nonsmokers married to
smokers. Lancet 1:1067.
Wald, N.J.; Nanchahal, K.; Thompson, S.G.; Cuckle, H.S. (1986) Does breathing other people's
tobacco smoke cause lung cancer? Br. J. Med. 293:1217-1222.
Wall, M.; Johnson, J.; Jacob, P.; Benowitz, N. (1988) Cotinine in the serum, salivia, and urine of
nonsmokers, passive smokers, and active smokers. Am. J. Public Health 78:699-701.
Wang, N.S.; Chen, M.F.; Schraufnagel, D.E. (1984) The cumulative scanning electron microscopic
changes in baby mouse lungs following prenatal and postnatal exposures to nicotine. J
Pathol. 144:89-100.
Ware, J.H.; Dockery, D.W.; Spiro, A., III.; Speizer, F.E.; Ferris, E.G., Jr. (1984) Passive smoking,
gas cooking, and respiratory health of children living in six cities. Am. Rev. Respir. Dis.
129:366-374.
Weber A.; Fischer T.; Grandjean, A. (1979) Passive smoking in experimental and field conditions.
Environ. Res. 20:205-216.
Weir, J.M.; Dunn, J.E., Jr. (1970) Smoking and mortality: a prospective study. Cancer 25:105-112.
Weiss, W.; Boucot, K.R.; Seidman, H.; Carnahan, W.J. (1972) Risk of lung cancer according to
histologic type and cigarette dosage. JAMA 222:799-801.
Weiss, S.T.; Tager, I.E.; Speizer, F.E.; Rosner, B. (1980) Persistent wheeze: its relation to
respiratory illness, cigarette smoking, and level of pulmonary function in a population
sample of children. Am. Rev. Respir. Dis. 122:697-707.
R-49
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Weiss, S.T.i Tager, I.E.; Munoz, A.; Speizer, F.E. (1985) The relationship of respiratory infections
in early childhood to the occurrence of increased levels of bronchial responsiveness and
atopy. Am. Rev. Respir. Dis. 131:573-578.
Weitzman, M.; Gortmaker, S.; Klein Walker, D.; Sobol, A. (1990) Maternal smoking and childhood
asthma. Pediatrics 85;505-11.
Wells, AJ. (1988) An estimate of adult mortality in the United States from passive smoking.
Environ. Int. 14:249-265.
Wells, AJ. (1990) Smoker misclassification does not account for observed passive smoking risk for
lung cancer. Submission to Indoor Air Quality and Total Human Exposure Committee,
Science Advisory Board, U.S. EPA, December 4, 1990.
White, J.R.; Froeb, H.F. (1980) Small airways dysfunction in nonsmokers chronically exposed to
tobacco smoke. N. Engl. J. Med. 302:720-723.
Whittemore, A.S. (1988) Effect of cigarette smoking in epidemiological studies of lung cancer.
Stat. Med. 7:223-238.
Wigle, D.T.; Collishaw, N.E.; Kirkbride, J.; Mao, Y. (1987) Deaths in Canada from lung cancer
due to involuntary smoking. J. Can. Med. Assoc. 136:945- .
Willatt, D.J. (1986) Children's sore throats related to parental smoking. Clin. Otolaryngol.
11:317-321.
Williamson, D.F.; Serdula, M.K.; Kendrick, J.S.; Binkin, N.C. (1989) Comparing the prevalence of
smoking in pregnant and nonpregnant women, 1985 to 1986. JAMA 261:70-74.
Woodward, A.; Douglas, R.M.; Graham, N.M.H.; Miles, H. (1990) Acute respiratory illness in
Adelaide children: breast feeding modifies the effect of passive smoking. J. Epidemiol.
Community Health 44:224-30.
R-50
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Woolcock, A.J.; Peat, J.K.; Leeder, S.R.; Blackburn, C.R.B. (1984) The development of lung
function in Sydney children: effects of respiratory illnesss and smoking. A ten year study.
Eur. J. Resp. Dis. 65(Suppl):l-137. :
Woolcock, A.J.; Peat, J.K.; Salome, C.M.; et al. (1987) Prevalence of bronchial hyperresponsivess
and asthma in a rural adult population. Thorax 42:361-368.
Woolf, B. (1955) On estimating the relationship between blood group and disease. Ann Human
Genet 19:251-253.
Wright, A.L.; Taussig, L.M.; Ray, C.G.; Harrison, H.R.; Holberg, C.J. (1989) The Tucson's
children respiratory study. II. Lower respiratory tract illnesses in the first year of life. Am.
J.Epidemiol. 129:1232-1246.
Wright, A.L.; Holberg, C.; Martinez, F.D.; Taussig, L.M. (1991) Relationship of parental smoking
to wheezing and non-wheezing lower respiratory tract illnesses in infancy. J. Pediatr.
118:207-214.
Wu, A.H.; Henderson, B.E.; Pike, M.D.; Yu, M.C. (1985) Smoking and other risk factors for lung
cancer in women. J. Natl. Cancer Inst. 74(4):747-751. . .
Wu-Williams, A.H.; Samet, J.H. (1990) Environmental tobacco smoke: exposure-response
relationships in epidemiologic studies. Risk Anal. 10:1.
Wynder, E.L.; Hoffmann, D. (1967) Tobacco and tobacco smoke: studies in experimental
carcinogenesis. New York: Academic Press.
Wynder, E.L.; Stellman, S.D. (1977) Comparative epidemiology of tobacco-related cancers. Cancer
Res. 37:4608-4622.
Wynder, E.L.; Gottlieb, S.; Wright, G. (1957) A study of tobacco carcinogenesis. IV. Different
tobacco types. Cancer 10:1206-1209.
R-51
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Wynder, E.L.; Mabuchi, K.; Beattie, E.J., Jr. (1970) The epidemiology of lung cancer: recent
trends. JAMA 213:2221-2228.
Xu, Z.Y.; Blot, W.J.; Xiao, H.P.; et al. (seven others) (1989) Smoking, air pollution and the high
rates of lung cancer in Shenyand, China. J. Natl. Cancer Inst. 81:1800.
Yarnell, J.W.; St Leger, A.S. (1979) Respiratory illness, maternal smoking habit and lung function
in children. Br. J. Dis. Chest 73:230-236.
Young, S.; LeSouef, P.N.; Reese, A.C.; Stick, S.M.; Landau, L.I. (1990) Factors predicting cough
and wheeze in the first 6 months of life. Am. Rev. Respir. Dis. 141:A901.
Zahm, S.H.; Cocco, P.; Blair, A. (1991) Tobacco smoking as a risk factor for colon polyps. Am. J.
Public Health 81:846-849.
Zetterstrom, O.; Osterman, K.; Machado, L.; Johansson, S.G.O. (1981) Another smoking hazard:
raised serum IgE concentration and increased risk of occupational allergy. Br. Med. J.
283:1215-1217.
R-52
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
APPENDIX A
REVIEWS OF EPIDEMIOLOGIC STUDIES ON ETS AND LUNG CANCER
-------
DRAFT—DO NOT QUOTE OR CITE
«
CONTENTS
A.I. INTRODUCTION A-l
A.2. AKIB A-l
A.3. BROW . • A-5
A.4. BUFF A-8
A.5. BUTL(Coh) A-10
A.6. CHAN '. A-13
A.7. CORR A-16
A.8. FONT A-19
A.9. GAO A-25
A.10. GARF (Case-Control) A-28
A.11. GARF(Coh) A-33
A.12. GENG '. A-38
A.13. HIRA(Coh) : A-40
A.14. HOLE(Coh) A-51
A.15. HUMB A-53
A.16. INOU A-56
A.17. JANE A-58
A.18. KABA A-63
A.19. KALA A-66
A.20. KATA A-71
A.21. KOO ... A-73
A.22. LAMT A-78
A.23. LAMW A-81
A.24. LEE A-84
A.25. LIU A-89
A.26. PERS A-92
A.27. SHIM • • • A-96
A.28. SOBU A-99
A.29. STOC : A-102
A* •
-11
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
CONTENTS (continued)
A.30. SVEN ....... . . ............ ...... ..... ........ ................ A_103
A-31- TRIG ...................... . ............................... . . A_106
'32- WU
A-110
A.33. WUWI ..................................... .... A-115
A-iii
05/15/92
-------
-------
DRAFT—DO NOT QUOTE OR CITE
A.l. INTRODUCTION
This appendix contains a review of each epidemiologic study based on the primary
references listed in Table 5-1. Descriptions of the four prospective cohort studies are
individualized according to the requirements of each study—for example, HIRA(Coh) has a long
history of controversy in the literature, so the main arguments are chronicled and discussed as part
of the review. Reviews of case-control studies follow a structured format, consisting of three
parts: (1) the author's abstract, which summarizes the most salient features and conclusions in the
author's opinion; (2) a study description based on the contents of a completed study format
designed around principles of good epidemiologic practice and features specific to environmental
tobacco smoke (ETS); and (3) a section of comments related to evaluation and interpretation of the
study. The author's abstract is, of course, entirely his words; the study description is intended to
portray accurately the reference article vis-a-vis items in the study format, so the author's words
are used when possible; the comments section is entirely our own assessment of characteristics
relevant to study interpretation and utility in this report.
Only an abstract is available for the case-control study by Stockwell et al., referred to as
STOC, which has not appeared in print yet. There is insufficient information on the study to
include it in the main body of this report. Similarly, only an abstract is available for the second
study of Kabat and Wynder, which is included in an addendum following the review of their first
study, KABA. The data for many of the studies reviewed have been extracted from a larger,
more comprehensive study that includes active smokers. The subjects and their data used for
investigation of an association between ETS exposure and lung cancer incidence are referred to as
"ETS subjects" and "ETS data," respectively.
A.2. AKIB
A.2.1. Author's Abstract
"A case-control study conducted in Hiroshima and Nagasaki, Japan, revealed a 50%
increased risk of lung cancer among nonsmoking women whose husbands smoked. The risks
tended to increase with amount smoked by the husband, being highest among women who worked
outside the home and whose husbands were heavy smokers, and to decrease with cessation of
exposure. The findings provide incentive for further evaluation of the relationship between
passive smoking and cancer among nonsmokers."
A-l
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
A.2.2. Study Description
This community-based case-control study was conducted in Hiroshima and Nagasaki,
Japan, in 1982. The data collected on passive smoking are part of a larger investigation of lung
cancer among atomic bomb survivors, the principal objective of which is to evaluate the
interactive roles of cigarette smoking and ionizing radiation. This article reports on married
female never-smokers, an unmatched subset of the data from the whole study.
The whole study includes a total of 525 primary lung cancer cases diagnosed between 1971
and 1980. Cases were identified from the Hiroshima and Nagasaki Tumor and Tissue Registries
and other records. Controls were selected from among the cohort members without lung cancer,
two per case in Hiroshima and three per case in Nagasaki. The controls were individually
matched to the cases with respect to year of birth (± 2 years), city of residence (Hiroshima or
Nagasaki), sex, biennial medical examinations, and vital status. The majority of cases were
deceased; those cases were matched to decedent controls by year of death (±3 years), in addition
to the other criteria. Controls were selected from causes of death other than cancer and chronic
respiratory disease. Face-to-face interviews were conducted for 81% (82%) of the eligible cases
(controls), but 80% to 85% of the interviews for both cases and controls were actually conducted
with the subject's next-of-kin. The mean age of cases at diagnosis is 72.1 years (range 36-94) for
males and 70.2 (range 35-95) for females, which is high for lung cancer in Japan. Fifty-seven
percent of the cases were pathologically confirmed; the remaining 43% were diagnosed by
radiological or clinical findings.
ETS exposure in adulthood was assessed by spousal smoking status, including the average
number of cigarettes smoked per day, age the spouse started smoking, and, for those who stopped
smoking, the age at cessation. For childhood exposure, a single question was asked regarding
whether the subject's mother or father or both smoked when the subject was living at home as a
child; responses were obtained for only two-thirds of the subjects. No specific information on
exposure to smoking by other household members' smoking or to smoking in the workplace was
obtained. ETS exposure data were checked by comparing smoking status with records from
RERF surveys in 1964-68 (self-reported by subjects when they were alive). Cases and controls
who had never married were excluded. Of the female cases exposed to spousal smoking, 16% had
squamous or small cell carcinoma, whereas no unexposed cases had those cell types. No
information was provided on location of the carcinomas.
The number of female cases exposed to ETS is 73 out of 94 (number exposed/total)
compared with 188 out of 270 female controls (crude odds ratio [OR] is 1.52 [95% C.I. =
0.88-2.63], by our calculations). Application of logistic regression to the whole study that includes
A-2
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
active smokers, gives an adjusted odds ratio of 1.5 (90% C.I. = 1.0-2.5), similar to the crude
analysis. It is not stated explicitly that matching variables were included in the logistic regression
model. Four additional analyses were conducted on the ETS data alone (i.e., without active
smokers). The authors stratified exposure by number of cigarettes smoked per day by husband (0,
1-19, 20-29, 30+) and obtained a marginally significant trend (p = 0.06). No dose-response
gradient was found in the association between the number of years the husband smoked cigarettes
and the risk of lung cancer in female never-smokers; the odds ratio decreases from lowest to
highest exposure level (2.1, 1.5, and 1.3). Stratified analysis according to recency of exposure to
husband's smoking (unexposed, exposed but not within the past 10 years, and exposed within the
past 10 years) shows a significant upward trend (p = 0.05). Further stratification of exposed
subjects by occupation found that lung cancer risk tends to increase across occupational categories
in the following order: housewife, white collar worker, blue collar worker. The highest odds
ratio occurred for women who had blue collar jobs and were married to men who smoked one or
more packs of cigarettes per day, but the number involved was small. It is reported that
additional analyses on the data indicated that factors for matching in the whole study have little
influence, but the details are omitted.
Limited histological information is provided. Among cases exposed to spousal smoking,
16% had squamous or small cell cancer, and 84% had adenocarcinoma or large cell cancer. All of
the unexposed cases had adenocarcinoma.
The authors conclude that there may be a moderate excess in lung cancer risk associated
with passive smoking. The odds ratio for lung cancer among nonsmoking women tends to increase
with amount smoked by their husbands, a trend seen among housewives, as well as among women
who work outside the home. There was little association with parental smoking or from passive
smoking that had ceased more than 10 years previously.
A.2.3. Comments
The larger study from which the ETS data are taken was primarily intended to investigate
the interaction of smoking and ionizing radiation in atomic bomb survivors of Nagasaki and
Hiroshima. The information on passive smoking has been collected posthumously in a large
percentage of the cases, requiring heavy use of proxy responses. The response rate was not high,
however, because of the refusal of next-of-kin to answer questions about deceased relatives and
lack of attempt to locate next-of-kin to answer questions about some subjects who had died or
moved away from Hiroshima or Nagasaki. The dependence on proxy respondents raises questions
about the validity of the exposure data for some measures, particularly in childhood, and about
A-3
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
detailed information such as the number of cigarettes smoked per day, duration of smoking habit,
and years since cessation of smoking. Information on childhood exposure was obtained for only
two-thirds of the subjects. The omission of data on subjects where the next-of-kin had refused
response or the subject had moved may be a source of bias. The diagnosis of lung cancer was not
pathologically confirmed in over 40% of the cases. Even if the data were complete and accurate
on all subjects, however, it is not clear that the subjects are representative of the target
population. They had been exposed to ionizing radiation to varying degrees, whatever implication
that may have; they are among the survivors, which may suggest selective characteristics; and their
age distribution is high, ranging from about 35 to 90 years of age with an average of 70 or more
years.
Only ever-marrieds are included in the ETS subjects, which is helpful in the analysis..
There is some ambiguity in the statistical analyses, however, in reference to Tables 2 through 6
(the main results). The tables contain odds ratios that are reported to be the result of logistic
regression with matching. The details regarding matching in the analysis are not given, but it is
reported that analysis of the crude data and matched logistic regression give similar values.
Regarding the analyses for trend, the outcome seems to be sensitive to the measure of exposure
used. The odds ratio are strictly increasing for stratification by number of cigarettes smoked per
day, but a different pattern emerges when ETS exposure is measured by the number of years the
husband smoked cigarettes.
In general, the conclusions are presented more strongly than the data warrant. The
assertions are somewhat tenuous that risks tend to increase with amount smoked by the husband,
are highest among those who worked outside the home and whose husbands are heavy smokers,
and decrease with cessation of smoking. Conversely, whereas little association between ETS
exposure in childhood and lung cancer is reported, relevant information was available for only
two-thirds of the subjects, and its accuracy is questionable because most of that information was
provided by proxies. Overall, the observed data suggest that ETS exposure may be related to risk
of lung cancer, but there is some potential for misclassificatioh and other sources of bias. Thus,
this study provides some useful information on lung cancer risk in passive smokers, but its
interpretation needs to be more conservative, taking into account the atypical characteristics of the
subjects and other concerns described above.
A-4
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
A.3. BROW
A.3.1. Author's Abstract
"The relation between various risk factors and adenocarcinoma of the lung was evaluated
in a case-control study. Subjects were selected from the Colorado Central Cancer Registry from
1979-82 in the Denver metropolitan area. A total of 102 (50 males and 52 females)
adenocarcinoma case interviews and 131 (65 males and 66 females) control interviews were
completed. The control group consisted of persons with cancers of the colon and bone marrow.
The risk estimates associated with cigarette smoking were significantly elevated among males
(odds ratio [OR] = 4.49) and females (OR = 3.95) and were found to increase significantly (p <
0.01) with increasing levels of cigarette smoking for both males and females. For adenocarcinoma
in females, the age- and smoking-adjusted odds ratios at different levels of passive smoke
exposure followed an increasing overall trend (p = 0.05). After additional adjustment for
potential confounders, prior cigarette use remained the most significant predictor of risk of
adenocarcinoma among males and females. Analysis restricted to nonsmoking females revealed a
risk of adenocarcinoma of 1.68 (95% confidence interval [C.I.] = 0.39-2.97) for passive smoke
exposure of four or more hours per day. Neither sex showed significantly elevated risk for
occupational exposures, although males bordered on significance (OR = 2.23, 95% C.I. = 0.97-
5.12). The results suggest the need to develop cell type-specific etiologic hypotheses."
A.3.2. Study Description
This study was conducted in Denver, Colorado, to evaluate the role of smoking, passive
smoking, occupation, community air pollution, and socioeconomic status in the etiology of
adenocarcinoma of the lung. Because subjects include active smokers, the data on ETS subjects
are part of a larger data set.
Cases and controls were drawn from the Colorado Central Cancer Registry. All subjects
were diagnosed with lung'adenocarcinoma between 1979 and 1982. Cases are white female
Denver residents of at least 6 months' duration. Controls are of similar description to the cases,
except that they were diagnosed with colon cancer or bone marrow cancer. Controls were
matched on a group basis to produce the same age and gender composition. It is not clear if
incident cases were used and whether control sampling was cumulative or density.
The subjects are not matched on smoking status, so the data on ETS subjects alone are
unmatched for all variables considered in the larger study. Face-to-face interviews were
conducted, blindly, on a total of 149 cases and 169 controls, after attrition in selection and follow-
up of 47 cases and 38 controls. The subject was interviewed in 31% of the cases and 61% of the
A-5
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
controls; the remaining interviews were conducted with a friend or relative. The mean age of the
female cases (controls) was 64.9 (68.2) years; no further details are provided. Clinical verification
of lung cancer diagnosis was conducted microscopically.
"Exposed" to ETS is used in two ways, depending on context: (1) the husband smoked
(presumably "ever-smoked" is intended, rather than "currently smokes," but that is not explicit);
(2) the subject was in the presence of tobacco smoke, from any source, 4 or more hours per day on
"average. Although there are two operational definitions of exposure, neither includes duration of
ETS exposure. Questions were apparently asked regarding exposure in both childhood and
adulthood, the latter including sources in the home and in the workplace. No indication was
found that the data collected from subjects was checked for internal consistency or against other
sources. No mention was found regarding the number of unmarried women in the study or what
assumptions may have been made regarding their exposure to ETS when spousal smoking is the
source considered (the first of the definitions given above).
The ETS subjects consist of 4 out of 19 (exposed/total) female cases and 7 out of 47
controls, when ETS exposure means the spouse smoked (Definition 1). For exposure from all
sources (Definition 2), the corresponding numbers for cases and controls are 4 out of 19 and 6 out
of 47, respectively. The crude odds ratio is 1.52 (95% C.I. = 0.39-5.96) for Definition 1 of ETS
exposure and 1.82 (95% C.I. « 0.45-7.36) for Definition 2 (data communicated from first author,
Brownson). A test for trend using hours per day as the exposure measure is conducted on the
whole data set for females including smokers (33 of 52 cases are smokers and 19 of 66 controls are
smokers; the two exposure categories, 4 to 7 and 8 or more hours per day of exposure to passive
smoke, contain a total of only 4 cases and 6 controls who are nonsmokers, but 19 cases and 7
controls who are smokers). The method of Miettenen is applied with stratification on age and
smoker status (p =* 0.05 for trend). The data for never-smokers alone were used in a multiple
logistic regression to compare subjects exposed 0 to 3 hours per day with those exposed from all
sources 4 or more hours per day (Definition 2 of ETS exposure). Adjustments were made for age,
income, and occupation. The reported odds ratio is 1.68 (95% C.I. = 0.39-2.97). (Note: It appears
that the upper confidence value may be in error. In view of the outcome for the crude odds ratio,
a value about twice what is shown might be anticipated.)
To summarize the statistical tests and authors' conclusions, no significant risk estimates
were shown when smoking by the spouse was considered as a dichotomous variable. When the
data for both active smokers and passive smokers were stratified according to level of passive
smoke exposure, a statistically significant trend in the risk estimates was shown for females
(p « 0.05) after adjustment for age and cigarette smoking. However, after adjustment by logistic
A-6
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
regression for age, income, occupation, and cigarette smoking, with the two exposure categories
for ETS combined (> 3 and 4+hours per day), no significant risk was detected.
A.3.3. Comments
The study is very small when reduced to the never-smokers alone. The measure of ETS
exposure used (hours/day from all sources) is not very specific to differentiate exposed from
unexposed persons, particularly exposure 20 to 30 years ago, which may be more relevant than
current exposure. Only 15% of the controls have a husband who smoked; only 13% of ETS
subjects are exposed from any source 4 or more hours per day. Thus, the cut-point selected by
the researchers for general ETS exposure (4+ hours/day) may be too high, resulting in a
substantial amount of exposure in the "unexposed" group. For either definition of ETS exposure,
however, the percentage exposed is extremely low. Details are lacking also in other areas that may
have a bearing (e.g., the treatment of unmarried subjects—whether they were present and, if so,
the assumption made regarding ETS exposure).
We experienced some difficulty with the statistical analyses. One of the adjusted
procedures is the trend test. Perhaps because the number of ETS subjects is so small, smokers,
were included in the analysis and then a method was used to attempt to adjust the effects of their
presence on the outcome. The only value that leaving the smokers in the analysis would serve,
that we can see, would be if one believes that they contribute to the evidence on lung cancer and
passive smoking. That seems doubtful. There are so few ETS subjects in the exposure categories
(see above) that it seems highly unlikely that a test for trend would be significant if based on the
ETS subjects alone (we did not have the number of ETS subjects by exposure group, however, so
we were unable to conduct the trend test to check the outcome).
_ When the two exposure categories were combined and only the ETS subjects used, the
results were not close to statistically significant (OR is 1.68; 95% C.I. = 0.39-2.97). We also had a
problem with that result. When a statistical procedure is used to determine which variables to
adjust for in using another procedure with the same data, it may distort the statistical
interpretation. There also may be a typographical error in the upper confidence limit because the
value shown is only about half the corresponding value for the crude odds ratio.
The remaining analyses are from the crude odds ratio, 1.52 (95% C.I. = 0-39-5.99) and 1.82
(95% C.I. = 0.45-7.36), which suggests a possible association between ETS exposure and lung
cancer, although it could easily be ascribed to chance in view of the wide confidence intervals.
The study has a very strict requirement for classification as exposed to ETS (4+ hours per day),
which is reflected in only 15% of the controls being designated as exposed (40-60% is more
A-7
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
typical). The 12% figure based on simply being married to a smoker, however, is no better. The
control subjects are unlikely to be a representative sample of the target population, or else the
classification of subjects exposed is too rigid. The crude odds ratio may be the preferred
statistical measure to represent the outcome of the data, but care should be exercised in using the
results from this study in conjunction with those of other studies.
A.4. BUFF
A.4.1. Author's Abstract
"A population-based case-comparison interview study of lung cancer was conducted from
1979 to 1982 in six Texas coastal counties—Orange, Jefferson, Chambers, Harris, Galveston, and
Brazoria—to evaluate the association of lung cancer with occupational and other environmental
exposures. Lung cancer mortality rates in these counties consistently have exceeded lung cancer
mortality rates observed for Texas and the United States from 1950-69 to 1970-75 for both sexes
and races (white and nonwhites).
Histologically and cytologically confirmed incident cases diagnosed during the interval
July 1976 to June 1980 among white male and female residents ages 30 to 79 years were
ascertained from participating hospitals in the six-county area. Both population-based and
decedent comparisons were selected and matched on age, race, sex, region of residence, and vital
status at time of ascertainment. The exposures of primary interest in the study of lung cancer are
those associated with occupation (employment in specific industries and occupations) in
conjunction with tobacco, alcohol, diet, and residential exposures."
A.4.2. Study Description
This population-based case-control study was conducted in six coastal counties of eastern
Texas to evaluate the association of lung cancer with occupational and other environmental
exposures. Those of primary interest are associated with occupation in conjunction with tobacco,
alcohol, diet, and residential exposures. The ETS subjects are part of this larger study that
includes active smokers.
Cases include males and females ascertained from hospital and state records during
1976-80, except for Harris County, which includes only females from 1977-80. All subjects are
white (including Hispanic) county residents of at least 6 months. Cases are incident, without
restriction to cell type, and histologically diagnosed to eliminate secondary lung cancers (there is
some inconsistency in the article on whether all diagnoses were by histology or whether some were
by cytology). Controls were selected from State and Federal records, group matched on age, sex,
A-8
05/15/92
-------
.DRAFT—DO NOT QUOTE OR CITE
race or ethnicity, county of residence, and vital status. The candidate sample size is estimated in
the report at approximately 1,650, including both sexes, of which just over 700 were lost to
attrition in selection or follow-up for various reasons. Face-to-face interviews were conducted, a
large number of which were with next-of-kin as necessitated by inclusion of decedent cases and
controls. For example, for females, the number of subject interviews is only 18% for cases
(81/460) and 24% (116/366) for controls. The distribution of ages is similar for cases and
controls, based on groupings of 10-year intervals.
"ETS exposed" means having ever lived with a household member who smoked regularly.
Exposure sources include the home environment during childhood and adulthood but excludes the
workplace. There is no mention of whether data on ETS exposure were cross-checked with other
interview questions or other sources. No indication was found regarding unmarried females in the
sample and how marital status may affect level of exposure to ETS. Some summary information is
provided on the distribution of tumors by cell type, but totals include smokers, so they are not
reproduced here. The ETS data for females consists of 33 out of 41 (exposed/total) cases and 164
out of 196 controls; for males, the respective figures are 5 out of 11 and 56 out of 90. For the
exposure definition given above, the crude odds ratio reported is 0.78 (95% C.I. = 0.34-1.81) for
females (direct calculation from the data yields a value of 0.81; Buffler apparently added 0.5 to all
cells to compensate for inclusion of no subjects in some cells). Little difference was found when
female smokers were categorized by number of years lived with a household member who smoked.
No adjusted statistical analysis is provided to account for variables used in matching for the study
as a whole, nor is there a test for trend. The authors conclude that no effect of passive smoking is
indicated for lung cancer. No attempt is made to evaluate whether exposure to ETS in childhood
or adulthood is a factor.
A.4.3. Comments
The potential relationship between ETS exposure and lung cancer risk was not a principal
issue in the design of this study. As described in the abstract, and more fully in the study
description above, other potential etiologic factors were of more central concern. There are
several limitations regarding the study's contribution to the epidemiologic evidence on ETS
exposure and lung cancer risk. For example, the interview question on exposure to ETS is not
very specific. "Having lived with a household member who smoked regularly" does not distinguish
between exposure in childhood and in adulthood, between substantial and only light exposure, or
between short-term and long-term exposure. One might expect a high percentage of persons to
qualify as "exposed" under such a broad definition, and that is what the study demonstrates: 84%
A-9
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
of the controls are classified as exposed. With such a high percentage, both cases and controls may
include a number of subjects who have experienced very light exposure to ETS. Another concern
in this study is the use of decedent subjects. The majority of both male (86%) and female (82%)
cases in the study (including smokers) were deceased. Consequently, a very high percentage of
interviews was by proxy (82% of cases and 76% of controls).
This study was conducted in a region with a significantly higher age-adjusted mortality
rate for lung cancer than for the United States in general. For all ages combined, the overall
excess lung cancer mortality in the Texas study area is approximately 30% to 40% and is
considerably higher for some age groups, according to the article. This was the apparent
motivation for the study, with emphasis on important occupational and industrial exposures for
residents of the Texas coastal area, including those associated with shipbuilding and repair,
chemical and petrochemical manufacturing, petroleum refining, construction, and metal
industries. If these nonsmoking factors affect the incidence of lung cancer, then they may be
confounding the attempt to detect an effect from passive smoking. Appropriate statistical
methods need to be applied to adjust the effect of each risk factor for the others.
Other factors may affect the ETS analysis also. Harris County, which is frequently
addressed in the article in distinction to the other five counties, was apparently added to the study
later (case ascertainment began 1 year later there and included only females; 10 of the 11 hospitals
that did not participate are in Harris County). Consequently, there are some regional differences
in the study as well as ethnic and racial differences (white and Hispanic). Although the authors
took care to match controls on these and other factors, the matching only applies to the whole
study (91% and 97% of male and female cases, respectively, are classified as having smoked
regularly), not to the ETS subject group specifically, and there is no adjustment for these factors
in the analysis. This potential confounding, the insensitive indicator of ETS exposure, and the
large use of decedent cases and proxy responses limit the value of this study toward detecting any
health effects associated with passive smoking.
A.5. BUTL(Coh)
This study was undertaken to explore the role of active and passive smoking in Seventh-
Day Adventists in California. Subjects were participants in a larger prospective cohort study of
factors affecting health in Adventists.
In 1974 the Adventist Health Study was initiated with the purpose of investigating the
associations of a number of lifestyle and nutritional factors with morbidity and mortality in
California Seventh-Day Adventists. Registered Adventist households were identified by
A-10
05/15/92
-------
DRAFT—DO NOT QUOTE OR GITE
contacting the clerks of all 437 California Adventist churches. A basic demographic questionnaire
sent to all households received a response rate of 58%. In 1976, all subjects aged 25 or older in
1974 were asked to complete a lifestyle questionnaire that included many demographic, medical,
psychological, and dietary variables. Over two-thirds of the targeted subjects responded. From
the non-Hispanic whites among these respondents, Butler and his colleagues drew two cohorts.
One consisted of 22,120 spouses married and living together at the time of completion of the
lifestyle questionnaire in 1976 ("spouse pairs") and the other of 6,467 individuals participating in
an Adventist Health Smog Study of air pollution and pulmonary disease (the "ASHMOG" cohort);
about two-thirds of the ASHMOG cohort was also included in the spouse-pairs cohort.
Subjects received annual forms for self-reporting of hospitalizations in the past year.
Medical records relating to reported hospitalizations were then reviewed. Mortality was traced in
four ways: linkage with California Death Certificate and National Death Index Systems, church
clerk notification of deaths entered in church records, and follow-up of hospitalization history
form responses (or nonresponses). Underlying and contributing causes of death were obtained
from death certificates. Death certificates were obtained for all reported fatalities.
For the spouse-pairs cohort, subjects were considered unexposed to ETS if their spouses
were either never-smokers or ex-smokers baptized into the Adventist church—which proscribes
tobacco usage—before marriage. Those whose spouses were ex-smokers with less than 5 years of
total smoking were also considered unexposed. All other subjects with ex- and current smoker
spouses were classified as exposed.
Incidence rates were calculated using person-years. In the spouse-pairs cohort,
age-adjusted lung cancer mortality rates for females married to past or current smokers were
higher than those for female spouses of never-smokers, yielding relative risks of 1.94 and 2.47 for
past and current smokers, respectively. Comparison of wives with ever- versus never-smoking
husbands yielded a relative risk of 2.0. The same age-adjusted relative risk resulted when
analyses were restricted to the 9,207 never-smoking females included in the spouse pairs.
Virtually identical risk estimates resulted from both Mantel-Haenszel and Maximum Likelihood
analyses. None of the relative risks was statistically significant at the 5% level.
In the ASHMOG cohort, the relative risk of lung cancer adjusted for age and past smoking
status among females was 1.16 for women who had lived with a smoker for at least 11 years
compared with women who had not lived with a smoker; no difference was observed for women
who had lived for less than 11 years with a smoker, although this group was only one-tenth as
large as the others. A similar pattern was seen among males who had lived for at least 11 years
with a smoker, with an adjusted relative risk of 1.17.
A-ll
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
In the spouse-pairs cohort, age-adjusted rates of smoking-related cancers (excluding lung
cancer) were only slightly higher among nonsmoking females married to smokers than among
nonsmokers (RR - 1.06); the relative risk rose to 1.22 when lung cancers were included.
In the ASHMOG cohort, age-adjusted rates using conditional maximum likelihood analysis
for all smoking-related cancers were higher among males who lived with a smoker (RR = 1.45 for
1-10 years; 1.74 for 11+ years) or worked with a smoker (RR = 2.62 for 1-10 years; 1.47 for 11+
years). Among females, in contrast, only one (at RR = 1.03) of the four exposed categories had a
higher rate than the nonexposed groups.
All lifestyle questionnaires were administered anonymously, thus reducing the potential
for inaccurate responses caused by fear of discovery; respondents to the special supplemental
ASHMOG questionnaire were assured of confidentiality but not anonymity.
Although causes of death were obtained from death certificates, review of medical records
revealed histological confirmation in 99% of the primary malignancies reported among the spouse-
pairs cohort. Thus, substantial misclassification of lung cancer deaths is unlikely. Subsequent
study of patients discharged from 1 of the 11 participating Adventist medical centers over a 6-
month period indicated that under 2% of study participants failed to report their hospitalizations;
serious underascertainment of cases thus also seems unlikely. Losses to follow-up by study's end
totaled only 1.2% of the original study cohort-a very low rate.
Comparing the results of the 1976 questionnaire with those of a supplemental
questionnaire given to ASHMOG subjects in 1987, 4.7% of male smokers now reported themselves
as "never-smokers" and 1.4% of never-smokers now reported themselves as nonsmokers.
Concordance of female responses was even higher. This concordance of responses does not
necessarily imply the degree of accuracy of responses, only their reliability.
Comparison of responses to the 1987 questionnaire by females revealed that about 6% of
those previously classified as not having a smoking spouse now reported having had one; the
converse was also true for 6% of the women. These data indicate a mild nondifferential
misclassification of exposure, which would push results toward the null.
Information is available on a large number of variables of possible interest as potential
confounders or risk mediators. Unfortunately, the modest number of total lung cancer deaths
among females in the spouse-pairs cohort (8) or among both sexes in the ASHMOG cohort (13)
discourages attempts to control for other potential confounders in addition to age in the analyses.
Separate consideration of the association between variables other than passive smoking and age-
adjusted lung cancer mortality among women indicated a high relative risk (RR > 4) for spousal
blue collar occupation. No other variables produced nearly as strong or consistent an association;
A-12
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
in fact, the only other consistent association was a relative risk of 1.3 to 1.6 for nonrural status.
Unfortunately, no breakdown of blue collar spousal status by exposure groups was presented.
By virtue of its basic design, the inherent minimization of sources of confounding
provided by its study population and the level of information available regarding potential
confounders, and other sources of bias, the Butler study has many of the key ingredients to
produce convincing results. Unfortunately, this potential goes largely unrealized because of the
low number of outcome events occurring during the follow-up period, which for the most part
renders stratification or control for multiple factors simultaneously impractical; even stratification
by several age or exposure levels produces unstable results.
Nevertheless, the findings of this study are quite consistent with the hypothesis that ETS
exposure of nonsmokers is associated with mildly elevated lung cancer, (active) smoking-related
cancer, and ischemic heart disease mortality. Insofar as the study data allow for consideration of
potential misclassification and confounding effects, neither misclassification nor confounding can
account for the observed association. Because of the limited number of outcome events, several
possible confounding factors could not be definitively or adequately addressed in the analyses and
the observed associations were not statistically significant, the study's findings must be viewed as
suggestive but not of themselves convincing.
A.6. CHAN
A.6.1. Author's Abstract
(Note: This study is described in two sources, both of which were used for the description below.
Chan et al. [1979] is the more complete description, but it contains considerable attention to active
smoking as a cause of lung cancer. Chan and Fung [1982] is a condensed version that specifically
addresses nonsmokers. The abstract given here is for the 1979 article. No abstract is provided in
the 1982 source.)
"Bronchial cancer is a disease of high and increasing annual incidence in Hong Kong,
especially in women, whose age-specific death rates from this cause are amongst the highest in the
world. A case-control study of the relationship of bronchial cancer with smoking was carried out
during 1976-77, taking particular note of the histological type of the tumour. Two hundred and
eight male and 189 female patients were interviewed, covering about half the total number of
cases of bronchial cancer registered as dead from the disease in Hong Kong during the period of
survey. The association with smoking was more evident in males than in females, and in
squamous and small-cell types, as a group, than in adenocarcinoma. Forty-four per cent of the
women with bronchial cancer were non-smokers, their predominant tumour being
A-13
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
adenocarcinoma, and in them no association could be detected with place of residence or
occupation. There was no strong evidence of an association with the use of kerosene or gas for
cooking; 23 did not use kerosene. The cause of the cancer in these nonsmoking women remains
unknown."
A.6.2. Study Description
(Note: This description is primarily based on Chan et al. [1979]. Chan and Fung [1982] are cited
when used as a reference.)
This study is the earliest of four from Hong Kong that consider ETS exposure as a
potential etiologic factor for lung cancer incidence in nonsmoking women. Here, however, that
objective is secondary to evaluation of the relationship of bronchial cancer with active smoking.
In the whole study, target cases are the lung cancer patients, male and female, in five
hospitals in Hong Kong during 1976-77 that were willing and able to be interviewed. Controls are
patients of the same general age groups from the orthopedic wards of the same hospitals as the
cases. No specific diseases are excluded. Cases are incident and control sampling is density. The
candidate sample size is 208 (189) male (female) cases and 204 (189) male (female) controls.
Attrition from selection or follow-up is not reported but appears high. Subjects were personally
interviewed, as possible. About half of the estimated number of lung cancer cases diagnosed in
Hong Kong during the study period were actually interviewed. Some patients were too ill to
answer questions, and more than expected were treated elsewhere than in the hospitals covered.
No interviews with next-of-kin were obtained for the cases interviewed.
The ETS subjects (never-smokers) alone include 84 (2) female (male) cases and 139 (30)
female (male) controls. The age distribution of the female cases (controls) is, by percentage, as
follows: age less than 40, 7 (5%); ages 40 to 49, 15 (15%); ages 50 to 59, 23 (30%); ages 60 to 69,
23 (22%); and age 70 or more, 32 (28%). Cases with a histological diagnosis were reviewed and
verified by reexamination of the pathological specimens. In the absence of a histological
specimen, cytological diagnosis was accepted. In some cases, on histological grounds, secondary
adenocarcinoma was suspected, and a few cases were rejected after detailed examination of the
clinical records. Of the cases, 46 (55%) were diagnosed by histology, 23 (27%) by cytology, and 15
(18%). by radiology and clinical means. Diagnoses by cell type were as follows: squamous or small
cell, 19 (22%); adenocarcinoma or large cell, 40 (48%); others and unspecified, 25 (30%). Of the
unspecified, 15 had no histological or cytological verification.
ETS subjects are never-smokers. Classification of a subject as exposed or unexposed to
ETS is based on the response to these questions: (1) If you do not smoke, have you been exposed
A-14
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
to cigarette smoke from other people at home or at work? (2) Does your husband/wife smoke? (If
"yes," how many cigarettes per day?) (The first question is included in Chan et al., 1979. The
second one is from a communication of Linda C. Koo.) No information is reported on the
distribution of tumors by central and peripheral location.
The ETS data on females based on question 1, above, consists of 50 out of 84
(unexposed/total) cases and 73 out of 139 controls. The authors state that "this is a rather
subjective approach to the problem." No statistical estimates are provided; our calculation of the
crude odds ratio is 0.75 (95% C.I. = 0.43-1.30). No clear conclusion is drawn regarding the
potential relationship between ETS exposure and lung cancer occurrence, but the authors imply
that no connection was found (which the odds ratio and confidence interval amply support). The
authors found no particular occupation as being dangerous. Their findings also do not support air
pollution as a factor, and they provide no strong evidence that cooking with various types of fuel
is relevant.
A.6.3. Comments
Although data on spousal smoking were collected along with an indication of the number
of cigarettes smoked per day, they are referred to only in the 1982 article, where the authors note
without further elaboration that more nonsmoking cases have nonsmoking spouses. It is reported
that answers to the question, "Are you exposed to the tobacco smoke of others at home or at
work?" gave no indication that other people's smoking was a risk factor for lung cancer in
nonsmokers, with 40.5% of cases and 47.5% of controls answering yes to this question. Why the
data for spousal smoking are not given and analyzed is unknown. The question about general ETS
exposure combines sources in the household and workplace and refers only to current exposure
without a measure of duration, which would likely affect any risk associated with passive
smoking.
Although it is reported that cases and controls are similar in age, occupation, and other
characteristics, comparability is questionable. The article cites a criticism of the whole study
(including smokers) for use of orthopedic patients as controls, on the basis that some patients may
be hospitalized with smoking-related diseases (e.g., osteoporosis). It was found that the controls
smoke more than a group representative of the population of Hong Kong. This would create a
bias toward negative association. Although these comments refer to smoking habits, they suggest
the potential for selection bias of controls that may extend to nonsmoking controls as well.
It is noted, also, that there are more cases from Hong Kong Island than would be expected
from the population distribution of Hong Kong as a whole, possibly due to more success
A-15
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
contacting cases in Hong Kong Island than in Kowloon. The authors caution about reaching any
conclusion about the distribution of cases within Hong Kong as a whole. The failure to follow up
on patients who were eventually treated at other hospitals or were too ill to be interviewed is, of
course, a potential source of bias itself.
Other differences are apparent between cases and controls. Among nonsmokers, a higher
percentage of cases than controls (1) are Cantonese (81 vs. 70) or (2) have ever cooked with
kerosene (73 vs. 60). It is speculatedthat the Cantonese diet, high in nitrite or nitrate content, may
be a factor in lung cancer incidence (Chan and Fung, 1982). More broadly, these comparisons
between cases and controls indicate differences in ethnic composition, lifestyle, and
socioeconomic status that are difficult to assess.
In summary, ETS subjects are not matched in the design, and an adjusted statistical
analysis is not conducted. Consequently, potential sources of bias and confounding are not
controlled. There is substantial basis to question the comparability of cases and controls, as
described above. Data quality is suspect because confirmation of primary lung cancer was limited
and cases were missed because patients were too ill to be interviewed personally of were
eventually treated at another hospital. Also, the question posed to subjects for classification as
exposed or unexposed to ETS is sufficiently general to invite subjective response. Overall,
methodological shortcomings hamper interpretation of this study's findings, rendering its
conclusions questionable.
The finding that spousal smoking appears to be more frequent in controls, mentioned in
the 1982 report, is noted to be at variance with the Hirayama study, which may have motivated
the authors to conduct this secondary analysis of ETS exposure using their previously collected
data. Whatever the motivation, the original study is rather limited as a source to evaluate passive
smoking. Overall, this study does not reflect as much care and attention to detail as would be
useful, limiting its value for assessing ETS exposure and lung cancer.
A.7. CORK
A.7.1. Author's Abstract
"Questions about the smoking habits of parents and spouses were asked in a case-control
study involving 1,338 lung cancer patients and 1,393 comparison subjects in Louisiana, USA.
Nonsmokers married to heavy smokers had an increased risk of lung cancer, and so did subjects
whose mothers smoked. There was no association between lung cancer risk and paternal smoking.
The association with maternal smoking was found only in smokers and persisted after controlling
for variables indicative of active smoking. It is not clear whether the results reflect a biological
A-16
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
effect associated with maternal smoking or the inability to control adequately for confounding
factors related to active smoking. This preliminary finding deserves further investigation."
A.I.2. Study Description
This study was conducted in Louisiana to investigate the relationship of smoking habits of
parents and spouses to lung cancer occurrence. Results of the study were published in 1983; some
clarifying details regarding study methodology were supplied in a 1984 paper addressing only the
effects of active smoking. The accrual period is not stated; cases are probably a mixture of
prevalence and incidence, and controls are cumulatively sampled. ETS subjects constitute a small
portion of the whole study, which includes active smokers.
Cases consist of patients diagnosed with primary lung cancer, exclusive of
bronchioalveolar carcinoma, from participating hospitals in several Louisiana parishes (counties),
predominantly in the southern part of the state. A total of 302 female and 1,036 male cases and
an equal number of controls are included in the whole study. Controls were selected from other
patients, excluding those diagnosed with emphysema, chronic bronchitis or obstructive pulmonary
diseases, or certain cancers (laryngeal, esophageal, oral cavity, and bladder). They were matched
to cases on hospital, age (±5 years), sex, and race. Information about active and passive smoking
was obtained by interview (presumably face to face and unblinded), with responses obtained from
next-of-kin in 24% of cases and 11% of controls; no information on refusals is provided. ETS
subjects were identified by exclusion of individuals who had ever smoked or had never been
married, which eliminated 279 female and 1,026 male cases. Removal of subjects with no spousal
smoking data eliminated one additional female and two male cases, leaving 22 female and 8 male
cases. Similarly, a total of 1,080 men and women were excluded from controls. No demographic
comparisons are given, either for the whole study or for the ETS subjects alone, nor is the number
of proxy responses provided for the ETS subjects. Histological confirmation was obtained for
97% of cases in the whole study, including ever-smokers.
"ETS exposed" is used in two ways, depending on the analysis given: (1) the spouse has
smoked at least 1 pack-year of cigarettes or (2) the spouse currently smokes. Units of exposure
are pack-years and current consumption is in cigarettes per day for (1) and (2), respectively. ETS
exposure in childhood means that at least one parent smoked during most of the subject's
childhood. Types of tobacco smoking other than cigarettes (e.g., cigars and pipes) are referenced
indirectly in regard to interview questions but are not included in the data analysis. Other sources
of exposure, either at home or in the workplace, are not considered. Never-married women are
excluded from ETS analysis, but no information is given on the number of nonsmoking widows
A-17
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
and divorcees and how they were handled with regard to ETS exposure. Adenocarcinoma
accounts for 54% of lung cancers in nonsmoking women, compared to 22% in women who actively
smoke. No further histological breakdowns are provided.
For the main analysis of spousal smoking, exposure constitutes one or more pack-years of
spousal cigarette consumption. ETS-exposed subjects include 14 (61) of 22 (133) female cases
(controls) and 2 (26) of 8 (180) male cases (controls). These data yield a crude odds ratio of 2.07
(95% C.I. = 0.81-5.25) for females (confidence interval was calculated by reviewers). Among
females, stratification by 0, 1 to 40, and 41 or more pack-years of exposure yields odds ratios of
1.0, 1.18, and 3.52, respectively, with the highest exposure category being statistically significant
at p < 0.05. No adjusted results are presented. It is, however, reported that analyses based on
current daily spousal cigarette consumption produced very similar results to the pack-year
analyses. In addition, it is reported that neither exclusion of proxy interview data nor restriction
to same-race subjects significantly alters the results. Analysis of parental smoking during
childhood embraces the combined population of smokers and nonsmokers, adjusting for smoking
status by logistic regression. Maternal smoking is associated with significantly increased estimated
risk of lung cancer (OR - 1.38, p < 0.05) but paternal smoking is not (OR = 0.83). No association
was noted among nonsmokers alone, but the authors note that small numbers preclude adequate
analysis of this group.
A.7.3. Comments
The study entails a major multicentric effort to assemble hospital-, age-, race-, and sex-
matched lung cancer cases and controls from Louisiana hospitals. Its use of trained local
interviewers familiar with the region's culture increases the probability of obtaining accurate
interview data for the nearly 3,000 subjects involved. Exclusion of active smokers to assess ETS
exposure, however, exacts a toll on the study's power and validity. Because the initial matching of
cases and controls did not include smoking status, the ETS subjects are unmatched in the analyses
of spousal and parental smoking. This potential problem is not addressed by the authors. The
lack of any demographic information on cases and controls leaves the comparability of these
groups uncertain.
The potential problem of misdiagnosis of primary lung cancer is minimized by the high
rate (97%) of histological case confirmations. Eligibility criteria for controls were intended to
exclude smoking-related diseases. Some 15% of the controls had cardiovascular disease, however,
which has been associated with both active and passive smoking. The authors also speculate that
the inclusion of adenocarcinoma, reportedly less smoking-associated than other lung cancers, may
A-18
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
have diluted the significance of their results, but they do not present analyses using their
extensive histological data to assess this question.
Restriction of the spousal smoking analysis to ever-married individuals eliminates
potential confounding by differences between lifelong single and married individuals.
Stratification by gender controls for any sex-related differences. Both race and proxy interview
were reported to have no effect on the spousal smoking results, and the spousal smoking
association was still observed after division of women into over and under 60 years of age. A
small number of nonsmoking ever-married cases (8 males and 22 females for this study) hampers
efforts to control statistically for confounders; nonetheless, direct adjustment for age and race is
needed. There are other potential confounders not controlled for, such as socioeconomic status,
diet, and other sources of smoke exposure.
It is concluded that females married to heavy smokers have an increased.risk of lung
cancer. A significant increase in risk for nonsmokers was found from maternal but not from
paternal smoking in childhood. The results for childhood exposure, however, use statistical
methods to adjust for the presence of active smokers instead of removing the data for them prior
to analysis. This gives the appearance of increasing an otherwise small sample size (the ETS
subjects alone) to attain significance, at the risk of biased results. The potential for bias in all of
the analyses, which could be in either direction and may or may not be of consequence, needs to
be kept in mind when using this study's results.
A.8. FONT
A.8.1. Author's Abstract
"The association between exposure to environmental tobacco smoke (ETS) and lung cancer
in female lifetime never-smokers was evaluated using data collected during the first three years of
an on-going case-control study. This large, multi-center, population-based study was designed to
minimize some of the methodological problems which have been of concern in previous studies of
ETS and lung cancer. Both a cancer control group and a population control group were selected in
order to evaluate recall bias. A uniform histopathologic review of diagnostic material was
conducted for case confirmation and detailed classification. Biochemical determination of current
exposure to tobacco and screening of multiple sources of information to determine lifetime nonuse
were employed to minimize misclassification of smokers as nonsmokers.
A 30% increased risk of lung cancer was associated with exposure to ETS from spouse, and
a 50% increase was observed for adenocarcinonia of the lung. A statistically significant positive
trend in risk was observed as pack-years of exposure from spouse increased, reaching a relative
A-19
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
risk of 1.7 for pulmonary adenocarcinoma with exposures of 80 or more pack-years. The
predominant cell type of the reviewed, eligible lung cancer cases was adenocarcinoma (78%).
Results were very similar when cases were compared to each control group and when separate
analyses were conducted for surrogate and personal respondents. Other adult-life exposures in
household, occupational, and social settings were each associated with a 40% to 60% increased risk
of adenocarcinoma of the lung. No association was found between risk of any type of lung cancer
and childhood exposures from father, mother or other household members."
A.8.2. Study Description
This study was initiated in 1985 in five major U.S. metropolitan areas to investigate the
association between exposure to ETS and lung cancer in female lifetime never-smokers. The
study was designed specifically to address this issue and includes only never-smokers. The results
reviewed are from an interim report, with the completed study expected to encompass an
additional 2 years of case accrual.
Patients were English-, Spanish-, or Chinese-speaking female residents 20 to 79 years of
age who have never used tobacco, have no prior history of malignancy, and have
histopathologically confirmed primary lung cancer. The lung cancers were originally diagnosed at
participating hospitals in Atlanta, Houston, Los Angeles, New Orleans, and the San Francisco Bay
area, between December 1, 1985, and December 31, 1988. Two control groups were assembled,
one from colon cancer patients and the other from the general population, with the same general
eligibility requirements as cases. The population control group, consisting of women selected
from the general population by random digit dialing and by sampling from Health Care Financing
Administration files, were frequency matched on age (< 50, 50-59, 60-69, 70+), with two controls
per case. The colon cancer controls were frequency matched to cases by 10-year age groups and
by race. The lung cancer group consists of incident cases, but there is no indication whether
density or cumulative sampling was employed for either control group. Exposure data were
collected in face-to-face, apparently unblinded, interviews.
Extensive efforts were made to include only never-smokers. For cases and colon cancer
controls, medical records were reviewed for tobacco use and physicians were contacted as
necessary. Eligible cases not previously excluded and all population controls were contacted by
telephone to screen for prior use of tobacco (no more than 100 cigarettes smoked or use of any
tobacco in any form for more than 6 months). Urinary cotinine was bioassayed to eliminate any
misreported current smokers.
A-20
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
A total of 514 eligible cases were identified, of which 83 were not interviewed for
unspecified reasons and 2 had urinary cotinine levels consistent with active smoking. Independent
histopathologic review by a pulmonary pathologist was performed for 84% (359/429) of the lung
cancer cases, resulting in nine exclusions. Only the remaining 420 cases are included in the study.
Colon cancers were not reviewed. Of 489 (1,105) eligible colon cancer (population) controls, 131
(311) were not interviewed and 7 (14) were excluded for high urinary cotinine. Proxies were
interviewed for 143 (34%) of the lung cancer cases and 35 (10%) of the colon cancer controls,
whereas no proxies were used for the population controls.
Cases and the two control groups all have similar age distributions, with the majority of
subjects between 60 and 79 (73%, 74%, and 74% of the cases, colon, and population groups,
respectively). The proportion of whites is similar across all groups (63-69%), but the control
groups contain a somewhat higher proportion of blacks and lower proportion of other minorities,
and a little higher percentage of high school graduates (76% and 79% vs. 68%). Cases and controls
are comparable by metropolitan size of adulthood and childhood residences and also by annual
income.
Four sources of adult ETS exposure are assessed: smoking by (1) spouse(s) and (2) other
household members while living with the subject, and reported exposure to ETS in (3)
occupational and (4) social settings. Three sources of possible exposure in childhood (up to 18
years of age) are considered: smoking by (1) father, (2) mother, or (3) other household member(s)
while living in the subject's home for at least 6 months. Subjects are characterized as ever- versus
never-exposed with a subahalysis by tobacco type (cigarette, pipe, or cigar). Years of exposure
are also tabulated. In addition, cigarettes per day for spouse and for other household sources and
pack-years for spouse(s) are calculated. No checks on exposure (aside from the cotinine
screening) are reported.
Adenocarcinoma is the dominant type of lung cancer among study subjects, representing
76% (311/409) of all cases included in the study (with the exception of 11 cases with "review
pending") and also 78% (281/359) of all independently confirmed primary bronchogenic
carcinomas among those cases. Other cell types include 12% (48/409) large cell, 7% (27/409)
squamous cell, 3% (14/409) small cell, and 2% (9/409) other cancers. No data on airway proximity
are provided.
The final study population (for this interim report) consists of 420 lung cancer cases, 351
colon cancer controls, and 780 population controls. Exposure to spousal smoking from all types of
tobacco is reported for 294 cases, 231 colon cancer controls, and 492 population controls, yielding
similar odds ratios (adjusted for age, race, area, income, and education) of 1.28 (95% C.I. = 0.93-
A-21
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
1.75) and 1.29 (0.99-1.69) using the respective control groups. Elevated but statistically
nonsignificant observed risks are also observed when cigarette, cigar, and pipe exposure are
assessed separately, with either control group. Restriction of analyses to the 281 independently
reviewed adenocarcinomas results in stronger associations, with adjusted odds ratios of 1.44 (95%
C.I. «• 1.01-2.05) and 1.47 (1.08-2.01) for all types of tobacco, and increased odds ratios for each
type of tobacco as well.
Odds ratios were also calculated for ETS exposure from cigarette smoking alone, with the
two control groups combined (the individual results using each control group are entirely
consistent). For all lung cancer types combined, the adjusted odds ratios are 1.21 (0.96-1.54) for
spousal smoking, 1.23 (0.97-1.56) for other household members, 1.34 (1.03-1.73) for occupational
environments, and 1.58 (1.22-2.04) for social exposure, the last two of which are significant (p <
0.05 and 0.01, respectively). The corresponding odds ratios for adenocarcinoma cases alone
continue to be uniformly higher: 1.38 (95% C.I. = 1.04-1.82), 1.39 (1.05-1.82), 1.44 (1.06-1.97),
and 1.60 (1.19-2.14). The odds ratio tends to increase over years of exposure for all carcinomas
combined and for adenocarcinoma alone, although not monotonically (without downturns). The
tests for upward trend are all significant or suggestive, with p-values ranging from < 0.001 to 0.07
(these p-values are half those reported, which apply to a trend in either direction). Finally, for
spousal smoking measured in pack-years, the upward trend is significant for adenocarcinoma
alone and for all lung cancers together (p < 0.005 and 0.04, respectively).
The authors interpret their findings as evidence of a causal relationship between ETS
exposure in adulthood and lung cancer in never-smoking women. In contrast to adulthood, ETS
exposure during childhood shows no association with lung cancer, for either all cell types
combined or adenocarcinoma alone. Adjusted odds ratios tend to be slightly (but not
significantly) below unity for all exposure sources.
A.S.3. Comments
This study is much larger than any other ETS case-control study. Over 400
never-smoking female lung cancer cases were enrolled in just over 3 years, in contrast to the 25 to
75 cases typical of most studies, and two control groups were formed, totaling over 1,200 subjects.
Additionally, the cases and controls are drawn from five widely dispersed metropolitan centers in
the United States, representing a population of approximately 18.5 million people, about 8% of the
U.S. population. This characteristic increases the generalizability of the study and diminishes the
potential for bias related to locale.
A-22
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Extensive efforts were made to achieve precision and validity, in evidence throughout the
study. Cases and controls are highly comparable. They are frequency matched on age and, for
colon cancer controls, on race as well. The distributions of other demographic variables—annual
income, childhood residence, and adult residence—are quite similar between cases and both control
groups. The control groups contain a little higher (lower) proportion of blacks (Asians and
Hispanics) and a higher percentage of high school graduates. These differences, however, should
not have influenced the reported associations because all odds ratios are adjusted for race and
education.
The use of incident cases reduces the potential for selection bias, and the implementation
of two control groups allowed for assessment of potential bias from comparison with cancer
patients or the general population alone. The similarity of results obtained from the two control
groups suggests little bias from choice of controls.
The use of a multistep procedure to eliminate inclusion of former or current smokers
reduces the potential for smoker misclassification as a source of upward bias. As a further
safeguard, urinary cotinine was bioassayed for all consenting persons to exclude those likely to be
current smokers. This is the only published study we are aware of to implement this precaution.
Attention to histopathology is also very thorough. Inclusion of only historically diagnosed
primary carcinoma reduces the likelihood of diagnostic error, which is further reduced by the use
of independent histopathologic review of most cases by a single pulmonary pathologist. The
study's histopathologic findings bring out two interesting points. First, comparison of cell type
diagnoses between hospital and independent reviewers revealed poor concordance for large (56%)
and squamous (67%) cell carcinomas, indicating that cell-type-specific analyses for these cancers
may be misleading, particularly if all diagnoses are not made by the same pathologist. The
histopathologic review also resulted in a net increase of adenocarcinomas from 244 to 281, 78% of
the total, a higher proportion than in most but not all other studies. The statistical results were
stronger when limited to cases of adenocarcinoma alone.
Exposure information was obtained in the most reliable way, by face-to-face interviews
with each interviewer trained and fluent in the subject's primary language. Information for a
substantial proportion of lung cancer cases (34%) was obtained from proxy respondents, but fewer
proxies were required for colon cancer controls (10%), and none were used for population
controls. The use of proxy respondents raises the possibility of information bias, but their
exclusion reportedly did not alter the study's findings. The apparent lack of blinding also raises
the possibility of interviewer bias, but it is unlikely that such bias (or recall bias, for that matter)
A-23 .
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
would restrict its effect to adenocarcinoma. Also, the same relationships hold whether the colon
cancer or population controls are used.
Particular attention is paid to all sources of ETS exposure, which is more informative than
addressing only spousal smoking, with four sources in adulthood and three in childhood evaluated
both individually and in combination. Additionally, subjects are counted as exposed to the ETS
of a spouse or other household smoker only while living with the source, giving a more accurate
account of exposure than simply determining whether a spouse or household member ever
smoked. Consequently, the measures of ETS exposure are more specific by source, and probably
more accurate, than in most studies. This reduces bias toward unity in the odds ratio arising from
poor distinction between exposed and unexposed subjects. Still, further accuracy might have been
achieved by stipulating that smoking must occur in the subject's household or presence, but this is
a minor point.
Most of the standard potential confounders—age, race, geographic area, income, and
education—are adjusted for in all" analyses and thus can be ruled out as sources of the observed
results. Although information on diet, occupational exposures, and "other exposures of interest"
were collected, these factors are not addressed in this interim report. Thorough treatment of the
issue of potential confounding by these factors will presumably be undertaken after subject
accrual is finished and published in the completed study.
To summarize, this study was designed specifically and solely to address the topic of ETS
as a potential lung cancer risk to nonsmoking women. Several issues were given special attention,
such as the potential misclassification of smoking status, histopathologic specificity, recall bias,
source of ETS exposure, and potential confounders and other risk factors. Histopathologic
specificity has not been convincingly demonstrated in prior studies, and the meaning of "exposed
to ETS" has differed widely between studies, even those addressing spousal smoking only. The
remaining issues are largely related to controlling potential sources of bias and confounding to
enhance validity. The qualitative rigor and completeness of detail in this study is impressive. In
addition, it is quite large, which increases precision of estimates and power to detect an
association, if it exists. Use of dietary, occupational, and other exposure data in the analyses,
along with an additional 2 years of subject accrual, will make the completed study for which this
constitutes an interim report even more valuable. As it stands, however, this study is already the
largest and most useful case-control study available. Its high quality and the reasonable
consistency of the evidence across sources of ETS exposure strongly support an increase in lung
cancer incidence associated with passive smoking.
A-24
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
A.9. GAO
A.9.1. Author's Abstract
"A case-control study involving interviews with 672 female lung cancer patients and 735
population-based controls was conducted to investigate the high rates of lung cancer, notably
adenocarcinoma, among women in Shanghai. Cigarette smoking was a strong risk factor, but
accounted for only about one-fourth of all newly diagnosed cases of lung cancer. Most patients,
particularly with adenocarcinoma, were life-long non-smokers. The risks of lung cancer were
higher among women reporting tuberculosis and other pre-existing lung diseases. Hormonal
factors were suggested by an increased risk associated with late menopause and by a gradient in
the risk of adenocarcinoma with decreasing menstrual cycle length, with a 3-fold excess among
women who had shorter cycles. Perhaps most intriguing were associations found between lung
cancer and measures of exposure to cooking oil vapors. Risks increased with the number of meals
cooked by either stir frying, deep frying or boiling; with the frequency of smokiness during
cooking; and with the frequency of eye irritation during cooking. Use of rapeseed oil, whose
volatiles following high-temperature cooking may be mutagenic, was also reported more often by
the cancer patients. The findings thus confirm that factors other than smoking are responsible for
the high risk of lung cancer among Chinese women and provide clues for further research,
including the assessment of cooking practices."
A.9.2. Study Description
This study was undertaken in Shanghai, China, during 1984-86 to explore reasons for the
high rates of lung cancer among women in Shanghai. Potential etiologic factors associated with
the high occurrence of adenocarcinoma among females in a population where few women smoke
cigarettes is of particular interest. Several potential risk factors, in addition to exposure to ETS,
are investigated. These are included in the abstract above. Smokers are included in the study as
well as nonsmokers.
A special reporting system for lung cancer linked with the area's medical facilities was set
up for the study period, integrated with the Shanghai Cancer Registry. Incident cases of lung
cancer occurring among 35- to 69-year-old female residents of urban Shanghai from February
1984 to February 1986 were interviewed by trained study personnel. Controls were women
selected from residents of the urban Shanghai community by stratified random sampling designed
to mimic the age distribution of Registry-reported lung cancer cases during 1980-81. It is not
clear whether cumulative or density sampling was employed.
A-25
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Face-to-face interviews were conducted with 672 cases and 735 controls. No cases refused
to be interviewed, but 93 died before interview and were therefore excluded; it is not mentioned
whether there were any refusals among potential controls. Nonsmokers composed 436 of the cases
and 605 of the controls. In the total subject population, distribution of age, education, and
marital status between cases and controls is described as similar, except for a larger proportion of
controls (32% vs. 20%) in the oldest age group (65-69 years). The age distribution in the ETS
population alone is not described.
ETS exposure is based on living with a smoker. For general exposure in childhood or
adulthood, exposed subjects are those who ever lived with a smoker. For spousal smoking alone,
however, women are ETS exposed only if they lived with a smoking husband for at least 20 years.
General ETS exposure sources include all household members but not coworkers. Verification of
exposure data was not mentioned. Based on the reported exposure criteria, widows and divorcees
would have been included in the spousal smoking data set, whereas never-married women would
have been excluded.
For ETS subjects, 246 (375) cases (controls) from the total of 672 (735) cases (controls) are
included in Table II of the article that lists the number of cases and controls by number of years
lived with a smoking husband. Presumably the 190 cases and 230 controls not included in the
table are unmarried (or never-married) and do not include women married and living with a
nonsmoker; no explanation is provided in the article.
Among nonsmoking women included in Table II, 189 out of 246 cases and 276 out of 375
controls had lived with a smoking husband for at least 20 years. These subjects were divided into
exposure categories of 20 to 29, 30 to 39, and 40 or more years for comparison with the
"unexposed" (< 20 years spousal smoking) subjects. The authors present no unadjusted analyses,
but calculations from their raw data yield an overall odds ratio of 1.2 and stratum-specific odds
ratios of 1.2, 1.3, and 1.1 for 20 to 29, 30 to 39, and 40 or more years of exposure, respectively.
Age and education-adjusted odds ratios increase with the number of years exposed: 1.1 (95%
C.I. - 0.7-1.8) for 20 to 29 years, 1.3 (0.8-2.1) for 30 to 39 years, and 1.7 (1.0-2.9) for 40 or more
years. The authors report an odds ratio of 2.9 (1.0-8.9) for squamous and oat cell cancer for 40
years of exposure or less but present no other type-specific results.
Information on cell type is available for the 542 (81%) study cases diagnosed by histology
or cytology; the rest of the cases were diagnosed by radiological or other means. Diagnostic
evidence was reviewed by a team of pathologists and clinicians. For the lung cancer cases
histologically typed, adenocarcinoma (61%) greatly predominates, followed by squamous (22%),
A-26
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
small cell (6%), and other (11%) types. No breakdowns of tumor type are provided for the ETS
group. ,
The authors conclude that ETS may account for some, but probably few, of the cancers
among nonsmokers, because there was little or no association with ever having lived with a
smoker. Among nonsmoking women married to smokers, however, there was an upward trend in
risk associated with increasing years of exposure. This latter finding is consistent with reports in
other parts of the world. Little evidence was found to implicate the type of fuels used for
cooking in lung cancer risk; occupational factors did not appear to be important, nor did familial
tendency to lung cancer. Our data suggest, however, that prior lung diseases, hormonal factors,
and cooking practices may be involved. Most provocative is the association with cooking oil
volatiles, and further investigations are needed to evaluate their contribution to the high lung
cancer rates among Chinese women in various parts of the world.
A.9.3. Comments
The number of ETS subjects for analysis is relatively large. Unfortunately, the study is
unmatched, with no demographic breakdown of the cases and controls, either for the whole study
or for the ETS subjects alone. Controls were selected to make their age distribution similar to that
expected for cases in the whole study, but the similarity may not apply to ETS subjects alone.
Consequently, there is little basis for evaluating the comparability of cases and controls. Age and
education were adjusted for in the analyses, which has some compensatory value.
The use of direct interview with all subjects without reliance on proxies to gather exposure
information should enhance the validity of the exposure comparisons. On the other hand, the
possible use of unblinded interviewers could have biased results. In light of the lack of association
noted for passive smoke exposure as a child or adult, however, it is unlikely that such a bias
produced the observed association between spousal smoking and lung cancer. For evaluation of
spousal smoking, the reference group can hardly be classified as "unexposed" to spousal smoking
because it includes women who lived with a smoking husband for up to 20 years. The
investigators probably selected the cutoff level of exposure for their spousal smoking reference
group to balance the numbers in each exposure category, as a practical matter. The reference
group contains an undisclosed number of women who may have been exposed to spousal smoking
for many years, potentially creating a substantial bias toward the null hypothesis (no association
between ETS exposure and lung cancer). Consequently, the odds ratios may be biased downwards.
The relative comparison across years of spousal smoking, however, is not affected. An increasing
trend in the odds ratio was observed, but no statistical test for trend is cited. In a similar vein, it
A-27
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
appears that active smokers may have been included in the data analysis of overall ETS exposure.
That factor, in combination with the use of ever- versus never-exposed classifications without
regard to degree or duration of ETS exposure in the analyses, may have reduced the likelihood of
detecting any positive association that may exist.
The study appears to have focused on potential risk factors other than ETS. Unfortunately
the effects of these other factors on the ETS results were not explored, even though many of these
appeared to be stronger risk factors than passive smoking. Some potential confounding factors,
such as age and education, were adjusted for in all analyses. Control for education should in turn
produce a degree of adjustment for factors related to socioeconomic status (e.g., dwelling size and
quality of diet).
Overall, the study presents evidence of a mild duration-dependent association between
lung cancer and spousal smoking that skirts statistical significance. Several sources of
misclassification bias are possible, but most would tend to bias the odds ratio downward. The
study was not, however, specifically designed to evaluate the ETS-lung cancer hypothesis.
Information was collected and analyzed on a number of other potential risk factors, but none of
these besides age and education were considered as potential confounders. Coupled with other
limitations, this omission reduces the weight of the study's results with regard to ETS, although
they support an increase in lung cancer risk with spousal smoking.
A.10. GARF (Case-Control)
A.10.1. Author's Abstract
"In a case-control study in four hospitals from 1971 to 1981, 134 cases of lung cancer and
402 cases of colon-rectum cancer (the controls) were identified in nonsmoking women. All cases
and controls were confirmed by histologic review of slides, and nonsmoking status and exposures
were verified by interview. Odds ratios (OR) increased with increasing number of cigarettes
smoked by the husband, particularly for cigarettes smoked at home. The OR for women whose
husbands smoked 20 or more cigarettes at home was 2.11 (95% C.I. 1.13, 3.95). A logistic
regression analysis showed a significant positive trend of increasing risk with increasing exposure
to the husband's smoking at home, controlled for age, hospital, socioeconomic class, and year of
diagnosis. Comparison of women classified by number of hours exposed a day to smoke in the
last five years and in the last 25 years showed no increase in risk of lung cancer."
A-28
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
A.10.2. Study Description
This study was undertaken in New Jersey and Ohio to investigate the relationship of
involuntary smoking to primary lung cancer. All data were collected specifically for this study,
and only nonsmokers were included as subjects. Cases are the lifelong nonsmoking women
histologically diagnosed with primary lung cancer during 1971-81 in four participating New
Jersey and Ohio hospitals. Controls selected from patients with colorectal cancer were matched
3 to 1 to a case on hospital and age (± 5 years). Subjects were not restricted to incident cases, and
controls were apparently cumulatively sampled. Exposure data were obtained by blinded,
face-to-face interviews with subjects or their relatives.
A total of 1,175 female lung cancer cases were initially identified from medical records.
Exclusion of women found to be current or former smokers or not to have histologically verified
primary lung cancer eliminated 1,041 of the identified cases, leaving 134 ETS subjects.
Interviews were conducted with patient, spouse, or child in about 75% of the subject population,
whereas the rest were conducted with another relative. The age distributions of cases and controls
are nearly identical.
ETS exposure includes pipe and cigar use as well as cigarette smoking. Three sources of
passive smoking are considered, which will be referred to as follows: "exposure to husband's
smoke" means having a husband or other related cohabitant who smokes more than occasionally,
either (1) anyplace or (2) at home; "general exposure" applies to the smoke of others at home,
work, or otherwise who have smoked more than occasionally during the past (1)5 years or (2) 25
years; and "childhood exposure" refers to experiencing ETS from any source during childhood.
Husband's smoking is quantitated as cigarettes per day and years smoked; general exposure is
given as average hours per day; and childhood exposure is treated as a dichotomous variable. Only
57 percent of the cases were women living with a husband at the time of diagnosis. No checks on
exposure status are described, and no classification of subjects by marital status was implemented.
Adenocarcinoma (87) predominates among lung cancer cases, followed by large cell (21), small cell
and miscellaneous (15), and squamous cell cancer (11); no data on airway proximity are provided.
Ninety of 134 cases were exposed to husband's (or other relatives') smoking at home,
compared to 245 of 402 controls, giving a crude odds ratio of 1.31 (reported 95% C.I. = 0.99-1.73;
C.I. calculated by reviewers is 0.87-1.98). For husband's smoking of 20 or more cigarettes per
day, the highest exposure category, the odds ratio increases to 2.11 (1.13-3.95). Husband's
smoking averaged 11.5 cigarettes per day for the exposed subject. For husband's smoking
anyplace, 91 of 134 cases and 254 of 402 controls were exposed, giving a crude odds ratio of 1.23
(0.94-1.60). At the highest exposure category, 40 or more cigarettes per day, the odds ratio is 1.99
A-29
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
(1.13-3.50). Cigar and pipe smoking alone yields odds ratios of 1.17 and 1.13 for husband's
smoking at home and anyplace, respectively. There are statistically significant trends for both
husband's smoking at home and for smoking anyplace when measured by cigarettes per day, but
not when evaluated by number of years smoked. The odds ratio for ETS exposure from husband's
smoke, both total and at home, is calculated by source of interview respondent for the categories
of "self," "husband," "daughter or son," and "other." It is readily apparent that the excess risk is
attributable to "daughter or son," with some contribution from "other." None of the excess risk is
attributable to "self or "husband."
General smoke exposure also shows an association with lung cancer. Exposure over the
past 5 and past 25 years yield odds ratios of 1.28 (0.96-1.70) and 1.13 (0.60-2.14), respectively.
The odds ratios do not increase with increasing level of exposure, however, and none of the
associations is statistically significant. No association was found between childhood smoke
exposure and lung cancer (OR = 0.9, 0.74-1.12). When the odds ratio is calculated by source of
respondent, "other" and "self account for the excess risk when smoking for 5 years is the measure;
for 25 years of smoking, "other" and "daughter or son" account for the excess risk.
Stratification by cell type reveals that husband's smoking is much more strongly associated
with squamous cell (OR - 5.00, both for smoking at home and anyplace) than adenocarcinoma
(corresponding ORs = 1.33 and 1.48); no association with other cell types was detected.
Stratification by age and socioeconomic status suggests little effect of these variables on the
results. The results, however, appear to be sensitive to whether the interview data were obtained
from the subject or a surrogate (offspring, relative, etc.), as noted above.
A logistic regression analysis including adjustment for age, hospital, socioeconomic status,
and year of diagnosis was undertaken for passive smoking. Cigarettes per day of husband's at-
home smoking is significantly associated with lung cancer, with an estimated relative risk of 1.7 at
exposure of 20 cigarettes per day compared to none. In contrast, husband's smoking outside the
home is not significantly associated with lung cancer, although the estimated relative risk is 1.26
for 20 cigarettes per day. General smoke exposure is not significantly associated with lung cancer,
for either the 5 years or past 25 years of exposure. Adjustment for type of respondent reportedly
had no significant effect on the logistic regression results.
A.10.3. Comments
The abundance of nonsmoking cases (134) and controls (402) in this study relative to most
ETS studies gives it above-average statistical power. Comparability of cases and controls appears
good based on their very similar age distributions, matching on hospital and age, and restriction to
A-30
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
nonsmokers. But the lack of further demographic comparisons means that divergence on some
other factor(s) cannot be ruled out.
A major difficulty in this study, however, arises from the extensive use of proxy
respondents. Only 12% (16 of 134) of the case interviews were with the patient. In the stratified
analysis, it was found that the husband's smoking at home is positively associated with lung cancer
only when the smoking information is provided by a son or a daughter rather than by the patient
or her husband. This leads to two possibilities. Either the son or daughter claimed that the
patient's husband smoked when he actually did not, thereby shifting cases from the nonexposed to
exposed category and increasing the odds ratio, or the patient or her husband claimed that the
husband did not smoke when actually he did, thereby shifting cases from the exposed to
nonexposed category and depressing the odds ratio. In general it is thought more likely that true
smokers are misclassified as nonsmokers more often than true nonsmokers are misclassified as
smokers (see, for example, Lee, 1986, and Machlin et al., 1989). Also, Machlin indicates that
proxies tend to misclassify smokers no more often than smokers themselves do. Thus, it may be
that the son or daughter data are better than the self or husband data. On the other hand, the
difference among the reporting sources may be due only to chance because the results in JANE on
self or proxy reports are quite the opposite of those in this paper, with the proxy reports (in this
case including the spouse) leading to lower odds ratios than the self-reports.
Another possible problem with this study is the use of colon and rectal cancer cases as
controls on the theory that these diseases are not smoking related. A recent paper, Zahm (1991),
notes that associations have been found between smoking and these cancers. If these associations
carry over to passive smoking, they might bias the result either higher or lower.
In general, the detailed results from the stratified analysis in Table 6 of the paper exhibit
considerable variation, probably caused by chance. Hence, the overall results in Table 5 of the
article, where all the cases and controls are used, may be the most reliable. They indicate an odds
ratio of 1.31 (1.24 after adjustment for smoker misclassification bias in the body of this report)
for exposure to all types of husband's smoking at home.
The study's exposure assessment methodology is strengthened by the attempt to maintain
blinding by not informing interviewers of the study hypothesis or the subjects' disease status.
This is impractical in most studies, but given the use of controls who also have cancer and a high
proportion of proxy interviews, effective blinding of interviewers and subjects may have been
largely achieved here. Detailed data on smoke exposure at home as well as elsewhere, including
pipe and cigar smoking, were collected. Pipe and cigar smoking are often not considered in ETS
studies, thus constituting a potential source of exposure misclassification, and smoking at home
A-31
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
should be a more meaningful index of smoke exposure than total smoking. What the authors
termed "husband's smoking" actually includes smoking by related cohabitants as well. Presumably
this was done both to increase subject numbers (by not excluding unmarried women) and to
enhance detection of passive smoke exposure. However, it could cause some oversight with regard
to classification of ETS exposure (e.g., a widow, living with a nonsmoking sister, whose husband
had been a heavy smoker). Less understandable is the failure to include smoking by unrelated
cohabitants and the inclusion of single women living alone. Diagnostic misclassification is
unlikely given the histological verification of all cases and controls.
Both husband's at-home and total cohabitant smoking are associated with lung cancer, the
association being stronger for at-home smoking. Both exposures show a statistically significant
general increase in association with level of smoking, with substantial associations only at high
levels. The adjusted association for at-home cohabitant smoking is much stronger (OR = 1.7;
p » 0.03) than that for smoking outside the home (OR = 1.3; p = 0.13), a pattern consistent with
home smoke exposure rather than some other smoking-related factor as the basis of the observed
results. General ETS exposure, in contrast, was inconsistently related to lung cancer in the
unadjusted analyses, with a stronger association for exposure within the last 5 years than within
the last 25 (possibly attributable to better recall). No dose-response pattern is evident, however,
and no association was found in the adjusted analyses.
The adjusted analyses include age, hospital, socioeconomic status, and year of diagnosis in
a logistic regression model, along with the passive smoking variable. This adjustment did not
significantly reduce the association between husband's smoking at home and lung cancer observed
before the adjustment, but it did eliminate any association with general ETS exposure. Thus, the
results for husband's smoking at home are probably not attributable to confounding by age,
socioeconomic status, hospital, or temporal variables. Dietary factors, heating and cooking
practices, and family history of cancer were not considered as potential confounders; thus, an
effect by one or more of these factors cannot be ruled out.
The heavy reliance on proxy respondents and their uncertain impact on the analysis leaves
some uncertainty to interpretation. On the favorable side of this issue, the authors' attempt to
blind subjects and interviewers to the study hypothesis lessens the likelihood of potential bias
from proxy response, and no effect due to respondent type was found in the adjusted analyses.
Some of the exposure categories seem vague, but this would tend to reduce the magnitude of the
observed association rather than to give rise to one. In summary, this study is suggestive of a
dose-dependent association between smoking in the home and lung cancer, with reservations due
to the use of proxies.
A-32
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
A.11. GARF(Coh) ,
A.11.1. Author's Abstract
"Lung cancer mortality rates were computed for nonsmokers in the American Cancer
Society's prospective study for three 4-year periods from 1960 to 1972 and in the Dorn study of
veterans for three 5-year periods from 1954 to 1969. There was no evidence of any trend in these
rates by 5-year age groups or for the total groups. No time trend was observed in nonsmokers for
cancers of other selected sites except for a decrease in cancer of the uterus. Compared to
nonsmoking women married to nonsmoking husbands, nonsmokers married to smoking husbands
showed very little, if any, increased risk of lung cancer." .
A.11.2. Study Description
This study examines the role of passive smoking in lung cancer among married women in
the United States. It uses data collected in a large prospective study initiated by Cuyler Hammond
of the American Cancer Society (ACS) in 1959. The ACS's objective was to evaluate the
association between potential cancer risk factors and cancer mortality. Although data were
collected on the smoking status of women and their spouses at the start of the study, Hammond
thought the study data should not be used to estimate lung cancer death rates in relation to amount
of passive smoking by female never-smokers. Specifically, Hammond notes that the study was not
designed for that purpose, and no special information on the subject was obtained; information
was available on the smoking habits of the husbands of many of the married women in the study,
but not on the smoking habits of the former husbands of women who were widowed, divorced,
separated, or married for a second time. More important is his statement that women in America
at that time were not generally barred from public and social gatherings where men were smoking,
and working husbands who smoked generally did much if not most of their smoking away from
home (Hammond and Selikoff, 1981). Similar reservations are expressed by Garfinkel, who also
notes'that 13% of the women nonsmokers who died of lung cancer in the ACS study reported that
they were previously married and that the classification of their exposure to their husbands'
smoking may not be pertinent (Garfinkel, 1981, p. 1,065).
A total of 29 ACS divisions encompassing 25 states took part in the study; participating
counties were in turn selected by division leaders based on feasibility. Data collection was
undertaken by networks of volunteers set up within participating counties. Recruitment of
subjects and subsequent follow-up monitoring were undertaken by volunteers who were
instructed to enlist qualifying acquaintances. Subjects were restricted to persons more than 30
years of age whose household contained at least one person over 45 years of age. Illegal
A-33
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
immigrants and persons who were illiterate, institutionalized, or itinerant were excluded. Detailed
questionnaires were distributed to subjects and all members of their household over 35 years of
age. These questionnaires covered factors such as diet, alcohol consumption, and occupational
exposures as well as smoking habits, but they did not address passive smoke exposure. Volunteers
who recruited subjects were given responsibility for tracing the subject's vital statistics for the
next 6 years and contacting living subjects again in 1961, 1963, and 1965 to complete a
questionnaire on changes in smoking habits. Alternate researchers were appointed as necessary to
replace volunteers who moved or quit. Finally, death certificates were obtained for subjects
reported deceased; where death due to cancer was indicated, verification was sought from the
certifying physician. Although follow-up initially ceased with 1965, in 1972 an additional follow-
up was initiated in 26 of the original 29 ACS divisions and terminated in September 1972.
A.I 1.3. Comments
The passive smoking study being described (GARF[Coh]) was undertaken by assembling a
subcohort of married women who reported that they had never smoked and whose husbands
completed a questionnaire including smoking habits. This subcohort totaled 176,739 women out
of the 375,000 never-smoking women enlisted by the ACS in 1959. Women were divided into
three exposure categories based on their husband's smoking status—nonexposed for never-smokers,
and low (high) for current smokers of less (more) than 20. Wives of former cigarette smokers and
men who smoked cigars or pipes rather than cigarettes were excluded (Garfinkel, 1984);
presumably these had already been excluded from the reported total (176,739). Mortality rates
were computed by 5-year age intervals for unexposed women (i.e., wives of nonsmokers), from
which the expected number of deaths for exposed women was estimated under the hypothesis that
spousal smoking does not affect lung cancer mortality. The ratio of observed to expected deaths
in the exposed group provides an age-standardized mortality ratio. This mortality ratio is 1.27
(95% C.I. - 0.85-1.89) for spousal smoking of under 20 cigarettes per day (low exposure) and 1.10
(0.77-1.61) for over 20 cigarettes per day (high exposure).
In a separate analysis, women healthy at the start of follow-up were divided into groups
matched on age (5-year grouping), race, education, urban or rural residence, and occupational
exposure of husband to dust, fumes, or vapors. Each of these matched groups was then
subdivided into zero, low, and high exposure categories. The proportion of observed deaths in
each category was multiplied by the proportion of subjects in the smallest category of the matched
group relative to that category. This "adjusted" number of deaths was then summed across all
groups with a given exposure and compared with the corresponding value for the unexposed (zero
A-34
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
exposure) category to provide a mortality ratio. In addition, we conducted a Mantel-Haenszel
analysis of mortality using data supplied by Garfinkel that yielded results similar to the author's
analyses. Ages 35 to 39 and 70 to 79 were excluded due to insufficient numbers. After stratifying
by age and correcting for time under study, the calculated lung cancer risk was greater in subjects
whose husbands smoked, but the predicted risk at low exposure was greater than at high exposure.
It is notable, however, that the lower risk at higher exposure is entirely attributable to the 50- to
59-year-old age group; otherwise, predicted mortality would be equivalent at the low and high
exposure (see Table C-l of this report).
The original ACS cohort study was a massive undertaking. By using it as the basis of his
cohort, Garfinkel was able to assemble a very large number (over 170,000) of never-smoking
married women. A cohort of this magnitude attains a number of lung cancer cases ordinarily
feasible only by means of a large case-control study, while avoiding the attendant pitfalls of
potential recall and interviewer bias associated with case-control studies. There are several
important limitations, however, that make the results of this study difficult to interpret.
The ACS study was not designed to yield a representative sample of the general
population. The sample of women is older (all at least 35 years of age, two-thirds between 40 and
59 at start of follow-up), more educated (only 5.6% were limited to a grade school education), and
contains a much smaller proportion of ethnic minorities (only 6.8% nonwhite) than the general
population (Stellman and Garfinkel, 1986). Although not representative of the population as a
whole, the relative homogeneity of the subject population does reduce the potential for
complications of interpretation that differences in ethnic or socioeconomic factors or both may
pose, and it increases efficiency by not including subjects belonging to age groups unlikely to
experience significant mortality during follow-up. Overall, the study population's
unrepresentativeness strengthens rather than undermines the study's conclusions. It would have
been useful, however, to confirm that exclusion of greatly underrepresented groups, such as
nohwhites and persons with no formal education beyond the eighth grade, had no effect on the
results.
Because the data on smoking habits were collected prospectively, no information on
exposures prior to 1959 was obtained. Exposure history for the years before 1959 may be as
important as for the 12 years of follow-up, however, if lung cancer has a long latency period,
such as 20 years or so. Inclusion of persons whose exposure status may have changed markedly by
1959 could be a biasing influence. Neither were changes in exposure status during the follow-up
period considered, despite the availability of data on smoking habits in 1961, 1963, 1965, and
1972. In fairness to the author, keep in mind that our comments are directed at evaluation of the
A-35
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
study for its contribution to the issues of passive smoking and lung cancer, although the ACS
study was not designed to assess ETS exposure. The only data collected on ETS exposure are
based on the spouse's current smoking habits at initiation of the study. If the ACS study had been
directed at evaluation of health effects of ETS, these issues would likely have been taken into
consideration to sharpen the classification of subjects with respect to ETS exposure. Overall, the
likely consequence of these factors is to .reduce the sensitivity of the study to detect an association
between lung cancer and ETS exposure, but the potential for bias in the direction of a false
positive cannot be ruled out. For example, if wives of smokers are more likely to become active
smokers during follow-up than wives of nonsmokers, these changes in smoking status could bias
results toward finding a positive association with passive smoking. (Relevant to this particular
example, the authors state that "very few" subjects reported a change in their smoking status, but
provide no further details. Also, 12 or fewer years is a short exposure to produce lung cancer. It
is thus probable that any bias introduced by active smoking would be minor; furthermore, the fact
that a stronger association was observed for low than for high levels of spousal smoking argues
against a confounder associated with spousal smoking. Nevertheless, potential sources of bias may
be present that influence the study outcome in either direction.)
During 1959-65, confirmation of primary lung cancer diagnosis was obtained from
physicians for 78% of all cancer cases. Among 203 cases of lung cancer in nonsmoking women
diagnosed by death certificate, confirmation attempts on an unspecified number of these cases
found 34 misdiagnosed as primary lung cancer, whereas 10 primary lung cancers were discovered
among cancers diagnosed as nonlung on death certificates. Thus, it appears that only about 85%
of the death certificate diagnoses of primary lung cancer were accurate, while a small percentage
of primaries were misdiagnosed as cancers of other cites. No confirmation of diagnoses was
undertaken during the period after 1965 when nearly two-thirds (119 out of 182, according to
data supplied to reviewers by Garfinkel) of the lung cancer deaths in the ETS study population
were reported. In light of the misdiagnosis rates found for 1959-65, it is likely that a substantial
percentage of the study's reported primary lung cancers in cases actually arose in other sites,
whereas a substantial percentage of reported cancers of other sites actually arose in the lung. The
resultant errors in subject classification probably bias the results toward no association (i.e., a false
negative conclusion), // a positive association actually exists.
Loss of subjects to follow-up is another source of potential bias. A subsequent report on
the ACS cohort (Garfinkel, 1985) states that, whereas more than 98% of the original cohort was
successfully traced through 1965, over 10% (3 of 29) of the original ACS divisions declined to
participate in the 1971-72 follow-up effort. In the study, now under review, Garfinkel reports
A-36
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
successful follow-up of 98.4% through 1965 and 92.8% through 1972, apparently not considering
subjects in the division who declined to participate in the extended follow-up as losses. It thus
appears that, whereas more than 98% of the original cohort was successfully followed up through
1965, less than 90% of the cohort was targeted for follow-up through 1972, and losses for this
targeted group approached 7%. Such losses not only reduced the number of observed deaths—and,
hence, the study's power—but introduced the possibility that differential loss to follow-up could
have distorted the study's results. A greater proportion of losses among exposed subjects than
among unexposed could partially mask a true positive association, whereas greater loss among the
unexposed could potentially create a spurious association.
Aside from the issues above, the study controls for basic sources of potential confounding.
Subjects were all of the same gender and marital status, and age was controlled for in all analyses.
Analysis by groups matched on race, education, residence, and occupation, along with age,
produced nearly identical results as the analyses standardized by age alone, indicating no
confounding due to these and unlikely confounding due to other socioeconomic, occupational, or
geographic factors.
In summary, this study predicts a weak positive association between spousal smoking at
levels of 1 to 19 cigarettes per day and lung cancer, but only slight association at higher exposure
levels; neither association is statistically significant. The lack of apparent dose-response pattern
further undermines the association, but the confidence intervals of the point estimates for the
high and low exposure groups overlap so broadly that the existence of a dose-response relationship
cannot be ruled out entirely. Meaningful interpretation of the results for the issue of ETS
exposure and lung cancer, however, is limited. As the study's objectives were directed elsewhere,
the data collected on ETS exposure is limited to the status of spousal smoking at the start of the
study. Past history and future changes in status are not well addressed. There is ample indication
that death certificate diagnoses are not a reliable source for the selection and classification of
subjects. Although a second 6-year follow-up period was undertaken to increase the follow-up
period to 12 years, its success was limited by incomplete participation and, perhaps, by
organizational difficulties related to long-term reliance on volunteers (who may relocate, change
interests, lose contact with the subjects originally enlisted over an extended period, etc.). Even if
the follow-up were entirely successful, however, 12 years of follow-up without regard to
exposure experience is not a particularly long period to evaluate the lung cancer potential for ETS
because the latency period associated with active smoking may be on the order of 20 years.
Although the ACS study has been an important contribution to its main study objectives,
A-37
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
the lack of information and potential sources of bias for the issue of passive smoking and lung
cancer leave its assessment in question.
A.12. GENG
A. 12.1. Author's Abstract
Not included in source.
A.12.2. Study Description
This study was conducted in Tianjin, where China's highest incidence of female lung
cancer occurs, to illustrate the relationship between cigarette smoking and female lung cancer.
The study explores both active and passive smoking, so the analyses for passive smoking apply to a
subgroup of the larger subject population. The source of the study's subjects and the time over
which they accrued is not specified. Subjects resided in Tianjin for over 10 years. The source of
controls is not given, but they consist of females pair-matched with cases on race, age (±2 years),
marital status, and birthplace. It is unclear from the article whether cases were incident or
prevalent and how controls were obtained. A draft summary description of this study (Liang and
Geng, undated) from Liang indicates, however, that hospitalized cases (96) were matched with
inpatient controls and general population cases (61) were matched with neighborhood controls.
The source of the study's exposure data is not clearly stated, but the draft from Liang
indicates that all identified cases and controls were interviewed. No information on collection or
verification of smoking or other data is provided. The authors state that cases and controls do not
differ significantly in age, education, occupation, race, marital status, birthplace, or residence,
but this refers only to the total study population of 157 cases and 157 controls that includes active
smokers; the same similarity may not hold for the 54 cases and 93 controls used in the passive
smoking analysis. Tumor types are provided for 85% of the total case population, but not
specifically for the passive smoking subpopulation; adenocarcinomas (36.9%) predominate, being
about twice as common as squamous (22.3%) or small cell (19.7%) tumors. Although nearly 85% of
the total cases were diagnosed histologically or cytologically, it does not appear that verification of
diagnosis or primary status of tumor was undertaken by the authors, and no information on tumor
distribution is supplied.
A nonsmoker (which usually means never-smoker) is ETS exposed if the spouse smokes.
Presumably women not currently married are excluded from the analysis, although they could
have been included with some assumption made regarding their exposure status. Information on
dose and duration of exposure was collected but not used in the passive smoking analysis, and it is
A-38
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
not indicated if cigar or pipe smoke was included. ETS exposure from parents and colleagues is
reported to have been evaluated. The parental smoking referred to is apparently in adulthood, as
cohabitants in the home, but that is not made explicit. Exposure during childhood was not
specifically addressed.
Among the ETS subjects, 34 out of 54 cases and 41 out of 93 controls were exposed. This
yields a statistically significant crude odds ratio of 2.16 (95% C.I. = 1.03-4.53) for husband's
smoking. No analyses adjusted for age or other factors are reported. On a rather confusing note,
an odds ratio of 1.86 is cited twice later, but that value is inconsistent with the odds ratio of 2.16
from the raw data. Whether this is an error or the product of an unspecified adjustment by
conditional logistic regression, which the authors employ for other purposes throughout the paper,
is unknown. The odds ratio increases with the number of cigarettes smoked per day by the
husband and with the duration of the husband's smoking. The odds ratios for smoking rates of
1 to 9, 10 to 19, and 20 or more cigarettes per day are 1.4, 2.0, and 2.8, respectively. For 1 to 29,
20 to 39, and 40 or more years of exposure, the odds ratios are 1.5, 2.2, and 3.3, respectively. No
tests for trend are cited, and the relevant data are not given. Consideration of ETS exposure from
smoking by father, mother, or "colleagues" reportedly yielded no results that are "quite
significant." No further details are provided, and it is not clear whether these results consider past
smoking status or apply only to current status.
The authors conclude that active and passive smoking are the most important risk factors
for female lung cancer in Tianjin. They attribute 35% to 42% of lung cancer occurring in their
nonsmoking female population to passive smoking. Female lung cancer is also found to be
associated with other factors, such as occupational exposure, with an odds ratio of 3.1 (95%
C.I. - 1.58-6.02); history of lung disease, with an odds ratio of 2.12 (95% C.I. = 1.23-3.63); and
cooking with coal, where the odds ratio increases with the duration of exposure from 1.5 to 5.5
(see Table 8 of this reference).
A.12.3. Comments
The quality of this study is difficult to assess given the dearth of details supplied by the
authors. Certainly the number of nonsmoking cases and controls included is more substantial than
in some other studies, and the reported association between passive smoking and lung cancer is
statistically significant. Questions regarding the mechanics of data collection and analysis,
however, remain unanswered.
Exposure and other data were obtained from hospitalized subjects at bedside and from
others in their homes. The extent (if any) to which information was obtained from proxy
A-39
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
responses, or interviews were denied, is unspecified. No blinding was employed, but that may
have not been feasible. Despite the reported similarity of the demographic characteristics of the
total case and control populations, dissimilarity cannot be ruled out within the subgroup used for
ETS analyses. Although the whole study, including active smokers, is matched on several
variables, the matching need not apply to the ETS subjects alone.
Lack of validation of diagnostic and exposure information may have led to substantial
misclassification, although the fact that 85% of the lung cancer diagnoses were obtained via
histology or cytology suggests that diagnostic misclassification would not have been extreme.
Lack of consideration of former smoking status is a potential problem. Inclusion of former
smokers among the nonsmokers, in combination with a tendency for former smokers to marry
smokers, could produce an upward bias in the odds ratios.
Finally, although the crude odds ratio of 2.16 for passive smoking is statistically
significant, it does not take into account even the most basic potential confounder—age. For the
larger case-control population (including smokers), occupational exposure (OR = 3.1), history of
lung disease (OR - 2.64), and cooking with coal (OR = 1.54-5.56, rising with cumulative
exposure) are statistically significant risk factors that the authors claim have joint effects with
smoking, yet the ETS analysis is not adjusted for these likely confounders. The anomalous odds
ratio of 1.86 given later in the results may have been adjusted for age or other factors, but there is
no way to tell.
In summary, the study's results are consistent with the hypothesis that passive smoking
increases the risk of lung cancer, but they are not definitive. More detail regarding the mechanics
of the study is needed to assess its general validity. If warranted, a clearer and more complete
analysis of the study's data regarding passive smoking, including consideration of the information
on dose, duration, and potential confounders already available, would then be useful. For the
current evaluation of epidemiologic evidence on ETS exposure and lung cancer, too many
questions remain about the design and execution of the study to properly interpret the data and
assess the authors' conclusions.
A.13. HIRA(Coh)
(Note: Because of the many publications relating to this study, a different format of presentation
is used.)
This cohort study and a later case-control study based on it were undertaken to explore the
relationship of passive smoking and other factors with lung cancer in Japanese women. Subjects
A-40
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
and data used in this study were, however, drawn from a larger study that was not designed to
investigate passive smoking.
An exploratory study of mortality determinants targeting adults at least 40 years of age
inhabitating 29 health center districts in Japan was initiated in 1965. In autumn of 1965, more
than 90% of the target population was interviewed to ascertain the status of lifestyle factors that
might affect health (e.g., cigarette smoking, alcohol consumption, and occupation). Individuals,
including husbands and wives, were interviewed separately. Follow-up of the interviewees was
conducted using a combination of an annual census of residents and death certificates to monitor
mortality. Mortality, as determined by death certificate, was the outcome variable. Hirayama
used this study population to examine the potential effect of passive smoking on lung cancer
mortality. In 1981, he reported the results derived from the first 14 years of follow-up (through
1979) in the British Medical Journal.
A total of 142,857 women were interviewed in 1965, of whom 91,540 were nonsmokers
whose husbands had also been interviewed regarding smoking status. Using their husbands'
smoking status as a surrogate for exposure to ETS, Hirayama calculated lung cancer mortality rates
for comparison of women married to smokers with women married to nonsmokers; rates were also
calculated using various strata of spousal smoking intensity (number of cig./day), as well as age
and occupation. A total of 346 lung cancer deaths occurred in this cohort during the first 14 years
of follow-up.
After standardization for age and occupation, it was found that women whose husbands
smoked daily had a higher annual rate of lung cancer mortality than did women whose husbands
were nonsmokers or only "occasional" smokers. The rate increased with the level of smoking (e.g.,
8.7/100,000 annually for no or occasional smoking, 14.0 for smoking 1-19 cig./day, and 18.1 for
20+ cig./day). Higher rates and a dose-response pattern were observed in women married to
smokers after stratification on either husband's age or agricultural work status. Mortality due to
two diseases associated with active smoking, emphysema and asthma, was also higher in wives of
smokers and increased with exposure. Conversely, mortality due to two cancers not linked to
active smoking, cervical and stomach cancer, was no higher in wives of smokers. Consideration of
husbands' drinking habits had no significant impact on mortality for lung cancer or other diseases
mentioned above.
Further study results appeared in the October 3, 1981, issue of British Medical Journal.
Among other things, results were presented by husband's age in 10- instead of 20-year intervals
and for 10 occupational categories instead of 2. These tabulations revealed a statistically
significant overall association between husbands' smoking and lung cancer mortality with a dose-
A-41
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
response pattern (1.00 RR for nonsmokers plus former smokers, 1.44 RR for medium smokers,
and 1.85 RR for heavy smokers). Also of interest was a breakdown of lung cancer mortality and
smoking habits in greater detail for both husband and wife. Notably, nonsmoking husbands with
smoking wives showed a higher lung cancer mortality rate (RR = 2.94) than did those with
nonsmoking wives. Because nonsmoking husbands with smoking wives were rather rare, however,
the numbers in this stratum were low (only seven deaths); thus, the observed association was not
statistically significant.
In 1984, Hirayama published results of an additional 2 years of follow-up of his cohort in
Preventive Medicine. The same basic associations reported after 14 years of follow-up for spousal
smoking and lung cancer remained after 2 additional years of follow-up. Mortality rates
increased with increasing exposure after stratification by age of husband, occupation,
geographical area, and time period during study; a trend had been reported after stratification for
age of wife at start of study only for ages 40 to 49 and 50 to 59. It was also reported that the
elevation of lung cancer mortality in nonsmoking women married to smokers was significantly less
among women who consumed green-yellow vegetables daily (e.g., for spousal smoking of 20+
cig./day, the RRs for disease mortality were 1.63 and 2.38). No such pattern was observed for
ischemia heart disease. In addition, a statistically significant excess of para nasal sinus cancer in
nonsmoking wives of smokers had been observed, which showed an apparent dose-response
relationship across four smoking categories, culminating in an RR of 3.44 for spouses of smokers
of more than 20 cigarettes per day. That effect dwarfed those related to social class and dietary
factors that were also examined.
In 1988 Hirayama reported the results of a case-control study nested within his cohort in
Environmental Technology Letters. To explore the relationship of women's age at marriage, as
well as husbands' smoking status with lung cancer mortality, lung cancer cases occurring among
nonsmoking women in the cohort study were contrasted with stomach cancer cases as controls.
Including only women under 59 years of age at the start of the cohort, the study divided husbands'
smoking into three categories—none, 1 to 19, and 20 or more cigarettes per day. Age at marriage
was also trifurcated in 19 or fewer, 20 to 23, and 24 or more years. Apparently as a result of
exclusion of women over the age limit or because of missing data, only 115 cases and 423 controls
were ultimately compared out of the 200 lung cancers and 854 stomach cancers among the
nonsmoking female cohort. Adjusting for woman's age and husband's smoking category resulted
in odds ratios for lung cancer of 4.95, 1.76, and 1.41 for the respective age-at-marriage groups;
the first two of these odds ratios were statistically significant. An additional comparison found
A-42
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
that among lung cancer cases, the mean age at first marriage to a smoking husband was nearly 8
years less than the mean age at start of smoking for active smokers.
A greatly expanded nested study was presented in the following year (Hirayama, 1989).
The study was designed to explore the potential for confounding of the relationship between lung
cancer and spousal smoking by dietary habits. A "baseline" sample of 2,000 nonsmoking wives
with known spousal smoking habits, aged 40 to 69 at the start of the cohort study, was randomly
selected from the available cohort of 90,458 for comparison with the 194 lung cancer cases
occurring in equivalent subjects within the cohort. After determining that the age distributions of
the case and baseline groups were very similar within smoking categories, the combined
population was stratified on daily versus less-than-daily consumption for each of five food types
(green-yellow vegetables, fish, meat, milk, and soybean paste soup) and wives with smoking and
nonsmoking husbands were contrasted to assess differences in dietary habits. After adjustment
for wife's age and husband's occupation, only daily meat consumption was significantly more
common among wives of smoking husbands, and this was limited to smokers of 20 or more
cigarettes per day. Calculation of odds ratios for dietary habits resulted in a "significant" elevation
only in daily fish consumers (OR = 1.365, 90% C.I. = 1.05-1.77; Table IV). A nearly significant
lowering of the odds ratio was found in daily meat consumers.
Finally, odds ratios were calculated for lung cancer adjusted by wife's age, husband's
occupation, and each of the dietary habit categories in succession. A dose-response pattern was
observed between lung cancer and husband's smoking that persisted after adjustment for any of
the five dietary factors. Odds ratios for the five dietary habit categories ranged from 1.42 to 1.69
for former smokers and smokers of 1 to 19 cigarettes per day, and from 1.66 to 1.91 for smokers
of 20 or more cigarettes per day compared with nonsmoking husbands. The observed trend was
highly statistically significant, regardless of which factor was adjusted for in the calculation.
A.13.1. Chronology of Controversy
Publication of Hirayama's initial 14-year follow-up results in 1981 provoked a sizeable
volume of commentary in the scientific literature. Following the release of updated results in
1983-84, the study attracted little controversy until the latter part of the 1980's, when criticisms
were directed at the study by a number of authors. This process reached its culmination in
response to the EPA's release for external review of the document Health Effects of Passive
Smoking: Assessment of Lung Cancer in Adults and Respiratory Disorders in Children, which
placed considerable emphasis on Hirayama's results. An author-by-author, letter-by-letter
consideration of the arguments regarding Hirayama's work would be dauntingly duplicative and
A-43
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
tedious. Instead, the most-discussed concerns will be highlighted, followed by an overall
assessment of the study as it stands today.
Chronology of Selected Events Relevant to the Hiravama Cohort Study
Jan. 7, 1981 Results of cohort study are published in British Medical Journal (282:183-
185).
Oct. 3, 1981
March 3-5 &
July 10-15, 1983
Dec. 17, 1983
1984
1985
1987
1988
1988
1989
Comments and letters to the editor by Kornegay and Kastenbaum (of the
U.S. Tobacco Institute), Mantel, Harris, and DuMouchel, and MacDonald
appear re: Jan. 7 article in British Medical Journal, along with the author's
reply.
Hirayama presents updated results for his study cohort incorporating an
additional 2 years of follow-up (for a total of 16 years) to the International
Lung Cancer Update Conference in New Orleans and the 5th World
Conference on Smoking and Health in Winnipeg, Canada.
Updated results of the cohort study are published in Lancet.
Results presented in conference of July 1983, and in summary form in
Lancet later that year, are published in full in Preventive Medicine.^ In
addition, Hirayama now reports a statistically significant increase in brain
tumors with husbands' smoking. In a round-table discussion published in
the same journal, Lee proposes that misclassification of active smoking
status may have biased Hirayama's results.
Another publication of results for the 16-year follow-up appears in Tokai
Journal of Experimental Clinical Medicine.
Hirayama includes previously published study data in a book chapter (Aoki
et al., 1987).
Uberla and Ahlborn publish an article from the Proceedings of the Indoor
Ambient Air Quality Conference in London (which is essentially the same as
an earlier presentation at the 1987 Tokyo International Conference on
Indoor Air Quality) criticizing the Hirayama study on several grounds.
Their primary assertion is that correction for the cohort's age distribution
removes the apparent effect of spousal smoking.
Hirayama publishes the results of nested case-control study based on cohort
study data in Environmental Toxicology Letters. Estimated risk of lung
cancer is reported to increase with earlier age of marriage to smoker.
Layard and Viren publish a paper presented at the Conference on the
Present and Future of Indoor Air Quality in Belgium. Making their own
projections of expected deaths and estimating losses to follow-up, they
conclude that mortality rates were anomalously low and follow-up losses
unacceptably high in the Hirayama study.
A-44
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
1989
Hirayama publishes nested case-control results in Present and Future of
Indoor Air Quality. Positive association of husband's smoking and lung
cancer with dose-response pattern reported after adjustment for dietary
variables.
A.13.2. Some Major Critical Works
A basic point raised by MacDonald (1981) and others soon after publication of Hirayama's
initial results concerns the selection of the study's sample population. It appears that the 29 health
centers included in the study were selected on grounds of convenience rather than to provide a
randomly sampled, representative cross-section of the Japanese population. The resultant sample
may thus be unrepresentative of the Japanese population as a whole.
A convenience sample may still produce a fairly good cross-section of the population, and
Hirayama replied in 1981 that "the satisfactory representativeness [of the study population]
. . . with regard to demographic and social indices was confirmed after the survey." He did not,
however, provide supporting data. MacDonald (1981) contends that the six prefectures from
which the sample was drawn are relatively industry-heavy (which does not necessarily contradict
Hirayama's contention); data presented by Hirayama (1983) showed that 40,390 of the cohort's
wives were married to agricultural workers, 19,264 to industry workers, and 31,886 to "others,"
which would indicate some overrepresentation of agricultural areas. Women aged 70 or more are
clearly underrepresented, composing less than 1% of the study's 40-and-older nonsmoking female
population. Thus, the cohort may be unrepresentative to some degree, but how unrepresentative is
unclear.
The key problem arising from an unrepresentative sample is that it may limit
generalizability of results derived from that sample to the population as a whole. In lieu of good
reasons to think that the association between exposure and disease would be different in the study
population and the general population, however, the possibility of an unrepresentative sample
assumes less importance. And, as will be seen in the subsequent discussion of possible
confounders, similar patterns of association were observed in a number of demographic sub-
groups.
Misclassification may occur in any epidemiologic study. Most of the critical commentary
has focused on potential misclassification of exposure status. Because the study relies on
interview data to establish smoking status, misreporting by interviewees may affect accurate
classification of both wives and their husbands' smoking habits. It has been argued that women
are especially likely to misrepresent their smoking habits because smoking is considered less
socially acceptable for women than for men, particularly in Asian societies. Such misclassification
A-45
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
would tend to reduce the degree of association between passive smoke exposure and its effect(s) if
women in the "exposed" and "unexposed" groups were equally likely to misreport their own
smoking. One of the most prominent criticisms leveled at the Hirayama study postulates a
differential misclassification of smoking status in women. Peter Lee (Lehnert, 1984) raised the
argument that if women married to smokers are more likely to be (or to have been) smokers than
are women married to nonsmokers, and a given percentage of smoking women claim to be
nonsmokers, then purportedly nonsmoking wives of spousal smokers will include a higher
proportion of active smokers than wives of spousal nonsmokers. This will cause bias in the
direction of a positive association. Arguments over the probable size of this bias have occurred
with estimated elevations in risk ranging from a few percent to around 50%, depending on
assumptions regarding the extent of misreporting, the risk inherent in active smoking, and the
degree of marital concordance between smokers (Lehnert, 1984; Wald et al., 1986; Lee, 1987a
and b).
Uberla and Ahlborn (1987) raised a number of points regarding the Hirayama study,
including those previously mentioned. Citing the "severe selection bias by age," the authors report
that the increase in risk with spousal smoking disappears when this bias is corrected for. The
study population in fact contained a very small proportion of women aged 70 or more (only about
1%)—so small that the rates generated by nonsmoking married women aged 70 or more are too
unstable to provide meaningful results. But by taking the negative results observed in this tiny,
unstable stratum of the cohort and weighting them to "correct" for the underrepresentation of this
age group, the overall association is made to disappear. Such a "correction" is meaningless. In
addition, Hirayama (1990) has noted that the authors inappropriately adjusted to the total female
population rather than to the population of currently married females, and characterized the
adjustment as "neither of scientific significance nor of creative value."
The authors also essentially take Lee's approach to the differential misclassification
problem and claim that a modest differential misclassification "leads to risk ratios of around
unity." As seen previously, this argument is plausible but purely speculative—and, of course,
potential biases toward the null are ignored in this and other "corrections." The authors conclude
that "the null hypothesis ... is consistent with the Hirayama data in the same way as the
alternative." But unless one applies the aforesaid "corrections," the Hirayama data is, in fact, more
*
consistent with the hypothesis of association than with the null hypothesis.
Layard and Viren (1989) estimated "projected" mortality rates for a cohort with the age
and time distribution found in the Hirayama cohort by applying "standard demographic life table
procedures" to year and age-specific life table data from United Nations and Japanese sources.
A-46
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
They concluded that female all-cause and lung cancer reported rates were only 76% and 85%,
respectively, of projected values. In a separate analysis, the authors also "calculated the numbers
of person-years that would have been observed in the cohort if there had been 100% follow-up"
from the reported numbers of deaths. The assumptions used in this calculation are unstated. The
authors then estimated, based on the difference between their person-years for 100% follow-up
and the reported person-years, and an assumption that 8 years of observation were lost on average
for each person lost to follow-up over the 16-year course of the study, that approximately 10% of
the cohort was lost to follow-up. Dismissing other possible causes of their estimated mortality
deficits, Layard and Viren conclude that "it is possible that biases exist in the data which might
invalidate an observed relationship between exposure to ETS and mortality."
Of course, acceptance of Layard and Viren's conclusions must start with acceptance of the
validity of their assumptions and calculations, not all of which are stated explicitly. Beyond that,
their rejection of alternative explanations for the difference between projected and reported
deaths is not convincing. For example, random sampling variation and regional variations in
death rates are both dismissed because neither could produce an effect as large as that observed,
although the authors' figures indicate that in combination they could well account for a sizeable
portion of the difference. Likewise the effect of admitting only (initially) "healthy" people to the
cohort is dismissed based on the observation of "still very substantial cohort deficits in the last
years of the study" without specification of how substantial such deficits were and ignoring the
fact that a pattern in which all-cause mortality is most affected and cancer mortality least, as their
calculations showed, is the expected pattern for an effect of selection of healthy individuals.
Finally, to produce a spurious association, a bias must operate differently on the exposed (smoking
spouse) and unexposed (nonsmoking spouse) groups, and no evidence is provided that supports
such a pattern. Lacking such a pattern, the most likely effect of loss to follow-up is a reduction
in the observed associations due to missing mortality events. The effect of selecting an
abnormally healthy cohort would in a strict sense limit generalizability of conclusions but would
not in itself produce an exposure-effect association when none actually existed.
A.13.3. Critique and Assessment
Hirayama's cohort is drawn from a study population assembled to explore the associations
between a number of potential health-influencing factors determined via interview and
subsequent mortality. Thus, the study was not designed to investigate passive smoking and lung
cancer specifically. Most of the weaknesses attributable to Hirayama's study derive from this fact.
A-47
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
The only indicator of ETS available to Hirayama was self-reported smoking status at time
of baseline interview. Thus, misclassification of spousal smoking status is possible and change in
status over time, modifiers of exposure to spousal smoking, and other sources of ETS exposure
cannot be determined.
As previously seen, an overrepresentation of current and former active smokers claiming
to be nonsmokers among wives of tobacco smokers probably biases the association between spousal
smoking and lung cancer in reported nonsmokers upward. Even the leading proponent of this
argument, however, states that unless this bias is much stronger than it appears to be in U.S. and
Western populations, it could not account for the major part of the observed results (Lee, 1990).
Lack of information regarding amount of smoking actually done in the home and in the presence
of the spouse, room size and ventilation, and other exposure modifying factors must lead to
imprecision in the estimates of exposure via spousal smoking. This imprecision would make an
actual ETS-lung cancer association more difficult to detect. The fact that spousal smoking
exposure, even if precisely measured, is an imperfect surrogate for total ETS exposure because
workplace and ambient environmental sources are not assessed introduces a similar effect. Both of
these problems would thus introduce a bias toward the null, suggesting that the study's results are
an underestimate of the real association.
Mortality information was derived from death certificate linkage. It has been contended
that lung cancer is routinely overdiagnosed as a cause of death on death certificates, thus
undermining the study's credibility. But the resultant misclassification of cause of death would
presumably be nondifferential and thus bias results toward the null. To cause overestimation of
the association, a greater proportion of women in the spousal smoking groups than in the
nonsmoking group would have to be falsely diagnosed as having lung cancer. Because the study
cohort was made up of nonsmoking women, there would be little reason for such a pattern.
(Unless, of course, all such cases came from women who falsely reported their initial smoking
status or took up smoking in the course of the study and the misclassification/smoking habit
concordance hypothesized by Lee were actually strongly at work.)
No information is given regarding whether the same interviewers interviewed both
husbands and their wives. Thus, interviewers may not have been blind to spousal smoking
characteristics of interviewees. This is likely to have been of little importance, however, because
the outcome—lung cancer mortality—was measured prospectively, and thus did not occur for some
time after exposure had been assessed. If information bias was to some extent operant in the
interview, the most likely scenario would find women whose husbands smoked being probed more
strongly for admission of their own smoking than were women whose husbands did not smoke.
A-48
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
This would tend to reduce underreporting of active smoking in the "exposed" group relative to the
"unexposed" group. The result would be to lower the observed association between husbands'
smoking and lung cancer mortality.
Hirayama's cohort includes only married, reportedly nonsmoking women who were at least
40 years of age and "healthy" at the start of the study. In addition, almost all of these women were
under 70 years of age, and agricultural families composed a relatively large part of the cohort.
Thus, the cohort does not present a representative cross-section of the Japanese population as a
whole. Nevertheless, there is little obvious reason why a relationship between spousal smoking
and lung cancer mortality found in this cohort should be dismissed on the grounds that it is not
generalizable to the greater Japanese (or other) population. In fact, one could argue that by
studying a more homogeneous population in the cohort, the possibilities for bias due to
differences between exposed and unexposed groups are reduced.
The possibility that confounding by other risk factors brought about an observed
association must be considered in any study. For lung cancer, of course, smoking, gender, and age
are major risk determinants. Restriction of comparison groups to same-gender nonsmokers avoids
potential confounding by gender or smoking (but see misclassification discussion regarding
smoking status). Age is only partially restricted in the study design, so its consideration in the
analysis is essential. Hirayama chose to control for husband's age in analyzing the cohort study's
results. All observed associations persisted after such adjustment. Spousal ages should be closely
correlated, but direct adjustment using the subject's own age rather than the age of their spouse
would clearly be preferable. One such analysis was supplied (Hirayama, 1983b), and in it a
significant association between spousal smoking and lung cancer mortality persisted.
Furthermore, in analyzing the nested case-control studies, adjustment for wife's age was used
throughout, which produced findings that confirmed the results of the cohort study.
The potential role of confounding by other factors in the observed results has received
considerable emphasis. A correlation between smoking and lower socioeconomic status with
concomitant lifestyle and environmental differences could be expected. Among these differences,
particular attention has been paid to the possible effect of dietary factors, particularly low beta
carotene intake, and occupational exposures, both of which, some hold, should correlate with
spousal smoking and thus could bring about the observed association even if spousal smoking and
ETS exposure has no effect. Yet, neither stratification on daily green-yellow vegetable
consumption—the best available surrogate for beta carotene intake in the data—nor on agricultural
versus nonagricultural occupation of husband eliminated the association between spousal smoking
and lung cancer mortality in the cohort study. Similarly, adjustment for husband's occupation and
A-49
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
any of five dietary habit characteristics, along with wife's age, yielded similar results in the case-
control approach. Thus, neither of the major proposed confounders satisfactorily accounts for the
observed results.
Because the data set does not contain the necessary information to examine confounding
due to differences in cooking practices (such as stir-frying), this cannot be ruled out, although
such practices might be expected to covary with some of the dietary factors considered in the
analyses. Similarly, use of coal for cooking or heating cannot be directly assessed, though a degree
of covariance with dietary habits or occupation is likely.
Husband's drinking habits were only marginally associated with lung cancer risk; mortality
rates stratified by both drinking and smoking would have been more useful (and stratification by
wives own drinking habits would have been more useful still).
When lung cancer mortality among wives is stratified by wife's age (in 10-year
increments) and husband's smoking category, a clear dose-response pattern is seen only in the 40
to 49 and 50 to 59 age strata, whereas a decrease in mortality with spousal smoking is seen in the
70 and older stratum. Given that the latter stratum includes less than 1% of the cohort and very
few deaths, its rates are too unstable to have much confidence in. The dose-response pattern does
become weaker with ascending age strata, however, which has led to conclusions of inconsistency
with an ETS-lung cancer connection and presence of confounding. Hirayama has proposed that
age-related increases in spousal mortality, smoking cessation, and decreased time spent in
husband's proximity during the follow-up period may account for the observed pattern
(Hirayama, 1990). The proximity effect seems questionable, because retirement of older husbands
would eliminate time spent away from the house at work, but the other arguments are plausible.
Alternatively, older women recently married to smokers may be more likely to die from
competing causes of death that increase with age before passive-smoke cancer develops.
Remarriage, possibly to a spouse whose smoking habits differ from those of the former spouse,
would also increase with age and could lead to misclassification of (former) exposure with a bias
toward the null. (It is unfortunate that history of former spouses' smoking habits and recency of
marriage were apparently not obtained in the baseline interview or the aforementioned problems
could be readily addressed.) Temporal trends in some risk modifiers, such as dietary factors,
could also play a role.
Confounding cannot entirely be ruled out in certain instances. But the underlying
question that must be raised in this regard is the following: // the spousal smoking group contains
a disproportionate number of individuals with risk-elevating factors such as poor diet, lack of
exercise, low socioeconomic status, and occupational hazard exposure, and these factors are
A-50
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
sufficient to produce an increase in lung cancer mortality relative to the spousal nonsmoking
group, despite an absence of any real smoking effect, why does this multitude of risk factors result
in elevations of established smoking-related diseases only and no substantial elevation of risk of
other causes of mortality (except brain cancer, which encompasses relatively few deaths)?
In considering the study's results in broader terms, Hirayama's findings are consistent with
the hypothesis that exposure of nonsmoking women to passive smoke via spousal smoking
increases risk of lung cancer. The observed association is statistically significant. And the
persistence of the association after stratification on numerous variables, the observation of a
parallel association in nonsmoking husbands of smoking wives, the appearance of associations with
other smoking-related diseases, the existence of a dose-response pattern in most analyses of strata
containing adequate numbers, and the production of similar conclusions by either cohort or case-
control approaches argues against attribution of results purely to chance or confounding.
Possible inclusion of active smokers among "nonsmoking" spouses of smokers through
misclassification bias or differential change in smoking status during follow-up remains the
study's greatest weakness. This problem could have been addressed by follow-up interviews or
questionnaires coupled with verification of smoking status by alternative means in a subsample of
the cohort, and still could be. In addition, losses to follow-up and failure to use more
sophisticated survival analysis techniques are weaknesses that probably reduced the study's power.
Overall, the Hirayama study provides supportive, although not definitive, evidence that
ETS exposure increases lung cancer risk.
A.14. HOLE(Coh)
This prospective cohort study was undertaken in the towns of Paisley and Renfrew,
Scotland. The primary objective was to explore the relationship between passive smoking and
cardiorespiratory symptoms and mortality, including lung cancer. The towns were selected
because they are situated in an area with a high incidence of lung cancer. All persons residing in
these towns between 45 and 64 years of age, inclusive, were visited between 1972 and 1976. Each
was asked to complete a self-administered questionnaire and to visit a cardiorespiratory screening
center where further interviews were conducted; 80% (15,399 persons) responded.
Participating households in which at least two "apparently healthy" subjects lived were
included in the study, yielding a study population of 3,960 males and 4,037 females. Data on
smoking habits were obtained from the questionnaire and verified by interview at the screening
visit. Mortality among subjects was traced using the Scottish National Health Service Central
Register and General Register offices (for death certificate linkage), as well as the national cancer
A-51
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
registry system. Results for follow-up through 1982 were published in 1984 (Gillis et al., 1984).
The primary results reported here are for follow-up through 1985, published in 1989 (Hole et al.).
In addition, the results of unpublished data extending follow-up through December of 1988 are
reported (personal communication from Hole to A.J. Wells).
Smoking habits were divided into three categories: persons who have never smoked,
former smokers, and current smokers. In addition, the number of cigarettes smoked per day was
obtained for current smokers. Both pipe and cigar smokers were excluded from the group who
had never smoked. Never-smokers with former or current smokers as cohabitants in their
household were classified as passive smokers; otherwise never-smokers were classified as
"controls." This classification yielded 1,538 passive smokers and 917 controls for both sexes
combined. The corresponding numbers for females alone are 1,295 and 489.
The number of lung cancer deaths among females occurring in the cohort during the
follow-up period is only six, too small to be of statistical consequence. The unpublished data
extending follow-up through 1988 includes one additional female lung cancer death that occurred
subsequent to 1985. The crude relative risk is 2.27 (95% C.I. = 0.40-12.7), which is in the
direction of a positive association between ETS exposure and lung cancer. The extremely wide
confidence interval is the result of the small number of cancer deaths being compared and
indicates that the data could easily arise when the true value of the relative risk is almost any
value. After adjustment for age and social class, the relative risk is 1.99 (95% C.I. = 0.24-16.72).
Lung cancer incidence was somewhat higher than mortality (10 cases vs. 7 deaths), yielding an
adjusted relative risk of 1.39 (95% C.I. = 0.29-6.61). The relative risks for adjusted mortality
(5.30) and incidence (3.54) were higher in males than in females but were based on even fewer
cases (four deaths, six incident cases).
Although the observed association could easily occur by chance, it is a useful contribution
to the pool of evidence on lung cancer and passive smoking. Consequently, it is worth noting that
the observed associations are not likely to be attributable to confounding by other factors, because
they persisted after control not only for age and gender, but for social class, diastolic blood
pressure, serum cholesterol, and body mass index. Thus, differences in lifestyle or environmental
factors such as diet, housing, and employment between passive-smoking households and
nonsmoking households is an unlikely source of the results. Specific adjustment for potential
occupational exposures or radon were not carried out, but these variables would presumably
covary with social class to a great extent.
As for other sources of bias, interviewer bias can be discounted because subjects were
"apparently healthy" at interview and supplied smoking information before cardiovascular
A-52
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
screening, and the investigators did not begin determining the passive smoking status of subjects
until 1983 (for the first published study on this cohort). The extent of loss to follow-up is not
specified, so one cannot tell whether this was a potential source of problems. However, linkage
was carried out through two registries for general mortality and an additional registry specifically
designed for cancers. Diagnoses of cancer mortality from death certificates were checked against
cancer registry records for verification, thus reducing potential inaccuracies attendant on use of
death certificates.
Some data regarding misclassification were collected in an additional questionnaire
administered to a portion of the cohort at some unspecified point in the study. Among controls,
5% said that their household contained a smoker-presumably someone who had not met the
inclusion criteria (e.g., age 45-64) for the study. Thus, a small portion of the control group was
actually currently exposed, which would produce a slight bias toward the null. Differential
misclassification of smokers as never-smokers resulting from concordance of smoking habits
among cohabitants cannot be assessed or ruled out, despite the authors' suggestion that persons
cohabitating with smokers may be more likely to falsely claim to be smokers themselves,
providing a bias toward the null.
In summary, this study appears well-designed and executed, but the number of ETS
exposed subjects is small. Although its influence may be relatively small, there are no apparent
methodological problems that would limit its usefulness otherwise.
A.1S. HUMS
A.15.1. Author's Abstract
"As part of a population-based case-control study of lung cancer in New Mexico, we have
collected data on spouses' tobacco smoking habits and on-the-job exposure to asbestos. The
present analyses include 609 cases and 781 controls with known passive and personal smoking
status, of whom 28 were lifelong nonsmokers with lung cancer. While no effect of spouse
cigarette smoking was found among current or former smokers, never smokers married to smokers
had about a two-fold increased risk of lung cancer. Lung cancer risk in never smokers also
increased with duration of exposure to a smoking spouse, but not with increasing number of
cigarettes smoked per day by the spouse. Our findings are consistent with previous reports of
elevated risk for lung cancer among never smokers living with a spouse who smokes cigarettes."
A-53
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
A.15.2. Study Description
This population-based case-control study was conducted through the New Mexico Tumor
Registry during 1980-84. The original purpose was to explain differing lung cancer occurrence in
Hispanic and non-Hispanic whites in New Mexico. The study questionnaire included questions on
spousal smoking and on indirect exposure to asbestos through a spouse's job. The current report
describes the risks associated with those exposures in smokers and nonsmokers. The data on ETS
exposure in nonsmokers is extracted from the larger study containing smokers.
For the whole study, a total of 724 eligible primary lung cancer patients were identified,
of which 641 were interviewed (89%). About half (48%) of the case interviews were conducted
with the subject. Information on the remaining subjects was obtained from surrogates, generally
the surviving spouse or a child. Cases were collected in two series, the first consisting of patients
with cancer incident in 1980-82. That group includes all cases less than 50 years of age and all
Hispanics, but not those exclusively. The number of cases was supplemented by a second series of
patients with cancer incident to a 1-year period beginning November 1983. Most of the controls
were selected by random telephone sampling, but some older subjects were randomly selected
from Medicare participants. The control group was frequency-matched to the cases for sex,
ethnicity, and 10-year age category, at a ratio of approximately 1.2 controls per case. Interviews
were held for 784 of the 944 eligible controls, with 98% of the responses from subjects.
The term "never-smoker" means not a cigarette smoker, where the latter is defined to be
someone who has smoked at least 6 months. The smoker classification is divided further into
current smokers and ex-smokers. The current smoker status includes smokers who have stopped
within 18 months prior to the interview; the ex-smoker status applies if smoking ceased more than
18 months prior to interview. Assuming that the minimum 6-month duration of smoking is
intended to apply to current and ex-smokers, never-smokers could have smoked previously for up
to 6 months.
An ETS-exposed subject is one ever-married to a spouse who smoked cigarettes,
regardless of the spouse's use of pipes or cigars. No information was obtained on exposure to ETS
from other sources, such as from other household smokers, in the workplace, or from parental
smoking during childhood. Measures of ETS exposure from spousal smoking include duration of
exposure (in years) and the average number of cigarettes smoked per day by the spouse. The ETS
subjects (never-smokers) include 20 (4) female (male) cases and 162 (130) controls (the article
reports eight male cases, the number used in much of the analyses, but four of those eight were
found to be smokers—personal communication from Humble). The age distribution for the female
cases (controls) is as follows: age less than 65, 5 (74); age 65 or more, 15 (88).
A-54
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
The odds ratio for the crude data on female never-smokers is 1.8 (90% C.I. = 0.6-5.4) for
spousal smoking of cigarettes only and 2.3 (90% C.I. = 0:9-6.6) when spousal smoking also includes
use of pipes and cigars. Based on mean cigarettes per day smoked by the spouse, the odds ratio of
1.2 at more than 20 cigarettes per day is somewhat lower than the odds ratio of 1.8 at the lower
rate, less than 20 cigarettes per day. For duration of exposure, the odds ratio increases from 1.6 at
less than 27 years to 2.1 at 27 or more years. It is reported that adjustment for age and ethnicity
did not alter these results from the crude analysis. A trend test is included for duration of spousal
smoking, but the sample sizes are too small to be meaningful. Application of logistic regression to
adjust for variables gives values very close to the odds ratios for the crude analyses shown above
for spousal smoking, for use of cigarettes only and also for combined use of cigarettes, cigars, and
pipes.
The distribution of cases by cell type is given, but only with males and females combined.
The ratios of ETS-exposed cases to the total, by cell type, are as follows: squamous cell (2/4),
small cell (1/1), adenocarcinoma (either 6/12, 7/12, or 8/12), and others (either 3/3, 2/3, or 1/3,
depending on correct ratio for adenocarcinoma).
The authors conclude that the results indicate increased risk from ETS exposure in never-
smokers but not in active smokers.
A.15.3. Comments
This study evaluates smokers as well as nonsmokers for increased risk of lung cancer from
spousal smoking. Not surprisingly, the number of smokers among the cases far outweighs the
number of nonsmokers. No evidence of added risk to smokers from passive smoking is found.
Such an evaluation, however, puts a great deal of faith in the exposure data and the power of
statistical methods to detect what may be only a marginal increase in risk from ETS on top of
active smoking.
Of more central concern to this review is the assessment of lung cancer from ETS exposure
in never-smokers. The ETS data are taken from a larger study, so the matching no longer applies,
although the adjustment for those variables (ethnicity and age category) in the analysis is
worthwhile. The article suggests that the high rate of proxy response for cases in the original
study (52%) may be due, at least in part, to inclusion of decedent cases. That topic is not
explicitly addressed, however, and controls were not matched to cases on vital status.
Never-smokers apparently may have a history of smoking, provided it is under 6 months'
duration. Whether any never-smokers actually have a short smoking history is not discussed, but
the never-smoker classification is less strict than in most studies.
A-55
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
The data are evaluated a number of different ways, consistently yielding an increased odds
ratio. The number of cases, however, is much too small (15 exposed, 5 unexposed) for the
observed odds ratio to be close to statistical significance. Although similar values of the odds
ratios might be observed in a larger study, that cannot be assumed. At most, the study outcome is
suggestive of a possible association between ETS exposure and lung cancer occurrence, in need of
additional support to be conclusive. Overall, this study is conducted well in many respects, but its
contribution to the pool of evidence for assessment of lung cancer and ETS exposure is tempered
by several weaknesses, as described above.
A.16. INOU
A.16.1. Author's Abstract
A case-control study on smoking and lung cancer in women was conducted in Kamakura
and Miura» both in Kanagawa prefecture, Japan. The two cities are distinctly different in social
environment; the former is a residential community, and the latter is a fishing village. After
stratification on city and age groups, the odds ratio of lung cancer in nonsmoking wives was
shown to be 1.58 when husbands smoked fewer than 19 cigarettes a day and 3.09 when husbands
smoked 20 or more cigarettes a day. For comparison, the odds ratio for active smoking is 5.50.
Although the study size is quite small, it provides additional evidence favoring the passive
smoking and lung cancer hypothesis. (Paraphrased from author's discussion; no abstract was
provided.)
A.16.2. Study Description
This study was conducted to assess the roles of active and passive smoking in the etiology
of lung cancer in women. It is unclear how subjects or diagnoses were obtained, but cases are
women who died of lung cancer in Kamakura or Miura in the time periods 1980-83 and 1973-81,
respectively. Controls, consisting of women who died of cerebrovascular disease during the same
time frames, are individually matched to cases on year of birth, year of death (± 2.5 years), and
district of residence. It is not clear whether incident cases were used.
Face-to-face interviews were conducted by public health nurses and midwives. ETS
subjects consist of the 28 nonsmoking cases and 62 nonsmoking controls remaining after
elimination of 9 cases and 12 controls who were smokers. Husband's smoking status was not
available for unspecified reasons in a total of 8 cases and 20 controls, but these figures include
smokers as well as nonsmokers. The exact number of nonsmokers for which spousal smoking
status was available is not specified but can be back-calculated from what is given (see below).
A-56
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
No information is given on the number of proxy respondents, the age distribution of the subjects,
or attempts to confirm diagnoses of primary lung cancer.
The term "nonsmoker" is not defined, so it is not clear whether it refers to persons who
never smoked or who do not smoke at present. Nonsmoking women whose husbands smoke at
least five cigarettes per day are classified as exposed to passive smoking. Considerations of
former smoking or marital status, ETS exposure at the workplace or in childhood, and duration of
exposure are not addressed. No attempts to verify the reliability or validity of the data are
mentioned.
The number of subjects is not delineated by case versus control and exposed versus
unexposed figures. They can be determined from the odds ratio and confidence interval,
however, as 18 out of 22 (exposed over total) cases and 30 out of 47 controls. For nonsmoking
women with smoking husbands, the crude odds ratio calculated by the reviewers is 2.55 (95%
C.I. = 0.74-8.78). (Note: OR = 2.25 is erroneously reported in the article. The OR value of 2.55
has been confirmed by Hirayama.) When husbands' smoking is divided into two strata (< 19
cig./day and 20+ cig./day), the odds ratios increase with exposure from 1.16 to 3.35, giving a
statistically significant trend (p < 0.05). Age-adjusted odds ratios of 1.39 and 3.16 are reported
for the two strata; adjustment for both age and district yields corresponding odds ratios of 1.58
and 3.09. (Note: The first OR value, 1.58, is incorrectly reported as 2.58. The value 1.58 has
been confirmed by Hirayama.) The authors conclude that, although the study size is quite small,
the results provide more evidence favoring the hypothesis that passive smoking causes lung cancer.
A. 16.3. Comments
The number of subjects remaining after active smoking and missing data exclusions is
small, guaranteeing poor power and lack of statistical significance in the absence of large odds
ratios. The details on study design are limited. The source of cases and controls is not mentioned,
for example, and it is unclear whether incident or prevalent cases were used.
Information regarding quality control and related concerns is equally sparse. Interviewers
used standardized questionnaires, which would help to promote consistency, but no mention is
made of blinding them to subject background or study question, the absence of which could
introduce interviewer bias (probably in a positive direction). Because cases and controls are stated
to have died during the study period, it is probable that proxy respondents were, required, but the
extent is unknown. In addition, neither duration of ETS exposure from spousal smoking nor
exposure from other sources, such as other cohabitants, was considered. The resultant inaccuracy
of exposure assessment probably biases the results toward the null. Lack of information on
A-57
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
former smoking status or verification of diagnosis may introduce biases of indeterminate
direction. Except insofar as the district acts as a surrogate for factors related to socioeconomic
status, no potential confounders other than age or district of residence were considered. The
meaning of "nonsmoker" is not given. Was that status left to self-classification? Is some degree of
past smoking acceptable? Is smoking history a factor at all (i.e., does nonsmoking refer simply to
the current status)? Accurate and meaningful segregation of never-smoking subjects is needed
for analysis, but there is no indication that that was accomplished.
Although a substantial odds ratio was observed for husband's smoking, these results are
based on a small sample with too few details provided to assess adequately either the evidence or
the study's design and execution. The numerous sources of potential bias are enhanced by the
omissions or sketchy descriptions of the study. The statistical uncertainty of the odds ratios given
is reflected in the extremely wide confidence intervals shown. The test for trend does not add any
additional information. It is basically a restatement of the significant comparison between the
heavily exposed group (husband smokes > 20 cig./day) and the unexposed group. Unfortunately,
the brevity of the description of this study in the source available severely limits its utility.
A.17. JANE
A.17.1. Author's Abstract
"The relation between passive smoking and lung cancer is of great public health ,
importance. Some previous studies have suggested that exposure to environmental tobacco smoke
in the household can cause lung cancer, but others have found no effect. Smoking by the spouse
has been the most commonly used measure of this exposure.
In order to determine whether lung cancer is associated with exposure to tobacco smoke
within the household, we conducted a population-based case-control study of 191 patients with
histologically confirmed primary lung cancer who had never smoked and an equal number of
persons without lung cancer who had never smoked. Lifetime residential histories including
information on exposure to environmental tobacco smoke were compiled and analyzed. Exposure
was measured in terms of "smoker-years," determined by multiplying the number of years in each
residence by the number of smokers in the household."
A.17.2. Study Description
This study was undertaken in New York State to clarify the role of exposure to tobacco
smoke in the household as a possible cause of lung cancer among nonsmokers. Interviews were
conducted with former as well as never-smokers initially (Varela, 1987), but because matching
A-58
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
was carried out on smoking status, only never-smoking case-control pairs were included in the
analyses for this article. The study includes both males and females, which are combined in all of
the analyses. There are 146 (45) female (male) pairs.
Cases are never-smokers aged 20 to 80 years newly diagnosed with lung cancer at 125
referral centers in New York from July 1, 1982, to December 31, 1984. Controls are cumulatively
sampled never-smokers identified from files of the New York Department of Motor Vehicles.
Controls are individually matched to cases on age (± 5 years), gender, and residence. In addition,
the same interview type (proxy or direct) was used for controls as for their corresponding cases.
Exposure data were collected face-to-face via standardized questionnaire by interviewers blind to
the subject's status.
From the 439 case-control pairs interviewed, 242 pairs containing former smokers and 6
pairs with a mismatch on the source of response were excluded. Of the remaining 191 pairs used
in the ETS study, interviews were conducted directly with the subjects in 129 (68%) and with
proxies in 62 (32%) (if a proxy was interviewed for a case, then a proxy was used for the matching
control as well). No demographic comparisons were provided for the ETS cases and controls. For
the whole study including smokers, the mean age of cases and controls is nearly identical (67.0 and
68.1, respectively; Varela, 1987). Histological verification of diagnosis was obtained for all but
five cases (for whom only clinical information was available) out of the initial population of 439.
Persons smoking no more than 100 cigarettes over the course of their lifetime qualified as
never-smokers for this study. Cigar or pipe smoking was apparently not considered. Exposure to
ETS was deemed to occur when a smoker lived in the subject's household at any time from
infancy to adulthood. Both total household smoke exposure and spousal smoke exposure were
determined. Preadult (before 21 years of age) and adult exposure were examined separately.
Exposures were computed in units of "smoker-years," the total number of years lived with each
smoker summed over smokers. In addition, pack-years were calculated for spousal smoking.
Workplace exposure also was estimated by smoker-years, whereas exposure in social settings was
estimated subjectively on a scale from 1 to 12 for each decade of life and summed. Exposure data
were not checked, and marital status was not considered in the analyses. No information on tumor
type or location was provided for the never-smoking population.
Preadult exposure to 24 or more smoker-years occurred in 52 (29) cases (controls) whereas
82 (94) were exposed to 1 to 24 smoker-years and 57 (68) were unexposed. Odds ratios were
calculated using matched-pairs regression analysis. Preadult passive smoking yielded increasing
odds ratio of 1.09 (95% C.I. = 0.68-1.73) for 1 to 24 smoker-years and 2.07 (1.16-3.68) for 25 or
more smoker-years. The odds ratios for adult exposure are low but also increase—from 0.64 (0.34-
A-59
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
1.21) at 1 to 24 smoker-years to 1.11 (0.56-2.20) at 75 or more smoker years. The odds ratios for
lifetime exposure increase from 0.78 (0.36-1.67) at 1 to 24 smoker-years to 1.80 (0.83-3.90) at 25
to 99 smoker-years and then dip to 1.13 (0.56-2.28) at 100 or more smoker-years. Spousal
smoking was not significantly associated with lung cancer. In fact, when results were stratified by
type of interview, proxy interviews yielded strong and, in some instances, statistically significant
negative associations for spousal smoking, with odds ratios between 0.20 and 0.68 for ETS
expressed in terms of present or absent, smoker-years, and pack-years of exposure. The odds
ratios for direct interviews, in contrast, range from 0.71 to 1.10 and are uniformly higher than the
odds ratios for corresponding proxy responses. Workplace exposure to 150 or more person-years
yielded an odds ratio of 0.91 (0.80-1.04), whereas a social setting exposure score of 20 led to a
statistically significant decreased odds ratio of 0.59 (0.43-0.81).
The authors conclude that they found a significant adverse effect of relatively high levels
of exposure to ETS during early life (before age 21). For those who were exposed to 25 or more
smoker-years in their first two decades of life, the risk of lung cancer doubled. By contrast, they
found no adverse effect of exposure to ETS during adulthood, including exposure to a spouse who
smoked. This lends further support to the observation that passive smoking may increase the risk
of subsequent lung cancer, and it suggests that it may be particularly important to protect children
and adolescents from this environmental hazard.
A.17.3. Comments
The number of never-smoking cases is relatively large, resulting in above-average
statistical power for evaluation of ETS effects. Controls were matched to cases on smoking status,
as well as the key demographic factors of age, gender, and neighborhood. Comparability of cases
and controls was likely good, as evidenced by the similar mean ages for the total population,
although no other comparative information is available. Interviews were ostensibly conducted
blindly, thus precluding interviewer bias, but in view of the use of population-based, basically
healthy controls, it is questionable that diagnostic blinding was effective. The study's matching on
smoking status with subsequent retention of matching and use of matched-pairs analysis for ETS
exposure effectively eliminates potential confounding by age, gender, or residence, and makes
confounding by related factors (such as socioeconomic status) less likely. A rare feature is the use
of matching on interview type (i.e., proxy or subject direct), thus eliminating potential
confounding by this source. Comparison of spousal smoking results for direct and proxy
interviews, however, indicates consistently lower estimated risks from proxies. This suggests that
use of proxy respondents did not merely lead to increased random misclassification but might
A-60
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
have biased the outcome toward a negative association. The authors posit that proxies of lung
cancer patients may be more likely to underreport exposure than those of control subjects.
Curiously, however, although the authors report that odds ratios "frequently differed according to
type of interview," they do not specify how the odds ratios differed for exposure other than
spousal smoking. Also, the composition of the proxy groups—relative proportions of spouses, other
relatives, and friends or associates—is never discussed, leaving unexplored the possibility that
misreporting by spouses of cases may lie at the heart of the observed discrepancy. It is also
interesting that the outcome of self- versus proxy responses in this study is in the opposite
direction of the findings in GARF. Diagnostic misclassification is unlikely, given the histological
verification of nearly all cases.
The restriction of subjects to persons smoking no more than 100 cigarettes in their lifetime
theoretically eliminates active smoking as a source of bias, although no verification of smoking
status was undertaken. Consideration of potential sources of ETS exposure is commendably
thorough, and the calculation of total years of living with smokers, regardless of relation to the
smoker, as an index of household smoke exposure minimizes the possibility that any source (e.g.,
roommates) is overlooked. In contrast, the index of exposure in social settings is highly
subjective, and persons more habituated to passive smoke may report a given exposure as less
severe than persons less accustomed to smoke, thus creating a negative bias. The proportion of
controls classified as exposed to ETS is 80%., which is high in comparison to other studies. This
suggests that some exposed controls may have only minor exposure to ETS, making detection of an
association (if present) less likely. Unlike almost every other ETS study, males and females are
combined in the analysis and only the joint results are reported. Because there are 45 (146) pairs
of males (females), the sample sizes are sufficient to warrant reporting odds ratios separately by
sex and to test the hypothesis of no difference due to gender.
Lung cancer odds ratios for adulthood, lifetime, and spousal smoking are consistently well
below one for low ETS exposure relative to nonexposure, as if exposure had a protective effect.
Thereafter, however, the odds ratios associated with increasing levels of exposure are suggestive
of an upward trend in response. Although we would not dismiss the occurrence of this outcome as
attributable to chance alone, it is consistent with the baseline lung cancer mortality rate in the
control population simply being higher than that of the case population for reasons other than
exposure to spousal smoking. A pervasive (systematic) negative bias linked with exposure could
also produce such an effect. Both of these contingencies are speculative, however, because there
is no evidence in the article to support either, aside from the outcome of the data. Further fueling
the speculation, however, are the markedly lower odds ratios obtained from surrogate responses,
A-61
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
indicative of some source of bias acting unequally on proxy and nonproxy sources. Also
speculative is the idea that using predicted responses from a model that fits the data poorly might
produce such an effect, but that level of detail is beyond the scope of most published articles,
including this one. Some explanatory discussion by the authors on these issues, as well as
separation of the analyses by sex, would enhance interpretation of results and facilitate
comparison with results of other studies on females.
The authors' finding that exposure during childhood and adolescence appears to influence
subsequent lung cancer risk more than exposure during adulthood raises some interesting
possibilities. More time may be spent in proximity to a household smoker (particularly the
mother), on average, in childhood than in adulthood. According to data presented by K.M.
Cummings (Roswell Park Memorial Institute, Buffalo, NY) at the Science Advisory Board meeting
on EPA's draft ETS report (U.S. EPA, 1990), on December 4-5, 1990, heavy childhood exposure
is a better surrogate for total lifetime exposure than is spousal exposure. Also, early exposure may
appear to become a risk, either due to a long latency period for lung cancer or, perhaps, due to
increased susceptibility at an earlier age. The results suggesting an effect from early exposure but
not from spousal smoking are more nearly atypical than reinforced by other studies, though, and
the number of exposure sources considered raises the possibility that the strength of association
seen for preadult exposure may be due to chance. However, after elimination of 78 pairs with
incomplete marriage or household exposure data, the association persisted and was strengthened
(OR - 2.59), arguing against chance as the major influence. It is unclear what role, if any,
negative bias due to proxy respondents may have had in the nonspousal analyses.
In summary, the findings for preadult exposure are not readily attributable to chance or
confounding, although some role of interviewer bias or unmeasured confounding factors such as
diet cannot be ruled out. No association with lung cancer incidence is observed for spousal
smoking. The authors conclude, however, that, spousal smoking aside, other sources of household
ETS exposure support the conclusion that exposure to ETS can cause cancer. That conclusion is
not unequivocal in our view. In general, the odds ratios (aside from preadulthood exposure) tend
to be low but trend upward with exposure, exhibiting more of a patterned response than one
might expect to see due to randomness. This is puzzling as there is no apparent source of bias and
the study appears to have been conducted with considerable forethought and thoroughness. The
only exception noted is the lack of separate analyses and comparisons of males and females. These
concerns notwithstanding, the study is a useful addition to the literature on ETS exposure and
lung cancer.
A-62
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
A.18. KABA
A.18.1. Author's Abstract
"Among 2,668 patients with newly diagnosed lung cancer interviewed between 1971 and
1980, 134 cases occurred in 'validated' honsmokers. The proportion of nonsmokers among all
cases was 1.9% (37 of 1,919) for men and 13.0% (97 of 749) for women, giving a sex ratio of 1:2.6.
Kreyberg Type II (mainly adenocarcinoma) was more common among nonsmoking cases,
especially women, than among all lung cancer cases. Comparison of cases with equal numbers of
age-, sex-, race-, and hospital-matched nonsmoking controls showed no differences by religion,
proportion of foreign-born, marital status, residence (urban/rural), alcohol consumption or
Quetelet's index. Male cases tended to have higher proportions of professionals and to be more
educated than controls. No differences in occupation or occupational exposure were seen in men.
Among women, cases were more likely than controls to have worked in a textile-related job
(relative risk = 3.10, 95% confidence interval 1.11-8.64), but significance of this finding is not
clear. Preliminary data on exposure to passive inhalation of tobacco smoke, available for a subset
of cases and controls, showed no differences except for more frequent exposure among male cases
than controls to sidestream tobacco smoke at work. The need for more complete information on
exposure to secondhand tobacco smoke is discussed."
A.18.2. Study Description
In 1969, the American Health Foundation began interviewing newly diagnosed lung cancer
patients with cancer at sites potentially related to tobacco use for a, case-control study that is still
ongoing (Wynder and Stellman, 1977). The current article considers the data on lung cancer in
nonsmokers alone collected from newly diagnosed lung cancer patients between 1971 and 1980.
Several factors are of interest: histology, demographic factors, residence, Quetelet's index, alcohol
consumption, previous diseases, occupation and occupational exposures, and ETS exposure. The
number of nonsmokers among the cases is small, so the authors consider the results to be
preliminary.
The study'from which the data on lung cancers in nonsmokers are extracted is a very large
effort that includes tobacco-related cancers at multiple organ sites and includes smokers as well as
nonsmokers. The cases are from approximately 20 hospitals in 8 U.S. cities (about one-third from
New York City). With reference to the lung cancer cases in that study, histologic type of lung
cancer was determined from pathology reports and discharge summaries. Secondary lung cancer
cases were excluded. Controls consist of hospital patients with diseases unrelated to tobacco use
who were pair-matched with cases on hospital, age (within 5 years), sex, race (with five
A-63
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
exceptions), date of interview (within 2 years), and nonsmoking status. Cases appear to be
incident, and control sampling is density. All subjects were interviewed while patients were in
the hospital. The questionnaire for the interviews was expanded in 1976. Questions on exposure
to ETS were not included, however, until an addendum to the questionnaire in 1978, which was
then modified in 1979.
The term "nonsmoker" applies to subjects who have smoked less than one cigarette, pipe,
or cigar per day for a year. The term "never-smoker" is used interchangeably. Independent of the
intended definition, however, subjects whose hospital charts indicated any record of smoking,
even in the remote past, were excluded from the nonsmoker classification. ETS subjects include
53 (25) females (males), after combined attrition of 22 (9 without primary lung cancer and 13 with
a record of smoking). The age distribution of the female cases (controls) is as follows: age less
than 50, 12 (15); age 50 to 59, 26 (24); age 60 to 69, 29 (34); age 70 or more, 30 (24). Histologic
data on lung cancer type are given for female cases: squamous cell (16), adenocarcinoma (60),
alveolar (12), large cell (4), and unspecified (5). The authors report that exposed cases did not
differ from the unexposed cases in the distribution of histologic type.
A person is "ETS exposed" (1) at home, if currently exposed on a regular basis to family
members who smoke, (2) at work, if currently exposed on a regular basis to tobacco smoke at
work, and (3) to spousal smoke, if the spouse smokes. There are data on 53 cases and their
controls for exposure at home and at work, but data on only 24 cases and 25 controls for spousal
smoking. This is because of the change in the questionnaire from 1978 to 1979 and because
spousal smoking was only applicable for women currently married. Because nonsmoking status
was a variable for matching, the 53 pairs of cases and controls for analysis of exposure at home or
at work are matched; the data for spousal smoking, however, are technically not matched. There
is no indication at all of an association between ETS exposure and lung cancer for women from
exposure at home, at work, or from spousal smoking. For ETS exposure at home, there are 16 out
of 53 (exposed/total) cases and 17 out of 53 controls; for exposure at work, the figures are 26 out
of 53 cases and 31 out of 53 controls; and for spousal smoking, the data are 13 out of 24 cases and
15 out of 25 controls. No statistical calculations are provided for females. From our calculations,
the odds ratio for spousal smoking is 0.79 (95% C.I. = 0.25-2.45). (Among male subjects, exposure
to ETS in the workplace was slightly significant, p = 0.05, as reported in the article.) For other
potential risk factors for lung cancer in women other than passive smoking, it was found that
cases were more likely than controls to have worked in a textile-related job (OR = 3.1; 95% C.I. =
1.1-8.6), but the significance of the finding was not clear. It was also found that more female
A-64
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
cases had a history of pneumonia compared to controls, but no interpretation could be attached to
the observation.
A.18.3. Addendum
Unpublished preliminary results of a study of ETS and lung cancer in never-smokers
conducted at the American Health Foundation have been reported at two meetings—The American
Public Health Association (APHA) 119th Annual Meeting, Atlanta, Georgia, November 10-14,
1991, and The Toxicology Forum, 1990 Annual Winter Meeting, Washington, D.C., February 19-
21, 1990. A completed report for our review was not available at the cutoff date for inclusion in
this document (personal communication with the first author, Dr. G.C. Kabat). Enclosed below is
the abstract for the APHA meeting.
RISK FACTORS FOR LUNG CANCER IN LIFETIME NON-SMOKERS
Geoffrey C. Kabat, Ernst L. Wynder
Risk factors for lung cancer in lifetime non-smokers (NS) were assessed in a hospital-
based case-control study carried out between 1983 and 1990. The study population
consisted of 41 male and 69 female NS cases, and 117 male and 187 female NS controls
matched on age, race, hospital, and date of interview. Evidence of an effect of exposure
to environmental tobacco smoke (ETS) was inconsistent. In males, there was no difference
between cases and controls in reported exposure to ETS (yes/no) in childhood, in
nonsignificant association with exposure in childhood (OR = 1.6, 95% C.I. 0.9-2.8), but no
association with exposure in adulthood at home or at work. Male cases were somewhat
more likely to have a smoking spouse (OR = 1.6, 95% C.I. 0.7-3.9), whereas there was no
difference in females. Cases and controls did not differ in reporting a history of previous
respiratory diseases. Female cases were more likely to report a history of radiation
treatment (OR = 4.3 95% C.I. 1.5-12.3). In females, but not in males, a significant inverse
association was observed between body mass index (based on self-reported weight 5 years
prior to diagnosis) and lung cancer risk.
A.18.4. Comments
Although the study contains more than 2,600 patients, only a small number of nonsmokers
are available because questions about ETS exposure were not included in the interview until 1978
and the questions were changed in 1979. It is not clear just how the questionnaire was changed,
although the general tenor of the article suggests care in study planning and execution. The
design for the larger study from which the ETS data are taken is pair-matched on numerous
factors of potential interest, including "nonsmoking status," which contributes favorably to the
analysis of ETS data alone. Cases with secondary tumors were excluded, histological type was
considered, and all subjects were personally interviewed. It appears that only the currently
married females were included in the question regarding exposure to spousal smoke, which
A-65
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
alleviates the need to make some approximating assumptions regarding exposure of widows,
singles, and so forth.
Two potential concerns about the analysis of ETS subjects have to do with the definition
of HETS exposure" and "nonsmoker." It is noted that duration of smoking was comparable in cases
and controls, but interview questions regarding exposure to ETS refer only to current exposure
(this is not explicit in the article but was confirmed by the first author). Also, this measure of
exposure has no units (e.g., number of cigarettes per day or pack-years smoked by spouse), which
might leave the question less subjective and perhaps help to dichotomize on ETS exposure more
sharply. Because lung cancer may have a latency period of 20 years or so, exposure in the past,
both in terms of duration and intensity, may be more meaningful than recent exposure. With
regard to the definition of nonsmoker, the requirement is less rigid than is often imposed. Ever-
smokers are included provided they did not smoke more than the equivalent of 1 cigarette per day
for 1 year (about 18 packs). Smoking may seriously confound ETS exposure, and it is difficult to
know what constitutes a "negligible" level of past smoking.
One of the factors of interest to the investigators is occupation, so cases and controls were
not matched on that variable. For ETS exposure, occupation could be a potential confounding
factor. Among females, the controls contain a higher percentage of professional and skilled
workers than do the cases (47 to 25), and a lower percentage of housewives (41 to 50). Some
differences are also apparent in religious preference between cases and controls that may bear
some influence through lifestyle or dietary practices. Variables such as these may need to be
taken into account in an adjusted analysis when more data become available.
As noted previously, this article is presented as a preliminary report, and it should be
interpreted in that light. The data set on ETS subjects is small. We expressed some reservations
about the operational meaning of "nonsmoker" and "ETS exposed," both of which could be more
strict. Nonsmokers may have a light history of smoking; exposed nonsmokers may have very little
history of exposure. Both factors may be sources of bias, the second one toward the null
hypothesis of no effect, and the first one possibly in either direction. This study contributes some
useful evidence for the epidemiologic evaluation of whether ETS poses a detectable lung cancer
risk, but the potential for bias and the uncertainty due to small sample size could be influential.
A.19. KALA
A.19.1. Author's Abstract
"A case-control study was undertaken in Athens to explore the role of passive smoking and
diet in lung cancer, by histologic type, in non-smoking women. Among 160 women with lung
A-66
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
cancer admitted to one of seven major hospitals in Greater Athens between 1987 and 1989, 154
were interviewed in person; of those interviewed, 91 were life-long non-smokers. Among 160
identified controls with fractures or other orthopedic conditions, 145 were interviewed in person;
of those interviewed 120 were life-long non-smokers. Marriage of a non-smoking woman to a
smoker was associated with a relative risk for lung cancer of 2.1 (95% confidence interval [CI] 1.1
— 4.1); number of cigarettes smoked daily by the husband and years of exposure to husband's
smoking were positively, but not significantly, related to lung cancer risk. There was no evidence
of any association with exposure to smoking of other household members, and the association with
exposure to passive smoking at work was small and not statistically significant. Dietary data
collected through a semi-quantitative food-frequency questionnaire indicated that high
consumption of fruits was inversely related to the risk of lung cancer (the relative risk between
extreme quartiles was 0.27 (CI 0.10 — 0.74). Neither vegetables nor any other food group had an
additional protective effect; furthermore, the apparent protective effect of vegetables was not due
to carotenoid vitamin A content and was only partly explained in terms of vitamin C. The
associations of lung cancer risk with passive smoking and reduced fruit intake were independent
and did not confound each other. Passive smoking was associated with an increase of the risk of
all histologic types of cancer, although the elevation'was more modest for adenocarcinoma."
A.19.2. Study Description
This study was undertaken in Athens, Greece, in 1987-89. It sought to explore the role of
passive smoking and diet in the causation of lung cancer in nonsmoking women. All data used in
the study were collected specifically for that purpose.
Gases are never-smoking women hospitalized in one of seven Greater Athens area
hospitals during an 18-month period of 1987-89 with a definite diagnosis of lung cancer from
histologic, cytologic, or bronchoscopic exam. Controls were selected from female never-smoking
patients in the orthopedic ward of the same seven hospitals and an orthopedic hospital. A control
was interviewed within 1 week of a corresponding case, thus essentially density-sampled but
otherwise unmatched. Cases were not specifically restricted to incident cancers. All subjects were
interviewed face-to-face by one of five trained interviewers; interviews were apparently
unblinded. A total of 160 lung cancer cases and an equal number of controls were initially
identified; 6 cases and 12 controls were too ill to interview, whereas 3 controls and no cases
refused to participate. After exclusion of smokers, 91 cases and 120 controls remained. The age
distributions of the cases and controls are very similar: for cases and controls, 16.5% (14.2%) were
less than 50 years of age, 19.8 (18.3%) were 50 to 59, 29.7 (25.8%) were 60 to 69, and 34.1 (41.7%)
A-67
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
were 70 or older. Current residence, level of education, occupation (housewife vs. other) and
marital status were also similarly distributed between cases and controls. Case diagnosis was
established by histology (48%), cytology (38%), or brbnchoscopy (14%), with exclusion of cancers
diagnosed as secondary.
Persons reportedly smoking fewer than 100 cigarettes in their lifetime are classified as
nonsmokers. No mention is made of pipe or cigar smoking. Several different sources of ETS
exposure are considered: husbands who smoke quantified in terms of years exposed and average
number of cigarettes smoked per day; household members other than husbands who smoke,
quantified by the sum of years exposed to each smoker; and coworkers who smoke, measured by
the number of smokers sharing the "same closed space" as the subject. Presumably childhood
exposure is included in the household exposure assessment. For spousal smoking, single women
are considered unexposed, whereas exposure of widowed or divorced women is based on their
married period. No attempts to verify exposure are mentioned.
For analysis of husband's smoking based on cigarettes per day, 64 out of 90 (exposed/total)
cases and 70 out of 116 controls gives a crude odds ratio of 1.6 for 90 cases and 116 controls; 64
cases and 70 controls were exposed. The authors present results stratified by four exposure
categories, which indicate no significant association (p = 0.16). Crude data for husband's smoking
stratified by five levels of smoking duration (never, < 20, 20-29, 30-39, and 40+ years) yield a
marginally significant increase in association with increasing duration (p = 0.07), with odds ratios
of 1.0, 1.3, 1.3, 2.0, and 1.9, respectively. No statistically significant association was noted for
ETS exposure from other household members (p = 0.60) or for exposure at work (p = 0.13), but
the crude odds ratios for these exposures were 1.41 and 1.39, respectively. Stratification by level
of intake for each of 16 food and nutrient groups yielded a significant negative (favorable)
association with cereals (p = 0.04) and a possible association with fruits (p = 0.11).
Multiple logistic regression was then used to adjust results for age, education, and
interviewer. An adjusted relative risk estimate of 1.92 (95% C.I. = 1.02-3.59) was obtained for
marriage to a smoker. After adjustment, trends for estimated lung cancer risk showed an increase
with duration of exposure (average 16% per .10 years) and packs per day (6% per pack), but these
were not statistically significant. No trend was observed for ETS in the household or workplace.
Adjustment for other sources of air pollution had no effect on the analyses. Adjustment of
dietary analyses for age, education, interviewer, and total energy intake indicated a significant
decrease in estimated risk between highest and lowest quartiles of consumption of fruit
(RR » 0.33; p - 0.02) and a nearly significant increase with consumption of retinol (RR =1.31;
p » 0.06), whereas beta carotene (RR = 1.01) and other dietary factors had no significant effect.
A-68
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Adding fruit consumption to the model for passive smoking increased the adjusted relative risk
for husband's smoking slightly, from 1.92 to 2.11. Stratification by lung cancer cell type yielded
somewhat lower adjusted estimated relative risks for adenocarcinoma (2.04) than for squamous,
small, and large cell cancer combined (2.58). No adjusted results were presented for other
household or workplace exposure.
The authors' conclusion is best reflected in their abstract (shown in full above). Marriage
of a nonsmoking woman to a smoker was associated with a relative risk for lung cancer of 2.1.
Number of cigarettes smoked daily by the husband and years of exposure to husband's smoking
were positively, but not significantly, related to lung cancer risk. There was no evidence of any
association with exposure to smoking of other household members, and the association with
exposure to passive smoking at work was small and not statistically significant. Dietary data
indicated that high consumption of fruits was inversely related to the risk of lung cancer. Neither
vegetables nor any other food group had an additional protective effect. The associations of lung
cancer risk with passive smoking and reduced fruit intake were independent and did not confound
each other. Passive smoking was associated with an increase of the risk of all histologic types of
cancer, although the elevation was more modest for adenocarcinoma.
It is noted that these findings are compatible with the relatively low incidence of lung
cancer in the Greek population—a population with the highest per capita tobacco consumption in
the world, but with a very high fruit consumption as well.
A.19.3. Comments
This study was generally well designed and executed. Set up specifically to address passive
smoking and diet as etiologicai factors in lung cancer, it includes sufficient numbers of
nonsmoking women to produce substantive results. Interviews were face-to-face and no proxies
were used, enhancing accuracy and comparability of responses, whereas the very low rate of
refusal minimizes potential bias due to volunteer selection. Cases and controls were very similar
demographically, were drawn from most of the same hospitals, and were matched temporally on
time of interview, so comparability seems high. Furthermore, the study hospitals' patient
population accounts for the majority of lung cancer and trauma patients seen in the Athens area,
enhancing generalizability of results. Most lung cancers were histologically or cytologically
confirmed, reducing chances for misclassification of disease status.
On the debit side, the apparently unblinded interviews could have been biased (although
what can be accomplished toward that end is limited). Adjustment for interviewer in the analyses
did not affect the results, however, and it is unlikely that all interviewers would share the same
A-69
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
bias. Determination of what constitutes workplace exposure is vague, and childhood exposure is
not clearly differentiated from adult household exposure; these were notably the passive smoking
categories, which showed the least association with lung cancer. ETS exposure in the workplace is
analyzed with regard to trend (Table 2), with levels of exposure represented by "housewife" (zero
exposure), "minimal," and "some," resulting in a p value of 0.13. Perhaps correctly, the authors
cautiously note the evidence that ETS exposure is associated with increased risk (referring to
Table 2 in general, not just exposure at work) but indicate that the differences are not large
enough to be interpretable without controlling for confounding effects. An analysis of exposed
versus unexposed for the workplace may have been useful, especially an adjusted analysis. Our
calculation of the crude odds ratio for a comparison of "minimal" and "some" exposure at work is
1.7, which is suggestive.
Methodological rigor and thoroughness are particularly evident in the treatment of
potential sources of confounding. Despite the demographic similarity of cases and controls, the
key demographic variables of age and education were nevertheless controlled for in the analyses,
along with interviewer identity. Air pollution, total energy intake, and other dietary factors were
also examined as potential confounders, and the impact of cancer type was evaluated. An
association of husband's smoking with lung cancer yielding an odds ratio of around 2 persisted
throughout. The authors claim to have taken special effort to exclude ex-smokers from
misclassification as never-smokers, taking account of this potential source of upward bias. No
discussion was found, however, of what measures were taken to control misclassification of
former smokers as never-smokers, beyond interviewing subjects about current and former
smoking habits.
In summary, this study presents evidence of a level- and duration-dependent association
between husband's smoking and lung cancer in a well-defined and highly comparable group of
Greek cases and controls. Positive but nonsignificant relationships with general home or
workplace passive smoking were observed, and there are indications that additional analysis of
workplace exposure may be worthwhile. No effect of air pollution was observed. With regard to
dietary factors, the large number of potential factors considered raises the issue of multiple
comparisons. Fruit consumption may be a significant factor, but further evidence is needed to
firmly establish this, particularly in view of the number of dietary factors explored. Dietary
factors, however, do not account for the results for ETS exposure in this study. The results
regarding spousal smoking cannot be readily attributed to bias, and they provide good quantitative
data on the issue of passive smoking and lung cancer. This well-conducted study makes a
valuable contribution to the evidence on lung cancer and ETS exposure.
A-70
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
A.20. KATA
A.20.1. Author's Abstract
"It is becoming noticeable in Japan that with increased incidence of lung cancer, there has
been an increase in pulmonary carcinoma in women. Active smoking by women is increasing,
while concern over passive smoking has been intensifying, and the effect of passive smoking on
carcinogenesis had become a social problem. Regarding this effect, immunological and public
health reports have appeared in Japan, but there have been few clinical reports, and detailed
analysis of patients has been inadequate. Lung cancer presents a variegated histological picture,
and presumably there are different carcinogenic factors for different histological types, although
there have also been few reports on this subject. The effect of passive smoking probably varies
depending on the regional environment and custom, and these factors should also be analyzed and
included in the investigation. The present report describes our findings regarding the effects of
smoking and familial aggregation of cancer in cases of pulmonary carcinoma in women."
A.20.2. Study Description
This study was undertaken in the Nara Prefecture, Japan, to investigate the effects of
smoking and familial aggregation of cancer in cases of pulmonary carcinoma in women. Active
smokers are included in the study, from which the nonsmokers are drawn for analysis. Matching
is retained, however, in the nonsmokers.
For the whole study, subjects were drawn from a hospital (presumably the Nara Prefecture
Medical University Hospital) during an unspecified period of time. Cases are female patients
with histologically diagnosed lung cancer; controls are female patients with "non-malignant"
disease, matched 2 to 1 with cases on age plus or minus 2 years. It is not clear if only incident
cases were used and if controls were density sampled. Case diagnoses were obtained from
histological exam results, whereas control diagnoses were presumably from medical charts. Other
information was collected from apparently unblinded "questioning," with an unspecified degree of
reliance on proxy responses from family members.
A total of 25 cases and 50 controls are included in the study; no information on refusals is
provided. Exclusion of active smokers leaves only 17 cases and, with retention of 1:1 matching,
17 controls. Mean ages for the total study population are 67.5 ± 8.8 years (67.6 ± 8.5 years) for
cases (controls). The age distribution of ETS subjects is not discussed. Nonsmokers are defined
by exclusion of "active smokers," with no delineation between former and current smokers. ETS
exposure is defined as exposure to smoking more or less daily through living with a smoker.
Three periods of ETS exposure are considered: current, past, and childhood, the last for those
A-71
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
"exposed since early childhood." Clearly these types are not mutually exclusive, although current
sources of exposure are omitted from the "past" exposure category, even if present for a long time.
ETS exposure is quantified as cigarettes per day smoked times number of years. No
mention is make of cigar or pipe smoking, nor of checks on exposure data. No distinction is made
regarding marital status. Tumors occurring among current passive smokers were mostly
adenocarcinomas (13/17), the remainder (4/17) being squamous or small cell cancers. Airway
proximity was not specified. Excluding active smokers, all 17 cases were current passive smokers,
compared to 14 out of 17 controls, for an odds ratio of 1.2, whereas past passive smoking
characterized 16 of 17 cases and 17 of 17 controls, for an odds ratio of 0.9 (these odds ratios
reflect the substitution of 0.5 for 0 in the exposure categories in which no subjects fall).
Childhood passive smoking was reported in 13 of 15 cases and 7 of 15 controls (apparently all
those for whom information was available), for an odds ratio of 7.4 (p < 0.1). None of the passive
smoking odds ratios was statistically significant at the 5% level. No definite conclusion can be
drawn from the present study, but there is a suggestion that passive smoking is associated with
development of lung cancer in the Nara region. The effect of passive smoking that continued to
the present time was especially marked, particularly in squamous cell carcinoma and small cell
carcinoma. With adenocarcinoma, an effect of passive smoking in the past is suspected. Along
with passive smoking, the association of some intrinsic factor (genetic tendency) to varying
degrees in the different histologic types of lung cancer in women, especially in adenocarcinoma, is
apparent.
A.20.3. Comments
The histological diagnosis of all cases, in combination with the apparent involvement of
the researchers in the diagnoses, virtually eliminates the potential pitfall of misclassification of
lung cancer cases. It also allows specific breakdowns by cell type. With regard to passive
smoking, however, limitations related to exclusion of active smokers greatly reduced the study's
potential.
In their initial analyses, the authors investigate passive smoking without excluding or
stratifying on active smoking and report statistically significant associations with lung cancer and
combined effects with family history of cancer. This is not a meaningful analysis, because the
effects of active and passive smoking cannot be separated and because passive smoke exposure
probably correlates strongly with extent of active smoking. Excluding active smokers greatly
reduces the available numbers of matched subjects and, in combination with the very high
exposure prevalence among qualifying controls, makes the differences between cases and controls
A-72
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
highly unstable for all comparisons except for that of childhood exposure. Even here, with an
estimated relative risk of 7.4, the results do not reach the 5% level of statistical significance,
notwithstanding the problem of multiple comparisons. This does not deter the authors from
attempting cell-type-specific analyses, but these too fail to yield significant results. The
extraordinarily high proportion of exposed present and past passive smoking controls is apparently
a fluke, because the proportion is not as high in the total control subject population (or childhood
passive smoking controls). Nevertheless, exposure was very common among controls. This
indicates that the exposure criteria may be too lax or, alternatively, that the control population
included a substantial proportion of persons with smoking-related diseases (controls being only
stipulated not to have malignant disease).
In light of the minimal utility of the study's passive smoking analyses, detailed
consideration of design strengths and weaknesses is unwarranted. Major points not already
mentioned relate to information ascertainment and confounding. Interviews were apparently
unblinded and, especially if conducted by the authors themselves, may thus have been biased
toward uncovering exposure among cases (although the high prevalence of exposure among
controls as well as cases argues against this). Furthermore, the extent of proxy interviews,
potentially decreasing accuracy of exposure assessment, is unclear.
All subjects are female and, although results are not age adjusted, matching on age was
retained for all analyses. No other potential confounders except family history of cancer were
considered, probably due to limited subject numbers, because much information on potential
confounders was collected. Moreover, family history was considered only in the nonmeaningful
analyses, which did not differentiate active and passive smokers. Thus, although the problems
with numbers and exposure misclassification probably reduced the study's ability to detect
whether an association exists, information bias and confounding could have biased results either
up or down.
In summary, this study's data are consistent with an association of passive smoking,
particularly childhood exposure, with lung cancer, but the results are too unstable and subject to
potential bias to carry much weight, and the quantitative results must be viewed with extreme
caution.
A.21. KOO
A.21.1. Author's Abstract
"Lifetime exposures to environmental tobacco smoke from the home or workplace for 88
"never-smoked" female lung cancer patients and 137 "never-smoked" district controls were
A-73
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
estimated in Hong Kong to assess the possible causal relationship of passive smoking to lung
cancer risk. When relative risks based on the husband's smoking habits, or lifetime estimates of
total years, total hours, mean hours/day, or total cigarettes/day, or earlier age of initial exposure,
were combined with years of exposure, there were no apparent increases in relative risk.
However, when the data were segregated by histological type and location of the primary tumor, it
was seen that peripheral tumors in the middle or lower lobes (or less strongly, squamous or small-
cell tumors in the middle of lower lobes) had increasing relative risks that might indicate some
association with passive smoking exposure."
A.21.2. Study Description
This study, the second of four from Hong Kong, is based on a secondary data set of
reported female never-smokers. The parent study from which the data on ETS subjects was
drawn includes ever-smokers in a matched case-control study of 200 cases and 200 controls (Koo
et al., 1984; also see Koo et al., 1983). Its objective is to assess the role of passive smoking as a
potential etiological factor in the high incidence rate of lung cancer among Chinese females in
Hong Kong. The current article emphasizes the quantitation of lifetime ETS exposure and the
histological profile of lung cancer in exposed never-smokers.
In the parent study, cases are from the wards or outpatient departments of eight hospitals
in Hong Kong during 1981-83. Controls are healthy subjects from the community, matched on
age (within 5 years), district of residence, and type of housing (public or private). The cases are
incident, and control sampling is density. Attrition due to selection or follow-up totals 26 (8 too
ill to interview and 18 with secondary lung cancers), leaving 200 cases for interview. Face-to-
face interviews of 1.5 to 2 hours were conducted directly with cases and controls. There was no
restriction of cases by cell type of lung cancer. The ETS subjects extracted from the parent study
include 88 cases and 137 controls. Of the 88 cases, 83 were confirmed by histology and 5 were
"confirmed malignant." The number of squamous cell and small cell cases combined is 32 (23 ETS
exposed; 72%); the corresponding figure for adenocarcinoma and large cell combined is 44 (31
ETS exposed; 70%); 12 cases are of another cell type, or otherwise unspecified. For the 86 cases
with available information, tumors were centrally located in 37 (25 ETS exposed; 67%) and
peripherally in 46 (34 ETS exposed; 74%).
The term "never-smoker" applies to persons who have smoked a total of fewer than 20
cigarettes. Interview questions regarding exposure to ETS include cigarette and cigar smoking in
the home during childhood, by the spouse and other cohabitants in adulthood, and workplace
exposure. "ETS exposed" is technically used in several ways. For the comparison of exposed with
A-74
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
unexposed ever-marrieds, it means the husband ever smoked in the wife's presence. For measures
of exposure in terms of duration or rate (e.g., total years, hours/day, total hours, and cig./day),
there is some variation. For example, total years of exposure is derived by adding the years
during which tobacco exposure occurred in the home or workplace. The total hours of exposure
are calculated by multiplying the average hours per day of exposure by the years of exposure from
each household smoker, or the amount of exposure at each workplace. The mean hours per day of
exposure are found by adding the hours per day of home and workplace exposures and dividing
this figure by the age of the subject. This figure is intended to approximate the average number
of hours of exposure per day experienced by the subject, over her lifetime. Cumulative exposure
is estimated by the total cigarettes smoked by family members, weighted by years of exposure.
When data are analyzed on the simple basis of whether a husband ever smoked in the
presence of the wife, the crude and adjusted odds ratios are 1.55 (95% C.I. = 0.94-3.08) and 1.64
(95% C.I. = 0.87-3.09), respectively. The crude analysis applies to ever-marrieds only, which
excludes three subjects. An adjusted analysis uses cigarettes per day smoked by the husband as
the measure of ETS exposure. Conditional logistic regression was applied with stratification on
district of residence and housing type (public/private); model parameters were included for age,
family history of lung cancer (yes/no), number of live births, and number of years since exposure
at home or in the workplace.
The crude and adjusted methods give very similar odds ratios and confidence intervals,
but the tests for trend differ substantially. The test for trend on the crude data is based on the
Mantel-Haenszel test, using midpoints of the intervals for cigarettes per day smoked by the
husband; the significance value is p = 0.10. The p value for trend in the adjusted analysis is 0.32.
For analysis of data by other measures of exposure, as described above, the estimated odds ratio
ranges between 1.0 and 4.1 across the three levels of the various measures of ETS exposure for
both the analyses of the crude data and the adjusted analyses by conditional logistic regression,
with two exceptions from analysis of the crude data for hours per day of exposure. The results
are not statistically significant in most cases, because the sample sizes at each exposure level are
small. The dose-response patterns observed are clearly sensitive to the measure of ETS exposure
used, with several exhibiting an apparent peak at a low exposure level. Although the authors
acknowledge that it was troubling to find the lack of a response pattern, no further explanation is
given.
The authors did not detect a significant trend in the crude or adjusted odds ratio for the
four lifetime measures of passive smoking (total years, hours, mean hours/day, cig./day).
Although the odds ratio for the intermediate level exposures of hours per day and cigarettes per
A-75
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
day was significant, the odds ratio at the highest levels of exposure for these two variables fell to a
nonsignificant 1.0 to 1.2. In fact, the odds ratio for the highest exposure levels for three out of
the four measurements were below all of those with lower exposures and ranged from a very weak
1.0 to 1.4. On the other hand, most of the crude and adjusted odds ratios were greater than 1.0.
Measurements based on increasing intensity of exposure, defined as increasing years (or hours, or
cig./day) by mean hours per day of exposure, also did not indicate a dose-response relationship.
The analysis of total years of exposure with age of exposure did not suggest that earlier age of
initial exposure and increasing years of exposure led to higher odds ratios.
It is concluded that when the lung tumors were segregated by histological type and
location, the resulting analyses showed that peripheral tumors in the middle or lower lobes, and
squamous or small cell tumors in the same lobes, exhibited better odds ratio patterns for passive
smoking in terms of consistency, strength, and dose-response. The odds ratio for total years,
hours, and hours per day measurements of squamous and small cell lung tumors indicated
consistently elevated risks with increasing exposure. This pattern was not found for any of the
adjusted odds ratios for adenocarcinoma or large cell lung cancers.
The cases are divided into two groups histologically, those with squamous cell or small cell
tumors, and those with adenocarcinoma or large cell malignancies. Although none of the crude or
adjusted analyses are found to be significant, it is concluded that an observed dose-response
pattern seems to be more apparent in the squamous or small cell group. With regard to tumor
location, some evidence suggests that peripheral tumors in the middle or lower lobes may be more
common in passive smokers.
A.21.3. Comments
As described above, the data employed in the current study were taken from a larger
retrospective study of female lung cancer in Hong Kong (Koo et al., 1984) that matched 200 cases
and controls on age, district of residence and housing type (private or public, an indication of
socioeconomic status). Attention to detail and accuracy is evident in most aspects of the parent
study. In particular, considerable effort was put into attempting to ascertain a better quantitative
measure of exposure than used in preceding studies of ETS. Records were apparently verified to
the extent possible to cross-check the accuracy of information collected, cancers were verified
histologically, and analyses investigated questions related to the histological types and sites of
tumors that may be related to passive smoking.
The never-smokers from the parent study, 88 cases and 137 controls, compose the
secondary data set on which the current article is based. The matching of the subjects, of course,
A-76
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
is no longer assured, leaving the comparability of the two groups uncertain. In addition, 60 (27%)
of the subjects are widows, with no information provided on the distribution between cases and
controls. Because spousal smoking is typically the, variable on which ETS exposure pivots, this
may have some bearing on the response. An adjustment is made in some analyses for years since
exposure to cigarette smoke ceased, but no information is provided to describe or support the
assumptions used to do that.
Some factors in the study itself may be contributing to the variable dose-response patterns.
First, the number of ETS subjects is fairly small. When the subjects are classified into finer
categories of exposure, the statistical variability is greatly increased (total cases and controls is
below 60, on average). Second, questionable measurements of ETS may be causing some
distortion. For instance, in the calculation of total years and total hours of ETS exposure, the
years and hours were not added for simultaneous exposure to more than one smoker or for
concurrent exposure in the home and workplace. Pipe smoking and the cigarette consumption
levels of coworkers were excluded from the weighted average of the total cigarettes per day
smoked by each household member. Additionally, the mean hours per day of exposure were
s
derived by adding the hours per day of home and workplace exposures and dividing this figure by
the age of the subjects. Thus, measurement appears to be based on the assumption that never-
smoking women were exposed to ETS evenly throughout their lives (the authors claim that only
subjects were used for which the exposure remained relatively regular during the lifetime,
although no mention was found of cases being omitted because of failure to satisfy this criterion).
Even if this assumption were valid, childhood and adulthood exposures are mixed as if the effects
of exposure are interchangeable. Interestingly, differences between exposure in childhood and
adulthood is one of the questions addressed in the article.
Although the objective is worthy, the attempt to quantitate exposure more precisely than
previous studies appears to obscure more than to clarify. Assumptions are not made very explicit
and their potential implications are not addressed well, which leaves some uneasiness about the
conclusions. The authors have published at least three articles before this study that have some
bearing on passive smoking and lung cancer, but their results are not discussed in the current
study, even when the data analyzed are from the same source (Koo et al., 1983; Koo et al., 1984;
Koo et al., 1985). Those articles, one of which describes the parent study (the 1984 citation),
appear to reach somewhat different conclusions from this study regarding the predominance of
histological type associated with passive smoking. Putting the current study's conclusions within
the context of related prior work would enhance their clarity and interpretation.
A-77
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Considering the reservations described above, the suggestion that the evidence indicates
some association of passive smoking with the location of tumors is an overinterpretation of the
data. A weaker conclusion is warranted, namely, that ETS exposure is associated with increased
lung cancer incidence. What may be of most value in this study is the analysis based on the
dichotomous classification of cases and controls as exposed or unexposed based on spousal
smoking. Two concerns, howeverj-wjll be reiterated. The ETS data are taken from a larger study
not matched on smoking status, so they are unmatched. The study includes 80 widows, without
mention of their distribution between cases and controls. In the adjusted analysis, an attempt is
made to take into account the number of years since last exposure, which would require some
assumption regarding the change of risk relative to cessation of exposure. Both of these concerns
are mitigated, however, by the similarity of the odds ratios and confidence intervals for the
unadjusted and adjusted analyses. The care and thoroughness of the study in general make the
results on the odds ratio for exposure to spousal smoke a useful contribution for evaluation with
other study outcomes.
A.22. LAMT
A.22.1. Author's Abstract
"In a case control study in Hong Kong, 445 cases of Chinese female lung cancer patients all
confirmed pathologically were compared with 445 Chinese female healthy neighborhood controls
matched for age. The predominant histological type was adenocarcinoma (47.2%). The relative
risk (RR) in ever-smokers was 3.81 (P<0.001, 95% CI = 2.86, 5.08). The RRs were statistically
significantly raised for all major cell types with significant trends between RR and amount of
tobacco smoked daily. Among never smoking women, RR for passive smoking due to a smoking
husband was 1.65 (P<0.01, 95% CI=1.16, 2.35^), with a significant trend between RR and amount
smoked by daily by the husband. When broken down by cell types the numbers were substantial
only for adenocarcinoma (RR=2.12, P<0.01, 95% CI=1.32, 3.39) with a significant trend between
RR and amount smoked daily by the husband. The results suggest that passive smoking is a risk
factor for lung cancer, particularly adenocarcinoma in Hong Kong Chinese women who never
smoked."
A.22.2. Study Description
This hospital-based case-control study was conducted in Hong Kong in 1983-86, to
investigate whether smoking is a major risk factor for lung cancer in Hong Kong Chinese women
and, if so, to determine the relationship between smoking and the histological types of lung
A-78 05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
cancer. Both active and passive smoking are of interest. The ETS subjects constitute only a
subset of the whole study, because it includes active smokers.
Eligible cases for the whole study are the 445 female patients with pathology-verified lung
cancer admitted into eight large hospitals in Hong Kong during 1983-86. Cases were interviewed
in person. Only a few eligible patients declined or were too ill to cooperate. An equal number of
healthy neighborhood controls were identified and interviewed by density sampling. Controls
were matched to cases on sex, age (±5 years), and place of residence. The cases and controls
include both never-smokers and ever-smokers, but smoking status was not used in matching.
"Never-smoker" means a person who never smoked as much as one cigarette per day, or its
equivalent, for as long as 1 year.
A woman is "ETS exposed" if her husband smoked for at least 1 year while they lived
together. If the husband was an ever-smoker, information on the type of tobacco and amount
usually smoked per day by the husband and the duration of exposure was obtained. No
information was collected on ETS exposure from other household members' smoking or smokers at
work. Single (never married) women were classified as nonexposed (6.8% and 5.2% in cases and
controls, respectively). The treatment of widowed and divorced subjects is not explicitly
addressed. Age and place of residence, as well as a series of other demographic variables, are
similar between cases and controls.
The distribution of lung cancer by cell type in ETS cases is as follows: squamous cell, 12
out of 27 (number exposed/total); small cell, 6 out of 8; adenocarcinoma, 78 out of 131; large cell,
7 out of 9; and others or unspecified, 12 out of 24. The corresponding crude odds ratios and 95%
confidence intervals are 0.85 (0.35-2.06), 3.00 (0.53-16.90), 2.12 (1.32-3.39), 3.11 (0.50-19.54),
and 1.08 (0.41-2.82), respectively. The odds ratio for all cell types combined is 1.65 (1.16, 2.35),
based on 115 out of 199 (exposed/total) cases and 152 out of 335 controls. The data for all cell
types together, and for adenocarcinoma alone, are both significant at p < 0.01. No information is
available on the airway proximity of tumors.
Trend tests were conducted for the amount smoked daily by the husband, categorized in
terms of cigarettes as "nil," 1 to 10, 11 to 20, and 21 or more. The odds ratios in the three
exposure categories are 2.18, 1.85, and 2.07, respectively, when all cell types are included. For
adenocarcinoma alone, the corresponding odds ratios are slightly higher (2.46, 2.29, and 2.89).
The dose-response relationship does not appear to increase between the lowest dose and the
highest dose, but a test for trend is significant (p < 0.01 for all cell types and p < 0.001 for
adenocarcinoma alone) when the "nil" group is included. No adjusted analyses are given.
A-79
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
The authors conclude that the significant trends observed between relative risk and
amount smoked daily by husband, for all cell types combined and for adenocarcinoma alone
supports the view that the observed association between ETS exposure and lung cancer is likely to
be causal.
A.22.3. Comments
This study is the fourth of the Hong Kong epidemiologic inquiries into tobacco smoke as a
possible etiological factor in the high rate of lung cancer, particularly adenocarcinoma, among
women. Active smoking was included as well as passive smoking because the previous studies in
Hong Kong were inconclusive. According to the authors, this led to the hypothesis that smoking
is not a risk factor for adenocarcinoma in Hong Kong Chinese women. Matching of controls to
cases was conducted for the whole study, including active smokers. It cannot be assumed,
however, that the never-smokers alone, who constitute 45% of the cases and 76% of the controls,
are matched.
Overall the study demonstrates care in planning and execution. The sample size of ETS
subjects is moderately large, providing higher statistical power than the previous Hong Kong
studies. All cases were pathologically confirmed as primary lung cancers, essentially eliminating
the potential for error due to disease misclassification. Odds ratios were calculated by histological
type for comparison. Cases and controls were interviewed personally, apparently with no proxy
respondents and very few refusals, which reduces the potential for response bias. The exclusive
use of incident cases helps to control potential selection bias, and density sampling of controls
contributes to comparability of cases and controls. For the whole study, including smokers,
healthy controls were matched to cases by sex, age, and neighborhood of residence. The mean and
standard deviation of ages are nearly identical in cases and controls. According to the authors, a
comparison by other demographic variables showed that, for the whole study, cases and controls
were also comparable in place of birth, duration of stay in Hong Kong, level of education, marital
status, and husband's occupation. Further attention to detail is evident in the clear definitions of
"never-smoker" and "ETS exposure," essential to accurate classification of subjects for analysis and
interpretation. Single women were treated as not exposed to husband's smoking, which could be a
source of bias because these women may be exposed from other household members. This
possibility was considered, however, because the article reports that similar results were obtained
when single women were excluded.
In summary, the crude odds ratios vary between 2.1 and 3.1 for small cell carcinoma,
adenocarcinoma, and large cell carcinoma, with adenocarcinoma significant at p < 0.01. The odds
A-80
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
ratios are consistently elevated at all three intensity levels of spousal smoking, varying between 1.8
and 2.9, with the odds ratio for adenocarcinoma alone somewhat higher than for all cell types
combined. There is no apparent upward trend, however, from the lowest smoking intensity (1-10
cig./day) to the highest (21+ cig./day). These statistical results are ostensibly suggestive of an
association between ETS exposure and lung cancer incidence, but they are based on only crude
data with cases and controls unmatched, even on ages. Nor are statistical methods used that could
adjust for matching variables, or other factors, in the data analysis (e.g., by stratification or
logistic regression). Although this study was carefully conducted in most respects, the disregard
for potential confounding effects leaves the authors' conclusion uncertain.
A.23. LAMW
(Note: This study is part of the thesis of LAM Wah Kit submitted to the University of Hong
Kong for the M.D. degree in 1985, entitled "A Clinical and Epidemiological Study of Carcinoma
in Hong Kong." The description given below is from Chapter 7 of the thesis only, entitled "Case-
Control Study of Passive Smoking, Kerosene Stove Usage, and Home Incense Burning in Relation
to Lung Cancer in Nonsmoking Females (1981-84)," which the author submitted in response to our
request. The abstract below was prepared by the reviewers, since none was available from the
author.)
A.23.1. Abstract
The study's objective is to investigate the hypothesis that an inhaled carcinogen may be
related to the high incidence of centrally situated adenocarcinoma of the lung observed in
nonsmoking female patients. Air pollution is probably not an important factor because it
presumably affects both men and women. Most women in Hong Kong either stay at home or join
the work force in commerce, services, or manufacturing, which are not associated with any known
risk factor for lung cancer. Three etiological activities, all predominantly in the home, are
considered in this study: passive smoking, kerosene stove cooking, and home incense burning. No
evidence was found to implicate exposure to,kerosene stove fumes or incense burning in centrally
located adenocarcinoma. There is suggestive evidence of an association between ETS exposure
from smoking husbands and occurrence of peripheral (but not central) adenocarcinoma. Why the
location tends to be peripheral instead of central is speculative.
A-81
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
A.23.2. Study Description
(Note: The details of the study are not complete in the material provided. Some useful
information, however, is available.)
The cases are all of the Chinese female patients admitted to the University Department of
Medicine, Queen Mary Hospital, Hong Kong, between January 1981 and April 1984 with
histologically and/or cytologically confirmed carcinoma of the lung of the four major cell types.
Care was taken to exclude patients with secondary carcinoma of the lung; otherwise, all patients
were included. The controls are Chinese female patients admitted to the orthopedic wards of the
hospital in the period 1982-84, comparable to lung cancer patients in age and social class. Patients
with pathological fractures due to smoking-related malignancies, or peripheral vascular disease-
related orthopedic conditions were excluded.
Both cases and controls were patients of the third-class general wards, mostly from the
lower income group. All subjects were interviewed in person. The questions covered dialect
group, occupation, smoking habits, passive smoking, domestic cooking with kerosene, and home
incense burning, in the form of a standardized questionnaire. For very ill patients, or for patients
who spoke a dialect other than Cantonese or Mandarin, the next-of-kin was interviewed, with the
patients as interpreter. The whole study, including active smokers, contains 161 cases and 185
controls, similar in age (median age is 67.5 [66] for cases [controls]), socioeconomic status (as
measured by occupation and years of schooling), and recent residence. The author considered it
unnecessary to stratify on these or any other variables.
The ETS subjects consist of 75 (144) cases (controls), including 16 (14) never-married
cases (controls). The distribution of cases by cancer cell type is as follows: squamous cell (7),
small cell (3), large cell (5), and adenocarcinoma (60). Questions related to ETS exposure include
details on each smoker in the home (husband, others, mother, and father), amount smoked per
day, hours of ETS exposure per day, and number of years smoked. Information about exposure in
the workplace includes size of the workplace, number of coworkers who smoke, exposure
time/day, and number of years of exposure at work.
Only the data for adenocarcinoma, the predominant cell type observed and the
pathogenesis of interest, are analyzed. The number of cases is 37 out of 60 (exposed/total), and
the number of controls is 64 out of 144, where ETS exposure refers to spousal smoking. The odds
ratio (calculated by the reviewers) is 2.01 (95% C.I. = 1.09-3.72). The author divides the cases by
location according to airway proximity, with 18 out of 32 (exposed/total) located centrally and 19
out of 28 in peripheral regions. The respective risk ratios are 1.61 and 2.64. Two tests were
conducted for significance, including the Bayesian risk ratio analysis and a test of the slope for
A-82
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
the exposure parameter in a simple logistic regression model. The significance levels are 0.11 and
0.19, respectively, for the central location, and 0.01 and 0.02, respectively, for peripheral tumors.
The test results differ widely for total passive smoking (home or workplace). For the central
location, the respective significance levels are 0.09 and 0.3; for peripheral locations, the
corresponding values are 0.03 and 0.15. It is suggested that the different outcomes for the two
tests applied to total passive smoking may be due to a nonlinear logistic dose-response curve or to
errors in assessing the level of exposure due to incomplete information. The apparent association
between passive smoking and peripheral adenocarcinoma (and not central tumors) in the cases was
unexpected. Based on the available raw data, exposure to a smoking spouse, cohabitant, and/or
coworker is associated with an odds ratio of 2.51 (95% C.I. = 1.34-4.67) for all cell types
combined. The author concludes that there is a suggestion of passive smoking associated with
peripheral adenocarcinoma, particularly passive smoking attributable to smoking husbands.
Kerosene and incense burning were not found to be associated with adenocarcinoma, either
central or peripheral.
A.23.3. Comments
Cases and controls appear to be comparable in age, socioeconomic status, and recent
residence for the whole study (including active smokers), although the study design is not matched
on these or other variables. Some discrepancies between cases and controls are apparent, however,
such as a higher percentage of cases than controls working outside the home (41% compared to
28%). The figures for nonsmokers alone (i.e., the ETS subjects) are not given, so .comparability is
uncertain for analysis of ETS exposure. Care has been taken to include only primary lung cancer
patients among the cases, essentially eliminating this potential source of bias. Subjects were
personally interviewed, with apparently only a small number of proxy respondents required,
although no figure is given. The interviews were apparently not blinded, but that may have not
been feasible considering the nature of the questions asked and the use of noncancer patients as
controls. Considerable attention is given to histological type of cancer and the location in terms of
airway proximity.
The author is particularly interested in the etiology of adenocarcinoma and focuses
discussion on the adenocarcinoma cases to the exclusion of others. Although the raw data
pertaining to other cell types are tabulated, more attention to those types in the analyses would
have been useful. The adenocarcinoma cases are categorized further by central and peripheral
location, which are analyzed separately. Again, a combined analysis would be useful (the
reviewers calculated the crude odds ratio for the combined data, which is given above). Although
A-83
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
logistic regression is employed as one of the two statistical tools for analysis, factors that may
differ between cases and controls are not included. Potential confounding variables need to be
controlled for, by logistic regression, poststratification, or otherwise. To claim that cases and
controls are similar in potential confounding characteristics does not alleviate the need to adjust
for them in the analysis, particularly when the ETS data are a subset of the larger data set to
which reference is made. Similarly, in testing three factors for an association with lung cancer
(passive smoking, cooking with kerosene, and burning incense), it would be useful to conduct an
analysis that will allow evaluation of the effect of each after adjustment for the other two.
The suggestive evidence that passive smoking is more likely associated with
adenocarcinoma in peripheral rather than central locations may be logical but is weak, especially
considering the lack of analytical rigor. The ratio of ETS-exposed cases of adenocarcinoma to the
total is 18 out of 32 (56%) for central locations and 19 out of 28 (68%) for peripheral locations.
This difference is not statistically significant (p = 0.26 by Fisher's exact test). Consequently, the
"apparent association" between passive smoking and peripheral adenocarcinoma (and not central
tumors) may well be due to chance alone. There is suggestive evidence in the data that passive
smoking may be associated with lung cancer (OR = 2.01, p < 0.03 for a one-sided test), but that is
based only on the crude odds ratio in unmatched data and needs to be confirmed by a more
thorough evaluation of the data that takes potential confounders into account. Overall, this study
provides some suggestive evidence for an association between passive smoking and lung cancer.
Potential confounders (including age) have not been controlled for, however, so attribution of the
elevated odds ratio to ETS exposure is uncertain.
A.24. LEE
A.24.1. Author's Abstract
"In the latter part of a large hospital case-control study of the relationship of type of
cigarette smoked to risk of various smoking-associated diseases, patients answered questions on
the smoking habits of their first spouse and on the extent of passive smoke exposure at home, at
work, during travel and during leisure. In an extension of this study an attempt was made to
obtain smoking habit data directly from the spouses of all lifelong non-smoking lung cancer cases
and of two lifelong non-smoking matched controls for each case. The attempt was made
regardless of whether the patients had answered passive smoking questions in hospital or not.
Amongst lifelong non-smokers, passive smoking was not associated with any significant
increase in risk of lung cancer, chronic bronchitis, ischemic heart disease or stroke in any analysis.
A-84
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Limitations of past studies on passive smoking are discussed and the need for further
research underlined. From all the available evidence, it appears that any effect of passive smoke
on risk of any of the major diseases that have been associated with active smoking is at most
small, and may not exist at all."
A.24.2. Study Description
This study was undertaken in England, essentially from 1979-83. Its stated objective is to
investigate the relationship between passive smoking and risk of lung cancer in nonsmokers. It is
an outgrowth, however, of a hospital-based case-control study to assess whether the risk of
cardiorespiratory disease associated with smoking varies by type of cigarette smoked. It was
initiated in 1977 in 10 hospital regions in England. In 1979, interviewers began gathering
information on passive smoking as well in four of the regions. Then in 1982, this case-control
study of the effects of passive smoking was begun using nonsmoking cases identified by the
ongoing cardiorespiratory effects study. For the new study, spouses of cases and specially selected
controls were interviewed regarding smoking habits. Previously collected data on passive smoke
exposure obtained from patients back to 1979 were used.
Basically, two substudies were conducted. One used the data obtained directly from
hospitalized cases and controls to address several sources of passive smoke, including spousal
(henceforward the "passive smoking" study); the second substudy used data obtained from the
spouses of cases and controls along with corresponding information from the patients themselves,
when available, to address spousal smoke exposure only (henceforward the "spousal smoking"
study). Cases for the passive smoking substudy were currently married lifelong nonsmokers
diagnosed with lung cancer (of any cell type), chronic bronchitis, ischemic heart disease, or stroke.
in one of four participating hospital regions. Controls were currently married lifelong nonsmoker
inpatients diagnosed with a condition definitely or probably not related to smoking and
individually matched on sex, age, hospital region, and, when possible, hospital ward and time of
interview. Thus, density sampling was used when possible. For the spousal smoking substudy,
previously married patients were excluded; the same criteria otherwise applied, except that
controls were now matched on sex, age decade, and—as far as possible—hospital and time of
interview.
Diagnoses were obtained from medical records. Exposure data were obtained through
apparently unblinded, presumably face-to-face interviews with inpatients and their spouses. A
total of 3,832 married cases and controls were interviewed regarding passive smoking through
1982; it is unclear how many potential subjects refused or died before interview. Only 56 of these
A-85
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
were married lung cancer cases meeting the spousal smoking study criteria. Spousal interview data
were obtained for 34 of these cases and 80 controls; interviews were refused by the remainder.
Although matching of cases and controls was initially carried out, it was not retained in the
analysis, and no demographic comparison of cases and controls used in the analyses is provided.
Diagnoses were apparently drawn from patients' charts, provisional diagnoses were used where no
final diagnosis was specified, no data on diagnostic technique(s) or histology is presented, and no
diagnostic verification is reported.
The patient population consists of never-smokers, defined as lifelong nonsmokers, which
presumably excludes cigar and pipe smokers. Exposure to ETS is approached in several ways.
The primary exposure is that of a spouse smoking manufactured cigarettes at some point over the
course of a marriage. Spousal smoking in the 12 months before interview was also assessed. In
addition, "regular" exposure to passive smoke in various situations (i.e., at home or work, during
travel or leisure) was assessed. The first two exposures were quantified in numbers of cigarettes
smoked per day, the others in terms of "not at all, a little, average, or a lot." Thus, it appears that
cigar and pipe smoking may not have been included in the spousal smoking exposures.
Comparison of individual responses regarding spousal smoking status by patients and their spouses
revealed a high degree of concordance (97%) for smoking during the last 12 months and a
substantial concordance (85%) for smoking during marriage. No other checks on exposure data
were reported.
The ETS patient data set includes 56 cases and 112 controls who met the initial study
criteria. Not all of these answered each passive exposure question, however, and not all met the
criteria for the spousal interview study. Similarly, spouses of 34 cases and 80 controls provided
exposure information of varying completeness. Thus the numbers involved in each analysis varied
considerably. For smoking during marriage, data obtained directly from spouses indicated that
for males and females combined, 24 of 34 lung cancer cases and 51 of 80 controls were exposed,
which yields a crude odds ratio of 1.4 for spousal smoking. With standardization for age, an odds
ratio of 1.33 (95% C.I. » 0.50-3.48) was reported. Data obtained from qualifying patients, in
contrast, revealed 13 of 29 cases and 27 of 59 controls to be exposed, yielding a crude and
adjusted odds ratio of 1.00 (95% C.I. = 0.41-2.44). Stratification by gender yielded adjusted odds
ratios from spousal interview data of 1.60 (0.44-5.78) and 1.01 (0.23-4.41) for females and males,
respectively, with corresponding odds ratios from patient interview data of 0.75 (0.24-2.40) and
1.5 (0.37-6.34). When spouses identified as smokers by interview with either source were
classified as exposed, an odds ratio of 1.00 (0.37-2.71) was obtained for female subjects. For the
larger inpatient passive smoking study population, age-standardized odds ratios for passive smoke
A-86
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
exposure at home, at work, during travel, and during leisure revealed no consistent associations,
with as many negative as positive relationships observed after adjustment for both age and
whether still currently married. The same inconsistency held true for spousal smoking during the
last 12 months and during the whole marriage. Adjustment for working in a dusty job reportedly
did not affect the conclusion that passive smoking was not associated with risk.
Spousal smoking was slightly negatively associated with chronic bronchitis, ischemic heart
disease, and stroke, whereas a combined ETS exposure index was negatively associated with heart
disease but positively associated with bronchitis and stroke.
The author concluded that the findings appear consistent with the general view, based on
all the available evidence, that any effect of passive smoking on risk of lung cancer or other
smoking-associated diseases is at most quite small, if it exists at all. The marked increases in risk
noted in some studies are more likely to be a result of bias in the study design than of a true
effect of passive smoking.
A.24,3. Comments
The heart of this study is the spousal interview investigation of lung cancer and spousal
smoking. Only 34 case spouses and 80 control spouses, and even fewer of the corresponding cases
and controls themselves, are included, which gives the study low statistical power. Because the
study began with hospital inpatient married lifelong nonsmokers, and matching on several key
factors was employed, good comparability of cases and controls would seem readily achievable.
No case-control demographics are provided, however, and matching is abandoned in the analyses.
Undoubtedly, the high rate of refusals and frequency of omitted responses (themselves a potential
source of selection and information bias) contributed to the decision to abandon matching, with
the aim of preventing further substantial reduction in numbers through exclusion of unmatched
subjects. The unfortunate result is that the comparability of the cases and controls is uncertain.
At least all are drawn from the same four hospital areas within a fairly limited timespan, which,
in combination with the other study criteria, reduces the likelihood of serious noncomparability.
Numerous opportunities for misclassification of disease and exposure status are present.
Current working diagnoses are apparently drawn from patient charts without verification, and
controls are selected from patients with diagnoses judged either probably or definitely not
associated with smoking by unspecified criteria. This creates considerable potential for
misclassification, both through inaccuracies in diagnoses generally and through inclusion of
smoking-related diseases in the control group particularly, which would produce a downward bias
in results. Exposure misreporting and recall problems would seem least likely where spouses are
A-87
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
interviewed directly about exposure within the last 12 months. Results for this situation are not
presented, although they are reportedly similar to those for smoking during marriage.
The larger inpatient study elicited smoking data from patients, and only for their first
spouse for patients who had remarried; thus, exposure occurring in subsequent marriages is not
addressed. In addition, no information on duration or level of smoking in marriage is used in any
of the spousal smoking analyses. The most likely result of these problems is nondifferential
misclassification resulting in a bias toward the null. For general estimated home, work, travel, or
leisure exposure to passive smoke, rough quantification is attempted by having patients categorize
their exposure as "not at all, a little, average, or a lot." By necessity, this is a very subjective
evaluation, and people more acclimated to smoke and tolerant of exposure might well tend to
characterize a given amount of exposure as less severe than would a person of less tolerance who
more actively avoids exposure. This tendency would produce a bias toward negative association.
Standardization for age and restriction of cases and controls to currently married lifelong
nonsmokers should control potential confounding by age, marital status, or active smoking,
although misreporting of current or former active smoking cannot be ruled out entirely. Dusty
occupation reportedly had no effect on the larger inpatient study results. Potential confounding
by race, socioeconomic status, diet, cooking habits, or any additional factors was not addressed.
One might expect the most accurate reporting of spousal smoke exposure when spouses are
interviewed directly regarding their own smoking habits, and the most inadvertent
misclassification when patients are queried about the smoking status of their first marital partner
only. Analyses along these lines yielded slightly positive associations with smoking for the former
and negative with the latter approach. No consistent pattern of association was seen for other
sources and lung cancer, although high combined exposure scores were associated positively with
chronic bronchitis and stroke and negatively with ischemic heart disease.
In summary, this study presents equivocal results that neither strongly confirm nor refute
the hypothesis that passive smoking mildly increases risk of lung cancer. The quality of the study,
however, is a limitation. The discrepant results for subject-supplied data (OR = 0.75) and spouse-
supplied data (OR » 1.60), varying degrees of completeness of information on subjects, the
subjective nature of questions regarding ETS exposure, and lack of information on intensity or
duration of husband's smoking do little to inspire confidence in the study's data and,
consequently, the results from analysis of those data.
A-88
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
A.25. LIU
A.25.1. Author's Abstract
. "In Xuanwei County, Yunnan Province, lung cancer mortality rates are among the highest
in China in both males and females. Previous studies have shown a strong association of lung
cancer mortality with indoor air pollution from 'smoky' coal combustion. In the present case-
control study, 110 newly-diagnosed lung cancer patients and 426 controls were, matched with
respect to age, sex, occupation (all subjects were farmers), and village of residence (which
provided matching with respect to fuel use). This design allowed assessment of known and
suspected lung cancer risk factors other than those mentioned above. Data from males and
females were analyzed by conditional logistic regression. In females who do not smoke, the
presence of lung cancer was statistically significantly associated with chronic bronchitis (odds
ratio [OR] = 7.37, 95% confidence interval [CI]: 2.40-22.66) and family history of lung cancer (OR
4.18, 95% CI: 1.61-10.85). Females' results also suggested an association of lung cancer with
duration of cooking food (OR 1.00, 9.18 and 14.70), but not with passive smoking (OR 0.77, 95%
CI: 0.30-1.96). In males, lung cancer was significantly associated with chronic bronchitis (OR
7.32, 95% CI: 2.86-20.18), family history of lung cancer (OR 3.78, 95% CI: 1.70-8.42), and
personal history of cooking food (OR 3.36, 95% CI: 1.27-8.88). In males a dose-response
relationship of lung cancer with smoking index (years of smoking/amount of smoking) was shown
by risks of 1.00, 2.61, 2.17 and 4.70."
A.25.2. Study Description
This study was undertaken in Xuanwei county of China's Yunnan Province, a county
whose lung cancer mortality rates are among the country's highest and wherein burning of smoky
coal indoors in unventilated pits is a common practice. The study sought to assess "the influence
of factors other than type of fuel on the occurrence of lung cancer in Xuanwei."
Cases of newly diagnosed lung cancer occurring among farmers at hospitals and clinics in
Xuanwei between November 1985 and December 1986 were identified as potential study subjects.
Up to five controls were identified for each case, depending on availability after matching on age
(±2 years), gender, and village of residence. A total of 112 cases were identified, from which 2
were excluded due to unknown addresses. Of 452 candidate controls, 26 were excluded due to
erroneous questionnaire responses. All subjects were interviewed face-to-face by trained
personnel using a standardized questionnaire, and blinding extended to both interviewers and
interviewees.
A-89
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
The final study groups consist of 54 (56) female (male) cases and 202 (224) female (male)
controls. Mean age is 52 years for both cases and controls, who are also similar in family size,
ethnicity, birthplace, dwelling type, and type of fuel used (smoky coal, wood). Separate
breakdowns for males and females are not provided. Very few of the cases (19/110= 17%) were
histologically or cytologically diagnosed, and no verification of diagnosis or exclusion of
secondary tumors was undertaken (except to monitor mortality among some of the cases).
Exposure to ETS was not evaluated for males. Among females, only one subject (a
control) reported ever having smoked, so the ETS population of females effectively consists of
never-smokers. Subjects were classified as exposed to ETS if their household contained at least
one smoker. Exposure is not quantified, and it is unclear whether former or only current
exposure is intended. No checks on exposure status or consideration of marital status are
mentioned, and no histological data are presented.
The ratio of exposed to unexposed female subjects is 45 out of 94 (176/202) for cases
(controls), yielding a crude odds ratio of 0.74. A conditional logistic regression analysis adjusted
for other risk factors (presumably the other factors referred to are age-began-cooking and years-
of-cooking) gives an odds ratio of 0.77 (95% C.I. = 0.30-1.96). No further analyses of ETS
exposure are provided.
Four non-ETS factors are significantly associated with lung cancer among females: family
history of lung cancer (OR = 4.18; 95% C.I. = 1.61-10.85), personal history of bronchitis (OR
«• 7.37; C.I. = 2.40-22.66), age-began-cooking (OR = 2.44-1.03, but with a reversing and
nonsignificant dose-response), and years-of-cooking (OR = 2.49-2.25, nonsignificant trend).
Among males, significant positive associations were noted for total smoking index, often-cooking-
own-food, family history of lung cancer, and history of chronic bronchitis, whereas age-began-
smoking, years of smoking, and intensity of smoking showed modest but nonsignificant
associations with lung cancer.
The authors conclude that "it is quite conceivable that the large amount of air pollutants
inhaled during indoor smoky coal burning in Xuanwei partly overwhelm the carcinogenic effect
of tobacco smoking" and "may also overwhelm the carcinogenic effect of passive smoking." "Our
results disclose important associations of lung cancer with factors other than fuel type and
therefore indicate that those factors must be considered in any comprehensive, quantitative risk
assessment of lung cancer in Xuanwei. Our results also confirm indirectly that smoky coal
pollution is an important determinant of lung cancer in Xuanwei."
A-90
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
A.25.3. Comments
This modestly sized study was not designed to test for effects of ETS exposure. Rather, it
is an hypothesis-generating exercise aimed at covering a broad range of possible risk factors.
Within that context, the study has considerable merit, but as an investigation of ETS it has
numerous flaws.
Restriction to farmers minimizes concerns with occupation and overall lifestyle, and
control selection, including matching on age, gender, and village, produced demographically
comparable case and control populations for males and females combined despite the enigmatic
exclusion criterion for controls. It is unknown, however, whether the groups remain comparable
after subdivision into males and females.
The use of newly diagnosed cases reduces potential selection bias due to inclusion of
prevalent cases, but the heavy reliance (83%) on clinical and radiological diagnosis and the absence
of independent confirmation or exclusion of secondary tumors introduces a strong potential for
misclassification of disease and precludes analyses by cell type. The observation that follow-up of
a number of lung cancer patients revealed that almost all died within 6 months of diagnosis does
little to confirm diagnostic validity, contrary to the authors' interpretation. Such presumably
random misclassification would make detection of an existing ETS-lung cancer association more
difficult.
Exposure data collection procedures, particularly the exclusive use of face-to-face
interviews without resort to proxies and the blinding of both interviewers and subjects, are
laudable. For ETS, however, the exposure measure used is nonspecific and nonquantitative.
Complications due to past exposure and differences in degree or duration could distort the
observed disease-exposure relationship, probably biasing results toward no effect. ,
Potential confounding is not adequately addressed in the statistical analysis. The authors
are particularly concerned with indoor smoky coal burning due to the known strong correlation
between smoky coal use and lung cancer mortality in Xuanwei. Wishing to focus their
investigations on factors other than smoky coal, they matched cases and controls on village, which
"provided effective matching on fuel type." But because age and a host of other demographic
factors, as well as smoky coal consumption, were comparably distributed in cases and controls (see
Study Description), these factors were not considered further in the data analysis. This is a
serious flaw, for pair matching was not retained in the analysis; thus, none of the above factors is
effectively controlled for. The conditional regression analyses do control for risk factors other
than those cited above, but exclusion of age, fuel type (e.g., smoky coal), and degree of exposure
to fuel fumes may produce misleading results.
A-91
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
The presence of other significant risk factors for lung cancer makes detection of an effect
from ETS, if present, less likely. Masking by the presence of smoky coal and other factors in the
study environment is probably a factor in the remarkably weak association between active
smoking and lung cancer among study males (adjusted OR = 1.36). If even an effect of active
smoking remains largely obscured under study conditions, it is unlikely that an effect of ETS
would be detected.
Overall, this study makes important contributions to its principal objectives but is not
helpful in assessing ETS and lung cancer. It is observed, for example, that persons in areas of
Xuanwei with high lung cancer rates (and high smoky coal consumption) may inhale more BAP
(benzo-[a]-pyrene) by spending 8 hours indoors than by smoking 20 cigarettes. Due to such
factors, the authors observe, "the effect of passive smoking on lung cancer may depend on local
environmental factors and results obtained in a given region may therefore not be applicable to
other regions." Avoidance of areas atypically rich in competing exposures and careful control of
potential confounders and interactive risk factors must be key objectives in studies of ETS and
lung cancer.
A.26. PERS
A.26.1. Author's Abstract
"The relation between passive smoking and lung cancer was examined by means of a case-
control study in a cohort of 27,409 nonsmoking Swedish women identified from questionnaires
mailed in 1961 and 1963. A total of 77 cases of primary carcinoma of the bronchus or lung were
found in a follow-up of the cohort through 1980. A new questionnaire in 1984 provided
information on smoking by study subjects and their spouses as well as on potential confounding
factors. The study revealed a relative risk of 3.3, constituting a statistically significant increase (p
< 0.05) for squamous cell and small cell carcinomas in women married to smokers and a positive
dose-response relation. No consistent effect could be seen for other histologic types, indicating
that passive smoking is related primarily to those forms of lung cancer which slow the highest
relative risks in smokers."
A.26.2. Study Description
This case-control study, undertaken to explore the role of passive smoking in lung cancer,
is based on cohorts of Swedish women assembled prior to 1963. Nonsmokers were drawn from
these cohorts to create matched case and control groups.
A-92
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Cases are nonsmoking Swedish women included in the Swedish National Census or Twin
Registry who responded to smoking status questionnaires in 1961-63 and who subsequently
developed primary lung or bronchial cancer by 1980. Two control groups were cumulatively
sampled from National Census or Twin Registry subjects who did not develop lung or bronchial
cancer. In group 1, two controls were matched to each case on year of birth (±1 year). In group
2, two controls were matched to each case (2:1) on year of birth (±1 year) and vital status in
1980. Thus, there were 58 cases and 232 controls from the National Census and 34 cases and 136
controls from the Twin Registry. A follow-up questionnaire that included questions on spousal
and parental smoking habits was distributed to each subject or the next-of-kin in 1984. Out of 92
cases of tracheal, bronchial, lung, or pleural cancer occurring by 1980, 15 cases in which a
diagnosis of primary cancer of the lung or bronchus was not established were excluded. Exclusion
of women indicated to be active smokers according to the 1984 questionnaire, or for whom ETS
exposure information was not available, eliminated a further 10 cases. Active smoking and lack of
exposure information eliminated 21 of the 368 controls initially assembled. Histological
confirmation was available for 64 of the 77 cases with primary lung or bronchial cancer; 12 cases
were cytologically confirmed; and the remaining case was verified at autopsy.
Never-smokers are subjects who report that they have never smoked any form of tobacco.
A woman is ETS-exposed if she has ever been married to a tobacco smoker; for women married
more than once, only the longest marriage is considered. Exposure to spousal smoking is
quantified in units of cigarettes per day or packs of pipe tobacco per week; parental smoke
exposure is defined as 0, 1,2, etc. (equal to the number of parents who smoke). No other sources
of ETS exposure are considered. Never-smoking status was checked by comparing the responses
to the 1961-63 questionnaires with those obtained in 1984. Data on sources of ETS were not
checked. Never-married women were classified as nonexposed to spousal smoke; widows and
divorcees were classified according to the smoking status of the former husband with whom they
had lived the longest. Of the never-smoking cases for whom passive smoking information was
available, squamous and small cell tumors constituted 20 cases, 13 of whom were exposed to
spousal smoke; of the other 47 cases, 20 were exposed to spousal smoke.
Responses to the ETS questionnaire were available for a total of 81 never-smoking cases
and 347 never-smoking controls. The 67 cases with primary lung or bronchial cancer constitute
the ETS study subjects. It is not clear how many of the 347 potential controls were employed in
each analysis. Presumably many (up to 4 for each excluded case from the original 81 never-
smoking cases) were not used in the matched analysis, whereas most or all were used in the
unmatched analyses described subsequently.
A-93
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
A total of 33 of the 67 cases were exposed to spousal smoking. Among the never-smoking
women, matched analyses indicate that the odds ratio for marriage to a smoker is 3.8 (95% C.I. =
1.1-16.9) for squamous or small cell cancer compared to control group 1, 3.4 (0.8-20.1) compared
to control group 2, and 3.3 (1.1-11.4) compared to both groups combined. For other cell types,
corresponding odds ratios are 0.7, 0.8, and 0.8, respectively. Subsequent analyses abandoned
matching and pooled all controls. For squamous and small cell cancer, high exposure to spousal
smoking (15 or more cig./day or at least one pack of pipe tobacco/week for 30+ years) is
associated with an age-adjusted odds ratio of 6.4 (1.1-34.7), whereas the lower exposure is
associated with an odds ratio of 1.8 (0.6-5.3). The estimated odds ratios for other types of cancer
are also elevated for the higher exposure, but not at the lower one. Odds ratios adjusted for age
and spousal smoking when at least one parent smokes as well are above 1 (1.9; 95% C.I. = 0.5-6.2)
for squamous and small cell types but not for other types.
Logistic regression analyses reportedly produced the same results as did the stratified
analyses. In addition, occupation, household radon, and urban or rural status had no significant
effect. It is notable, however, that for all cancers combined, the odds ratio for radon exposure is
1.4 (0.4-5.4), the odds ratio for spousal smoking is 1.2 (0.6-2.6), and the odds ratio for radon and
spousal smoking combined is 2.5 (0.8-8.5). No separate analyses for squamous and small cell
cancer are provided for radon and other potential confounders. The authors conclude that
exposure to ETS is related primarily to the forms of lung cancer that show the highest relative
risks in smokers. The results are internally consistent.
A.26.3. Comments
Although based on cohorts assembled for other purposes, this case-control study was
specifically designed to investigate passive smoke exposure. Thus, all participants are ETS
subjects, which are matched. Matching criteria are rather modest—birthdate (±1 year) for control
group 1 and birthdate and vital status for control group 2. Because the study targeted all cases
detected in the same cohorts from which matching controls were randomly drawn, good
comparability of cases and controls is likely. No demographic comparisons of cases and controls
for whom ETS information was available—and thus who constituted the analytical subjects—were
provided to confirm this, however. Data on active smoking among subjects were collected both at
the start and after the end of mortality monitoring, providing an opportunity to verify the
nonsmoking status over time and exclude individuals whose status had changed (apparently those
reported in 1984 to have smoked daily for at least 2 years were so excluded). Thus, the
probability of significant misclassification of active smoking status is low. Data on passive
A-94
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
smoking were collected only after the end of mortality monitoring and by necessity employed
proxy respondents extensively, so some misclassification of exposure is likely. Self-administration
of questionnaires eliminates interviewer bias as a source of error, making misclassification less
likely to be systematic, but preferential recall of smoke exposure by relatives of cancer victims
could have produced a bias. Misclassification of disease is unlikely to have been a problem
because most cases were histologically diagnosed and secondary lung cancers were excluded.
Consideration of spousal smoke exposure only in their longest marriage among women
married more than once means that some of the unexposed group probably had substantial
exposure to spousal smoking, creating a bias toward no association. Classification of all
never-married women as unexposed despite possible smoking by cohabitants creates the same bias.
Few subjects (less than 20%) were single, but the frequency of remarriage is unknown; therefore,
it is unclear how important this bias might have been. Lack of consideration of workplace smoke
exposure may also have contributed a bias toward the null hypothesis of no association.
The authors addressed a number of potential confounders. Restriction of subjects to
women eliminates potential confounding by gender, and age is addressed by retaining
age-matching or, alternatively, adjusting for age in all analyses. Reportedly neither occupation,
radon, nor urban residence had significant confounding effects, which makes confounding by
other factors related to socioeconomic status or lifestyle unlikely, too. An analyses of parental
smoking controlled for spousal smoking. The authors do, however, present evidence that the odds
ratio for simultaneous exposure to radon and spousal smoke approximately equals the sum df the
separate odds ratios for radon and spousal smoke, consistent with additivity of the effects. But,
perhaps due to limited numbers, they report results only for all cancers combined rather than for
the squamous and small cell subgroup in which the only significant spousal smoking association
was observed.
In summary, this study reports a consistent, dose-related, and (for high exposure levels)
statistically significant positive association between exposure to spousal tobacco smoke and
squamous and small cell carcinoma of the lung; a positive but nonsignificant association was also
observed for parental smoke exposure. No significant associations were observed for other cell
types. The observed associations apparently are not due to confounding by other major risk
factors, although dietary and smoking habits were not directly addressed. A possible recall bias
cannot be ruled out but seems unlikely given the negative results obtained for cancers other than
squamous and small cell. The study provides a useful contribution to investigation of the
relationship between ETS exposure and lung cancer.
A-95
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
A.27. SHIM
A.27.1. Author's Abstract
"A case-control study of Japanese women in Nagoya was conducted to investigate the
significance of passive smoking and other factors in relation to the etiology of female lung cancer.
A total of 90 nonsmoking patients with primary lung cancer and their age- and hospital-matched
female controls were asked to fill in a questionnaire in the hospital. Elevated relative risk (RR) of
lung cancer was observed for passive smoking from mother (RR=4.0; p<0.05) and from husband's
father (RR=3.2; p<0.05). No association was observed between the risk of lung cancer and
smoking of husband or passive smoke exposure at work. Occupational exposure to iron or other
metals also showed high risk (RR=4.8; p<0.05). No appreciable differences in food intakes were
observed between cases and controls."
A.27.2. Study Description
This study was undertaken in Nagoya, Japan, during 1982-85 to investigate the
significance of passive smoking and other factors such as occupational history, domestic heating
system, and dietary habits in the etiology of lung cancer in nonsmoking Japanese women. All data
were collected specifically for this study, which was limited to never-smokers.
All subjects were obtained from four hospitals in Nagoya. Cases are women with primary
lung cancer (of any type) treated in these hospitals between August 1982 and July 1985 who
reported themselves to be never-smokers and consented to interview. Controls are women with a
diagnosis other than lung cancer from the same or adjacent wards with controls matched 2:1 with
cases on age (±1 year), hospital, and date of admission. Cases were not restricted to incident
disease, but controls were essentially density-sampled by admission date. Data collection was by
self-administered questionnaire; no attempt at blinding is described. Of 118 female lung cancer
cases treated during the study period, four refused to participate in the study and 24 were
excluded as current or former smokers. Only a single matching control could be found for 17 of
the cases. No other information on loss of potential controls is provided. There is a total of 90
(163) cases (controls), with 52 (91) currently married to a smoker. Cases and controls share
identical age ranges (35-81 years) and have nearly identical mean ages (59 years for cases, 58 for
controls). All cases were histologically diagnosed, excluding secondary lung cancers.
All study subjects are self-reported never-smokers. A number of individual sources of
ETS in the home are considered, including smoking by mother, father, husband, father-in-law,
mother-in-law, offspring, and siblings. For each of these sources, smoking in the home at any
time constituted exposure. Workplace exposure was characterized simply as presence or absence;
A-96
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
for other exposures, the number of cigarettes per day was obtained. In addition, data on length of
marriage, time spent in the same room as the wife, and total number of cigarettes smoked were
obtained for husbands. Exposure data were not checked, and marital status was not considered in
the design or analysis of the study. The predominant type of lung cancer is adenocarcinoma (69
out of 90 cases), followed by squamous (13), large cell (4), small cell (3), and adenoid cystic
carcinoma (1). No data on airway proximity are provided.
Logistic regression was used to estimate the relative risk for each source of ETS exposure.
No significant association with lung cancer was noted for smoking by the husband (RR = 1.1),
father (RR = 1.1), husband's mother (RR = 0.8), offspring (RR = 0.8), or siblings (RR = 0.8);
smoking by the subject's mother (RR = 4.0) and by the husband's father (RR = 3.2), however, are
significant (p < 0.05). None of eight dietary factors, including green-yellow vegetable and fruit
intake, demonstrated a significant association, nor did type of cooking fuel or frequency of
cooking oil use. Occupational history of exposure to iron or other metals shows a moderately
strong but nonsignificant association (RR = 2.8), whereas for use of kerosene, coal, or charcoal
heating there is a mild.association (RR = 1.6-1.7).
Simultaneous stratification by father-in-law's and mother's smoking indicates that the
effects of the two exposures are not additive. Smoking by father-in-law, smoking by mother, and
occupational metal exposure were included simultaneously in a logistic regression model. After
adjusting the effect of each variable for the other two, the relative risk for maternal smoking,
father-in-law's smoking, and metal exposure are 2.1, 3.2 (p < 0.05), and 2.4, respectively. The
authors conclude that the exposure to tobacco smoke from household members (i.e., mother or
husband's father) could be associated with female lung cancer. As the precise situation of passive
smoking in the home or other places is still unclear, however, they find that further studies are
needed to clarify the significance of passive smoking in relation to the etiology of lung cancer in
Japanese women.
A.27.3. Comments
This study employs a moderate number of well-matched cases and controls. Their
comparability appears good, as supported by the identical age ranges and similar mean age and
occupational categories for the two groups. A further strength of the study is its lack of reliance
on proxy information with attendant potential for inaccurate recall. Exposure information was
obtained from self-administered questionnaires, which eliminates the possibility of interviewer
bias but may lead to inaccuracy due to misinterpretation of questions or varying care in their
completion. Such problems with exposure information would tend to mask any actual association.
A-97
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Lung cancer was histologically diagnosed in all subjects and secondary lung cancers excluded, so
diagnostic accuracy appears good for cases. Control diagnoses, however, were not validated, so
some smoking-related disorders (in addition to the heart conditions noted in 3% of controls) may
be included among the controls, a problem that once again would tend to reduce any observed
association. '
Restriction of subjects to never-smokers maximizes efficiency because effects of passive
smoking would likely be dwarfed by active smoking. But it is unclear precisely what subjects
were asked about their smoking status. Were any cut-points regarding pack-years or cigarettes
per day specified? Was former smoking specifically questioned? Thus, some misclassification of
smoking status may have occurred, and if a greater proportion of persons with smoking family
members misreport themselves to be never-smokers, this would create an upward bias.
The authors restrict their assessment of exposure from relatives to at-home smoking,
which may be more meaningful than total smoking as a potential source of passive smoke
exposure. Furthermore, they collected data on smoking habits of all relatives, not just spouses or
parents, thus reducing the chance of missing an exposure source. On the other hand, there is no
consideration of total household smoking (all sources combined), cumulative exposure (except for
husbands), or of pipe or cigar smoking; nor is there differentiation of current and former
exposure—all potential sources of exposure misclassification, which would tend to make an
association more difficult to detect.
Of the several sources of ETS exposure at home, only the relative risks for smoking by the
mother and by the father-in-law are suggestive, and both of these are significant (p < 0.05).
When these sources are considered simultaneously, however, and the effect of each is adjusted for
the other, smoking by the husband's father remains significant (RR = 3.2; p < 0.05) but the effect
of mother's smoking is diminished (RR = 2.1) and is not statistically significant. The authors'
emphasis on the significance of exposure in childhood from maternal smoking appears misplaced.
Exposure from the father-in-law is, of course, in adulthood. There is no evidence of an effect
from husband's smoking (RR = 1.1), however, and these exposure sources were considered
simultaneously so that the effect of one could be adjusted for the other. The large number of
comparisons (e.g., eight groupings of passive smoke exposure, alternative spousal exposure
measures, and several occupational and eight dietary factors) increases the likelihood that an
observed relative risk will appear to be significant by chance alone (the effect of multiple
comparisons).
Another aspect of the statistical analysis worth noting is that, although cases and controls
appear well matched on age, hospital, and hospital admission date, these factors and other
A-98
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
potential confounders are not included in an adjusted analyses of the data (aside from the example
with three sources of exposure described above). Consequently, possible confounding cannot be
ruled out, although the demographic similarities between cases and controls make severe
confounding less likely.
In summary, this study presents some interesting results. It finds a strong (adjusted
RR = 3.2) and statistically significant association between father-in-law's smoking at home and
lung cancer and associations for maternal smoking and occupational metal exposure as well. The
lack of association for any of the other sources of ETS examined could be due to problems with
exposure assessment and control disease criteria. Equally, however, given the unclear treatment of
matching factors in the analysis, and the number of variables explored, the few substantial
associations noted might be due to chance, confounding, or both. Were potential confounders
clearly treated in their analyses, this study would have made a stronger contribution. As it stands,
the study's data are of moderate utility, providing the number of comparisons and limitations
regarding bias are kept in mind.
A.28. SOBU
A.28.1. Author's Abstract
"A hospital-based case-control study among non-smoking women was conducted to clarify
risk factors in non-smoking females in Japan. Cases consisted of 144 non-smoking female lung
cancer patients, and these were compared to 713 non-smoking female controls. The odds ratios
(95% confidence interval) for use of wood or straw as cooking fuels when subjects were 30 years
old was estimated as 1.77 (1.08 to 2.91). For those whose household members, other than
husbands, had smoked, the odds ratio was estimated as 1.50 (1.01 to 2.32). For those whose
mothers had smoked, the odds ratio was estimated as 1.28 (0.71 to 2.31). Use of heating
appliances did not show an elevated risk. Some points to be noted in this study of low-risk agents
for lung cancer are discussed."
A.28.2. Study Description
This study was conducted in Osaka, Japan, to clarify risk factors for lung cancer in
nonsmoking females in Japan. Of interest are the roles of both active and passive smoking and
other indoor air pollutants, particularly smoke or fumes from sources of indoor cooking and
heating. This article reports only on female nonsmokers in the study, which is not matched on
any variables. A very similar article presenting interim results and using slightly fewer subjects
than the one described here is by Sobue and coworkers (1990).
A-99
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Cases consist of all newly admitted lung cancer patients in eight Osaka hospitals between
January 1986 and December 1988. Controls were collected from newly admitted patients in one or
two other wards of the same hospitals during that period. Almost 90% of the controls were
admitted as cancer patients, about half of which were diagnosed with breast cancer.
Self-administered questionnaires designed for this study were completed by both cases and
controls at the time of hospital admission. Cases are incident and control sampling is density,
unmatched aside from the time of hospital admission (within 1.5 years). The entire study,
including active smokers and males, consists of 295 (1,079) female (male) cases and 1,073 (1,369)
female (male) controls. Nonsmoking females compose 156 cases, of which there was missing
information on 12. The resultant number of ETS subjects is 144 (731) female nonsmoking cases
(controls). The age distribution of the cases (controls) is as follows: 40 to 49, 20 (238); 50 to 59,
34 (229); 60 to 69, 41 (186); and 70 to 79, 34 (78). The corresponding percentages are 14 (33), 34
(31), 28 (25), and 24 (11), which indicates that controls tend to be younger than cases. Also, the
mean age of cases (controls) is 60 (56). There was no systematic review of histological diagnosis.
AH original diagnoses were confirmed microscopically, however, and all the pathologists involved
in the eight participating hospitals were experienced specialists in lung cancer. Thus, the
likelihood of secondary lung cancers among the cases should be small.
Several sources of ETS exposure are included, all of which occur in the home. Exposure
in adulthood is expressed by two measures—smoking by the husband and by other household
members (the last category consists chiefly of households where the husband's father and/or sons
smoke). Three sources of exposure in childhood are considered—father smokes, mother smokes,
and other household members smoke. No information is provided on how exposure to spousal
smoking is handled for unmarried women (single, divorced, or separated). The entire complement
of cases and controls is included in the summary data for each of the five sources of exposure
given above. If only married women were included in the study, no mention of it was found.
The histological data for ETS subjects are not classified by exposure to ETS, but the
percentage of cases by cell type are given: squamous cell (8), small cell (5), adenocarcinoma (78),
large cell (5), and other (4). The ETS data on spousal smoking consists of 80 out of 144
(exposed/total) cases and 395 out of 731 controls, for an odds ratio of 1.13 (95% C.I. = 0.78-1.63).
(Our calculations give 1.06 [0.74-1.52].) The odds ratio for ETS exposure from other household
members in adulthood is 1.57 (95% C.I. = 1.07-2.31). (Our calculated values are 1.77 [1.21-2.58].)
For ETS exposure in childhood by the father, by the mother, and by other household members,
the respective odds ratios are 0.79 (95% C.I. = 0.52-1.21), 1.33 (95% C.I. = 0.74-2.37), and 1.18
(95% C.I. - 0.76-1.84). Tests were conducted by the Mantel-Haenszel procedure, with
A-100
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
stratification by age and education (two levels). Analysis by logistic regression, adjusted for age
at time of hospitalization, was conducted for two of the exposure measures described above with
similar outcomes. Based on this evidence, the author concludes that for childhood exposure, a
slight increase of risk was suggested for those with smoking mothers, although statistical
significance was not observed. For exposure in adulthood, an elevated risk was estimated for
those with smoking household members other than husbands.
The statistical analysis includes exposure to sources other than ETS, namely, the use of
wood or straw as cooking fuel, the use of heating equipment that pollutes the room with
combustion products, and the use of charcoal foot warmers. All exposures considered, including
ETS, are smoke or fumes from products burned indoors. It is concluded that significantly
elevated risks were observed for subjects who had used wood or straw as cooking fuels at 30 years
of age (OR = 1.89; 95% C.I. = 1.16-3.06). No elevated risks were found for sources of indoor
heating (use of kerosene, gas, coal, charcoal, and wood stoves without chimneys). Similarly, no
significance was found for the use of charcoal foot warmers, a practice that was popular until the
1960's.
A.28.3. Comments
With 144 cases and 731 controls, the sample size is larger than many of the other
case-control studies on ETS. Information on cases and controls was obtained by self-administered
questionnaire, which is generally considered less reliable than face-to-face interviews. The
questionnaires were presumably completed by the subjects themselves in all cases, however, which
is preferable to proxy-supplied information. The information supplied was not verified from
other sources, as noted by the authors in reference to testing for biomarkers of exposure to
tobacco smoke (they note that laboratory tests can only detect recent exposure, but they could still
be useful in eliminating current smpkers who may misreport themselves as never-smokers).
Although cases and controls were newly diagnosed patients within a short time period in the eight
participating hospitals and were supplied with the same questionnaire, there are still some
questions regarding the comparability of cases and controls and their representativeness of the
target population.
Controls tend to be younger than cases: Mean ages are 56 and 60, respectively, and 33% of
controls, compared to 14% of cases, are below the age of 40. Controls also tend to be more
educated than cases, with 69% of controls having completed 10 or more years of education
compared to 52% of cases. Differences in age and educational level further reflect differences in
lifestyle and socioeconomic status that may affect risk of disease. Also, the controls are
A-101
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
predominantly cancer patients too, almost half with breast cancer. Although the diseases of the
controls may not be known to be related to tobacco use, controls may be a biased sample (as noted
by the authors). Furthermore, the statistical analysis stratifies on age and education so, even
though cases and controls were not strictly matched on these variables, the reported results should
not be due to confounding by either of these factors. On the other hand, exclusion of breast
cancer controls reportedly leaves the results unchanged.
Although some of the issues and reservations described above are methodological in nature
and apply to the study throughout, others are specific to the ETS data alone. For example, one
might expect a question regarding the use of cooking with wood or straw at age 15 and at age 30
to be open to little subjective interpretation or error in recall, presuming that methods of cooking
persisted for several years between changes within a household. Although there is some suggestive
evidence of increased lung cancer from ETS exposure (the reservations above set aside for the
moment), the statistical evidence may be stronger for an association between lung cancer
prevalence and use of wood or straw for cooking at age 30. Further support is provided by the
observation that among those who had used wood or straw for cooking at age 30, 90% had also
used those fuels at age 15, suggesting extended exposure in most cases. The age distribution of
those exposed to wood or straw cooking is not given, but exposure at 30 years of age and before
would allow for the long latency expected for lung cancer because 86% of the patients are at least
50 years of age.
The smoke from cooking sources may obscure or distort any impact of ETS exposure
because the two sources probably contain some of the same carcinogens. The temporal dimension
of exposure may also be a factor because indoor smoke from cooking may be less common at
present than 30 years ago in comparison to ETS exposure. Further statistical analysis to adjust the
effect of ETS exposure for the presence of smoke from cooking might aid interpretation of the
results in this study, depending on the extent of confounding present.
A.29. STOC
A.29.1. Author's Abstract
(Note: This study has not been published. Only the abstract is available, which is given below.)
"Risk factors for lung cancer among women who had never smoked cigarettes were
examined in an ongoing, population based, case-control study conducted in Florida. One hundred
and twenty-four primary carcinomas of the lung, and 241 control women who had never smoked
were included. Results suggest that childhood and adult exposures to environmental tobacco
smoke may increase the risk of lung cancer among women who never smoked cigarettes. Having a
A-102
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
husband who smoked cigarettes resulted in a statistically significant increase in risk of lung cancer
among women who had never smoked, with an odds ratio of 1.8 (95% C.I. 1.1-2.9). A 40%
increase in risk was observed among women with less than 25 years of exposure to a spouse who
smoked, when compared to women who reported their spouse had never smoked, with the risk
increasing to 60% among women exposed 25 years or longer.
When exposure to tobacco smoke in childhood was considered, the data were less
consistent. Having a parent who had smoked during the respondent's childhood did not increase
the risk of lung cancer. However, among those respondents with high levels of exposure to
parental smoking, an excess risk, although not statistically significant, was observed. Never
smoking women who accumulated 25 or more exposure years experience a 70% increase in risk
(OR = 1.7, 95% C.I. 0.8-3.6) of lung cancer compared to women who reported neither parent had
smoked cigarettes."
A.30. SVEN
A.30.1. Author's Abstract
"In a population based-case control study the association between female lung cancer and
some possible etiological agents was investigated: 210 incident cases in Stockholm County, Sweden,
and 209 age-matched population controls were interviewed about their exposure experiences
according to a structured questionnaire. A strong association between smoking habits and lung
cancer risk was found for all histological subgroups. Relative cancer risk was found for all
histologic subgroups. Relative risk for those who had smoked daily during at least one year
ranged between 3.1 for adenocarcinoma to 33.7 for small cell carcinoma in a comparison with
never-smokers. All histological types showed strong dose-response relationships for average daily
cigarette consumption, duration of smoking, and cumulative smoking. There was no consistent
effect of parental smoking on the lung cancer risk in smokers. Only 38 cases had never been
regular smokers and the risk estimates for exposure to environmental tobacco smoke were
inconclusive. The high relative risks of small cell and squamous cell carcinoma associated with
smoking may have relative implications for risk assessments regarding passive smoking."
A.30.2. Study Description
This study was undertaken in Stockholm County, Sweden, from 1983 to 1986 to investigate
the association between female lung cancer and some possible etiologic agents, particularly active
and passive smoking. Because active smoking was an exposure of interest, cases and controls were
not matched on smoking status; thus, the ETS study population is unmatched.
A-103
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Cases are Swedish-speaking women with primary lung cancer from three Stockholm
County hospitals who were willing and able to be interviewed between September 1983 and
December 1985. Cases with carcinoid tumors were excluded from the ETS analysis. Both
population and hospital-based control groups were assembled. Population controls were women
randomly selected from the county population register, matched to a case on birthdate and
interviewed between September 1983 and December 1986. Hospital controls were subjects
originally interviewed as potential lung cancer cases but subsequently diagnosed with
nonmalignant conditions. Population controls were enlisted and interviewed as soon as a case's
diagnosis was confirmed, but because this confirmation took as long as a year after the interview,
controls were not density sampled. Unblinded interviews were conducted face to face with all
cases (and hospital controls) and 58% of the total population controls; the remainder were
interviewed by telephone.
After exclusion of 21 potential cases due to initial diagnostic uncertainty, refusal, or illness
precluding interview, 210 confirmed cases remained. Elimination of 172 ever-smokers and four
subjects with carcinoid or not-microscopically-confirmed tumors left 34 never-smoking cases.
Similarly, 209 population and 191 hospital controls were included in the total study, but a
combined total of only 174 were never-smokers. The total case population averaged 62.5 years of
age, but no other demographic information regarding cases or controls is provided. All cases used
in the ETS analyses were histologically or cytologically confirmed primary lung cancers.
Daily smoking for at least 1 year is the criterion for a smoker; all other persons are
considered never-smokers. Pipe and cigar smoking are never specifically addressed. Exposure to
ETS is calculated for four sources: mother, father, home, and work. Having a smoking mother or
father (at any time during ages 0-9 years) constitutes exposure to that particular source, whereas
the presence of a smoker at home and work constitutes exposure. Adulthood and total lifetime
exposure are considered separately for home and workplace exposure. Exposure levels are
arbitrarily scored 1 for nonexposure, 2 for exposure to one source, and 3 for exposure to both
sources in trend analyses of never-smokers, where exposures are considered in pairs (i.e., maternal
and paternal smoking, home and workplace exposure). No other units of ETS exposure are used.
Adenocarcinomas constituted 22, squamous cell 5, and small cell 2 of the 34 lung cancers
occurring among never-smokers in the ETS population; no further histologic details regarding the
ETS study population are provided.
To maximize available case numbers, parental smoking was first analyzed among all cases
and community controls using stratification to adjust for active smoking (cig./day) and age. A
risk of 1.8 (95% C.I. = 0.5-7.0) was estimated for maternal smoking and 0.8 (0.3-1.4) for paternal
A-104
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
smoking. A trend analysis in which maternal, paternal only, and no parental smoke exposure were
scored as 3, 2, and 1, respectively, revealed no indication of trend (p = 0.9). Analyses restricted to
never-smokers used both community and hospital-based controls combined. Among cases
(controls), for childhood up through 9 years of age, 3 (5) had smoking mothers, 12 (71) had
smoking fathers (but not mothers), and 19 (98) were unexposed. This yielded atfage-adjusted risk
estimate of 3.3 for maternal smoking (with or without paternal smoking) and 0.9 for paternal
smoking during childhood. Adult exposure at home and at work yielded an estimated risk of 2.1,
whereas exposure at home or work yielded a risk of 1.2. For lifetime exposure, the estimated risks
for exposure as both a child and adult and as either a child or an adult were 1.9 and 1.4,
respectively. None of these associations were statistically significant, and no significant trends
were observed. The authors conclude that the results pertaining to ETS in the present study were
not conclusive. The small number of never-smokers among the cases could be one important
reason. It should be noted, however, that most of the point estimates of relative risk were greater
than unity, which agree with results from previous studies on ETS exposure and with risk
estimates concerning active smoking.
A.30.3. Comments
This study was undertaken to explore the role of active as well as passive smoking in lung
cancer. After exclusion of active smokers, the available number of cases is too small to yield
much statistical power.
Cases and population-based controls were initially matched on date of birth, but this
matching was abandoned in the ETS analysis; furthermore, unmatched hospital-based controls are
combined with the population-based controls in most analyses to boost available numbers. The
comparability of these groups is thus unclear, and the authors provide no demographic
comparisons to facilitate assessment of this potential problem. The reported similarity of results
using only population-based controls is reassuring, but no details are provided as to how similar
results actually were.
Diagnostic misclassification of cases is unlikely, given the histological or cytological
confirmation of all cases and exclusion of secondary cancers. All cases were interviewed face to
face, but 42% of controls were interviewed by telephone. The accuracy of responses may thus be
lower for controls than for cases. And because interviews were not conducted blindly, inflation of
estimated associations through interview bias is possible. A potential bias is also introduced by the
rather large amount of active smoking required for classification as an ever-smoker. This allows
considerable active smoking among persons in the never-smoker group, the effect of which could
A-105
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
mask an effect of passive exposure, or, if covarying positively with passive smoking, cause
overestimation of association.
The first set of analyses of paternal and maternal smoking includes ever-smokers while
attempting to adjust for active smoking on the basis of average daily cigarette consumption. The
adequacy of this adjustment is questionable given the large estimated risks associated with active
smoking relative to those posited for passive smoking, so the elevated estimated risks for maternal
smoking obtained in these analyses are of questionable validity.
Restriction of the analyses to never-smokers similarly produces an elevated odds ratio for
maternal smoking of 3.3, but the numbers involved (three cases and five controls) are so small that
this value is quite unstable. A pattern of increasing estimated risk with increasing sources of
exposure (at home or at work) as an adult and increasing periods of exposure (in childhood or
adulthood) over the lifetime is suggestive of an association between lung cancer and ETS, but
again small numbers preclude statistical significance of these results.
Restriction of the study population to females rules out the possibility of confounding due
to gender. The likelihood of an ethnicity effect is reduced by restriction to Swedish-speaking
residents of Stockholm County, and age is reportedly controlled for in all analyses. No other
potential confounders are addressed. For example, marital status is not considered in the analyses
of spousal smoking, leaving open the possibility that nonsmoking-related differences between
married and unmarried women contributed to the observed association. The reported similarity of
results when only population controls were used instead of hospital and population controls
combined provides a general argument against confounding, although no specifics regarding the
degree of similarity were supplied.
In summary, this study presents consistent evidence of associations between lung cancer
and maternal, home, and workplace passive smoking exposure. Limited numbers preclude
statistical significance and interviewer bias or confounding due to dietary or other factors cannot
be ruled out as contributors to the observed results. Bearing these limitations in mind, the study's
results are inconclusive but (excluding the analyses that include active smokers) do make a useful
contribution to the pool of information available regarding ETS and lung cancer.
A.31. TRIG
A.31.1. Author's Abstract
"Fifty-one women with lung cancer and 163 other hospital patients were interviewed
regarding the smoking habits of themselves and their husbands. Forty of the lung cancer cases
and 149 of the other patients were nonsmokers. Among the nonsmoking women there was a
A-106 05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
statistically significant difference between the cancer cases and the other patients with respect to
their husbands1 smoking habits. Estimates of the relative risk of lung cancer associated with
having a husband who smokes were 2.4 for a smoker of less than one pack and 3.4 for women
whose husbands smoked more than one pack of cigarettes per day. The limitations of the data are
examined; it is evident that further investigation of this issue is warranted."
A.31.2. Study Description
This study was undertaken in Athens, Greece, to investigate the relationship of spousal
smoking and lung cancer. All female Caucasian Athenian residents admitted to one of three chest
or cancer hospitals in Athens and assigned a final diagnosis of lung cancer other than
adenocarcinoma and alveolar carcinoma from September 1978 through June 1980 were
interviewed by a physician. Controls were gathered from nonsmoking female Caucasian Athenian
patients hospitalized during the same time period in the Athens Orthopedic Hospital. Some
prevalent cases were thus presumably included, so control sampling probably approximated a
density approach but did not strictly conform to one.
Diagnostic information was obtained from patients' charts. Exposure information was
obtained by face-to-face unblinded interviews conducted by the same physician for all subjects.
A total of 51 cases and 163 controls were interviewed. Of these, 11 cases and 14 controls reported
themselves to be active smokers, leaving 40 cases and 149 controls as ETS subjects. No interview
refusals are reported. Mean age of cases (controls) is 62.8 (62.3) years. Husband's education was
marginally higher in controls than cases with 63% and 58% of spouses having completed primary
school, respectively. No other demographic comparisons are reported for the ETS subjects alone.
For the sample population including smokers, factors such as age, duration of marriage,
occupation, education, and urban versus rural residence are all similar for cases and controls,
except once again educational level is slightly higher for controls. There is no indication that
verification of diagnosis or exclusion of secondary lung cancers was undertaken in cases. Of the
51 total cases, 14 were diagnosed histologically, 19 cytologically, and 18 by radiological or clinical
means. No breakdown is given for the ETS subjects alone.
The study classifies as nonsmokers both reported never-smokers and former smokers who
quit more than 20 years ago. It is not mentioned whether cigar and pipe smoking are considered
as sources of exposure. Nonsmoking women are considered exposed to ETS if they are married to
a man classified as a smoker. The average number of cigarettes smoked per day by the husband
and the number of years of marriage are used to estimate the total number of cigarettes smoked by
the husband during marriage. No data on childhood or nonspousal ETS exposure were collected.
A-107
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Single women are grouped with women married to a nonsmoker and are thus considered
unexposed. Widowed or divorced women were classified according to their former husband's
smoking status on the assumption that smoking stopped at death or divorce. No checks of
exposure information are reported.
For ETS subjects, the number of cases (controls) exposed over the total is 29 to 40
(78/149). The crude odds ratio calculated by the reviewers is 2.4 (95% C.I. = 1.12-5.16). The
results presented in the article are all stratified by level of husband's smoking. The odds ratios are
1.8, 2.4, and 3.4 when the husband is a former smoker, smokes 1 to 20 cigarettes per day, and
smokes 20 or more cigarettes per day, respectively. No confidence intervals are given, but a test
for upward trend was statistically significant (p < 0.02). When ETS exposure is estimated by total
number of cigarettes smoked during marriage, odds ratios (1.3, 2.5, and 3.0) increase with
cumulative exposure (1-99, 100-299, and 300+ thousand, respectively). The upward trend remains
statistically significant at p < 0.02. No analyses adjusted for age or other potentially confounding
variables. With regard to age and other demographic variables, the authors conclude from the
similarity of cases and controls that it is not necessary to stratify for these variables in the
analysis, particularly because none is significantly associated with smoking in the study.
The authors note that this study has obvious limitations and is offered principally to
suggest that further investigation of this issue should be pressed. Most seriously, the numbers of
cases are small. Nevertheless, the association is in the direction expected if passive smoking is
related to lung cancer, and the outcome is unlikely to be due to chance. Other limitations noted
include the high percentage (35%) of cases lacking cytology and the selection of controls from a
hospital different from those of the cases; it is argued, however, that neither of these appears to
be consequential. The observation is made that it is potentially easier to detect an effect of
passive smoking in the Greek population than in most Western populations, because in the latter
groups, the overwhelming effects of active smoking, together with the high correlation between
smoking habits of spouses, would tend to confound and conceal the lesser effects of
passive smoking.
A.31.3. Addendum
In a letter to the editor of Lancet in 1983, Trichopoulos et al. released a data table derived
from extension of subject collection through December 1982. This nearly doubled the sample size
used in the 1981 publication, yielding 77 nonsmoking cases (102 total) and 225 smoking controls
(251 total). The crude odds ratio calculated by the reviewers is 2.08 (95% C.I. = 1.20-3.59). The
results for the expanded study show very little change; (estimated) relative risks when husbands
A-108
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
are former smokers, (1-20 cig./day and > 20 cig./day) compared to nonsmokers are 1.95, 1.95, and
2.54, respectively. The test for upward trend in the dose-response is significant (p = 0.01). No
other analyses are presented.
A.31.4. Comments
This study was conceived and undertaken to explore the association of spousal smoking
with lung cancer and does not rely on a preexisting data set. Thus, the investigators were in a
position to design their selection and data collection to maximize the strength of their findings.
This did not, however, prevent the appearance of some design and analytical flaws.
Demographics of the total case and control populations are very similar. All subjects in
the spousal smoking analysis are resident Athenian nonsmoking women hospitalized in the same
area of Athens; case and control groups have very similar mean ages, and their husbands are
comparable in education. Thus, the groups probably have good demographic comparability,
although it would have been helpful if the detailed demographic comparisons were focused on the
nonsmokers alone. Most of the controls (108 out of 163) were being treated for fractures, a
relatively minor and nonchronic illness compared to lung cancer, which may make them more
representative of the general community than of hospitalized patients as a whole. This should
reduce the problem of inclusion of smoking-related illnesses in the control group.
Although the researchers sought to exclude adenocarcinomas and alveolar carcinomas,
presumably considering these would be less smoking-related, nearly two-thirds of the cases were
not histologically confirmed, so an indeterminate number of these cell types was probably
included. More important, the infrequency of histologic confirmation and lack of mechanisms to
verify diagnoses or primary tumor status introduces potential for misclassification. The likely
effect is a bias toward no association.
The researchers clearly devoted considerable thought to the smoking and exposure criteria,
particularly with regard to changes in smoking and marital status over time. Single women were,
however, automatically classified as unexposed. The authors contend that this is warranted by the
traditional nature of Greek society and report that analyses restricted to married women result in
similar, and still statistically significant, associations, although with somewhat lower estimated
risks. There is a small reduction in the odds ratios after exclusion of single women, however, and
the restriction of the full analyses and results to married women may have been useful.
Another issue related to exposure concerns inclusion of former smokers in the study,
provided they had not smoked for at least 20 years. Active smoking 20 to 30 years before the
onset of lung cancer may be of etiological relevance, however, in view of a long latency period for
A-109
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
lung cancer. Although use of the same interviewing physician for all subjects eliminates the
problem of interobserver variability, it magnifies the potential problem of interviewer bias in
exposure assessment, presumably toward a positive association, because the interviews were
apparently conducted unblinded (virtually unavoidable with regard to diagnosis, given that
controls were drawn from orthopedic trauma and rheumatology wards).
A larger concern, however, is the issue of potential confounders. It is contended that the
similar distribution of demographic variables between cases and controls eliminates the need to
consider these variables in the analyses, but similarity between cases and controls does not
preclude confounding from an independent risk factor differentially distributed by exposure.
More convincing is the contention that these variables were not significantly associated with
smoking in these data, although no specifics are included. Potential confounders such as diet,
cooking, and heating practices are not addressed. The appearance of a statistically significant
trend, for ETS exposure measured by either current spousal smoking or cumulative cigarette
consumption during marriage, supports an association between spousal smoking and increased lung
cancer incidence.
Overall, the issues addressed above would probably produce a conservative bias, resulting
in an underestimate of the degree of association. The study's basic design is sound. It provides
statistically significant evidence of dose-response, and although the limitations described above
should be borne in mind, it provides useful data for assessment of the relationship between ETS
and lung cancer.
A.32. WU
A.32.1. Author's Abstract
"A case-control study among white women in Los Angeles County was conducted to
investigate the role of smoking and other factors in the etiology of lung cancer in women. A total
of 149 patients with adenocarcinoma (ADC) and 71 patients with squamous cell carcinoma (SCC)
of the lung and their age- and sex-matched controls were interviewed. Personal cigarette smoking
accounted for almost all of SCC and about half of ADC in this study population. Among
nonsmokers, slightly elevated relative risk(s) (RR) for ADC were observed for passive smoke
exposure from spouse(s) [RR = 1.2; 95% confidence interval (CI) = 0.5, 3.3] and at work (RR =
1.3; 95% CI =* 0.5, 3.3). Childhood pneumonia (RR = 2.7; 95% CI = 1.1, 6.7) and childhood
exposure to coal burning (RR = 2.3; 95% CI « 1.0, 5.5) were additional risk factors for ADC. For
both ADC and SCC, increased risks were associated with decreased intake of /J-carotene foods but
not for total preformed vitamin A foods and vitamin supplements."
A-110
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
A.32.2. Study Description
This study was undertaken in California during 1981 and 1982 to investigate the role of
smoking and other factors in the etiology of lung cancer in women. These other factors included
prior lung disease, coal heating and cooking, diet, and occupation. Both active and passive
smokers are included; some of the ETS analyses retain active smokers while attempting to adjust
for smoking status.
Cases are white female English-speaking Los Angeles County residents under 76 years of
age at time of diagnosis with primary adenocarcinoma or squamous cell cancer of the lung
between April 1, 1981, and August 31, 1982. Cases are restricted to U.S.-, Canadian-, or
European-born individuals with no history of prior cancer other than nonmelanoma skin cancer.
Controls are density sampled, matched individually on neighborhood and age (± 5 years), and
meet all case criteria (except, of course, diagnosis of lung cancer). The L.A. County tumor
registry was used to identify incident cases for inclusion in the study, whereas controls were
recruited house to house. Interviews to obtain exposure data were conducted by telephone with
participating subjects, apparently unblinded.
A total of 490 eligible cases were identified; 270 were not interviewed because they were
too ill or had died (190), their physician refused permission to contact them (28), they could not
be located (8), or they refused (44). Those not interviewed did not differ significantly from those
interviewed with regard to age or their marital, religious, or smoking status as recorded on
registry records. Refusals eliminated 70 potential controls. The case and control populations had
nearly identical mean ages for adenocarcinoma, 59.7 versus 59.5 years, respectively, and for
squamous cell cancer, 61.4 versus 61.1 years. No other demographics are provided. Histologic
diagnoses were obtained for all cases.
For spousal smoking, exposure constitutes having a spouse who smoked while living with
the subject. For workplace smoke, exposure is based on the opinion of the subject. It is not clear
whether for the lung cancer analyses, parental smoking refers only to adult life (as for spousal and
workplace exposure) or to the childhood and teen years (as was stipulated for coal and preadult
lung disease exposures). Adult life seems most probable. Units of exposure for spousal and
parental smoking are cigarettes per day and years of exposure, apparently entered into a regression
model as a combined variable; for occupational exposure, units are in years of exposure. Exposure
data were apparently not checked, treatment of cigar and pipe smoking is never mentioned, and
no results are reported for household smoking aside from spouse and parents, although
information on this exposure was collected. Never-married women were excluded from the
spousal smoking analysis, but marital status was not otherwise considered in the analyses. The
A-lll
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
only histologic or airway proximity information provided for the ETS subjects is that 29
adenocarcinomas occurred among nonsmokers, 12 of which were bronchoalveolar.
The total study population includes 220 cases and an equal number of matched controls.
Of the cases, 149 are adenocarcinoma and 71 are squamous cell. Nonsmokers constituted 29 of the
adenocarcinoma cases and 62 of the corresponding controls, while composing 2 of the squamous
cell cases and 30 of the controls. No raw data are presented regarding passive smoking and lung
cancer. Logistic regression analysis of matched pairs was used in all calculations. Results
restricted to nonsmokers are presented only for adenocarcinoma. An estimated relative risk of 1.2
is found for spousal smoking, 1.3 for workplace exposure, and 0.6 for smoking by either parent.
None of these estimates was statistically significant. Exposure from spouses and at work,
however, show a dose-response trend with years of exposure, yielding estimated relative risks of
1., 1.2, and 2.0, for 0, 1 to 30, and 30 or more years of exposure, respectively.
Analyses that include active smokers but attempt to adjust for them by including the
number of cigarettes smoked per day and age at start of smoking in a logistic regression model are
presented for both lung cancer types. For adenocarcinoma, estimated relative risks for maternal,
paternal, spousal, and workplace exposure of 1.7, 1.3, 1.2, and 1.2, respectively, were obtained.
For squamous cell cancer, maternal, paternal, spousal, and workplace relative risks are 0.2, 0.9,
1.0, and 2.3, respectively. None of these estimates is statistically significant.
History of lung disease at least 5 years prior to diagnosis of lung cancer reportedly had no
significant association with lung cancer. History of lung diseases before age 16 yielded a
significant association for pneumonia (RR = 2.7 [95% C.I. = 1.1-6.7] for adenocarcinoma and
RR « 2.9 [95% C.I. « 0.5-17.4] for squamous cell cancer) but not for six other diseases.
Heating or cooking with coal during the childhood and teenage years is also significantly
associated with lung cancer (RR = 2.3 [95% C.I. = 1.0-5.5)] for adenocarcinoma and RR = 1.9
[95% C.I. - 0.5, 6.5] for squamous cell). Among dietary factors, low beta carotene consumption is
significantly associated with adenocarcinoma (RR = 2.7) and mildly associated with squamous cell
(RR ~ 1.5). Diets low in dairy products and eggs have similar relative risk values. No significant
associations were noted for vitamin A consumption, occupation, or other health history factors not
previously considered.
The authors conclude that the etiology of squamous cell carcinoma can be explained almost
entirely by cigarette smoking. Cigarette smoking, however, explains only about half of the
adenocarcinoma cases. On the basis of this study, childhood lung disease and exposure to coal
fires in childhood explain at least another 22% of adenocarcinoma cases. Passive smoking and
A-112
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
vitamin A may be involved, but more research is needed to clarify their roles in lung cancer
etiology.
;
A.32.3. Comments
This study took particular care with its treatment of case and control assembly. Extensive
inclusion criteria extending to both groups, matching not only on age but neighborhood of
residence, and retention of matching through analysis all bode well for comparability of cases and
controls. The virtually identical mean ages of cases and controls indicate the success of these
efforts. In addition, exclusive use of incident cases reduces the potential for selection bias, and
density sampling of controls reduces potential problems with temporal variation. The only real
fault in the treatment of cases and controls is the failure to provide any demographic comparison
other than for age, thus denying concrete confirmation of high case-control comparability.
Case diagnoses are likely to be accurate, because all were histologically diagnosed, making
misclassification unlikely and making cell-type-specific analyses possible. Although no one
pathologist or team verified these determinations, the authors note that there is generally good
interobserver agreement for the cell types included in this study. Potentially eligible cases not
interviewed due to illness, refusal, or other reasons did not differ significantly in demographic or
smoking status from those actually interviewed, again arguing against biased selection.
No proxy interviews were used and all subjects were English-speakers, enhancing the
chances of obtaining accurate exposure information. On the other hand, interviews were by
telephone—possibly decreasing accuracy relative to face-to-face interviewing--and apparently
unblinded, thus introducing possible interviewer bias toward positive results.
Collection of exposure data seems generally adequate, except that treatment of pipe and
cigar smokers is not described. Uncertainty on this point extends to the analysis and is coupled
with a vague treatment of parental smoking (current only? childhood only? or both?) and lack of
treatment of household smokers other than parents or spouses, despite collection of data on this
point. These uncertainties probably translate into nondifferential exposure misclassification,
biasing results toward the null.
The analyses themselves suffer from the common problem of restricted numbers of
nonsmoking cases—29 for adenocarcinoma and only 2 for squamous cell. Some factors examined
are restricted to nonsmokers alone for adenocarcinoma, but for most analyses, an adjustment for
active smoking by logistic regression modeling was attempted. The adequacy of such adjustment
may be questionable. For adenocarcinoma, however, the results for passive smoking were very
similar, regardless of whether restriction or adjustment was used. Further, a dose-response
A-113 .
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
pattern was seen for cumulative years of spousal and workplace exposure among nonsmokers. The
utility of the smoking-adjusted cell analyses is nevertheless questionable, given the paucity of
nonsmoking cases.
The findings of substantial associations between lung cancer (or, at least, adenocarcinoma)
and childhood pneumonia and coal burning are of interest. It must be borne in mind that seven
adult respiratory diseases (including pneumonia) as well as six other childhood respiratory diseases
were examined, so the possibility that the pneumonia association was an artifact of multiple
comparisons cannot be ruled out. History of hysterectomy and multiparity showed nearly
significant associations with adenocarcinoma, but it is not clear how many other health history
factors were also considered. Coal burning has been associated with lung cancer in several other
studies. Similarly, as in several other studies, one found an association with low beta carotene
intake, but there was no evidence of a dose-response gradient, and no significant association was
found for preformed vitamin A. The strongest association with a dietary factor was actually that
for low intake of dairy products and eggs, which showed a consistent dose-response pattern. The
use of a matched-pair analytical approach controls for possible confounding due to age or
neighborhood, which also reduces the likelihood of neighborhood-related factors such as
socioeconomic status as major sources of bias. Confounding due to active smoking can be ruled
out in the passive smoking results for adenocarcinoma and is not likely in regard to other factors
given adjustment for this variable in all analyses. Likewise, the authors report that adjustment for
childhood pneumonia, coal burning, and beta carotene intake did not alter their results. Strangely,
however, no adjustment for dairy product and egg intake—the dietary factor with the most
convincing association with lung cancer in their data—was carried out.
Overall, this study's results are consistent with a mild association between spousal and
workplace ETS exposures and lung adenocarcinoma, although they support no such association for
parental smoking. In addition, the study it raises childhood pneumonia, coal burning during early
life, low intakes of beta carotene, and low intake of dairy products and eggs as potential moderate
risk factors that should be considered by future studies. The results for squamous cell carcinoma
are uncertain given the small number of nonsmoking cases available, and in all instances, they lack
statistical significance due to sample size limitations. Thus, the study provides useful information
on the relationship of adenocarcinoma of the lung with ETS and a number of other factors:
information regarding squamous cell cancer is of much lower utility;
A-114
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
/
A.33. WUWI
A.33.1. Author's Abstract
"A case-control study of lung cancer involving interviews with 965 female patients and
959 controls in Shenyang and Harbin, two industrial cities which have among the highest rates of
lung cancer in China, revealed that cigarette smoking is the main causal factor and accounted for
about 35% of the tumors among women. Although the amount smoked was low (the cases
averaged eight cigarettes per day), the percentage of smokers among women over age 50 in these
cities was nearly double the national average. Air pollution from coal burning stoves was
implicated, as risks of lung cancer increased in proportion to years of exposure to Kang and other
heating devices indigenous to the region. In addition, the number of meals cooked by deep frying
and the frequency of smokiness during cooking were associated with risk of lung cancer. More
cases than controls reported workplace exposures to coal dust and to smoke from burning fuel.
Elevated risks were observed for smelter workers and decreased risks for textile workers. Prior
chronic bronchitis/emphysema, pneumonia, and recent tuberculosis contributed significantly to
lung cancer risk, as did a history of tuberculosis and lung cancer in family members. Higher
intake of carotene-rich vegetables was not protective against lung cancer in this population. The
findings were qualitatively similar across the major cell types of lung cancer, except that the
associations with smoking and previous lung diseases were stronger for squamous/oat cell cancers
than for adenocarcinoma of the lung."
A.33.2. Study Description
The objective of this study was to evaluate the role of potential risk factors for lung
cancer in Harbin and Shenyang, two cities among those with the highest mortality rate for lung
cancer in China. Active smokers are included in the cases, so data on ETS subjects constitute a
subset of the whole study.
Cases consist of female residents under age 70 newly diagnosed with primary lung cancer
in about 70 participating hospitals in Harbin and Shenyang between 1985 and 1987. Controls are
female residents randomly selected from the general population of these cities and frequency
matched by 5-year age group to the age distribution of female lung cancer cases reported in the
cities in 1983. Trained interviewers collected information on smoking habits, diet, cooking and
heating practices, and other factors from subjects in face-to-face unblinded interviews.
A total of 1,049 qualifying cases were found, including both ever-smokers and
never-smokers, of which 405 were diagnosed by histology, 309 by cytology, and 351 by radiology
or clinical means. (Note: These diagnostic numbers do not total 1,049. The 351 figure may be
A-115
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
intended to be 251, which would give a total of 965 diagnoses, about the number of cases
interviewed.) Of these, 85 either died prior to interview, refused to participate, or could not be
located. Mean age of participating cases was 55.9 years, whereas that of the 959 controls was 55.4
years. Nonsmokers compose 417 of the interviewed cases and 602 of the controls.
A smoker is defined as a person who has smoked cigarettes for 6 months or longer, so a
nonsmoker apparently may have smoked up to 6 months. Information on all types of tobacco
products smoked was collected. Sources of ETS exposure include smoking by any household
cohabitant and smoking by individuals (spouse, mother, and father) over the course of the
subject's lifetime. Exposure at the workplace is also addressed. ETS exposure in the home is
expressed in terms of cigarettes per day and number of years smoked; no units of measurement
are used for workplace smoking. No checks on exposure data were undertaken. Marital status of
subjects is not discussed. Of the cases with histological or cytological data, adenocarcinomas
compose 310 (41.7%), squamous cell cancers 201 (28.9%), small and oat cell cancers 117 (16.8%),
and large cell or unspecified types 66 (9.5%). No data on airway proximity or diagnostic
breakdowns limited to nonsmokers are provided.
Statistical analyses of potential risk factors, including ETS, largely include data on active
smokers and then adjust for the effect due to smoking by logistic regression, along with other
potential confounders such as age, education, and location. These analyses indicate no increase in
risk from household sources of ETS, with estimated relative risks of 0.8 (household cohabitants),
0.9 (spouse), 1.0 (mother), and 1.0 (father). The estimated risk for workplace exposure is
nonsignificant (RR - 1.2). Restriction of analyses to ETS subjects alone (i.e., only the
nonsmokers) produced similar results, with estimated relative risks of 0.7 for general cohabitant,
0.7 for spouse, 0,9 for mother, 1.1 for father, and 1.1 for workplace exposure. The ETS exposure
from spousal smoking is significantly low (i.e., associated with a decrease in lung cancer by this
analysis, as apparent from the confidence interval; RR = 0.7; 95% C.I. = 0.6-0.9).
The smoking-adjusted analyses indicate associations with lung cancer for several types of
heating devices, including kangs (brick beds heated by pipes from the stove or by burners directly
underneath), coal stoves, and heated brick walls or floors. The risk associated with the use of
burning kangs (those heated by stoves underneath) shows an upward trend with years of use,
becoming statistically significant at 21 or more years of use (RR = 1.5; 95% C.I. = 1.1-2.0).
Significantly elevated risks are also associated with use of heated brick walls or floors (RR =1.5
[1.1-2.1] for 1-20 years of use; RR =1.4 [1.1-1.9] for > 20 years). Nonsignificant increases in risk
are noted for use of kangs of all types, coal stoves, and coal burners; nonsignificant reductions in
risk are indicated for noncoal stoves and central heat. Use of deep frying at least twice a month
A-116
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
and eye irritation during cooking are both significantly associated with lung cancer, as are regular
intake of animal protein and fresh fruit. (Note: Multiple comparisons may be a factor for the
apparent significance of some items, as discussed further in the next section.)
The authors find no overall association between lung cancer and ETS exposure. On the
other hand, coal burning, exposure to cooking oil fumes, and chronic lung disease may all be risk
factors. Consumption of beta carotene shows no evidence of a protective effect. Overall, active
smoking is the major cause of lung cancer among women in the regions sampled.
A.33.3. Comments
The sample size is impressive, with ETS exposure data available for nearly 1,000 cases
including smokers and over 400 cases when restricted to nonsmokers, thus providing substantial
statistical power. All subjects are women recruited from two industrial cities in northeast China,
reducing potential for complications due to regional or urban-rural differences. Nearly all of the
hospitals in these cities were involved, all cases occurring in these hospitals were targeted, and the
rate of participation among eligible cases was high; thus potential for selection bias is minimized.
The effective case recruitment in combination with the use of general population controls
maximizes generalizability of the study's results for northeast China. It would have been useful,
however, to present the results for the two component study locations separately. Although
coordinated in planning and execution, there are two separate study locations and the sources of
heterogeneity between them tends to be obscured when results are combined.
Unfortunately, the study's results with regard to ETS are more limited than the strengths
listed above might suggest. The inclusion of age, education, and location as control variables in all
analyses is laudable, thus eliminating three sources of potential confounding. The attempt to
control for potential sources of confounding that may be causally related to lung cancer by
statistical methods, however, is less certain. Although some analysis was conducted with data for
active smokers included, to the authors' credit they also analyzed data for ETS subjects alone (i.e.,
with the data for active smokers removed), which is the surest way to control for confounding by
active smoking. Other potential causes of lung cancer (e.g., air pollution from coal-burning
stoves, smokiness during cooking, and deep-fat frying foods) also need to be taken into account in
an analysis of ETS. This cannot always be accomplished effectively by statistical methods,
particularly when there are multiple risk factors to be taken into account that are variable, poorly
measured, and possibly more potent risk factors than ETS may be.
At the risk of belaboring this point, as the reader is aware, a case-control study is ideally
designed and executed under conditions where cases and controls are as comparable as possible
A-117
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
aside from the factor of interest, such as ETS exposure. The presence of other risk factors may
tend to pollute and obscure, much like the contamination of a laboratory experiment. In this same
sense, the presence of indoor sources of smoke other than ETS may contaminate an environment
for measuring ETS effects because the non-ETS smoke likely contains many of the same
carcinogens as ETS, and possibly in much larger quantities, depending on the relative levels of
exposure. Other factors outside the home, such as workplace exposure to coal dust and to smoke
from burning fuel that was reported more often in cases than controls, contribute to the potential
confounding in a similar way. Consequently, a credible analysis of ETS requires being able to
adjust for these likely confounding factors satisfactorily, and the ability to do that depends on
reliable measures of exposure and the extent of confounding. That kind of statistical analysis is
not given in the article, and it does not appear to have been possible, based on conversations with
the authors (Wu-Williams and Blot) and the text of the article: "Despite the large size of our
study, we were unable to clarify the magnitude of risks due to passive smoking, recognized as a
cause of lung cancer around the world (U.S. DHHS, 1986). Perhaps in this study population the
effects of environmental tobacco smoke was obscured by the rather heavy exposures to pollutants
from coal-burning Kang, other indoor heating sources, and high levels of neighborhood air
pollution (Xu et al., 1989)."
The multivariate analysis reported in the article reinforces the viewpoint that any ETS
effect may be dominated by the presence of other risk factors. In that analysis, variables were
allowed to enter a logistic regression model in the order of their explanatory value (a stepwise
regression exercise in statistical terminology). The order of entry into the model is deep frying,
eye irritation, pneumonia, household tuberculosis, burning kang, self-reported occupational
exposure to burning fuel, passive smoking, and heated brick wall or floor. Passive smoking, in
this exercise, is significant (p < 0.05) but in the direction of reducing lung cancer, not
contributing to it. The 0.05 value, however is not fully meaningful as a significance level for
ETS, because of the stepwise procedure used (the same data used in the construction of a model is
used for testing variables in the model) and because of the likely confounding between ETS and
other variables. Note, for example, that passive smoking entered the model ahead of heated brick
wall or floor, which is highly significant when analyzed alone, whereas passive smoking is not.
The evidence for association of lung cancer with burning coal and deep frying foods is
particularly provocative, as it indicates two factors that may play a substantial role in the etiology
of lung cancer in northeast China and, hence, in other areas as well where such practices occur.
The associations noted with other factors are also of interest, but their importance is undermined
by the problem of multiple comparisons. In the table presenting results for dietary factors, for
A-118
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
example, 26 risk estimates are computed, 4 of which are significant at the 5% significance level
(for a two-sided test, 2,5% level for the test of an effect), only one more significant finding than
expected due to chance alone.
Being somewhat speculative, the use of cases age 70 and below may be a factor. Wells
(1988) showed that about half of the female passive smoking deaths occur after age 70, for the
studies included in that reference. If ETS is a risk for lung cancer and if individual susceptibility
to lung cancer is a factor, some of the stronger risk factors such as coal burning and cooking oil
may have caused lung cancer in the more susceptible subjects before passive smoking had a
chance to exert itself.
In summary, this large and basically well-executed study observed no significant
association between exposure to ETS from cohabitants, spouse, parents, or workplace and lung
cancer. Lack of control for a number of other significant risk factors identified in the study
undermines these results, however. The associations with coal burning for heat and oil frying are
particularly notable. Use of the heating devices most strongly linked with lung cancer is
presumably more common in colder northern regions, whereas stir frying may be more widespread
in Asian communities, without regard to climate. Thus, this study was exploratory, designed to
generate hypotheses rather than to test the specific hypothesis that ETS exposure is associated with
lung cancer. It identifies a number of potential risk factors for consideration in future studies.
The prevalence of these factors in the study population combined with the lack of analysis of their
association with ETS exposure, however, renders the results for ETS inconclusive.
A-119
05/15/92
-------
-------
DRAFT-DO NOT QUOTE OR CITE
APPENDIX B. METHOD FOR CORRECTING RELATIVE RISK
FOR SMOKER MISCLASSIFICATION
B.I. INTRODUCTION
The purpose of this Appendix is to present the details of the method used in Section 5.2.2. to
correct observed passive smoking relative risks for the systematic upward bias caused by
misclassification of some smokers as never smokers. The method used is that proposed by A. J. Wells
and W. F. Stewart (Wells, 1990). This Appendix covers: Section B.2) the principles of the method;
Section B.3) how the method differs from those previously used by the National Research Council
and P. N. Lee; Section B.4) the data used to calculate the misclassification factors and other
i
parameters; Section B.5) the mathematical model used to calculate the corrected relative risks; and
Section B.6) a numerical example to show how the method is applied in a practical case. Evidence is
also presented indicating that the true downward corrections for smoker misclassification bias may be
even-smaller than those used in Section 5.2.2.
There is considerable literature on this topic and a history of controversy regarding the
magnitude of the bias and whether it may explain the observed increase in lung cancer mortality due
to ETS exposure. The NRC report on the health effects of passive smoking (NRC, 1986) delves into
this topic in considerable detail. It concludes that bias is likely and estimates an adjustment for the
summary relative risk from the combined results for all ETS studies. The NRC report further
concludes that smoker misclassification does not account for the observed passive smoking risk. On
the other hand, Lee in various publications (Lee, 1987b, 1988, 1990, 1991) has claimed that the
smoker misclassification bias is large enough to explain most or all of the observed passive smoking
lung cancer risk.
Approaches to estimation of misclassification bias have used mathematical modeling with
parameters estimated from a variety of sources that have not always been consistent. The procedure
described below attempts to rectify some previous sources of misunderstanding on this topic and
utilizes the extensive data sources now available to improve parameter estimates and tailor refinements
to individual populations.
B-l
05/21/92
-------
DRAFT-DO NOT QUOTE OR CITE
Mathematical model:
Calculate the corrected risk directly—that is, do not first calculate a bias assuming no
passive risk and then divide the observed risk by that bias to get a corrected risk.
B-2
05/21/92
-------
DRAFT-DO NOT QUOTE OR CITE
Subjects found to be misclassified as nonsmokers are categorized according to their true
smoking status-former or current. Current smokers are further classified as "regular" or "occasional",
according to. cotinine levels observed. "Regular" means the cotinine level is above 30% of the self-
reported smoker mean; "occasional" applies to the range 10-30%. Cotinine levels are not informative
for misclassified former smokers, who tend to be long term abstainers (10+ years, according to Lee
(1987b) and Wald et al. (1986)). The two studies with detailed cotinine levels on female current
smokers (Lee, 1986 and Haddow et al., 1986, in Table B-l) indicate that about 10% of the current
smokers are occasionals.
B.3. DIFFERENCES FROM EARLIER WORK
The Wells-Stewart method differs from the method used by the NRC (1986), which is also
described by Wald et al. (1986), in that the NRC method failed to separate the misclassified smokers
into regular, occasional, and exsmokers, (and 'they failed to account for the effect of smoker
misclassification on active smoker risk. The NRC made an overall correction to the aggregated
passive relative risk using United Kingdom smoking prevalence and risk rather than making the
corrections study by study, with appropriate smoking prevalences and risk for each study's time and
locale, and they mixed male data with female data in arriving at misclassifieation factors. Their
calculated bias of 1.34/1.25 = 1.07, or 7%, for the combined worldwide studies is substantially higher
than the 2% overall bias that would result if. the biases in Table 5-7 were aggregated. The discrepancy
is largely due to NRC's use of U.K. parameters for all of the studies regardless of locale, plus some
overestimation of the impact of misclassified occasional and exsmokers.
Lee's methods have evolved over the years in three stages. In Lee (1987b, 1988) he improved
on the NRC method in that he divided the misclassified smokers into exsmokers and current, regular
and occasional smokers, and he corrected the smoker risk for misclassification. However, all of the
fiye principles listed above were violated, to some degree resulting in about a twelve-fold
overestimation of the bias. The Lee (1990) paper correctly limits misclassifieds to never-smokers,
relates misclassified smokers to smokers, not to never smokers, and treats each study separately, but
still mixes male input data with female data for use in calculating bias for females. Furthermore, his
(Lee, 1990) mathematical model still relies on the assumption of no passive risk, which results in
increased estimates of the bias as the observed relative risk increases. In addition, Lee (1990) has
changed from separating the misclassified smokers into three groups in favor of the (less useful) '
B-3
05/21/92
-------
DRAFT-DO NOT QUOTE OR CITE
overall category of ever smokers. Most recently Lee (1991) presents a more complex mathematical
model that includes a term for passive risk, but the method still has the other shortcomings noted for
Lcc (1990). A comparison of the most recent Lee bias estimates with those in Table 5-7 is shown in
Table B-2 for the five U.S. studies with the greatest statistical weight. When Lee's inputs are used
with the Wells-Stewart mathematical model, the calculated biases are if anything somewhat larger than
when using Lee's most recent model. Therefore, the difference between Lee's most recent estimates
of bias and those shown in Table 5-7 are in practical terms due almost entirely to differences in input
parameters. The input parameters we have chosen are developed in the next section, and comparison
with the Lee parameter estimates are shown as footnotes to Table B-2.
B.4. PARAMETER ESTIMATES
The key input into these calculations is the proportion of misclassified regular current smokers
who claim they have never smoked. Our definition of misclassified regular current smokers, first
suggested by Lee (1987b), produces a mean cotinine level approximately equal to that of all self-
reported current smokers. Detailed data from three large cotinine studies have been assembled for use
herein with the cooperation of their principal investigators (Coultas, Curnrning, and Pierce in Table B-
3). The data identify individual nonsmokers with cotinine values greater than 10% of the mean for
self-reported smokers, by sex and self-reported smoking status (never or former). Data on nonusers
are also available from several other studies (the lower portion of Table B-3). Since the numbers of
misclassified smokers are small, the proportions of misclassified smokers who would have said "never"
versus "former" are estimated using the proportions observed in the first three studies. Data sets.not
differentiating outcomes by sex have not been used. Also the large 1986 study by Haddow and
colleagues has not been used for this purpose on the advice of one of the authors (private
communication from GJ. Knight).
The number of self-reported never- and former smokers with sufficiently high cotinine levels
to be reclassified as current smokers is shown by study in Table B-3. As described above, those with
coiinine levels in the 10-30% range are considered to be occasional smokers while those above 30%
arc treated as regular smokers. If it is assumed that 90% of 1,525 self-reported current smokers, or
1,372, are regular smokers, leaving 10%, or 153, as occasionals, then the percentage of current regular
smokers misclassified as never-smokers totalled over all studies in Table B-3 is 14/1,372 or 1.02%.
The percentage is almost the same if the number of true, i.e., self-reported plus misclassified current
B-4
05/21/92
-------
DRAFT-DO NOT QUOTE OR CITE
regular, smokers is used. For the occasional smokers only, the misclassification rate is much higher,
about 20% (15%) of observed (true) occasional smokers. It is possible, however, that the subjects
classified as occasional smokers based on cotinine levels in the range 10-30% may contain some true
never-smokers that are just highly exposed to passive smoke.
The studies in Table B-4 provide data on discordant answers, i.e., reported never-smokers who
have called themselves smokers on one or more previous occasions. Based on those data, the
estimated percentage of former smokers misclassified as never-smokers is about 12% (11%) of the
observed (true) number of former smokers. As mentioned previously, evidence suggests (Wald et al.,
1986; Lee 1987b) that most former smokers misclassified as never-smokers have been nonsmokers for
an extended period, such as 10+ years, and may have been light smokers on average. Accordingly, we
have used a weighted average of the data of Alderson et al. (1985), Lubin et al. {1984), and Garfinkel
and Stellman (1988) for 10+ year abstainers to estimate former smoker relative risk, namely, an excess
risk that is 9% of current smoker excess risk.
Some confusion and misleading conclusions on smoker misclassification have resulted from
the practice of expressing the number of smokers misclassified as never-smokers as a percentage of the
total number of (either true or observed) never-smokers, rather than as a percentage of the number of
smokers. That leads to a higher expected percentage of smokers misclassified as never- smokers
among cases than controls because lung cancer cases are much more likely to have been smokers than
never-smokers. Some people have interpreted a higher percentage of observed never-smokers later
found to be misclassified smokers among the cases as evidence that smokers with lung cancer are
more apt to claim falsely to be never-smokers than persons without cancer. That conclusion, however,
appears to be an artefact of treating the misclassification rate as a percentage of the number of never-
smokers rather than as a percentage of the number of smokers. The study data summarized in Table
B-5 do not support that conclusion. If anything, it is more supportive of the conclusion that ever-
smokers in lung cancer studies may be less likely to misrepresent themselves as never-smokers than
members of the general public who are questioned in community surveys. The one percent average
misclassification rate shown in Table B-5 for the lung cancer cases suggests that estimates such as the
5.7% from the general population studies (Table B-5) or the near four percent of ever-smokers (Table
B-4) that we have used may be much too high.
Further corroboration that the misclassification rates from the community studies are too high
relative to those in the epidemiologic studies is found in the recent study by Fontham et al. (1991).
B-5
05/21/92
-------
DRAFT-DO NOT QUOTE OR CITE
After eliminating possible smokers among the self-reported never-smokers by the usual .epidemiologic
techniques, the investigators found by cotinine measurements that only two probable occasional
smokers and no probable regular smokers were left among the 239 never-smoking lung cancer cases
for which cotinine measurements were made. Assuming 45% ever-smoking among controls and an
ever-smoker relative risk of 8 for regular smokers and 2.4 for occasional, there would have been
1,456 smoker cases, consisting of 1,409 current smokers and 47 occasional smokers. It is seen that a
misclassification rate of 0/1,409 = 0.00% for regular smokers is well below the 1.0% that we have
used from the community surveys, and 2/47 = 4.3% for occasional is also well below the 19.6% for
occasionals that we have used.
Another indication that the estimates based on community surveys may be too high comes
from analysis of male data. The observed percentage of never-smokers is typically much lower for
males (17% to 35%) than females (41% to 86%). To correct for smoker misclassification we set up a
deletions table analogous to Table B-15 where the number of current and former smokers misclassified
as never-smokers are subtracted from the reported number of never-smokers. When the
misclassification rates generated from community surveys are applied to the male data, the outcome is
not credible—the number deleted for misclassification exceeds the total number of reported never-
smokers in three of the eleven examples of which we are aware and drives the corrected relative risk
well below unity in four more. This outcome indicates that the misclassification rates derived from
the community surveys are too high. It is probable that the true smoker misclassification bias is on
the order of one-fourth to one-half of the values shown in Table 5-7.
It is also said that East Asian women misclassify themselves at much higher rates than
Western women. The data from the International Agency for Research on Cancer in Table B-3 do not
support that claim, however, because the East Asia (Hong Kong, Japan, and China) misclassification
rate for current regular smokers is 1/77 = 1.3%, not much different from the overall rate of 1.0%.
The main proponent of the idea that smoker misclassification accounts for most or all of the
observed passive smoking risk has been P.N. Lee (1986, 1987b, 1988). He has estimated the bias for
females to be as high as 1.24. However, his methods are open to considerable question. He used
"nonuscr" cotinine data, which includes people who said or would have said they were former
smokers, rather than using only data on people who said they never smoked. This would about double
the calculated bias. He averaged high male misclassification rates into low female misclassification
rates. He made an overall correction to the combined risk using modem U.K. smoker risk and
B-6
05/21/92
-------
DRAFT-DO NOT QUOTE OR CITE
smoking prevalence rather than making the corrections study by study, as is done here with smoker
risks and prevalence appropriate for each study. He transferred misclassification rates from the
cotinine and discordant answer studies using percent of never-smokers rather than percent of smokers.
He also used as an input the data from the large Haddow study (Haddow et al., 1987) when the
authors state (private communication from Dr. George Knight) that the data from the study should not
be used for misclassification studies. Also Lee's mathematical method tends to overstate the bias for
passive risks greater than about 1.3. At a risk of two, his method overstates the bias about 100%.
In conclusion, it would appear that the bias introduced by misclassification of smokers as
never-smokers is not a serious problem. It probably increases perceived relative risks on a worldwide
basis by 1% to 2%, with the effect being about three times as large for combined U.S. studies.
B.5. MATHEMATICAL MODEL
The proportion of observed smokers, mho. misclassified as never-smokers is estimated
separately for former smokers (m10), occasional smokers (m20), and regular smokers (m30). Similarly,
the proportion of observed current smokers, mhl, misclassified as former smokers is estimated
separately for occasional smokers (m21) and regular smokers (m31). These estimates are given in
Tables B-3 and B-4. It is assumed that there is ho misclassification of true never-smokers as current
or former smokers or of observed former smokers as current smokers. Also these misclassification
factors are used for all the studies unless otherwise noted. We suspect that misclassification rates
probably vary from study to study. That variability, however, would tend to cancel out as the
individual study results are combined.
Let cijk designate the observed proportionate distribution of controls (i = 0) and cases (i = 1)
by, their smoking status (j = 0,1,2,3) and the smoking status of their husbands (k = 0,1) as illustrated in
Table B-6. Following the notational convention that a dot in the subscript position means summation
on that subscript, then CQ.. = c^. = 1.
The observed cijlc's are corrected for misclassification of the wife's smoking status by first
specifying a 4 x 4 matrix of proportionate distribution (Table B-7), where Phj (h,j = 0,1,2,3) is the
probability that a subject with true smoking status h will also be observed to have smoking status j.
The subscripted notation is shown in Table B-7 for easy reference. P.. is equal to unity.
For passive smoking, we are interested only in correcting the ci0k values that are for the
observed never-smokers. It is assumed.that the Phj's are the same for cases and controls
B-7
05/21/92
-------
DRAFT-DO NOT QUOTE OR CITE
(nondifferential misclassification). For given values of wife's subject status (i) and husband's smoking
status (k), the correction when the wife's observed smoking status is "never" (j = 0), is:
Ci0k = ci
iOIC
(B-l)
h=j=i
where C^ is the corrected form of the element of the element Ci0k. Then the corrected passive risk,
RR(c), becomes:
RR(C) — C10i X COOQ /Clod X
(B-2)
The values of Cojk in Table B-6 are from prevalence data in the study itself or from a related study,
from concordance data, and from each study's data on the smoking prevalence of the never-smokers'
husbands. If necessary, the number of former smokers can be estimated from the ever-smokers based
on data from nine studies known to us where the percent of both current smokers and former smokers
is known (see Table B-16). These data indicate a time trend in ndntraditional societies, from 17%
former smokers relative to ever-smokers in 1960 to 45% in 1985; we estimate a 20-year lag for the
traditional societies such as Hong Kong, China, Japan, and Greece. However, there are no data to
support this assumption.
To calculate the individual elements, c^, of Table B-6, it is necessary to establish concordance
factors-that is, the cross products in 2 x 2 tables of smoking status of husbands and wives by smoking
level of the wives. Using data from Sutton (1980), Lee (1987b), Akiba et al. (1986) and Hirayama
(1984) and the detailed data in Lee (1987b) on never-smokers, current smokers, and former smokers,
we have calculated that an appropriate average concordance factor for current smoking wives and ever-
smoking husbands versus never-smoking wives and never-smoking husbands is 3.2; for ever-smoking
wives and husbands versus never-smoking wives and husbands, it is 2.8, and for former smoking
wives and ever-smoking husbands versus never-smoking wives and husbands, it is 2.2. These
concordance factors can be expected to vary from study to study, but the effect of the variability
3
should tend to cancel out as the studies are aggregated. The element CM. and a quantity s0 = >p c0j.
Bo
-o
05/21/92
-------
DRAFT-DO NOT QUOTE OR CITE
are obtained from smoking prevalence data in the study itself, in a related study on the same cohort, or
as a last resort from national statistics. .The elements c^. and c^. + c03. are taken from the study or are
estimated from Table B-16. Element GO* is estimated to be 10% of (c02. + c03.); c03. is 90%. Elements
COOQ and c^ are obtained from cm. and the proportion of never-smoking controls in the study who are
married to either never-smokers or ever-smokers. Elements %„ arid cou are obtained by solving the
3 3
p c0j0 and s01 = \
equations c010 + cou = c01. and CQQO x cou /cmi x Cp10 = 2.2. Terms s^ =
c0jl
J=i
are obtained from the equations SQO + s01 = s0 and s01 x CMO/CM x s^ = 2.8. Then c020 + c030 = s^ - c010
and c021 + c031 = s01 - c011. The values of c020 and c02l are then assumed to be 10% of c020 + c030 and
c02i + c031, respectively, and c030 and c03i are assumed to be 90%.
To obtain the elements for the subject cases (i = 1) in Table B-6, it is necessary first to set up
relative risks for the passively exposed (k = 1) and not passively exposed (k = 0) wives by observed
smoking status (j = 0,1,2,3). These risks are shown in Table B-8. - . . •
In most instances, the relative risk, RR(e), for female ever-smokers can be obtained from the
study itself or from a related paper (Table B-9). In a few instances, it is necessary to estimate RR(e)
from other studies similar in time and locale. In some papers, a current smoker risk also is given. We
assume (based on cotinine measurements) that the misclassified regular smoker risk, RR(a)3, is equal to
the self-reported current smoker risk. Where only RR(e) is available, RR(a)3 can be assumed to be
equal to 1.24 x RR(e) based on the data in Table B-17. Because occasional smokers have cotinine
levels that are 10% to 30% of those of regular smokers, it is assumed that RR(a)2 - 1 = 0.20(RR(a)3 -
1), and because the former smokers (j = .1) are said to be. on average, long term (Wald et al., 1986;
Lee, 1987b), we have averaged the data of Alderson et al. (1985), Lubin et al. (1984), and Garfmkel
and Stellman (1988) for 10+ year former smokers, namely, that RR(a)j - 1 = 0.09 (RR(a)3 -1).
The elements RR^ and RRgj are obtained from the observed passive relative risk in the study
and the never-smoking population weights for controls in Table B-6 by solving the equations
(RRoo x COQO) + CRRoi x CQOI) = 1-00
and
B-9
05/21/92
-------
DRAFT-DO NOT QUOTE OR CITE
! /RRoo = RR(p)0.
Other assumptions regarding passive risks can be used for j = 1,2, and 3. We have assumed,
based on the data in Varela (1987) who found that 242 long-term former smokers had essentially the
same passive risk as 197 never-smokers, that the passive risk for former smokers to be the same as for
never-smokers, namely, that RRCpX = RR(p)0. It is also assumed that there is no passive risk for
current or occasional smokers so RR(p)2 and RR(p)3 are unity.
Crude versions of the elements cljk (i = 1 for cases) are obtained by multiplying each element
Ccju by its respective RRjIc. These are then normalized to give
_ cojkRRjk
~
j=0 k=0
The next step is to set up Table B-7, which is the table of proportionate distribution. This is
done by multiplying the observed misclassification rates (Pho/P.j) from footnotes 2 and 4 in Tables B-3
and B-4, respectively, by the appropriate elements from Table B-6. For example, P10 = C01.(Pio/P.i).
An attempt was made to use the true misclassification rates from Tables B-3 and B-4 on the theory
that they would exhibit less variability in being transferred from the cotinine and discordant answer
studies to the passive smoking calculations. However, the method is laborious and, as is shown in the
Correa example below, does not lead to increased accuracy.
The next step is to develop a deletions table to implement Equation B-l, above, using the
control and case smoking prevalences in Table B-6 and the proportionate distribution in Table B-7.
Each observed element, c^, in Table B-6 is multiplied by its appropriate observed misclassification
factor, EM /P.j, where h = j, to yield a deletion element to be subtracted from the appropriate observed
wives' never-smoking-status elements: COOQ, c^, c100, and c101, to obtain corrected elements C000, C001,
3
C,w and C101. Thus, Coo, = c^ - £ c0j0 P«/P.j, etc.
Once these corrected never-smoker elements are obtained, the relative risk corrected for smoker
misclassification is obtained from Equation (B-2); RR(c)0 = Qoi'x COQO/CJOO x C001, and the bias
becomes RR(p)0 /RR(c)0.
B-10
05/21/92
-------
DRAFT-DO NOT QUOTE OR CITE
B.6. NUMERICAL EXAMPLE
Using the Correa study as an example, the study tells us that 52.8% of the wives never
smoked .and that 45.9%.of the never-smoking wives were exposed to their spouses' smoke. This
establishes CQQ. as 0.528 and c^ and c^ as 0.286 and 0.242, respectively. The quantity s0., the
proportion of ever-smokers, by difference is 0.472. Assuming (from Table B-16) that the former
smokers are 35.5% of the ever-smokers, the former smokers, c01., become 0.167, and the current
smokers (c^,. + c03.) become 0.305. The current smokers are divided into current regular smokers at
90% (c03. = 0.275) and current occasional smokers at 10% (c02. = 0.030). These data are shown in the
bottom line of Table B-10.
Using the concordance factor of 2.8 for ever-smokers versus never-smokers, it is possible to
show algebraically that 33.2% of the females in the Correa study would be ever-smoker wives with
smoking husbands (s01) and that 14.0% would be ever-smoker wives with never-smoking husbands
(SOQ). Similarly, using the concordance factor of 2.2 for former smoking wives and ever-smoking
husbands versus the never-smokers, the former smoking wives married to ever-smoking husbands (con)
would be 10.9% of the total and those married to the never-smoking husbands (c010) would be 5.8%.
Then by difference, exposed current smoking wives (c021 + c03i) would be 22.3%, to be split into
20.1% regular smokers (c031) and 2.2% occasional smokers (c021), and the nonexposed current smoking
wives (c020 + Co30) would be 8.2%, split into 7.4% regular smokers (c030) and 0.8% occasional smokers
(c020). These data now supply all the elements needed in Table B-10 and the control part of Table
B-6.
. The relative risk for passive smoking, RR(p)0, for females is 2.07 (Correa et al., 1983). The
age- and sex-adjusted relative risk for current smoking from a related paper (Correa, 1984) is 12.6.
The ratio of female smoking crude risk to the average for males and females is about 80%, indicating
an age-adjusted current female risk of about 10. (Note: This is different from the current smoker
relative risk that would be calculated from the crude ever-smoker risk of 12.4 used in Table 5-7 [of
this report] and Table B-3. The adjusted risk is used here simply as an example.) With these inputs
and the weights of controls in the study, the various exposed and nonexposed relative risks are those
shown in Table B-ll. The weighted average risk for the .occasional smokers is calculated as 0.20
(current regular risk - 1) + 1, which for this example is 0.20 (10 - 1) + 1 = 2.80. The weighted
average risk for former smokers is 0.09 (current regular risk - 1) + 1, which is 0.09 '(10 - 1) + 1 =
1.81. The weighted average risks are split between never-smoking and ever-smoking husbands by
B-ll
05/21/92
-------
DRAFT-DO NOT QUOTE OR CITE
using the passive risks and the population weights. A crude case prevalence table is then made up
(Table B-12) by multiplying each cojk by its respective RRjk. This table is then normalized by dividing
through by 3.665 to yield Table B-13, which is the lower half of Table B-6 for this example.
The proportionate distribution table (Table B-14) is developed, as described above, from the
misclassification factors in Tables B-3 and B-4 and the bottom line of Table B-10. For example, to
arrive at element (h = 3, j = 0) the observed P.3 of 0.275 is multiplied by an observed misclassification
factor of 0.0102 (from Table B-3 of this report) to yield 0.00281, which rounds to 0.003. To explore
the value of using the true misclassification factors instead of the observed ones, the true and observed
m's were carried to five decimal places. An approximation procedure to determine the true smoking
probabilities P0., PI- P2., and P3. was carried through four stages. The resulting total true distribution
of smoking status, 0.49987, 0.18040, 0.03893, and 0.28081, rounded to three decimal places is
essentially identical to the distribution shown in the bottom line of Table B-14. Similarly, any
differences in the individual elements were very small and beyond the accuracy of the underlying data.
The Correa study was chosen as our example because the female ever-smoking prevalence is
reasonably high (47.2%) and the female current smoker lung cancer relative risk is high (10), both
factors that should lead to a greater rather than a smaller correction to the passive risk.
We now can set up a deletions table, Table B-15, which is the equivalent of Equation 1 above,
by multiplying the control and case elements in Table B-10 and B-13 by the appropriate observed
misclassification rates PM /P.j (h = j), namely, P10/P.i = 0.117, P20 /P.2 = 0.196, and P30 /P.3 = 0.01020.
For example, to get 0.00678, one multiplies 0.058 from Table B-10 by 0.117. Then the first three
columns are summed horizontally to get the fourth column which is then subtracted from the elements
in the "never" columns of Tables B-10 and B-13 (column 5) to get the "corrected never" elements
(column 6).
The corrected passive risk is now obtained by taking the cross product from the "corrected
never" column: 0.07516 x 0.27690/0.04705 x 0.22308 = 1.984, which is to be compared with the
observed risk of 2.07. The bias is then 2.07/1.984 = 1.044. It is interesting to note how sensitive the
bias is to the smoker relative risk that is assumed. When the crude smoker risk (no age adjustment) of
12.4 for ever-smokers, equivalent to about 15.4 for current regular smokers, is assumed, the corrected
passive risk is 1.90, and the basis is twice as great at 1.09.
B-12
05/21/92
-------
DRAFT-DO NOT QUOTE OR CITE
Table B-l. Observed ratios of occasional smokers to current smokers (based on cotinine studies)
\. - - - L, ^ "" ' Females,
/ StBdy - Oce' Current Octfl/CitrreBt ,
1
Lee (1986) 4 72 0.056
Coultas
et al. (1988)
Haddow 10 64 0.156
etal. (1986)
Feyerabend
(1982)2
Jarvis (1987)
Pojer(1984)
Wald et al.
(1984)
Overall 14 136 0.103
-
Oco-l1 Current -
12 176
59 278
7 82
12 90
25 187
. 13 ,131
128 944
Both $exes?
Oco*l Current
0.068
0.212
0.085
0.133
0.134
0.099 ...
0.136
1 Occasional smokers are defined as persons who have cotinine levels in body fluids that are
between 10% and 30% of the mean of all self-reported current smokers.
2 The Feyerabend (1982) data are for nicotine.
3 The "Both Sexes" data are shown to indicate that the female value of 10.3% is not unduly high.
B-13
05/21/92
-------
DRAFT—DO NOT QUOTE OR CITE
§
a
E
8
-a
A en
O "^
jf
(9 U3
U di
§ I?
.s a
I
.
«?
M
^
.Q
o
"o
•o
1
•H
S
1
i
r^
">?
*§
^
S
T-H
¥
r-J
O
1 •
b
w
*-*
C is
(§ «n
3
CL,
C
i — t
t?*
W
J2
S
g
c3
p
e*»
I
£
T-H
ON
ON
s
2
o
a
j_,
5
3
t
x— v
T"H
8}
t-H
\~s
8
'"'
P
2
CO
S
2
o
o
OH
w
2
o
s
sf
a
2
y
S
O
^j
60
'I
P
T-H
VO
rH
0\
T-H
JQ
t-H
(*1
T-H
T-H
en
en
t-H
T— t
r-H
t— <
CS
en
T-H
en
fc
O
P
T-H
^o
t-H
r-H
T-H
ON
T-H
^.
O\
ON
O
T— 1
IT)
t-H
T-H
P
T-H
^
T-l
'-*
04
I
fe
o
s
•^j-
T-<
en
i — i
£;
< — i
VO
0
i— i
en
cs
T-H
en
•
P
» — i
en
cs
T-H
IT)
1-H
rb
0
o
T-H
y—t
*O
OO
O
VO
OO
o
cs
vo
VO
C3
C3
m
i — i
i — i
VO
O
in
r-
o
o
1 — 1
H
.1
ON
O
T-H
c^
O
ON
1 — 1
o
cs
5
i — (
^
^*-
-^-
T— t
CS.
VO
T-!'
en
vq
t — i
^
.O
CS
en
%
O
u.
.— H 5G 11
*— ( TO 11
'O w 52
O *• , cs
^ A* CQ
|« $
r1 ?r~l
4—t ^ *55
U w 73
S Pi 'p
c -S n
*-sl
jas
4-> . s-i
~^ O ^^
"^3 CH >rt
QJ 44 Q
ill
;=) & g
J=! 2 ^
t3 o G
| ||
>] , 'O
^ Jxj <8 I
c/a c^ O
2 'cS tS
'<"> W
^ £2 1
1!!
•j— »
o
en
•o
o
Multiplica
e o
"0
So
•
•g
u.
0 0
-------
DRAFT—DO NOT QUOTE OR CITE
"S
oo
ON
%
<&
oo O
•o
o
00
I
VO
OT
oo
i
jy
1
&
gf
3
ft} V
•§
o
cs
CO
o
4)
00
i
*—H
c\
T-H
•o
i
•a
43
ON
.2
cL
S
O.-0
i
I
.On
o
E
O
I
I
a,
I
<3
o
'^
o
s
oo
S
ued
*?
CQ
~
3
^2
S
g
CO
oo
CO
'o
is
s
w CD
4) OO
TJ
s
OT
•a
4-1
O
•io
"C
i-,
4>
•g
g
w
ciS
00
s
EPA es
i -^
n w
S >->
W S
•S 1
00 CC
gs
fi s
«2 ^
S «>
•3 ^
r^ Wi
•r3 O
i y
e
00 0
EPA
B-15
05/15/92
-------
DRAFT-DO NOT QUOTE OR CITE
Table B-3. Misclassification of female current smokers
t
1
-------
DRAFT-DO NOT QUOTE OR CITE
Table B-3. (continued)
1 Cotinine levels are in units of percents of the mean of self-reported smokers for each study 30+%
are defined as current regular smokers, 10-30% are occasional smokers.
2 The observed current smokers are assumed to be 90% regular (1,372) and 10% occasional (153)
smokers. For regular smokers, misclassification as never-smokers is 14/372 = 1 02% of observed
current regulars or 14/(1,372 + 14 + 15) = 1.00% of true current regulars. For occasional smokers,
rmsclassification is 30/153 = 19.6% of observed current occasional or 30/(153 + 30 +16) = 15 1% of
true current occasional. For current smokers misclassified as former smokers the factors are 15/1 372
= 1.09% for observed and 15/1,401 = 1.07% for true regular smokers, and 16/153 = 10 5% for '
observed and 16/199 = 8.0% for true occasionals.
* For Lee (1986), Haddow et al. (1986), Haddow et al. (1988), and Riboli (1991), there was no
breakdown given between "Never" and "Former", because the numbers are small, an estimate was
made based on the subtotal distribution. The number of smokers had to be estimated in some cases.
4 New Orleans, Los Angeles, and Honolulu.
5 China (Shanghai), Hong Kong, and Japan (Sendai).
6 Athens.
B-17
05/21/92
-------
DRAFT-DO NOT QUOTE OR CITE
Table B-4. Misclassification of female former smokers reported as never-smokers based on discordant
answers
Study
Kabat and Wynder
(1984)5
Controls
Cases
Machlin
et al. (1989)
Krall et al. (1989)2
Britten (1988)3
Lee (1987b)
Akiba et al. (1986)
Overall4
x -I S"" ""'"•''
•> "> s ^"V "" ^ °lVl "* **/ \^'~
"&"l " Former "m
Locale f\ Smokers ' they toad snaked1
BVif-
Smofceri,^"
319
652
687
30
878
243
38
2847
, N
0
7
52
1
38
13
0
111
Percent
0.0
1.1
7.6
3.3
4.3
5.5
0.0
3.9
percent '
0.0
3.2
26.8
9.1
11.9
15.3
0.0
11.7
1 Number of former smokers and ever-smokers had to be estimated in some cases.
2 Krall data are based on 20-year recall.
3 Britten data include only those persons who said they never smoked but actually had smoked
regularly one or more cigarettes per day.
4 For former smokers, misclassification as never-smokers would appear to be 111/949 = 11.7% of
Observed former smokers or lll/(949 + 111) = 10.5% of true former smokers, but from Table B-3
16 + 15/(16 + 15 + 975) = 3.08% of former smokers are really current smokers, so the 949 +
111 = 1,060 should be reduced by 3.08% to 1,027 as the number of true former smokers. Then
111/1,027 = 10.81% based on true former smokers.
5 Dr. Kabat (private communication) advised that of 13 misclassifieds, 8 were females, 1 of whom
used snuff.
B-18
05/21/92
-------
DRAFT-DO NOT QUOTE OR CITE
Table B-5. Misclassification of female lung cancer cases
V/"*TMT'/»J»
OyW»"V'W ""
'•'.•*"'
CHAN
Chan et al. (1979)1
KABA
Kabat and Wynder (1984)2
AKIB
Akiba et al. (1986)
PERS
Pershagen et al. (1987)
HUMB
Humble et al. (1987)3
Total
General Population4
N«IB&«F of JEver-smokers
12
652
. 38, . . .,
179
223
1,104
1,838
Ntet^ttiffMitai *
1
7
. 0
2
1
11 (1%)
104 (5.7%)
1 Chan sampled five Type I and II never-smokers, one of whom was said by a relative to have
smoked a few hand-wrapped cigarettes for a year at age 71. The ratio of smoking to nonsmoking
cases for Types I and II was 44/19, which, multiplied by 5, leads to 12 estimated ever-smokers.
2 Dr. Kabat (private communication) advised that of 13 misclassifieds, 8 were females, 1 of whom
used snuff.
3 Of the four misclassifieds found, Dr. Humble (private communication) has advised that most if
not all were males. We have assumed one female.
4 The general population data are taken from the four nonlung cancer cohorts in Table B-4,
namely, Machlin (1989), Krall (1989), Britten (1988), and Lee (1987b).
B-19
05/21/92
-------
DRAFT-DO NOT QUOTE OR CITE
Table B-6. Notation for proportionate distribution of reported female lung cancer cages and controls
by husband's smoking status
- , :' '.••
Wife's
Subject
Status (i)
Control
0 = 0)
Case
0=1)
"rr " s"
Husband's " ^ ""\ u
Sraoking ' - " sJ&$Ve
Status (k> s ^ lcv"~ " (I, ^-' C
Never (k = 0) COQQ
Ever (k = 1) CQQ!
Total CQQ.
Never (k = 0) c100
Ever (k = 1) c101
Total c10.
, Wife's Observed Siaoktag Status; (f
^vv, ,*,-
f^ * s Ex OOC'l ' ^eg
)},;----
-------
DRAFT-DO NOT QUOTE OR CITE
Table B-7. Proportionate distribution notation for subjects by observed and true smoking status
'' ' ' - , ' life's True Smajdng States T>'
^10 r20 ^30 ' r.Q
Pll P21 P31 P-! .
Pl2 P22 P32 P-2
P13 P23 P33 P-3
Pi" P2- P3v P..(= 1)
B-21
05/21/92
-------
DRAFT-DO NOT QUOTE OR CITE
Table B-8. Observed lung cancer relative risks for exposed and nonexposed wives by the wife's
smoking status using average never-smoking wives as the reference category
Husband's
Smoking Status
Never (k = 0)
Ever (k = 1)
Weighted avg.
active risk
Passive risk1
RR(p)j =
RRjl/RRjo
,
Never „,"" I
(J * 0> , -
RRoo
RR,
RR(a)0 = 1.00
RR(P)0
\"ira,. Wife's SmoMng States '"
o, former -;\1; ',"boc*l ,;, Heg+
RR-io RRso KRao
RRU KR2l RR31
RR(a)! . RR(a)2 RR(a)3
KRfp\ RR(P)2 RR(P)3
Observed passive risk--the ratio of the exposed risk to the unexposed risk in each column.
B-22
05/21/92
-------
DRAFT-DO NOT QUOTE OR CITE
Table B-9. Prevalences and estimates of lung cancer risk associated with active and passive smoking33
s
Case-
Coiitrol
MMUMB
AKIB
BROW
BUFF
CHAN
CORR
FONT34
GAO
GARF
GENG
HIRA8
HUMB
INOU
Ever-smofcers.
% , % Prev. " Grade -
21 2.38
(1.67,3.39)
29 4.3023
(2.24, 8.24)
59 7.0615
(5.18, 9.63)
26 3.48
(2.42, 4.99)
47 12.40
(8.35, 18.4)
.4221 8.021
18 2.54
(2.06, 3.12)
* , * :,
41 2.7727
(1.89, 4.07)
16 3.200
(2.67, 3.83)
41 16.3
(10.5, 25.1)
16 1.66
(0.73, 3.76)
Prev. of
70 ,
15
12
84
47
46 , .
63
66
64
74
61
44
77
56
64
H&vejr-stKolc-ers
"'_ Oracle
1.52
(0.96,2.41)
1.5223
(0.49, 4.79)
1.8223
(0.45, 7.36)7
0.8115
(0.39, 1.66)
0.75
(0.48, 1.19)
2.0724
(0.94, 4.52)
1.37
(1.10, 1.69)
1.21
(0.94, 1.56)
1.32
(1.08, 1.61)
1.19
(0.87, 1.63)
1.31
(0.93, 1.85)
2.16
(1.21, 3.84)
1.53s
(1.10, 2.13)
2.34
(0.96, 5.69)
'. 2.5516
(0.90, 7.20)
-
Adj;
1.5
(1.0, 2.5)
*
1.6823
(0.39, 6.90)7 •
*
*
*
1.29
(1.03, 1.62)
1.28
(0.98, 1.66)
*
1.345-6
' 1.7026
(0.98, 2.94)7
*
1.64s
*
2.2
(0.9, 5.5)
2.545-10
*
B-23
05/21/92
-------
DRAFT-DO NOT QUOTE OR CITE
Table B-9. (continued)
Case-
Control
JANE
KABA28
KALA
KATA
KOO
LAMT
11
LAMW
1
ij
LEE
LIU
PERS
SHIM
SOBU
SVEN
TRIG
Ever-smokers "
Prev>- " Crude , ,,
(%)* ' , ^ " ,£&*._. J,""'E
4621
42
17
28
32
24
22
6029
0.05
3721
2121
21
43
10
8.021
5.90
(4.53, 7.69)
3.32
(2.12, 5.22)
1.21
(0.50, 2.90)
2.77
(1.96, 3.90)
3.77
(2.96, 4.78)
4.12
(2.79, 6.08)
4.6129
*
4.221
2.821
2.81
(2.22, 3.57)
5.97
(4.11, 8.67)
2 8 125
(1.69,4.68)
J&eV'of
xp&sed' <$)3
80
60
60
82
49
45
56
68
87
43
56
54
66
52
,^Mever-smoksrs
C/rode
0.86
(0.57, 1.29)
0.79
(0.30,2.04)
1.6212
(0.99, 2.65)
1.41
(0.78, 2.55)
*19
1.55
(0.98, 2.44)
1.65
(1.22, 2.22)
2.5120
(1.49, 4.23)
1.03
(0.48, 2.20)
0.74
(0.37, 1.48)
1.28
(0.82, 1.98)
1.0830
(0.70, 1.68)
1.0612
(0.79, 1.44)
1.77
(1.29, 2.43)
1.2614
(0.65, 2.48)
2.0825
(1.31, 3.29)
Af 4 17
0.93/0.4411
*
1.92
(1.02, 3.S9)7
*
1.64
*
*
0.75/1. 6013
0.77
^ (0.35, 1.68)
1.2
(0.7, 2.1)7
*
1.1312
(0.78, 1.63)7
1.57
(1.07, 2.31)7
1.414
*
B-24
05/21/92
-------
DRAFT-DO NOT QUOTE OR CITE
Table B-9. (continued)
Case.-
Control,
WU
WUWI
BUTL
(Coh)
GARF
(Coh).
HIRA
(Coh)
HOLE32
(Coh)
Bver-smokers-
Never-smoteets
-Prey. Crude Prev, of Crude
l%)1 ' •• RR2 s Exposed (%)3 "RR2-i? -•"
58 4.38
. (2.97, 6.47)
37 2.24
(1,92, 2.62)
1421 4.021
3322 3.522
16 3.209
(1.96, 3.90)
56 4.221
60 1.4118
(0.63, 3.15)
55 0.79
(6.64,0.98)
* 2.4531
72 *
77 1.38
(1.03, 1.87)
73 2.27
(0.40, 12.7)
f j. f
< Adj,
HR^4'IT
- 1.2
(0.6, 2.5)7
0.7
2.02
(0.48, 8.S6)7
1.17s
(0.85, 1.61)7
1.61
*
1.99
(0.24, 16.7)7
Percent ever-smokers in controls of whole study (or parent study).
Parentheses contain 90% confidence limits, unless noted otherwise. Crude ORs and their
confidence limits were calculated by the reviewers wherever possible. Boldface indicates values
used for analysis in text of this report. OR for case-control studies; relative risk (RR) for cohort
studies. The reference category for active smoking is all never-smoking, for passive smoking, it is
unexposed never-smokers.
Percent of never-smoking controls exposed to spousal smoking, unless noted otherwise.
Calculated by a statistical method that adjusts for other factors (see Table 5-5).
Composite measure formed from categorical data at different exposure levels.
For Gao, data are given as (number of years lived with a smoker, adj. OR): (< 20, 1.0), (20-29,
1.1), (30-39, 1.3), (40+, 1.7).
95% confidence interval.
Case-control study nested in the cohort study of Hirayama. OR for ever-smokers is taken from
cohort study (shown in table below). This case-control study is not counted in any summary
results where HIRA(Coh) is included.
Crude OR is calculated from prospective data in Hirayama (1988). Adjusted OR for ever-
smokers given there is 2.67 (no confidence interval [C.I.]).
For Inoue, data are given as (number of cig./day smoked by husband, adj. OR): (< 19, 1.58), (20+,
3.09).
From subject responses/from proxy responses. '
For the first value, "ETS exposed" means the spouse smokes; for the second value, "ETS exposed"
means a member of the household other than the spouse smokes.
13 From subject responses/from spouse responses.
14 Exposure at home and/or at work.
10
12
B-25
05/21/92
-------
DRAFT-DO NOT QUOTE OR CITE
Table B-9. (continued)
15
Exposure to regularly smoking household member. Differs slightly from published value of 0.78,
wherein 0.5 was added to all exposure cells.
16 OR reported in study is 2.25, in contrast to the value shown that was reconstructed from the
confidence intervals reported in the study; no reply to inquiry addressed to author had been
received by press time.
17 ORs for never-smokers applies to exposure from spousal smoking, unless indicated otherwise.
18 Raw data for WU is from Table 11 of the Surgeon General's report (U.S. DHHS, 1986). Data
apply to adenocarcinoma only.
19 Odds ratio is not defined because number of unexposed subjects is 0 for cases or controls.
Table entry is for exposure to smoking spouse, cohabitants, and/or coworkers; includes lung cancers
of all cell types. The OR for spousal smoking alone is for adenocarcinoma only: 2.01 (90% C.I. =
1.20, 3.37).
From other studies similar in location and time period (see Table 5-7).
Prevalence is calculated from figures in Stellman and Garfinkel (1986) and includes all women
except those who "never smoked regularly." RR is from U.S. Surgeon General (U.S. DHHS,
1982).
Adenocarcinoma only. Data and OR value communicated, from author (Brownson).
Excludes bronchioalveolar carcinoma. Crude OR with bronchioalveolar carcinoma included is
reported to be 1.77, but raw data for calculation of confidence interval are not provided.
Known adenocarcinomas and alveolar carcinomas were excluded, but histological diagnosis was not
available for many cases. Data are from Trichopoulos et al. (1983).
26 Estimate for husband smoking 20 cigarettes per day.
Crude OR reported in study is 3.05 (95% C.I. = 1.77, 5.30); adjusted OR is 2.6(95% C.I. = 1.4,
4.6).
For second KABA study (see addendum in study description of KABA), preliminary unpublished
data and analysis based on ETS exposure in adulthood indicate 68% of never-smokers are exposed
and OR = 0.90 (90% C.I. = 0.51, L58), not dissimilar from the table entry shown.
From Alderson et al. (1985).
30 From crude data estimated to be the following: exposed cases 52, exposed controls 91, unexposed
cases 38, unexposed controls 72.
RR is based on person-years of exposure to spousal smoking. Prevalence in those units is 20%.
RR values under never-smoker are for lung cancer mortality. For lung cancer incidence, crude RR
is 1.51 (90% C.I. = 0.41, 5.48) and adj. RR is 1.39 (95% C.I. = 0.29, 6.61).
33 Values used for inference in this report are shown in boldface. * means no information available.
34 The first, second, and third entries are calculated for population controls, colon cancer controls, and
both control groups combined, respectively. For adenocarcinoma alone, the corresponding ORs,
both crude and adjusted, are higher by 0.15 to 0.18.
20
21
22
2S
27
28
29
31
32
B-26
05/21/92
-------
DRAFT-DO NOT QUOTE OR CITE
Table B-10. Observed smoking prevalence among the controls-Correa example
Husband's -
Smoking Status
Never
Ever
All
' 4", '
- Kever
0.286
0,242
0.528
-
' Former
0.058
0.109
0.167
Wile's SmoMog Status '
Oc«io«*'
0.008
0.022
0.030
'
Regular >Ul
0.074 0.426
. 0.201 0.574
0.275 1.000
Table B-ll. Observed relative risks—Correa example
s Wjlc- S- WX&OOI3& •O'tSt&S- ^
Husbands' ^ /t t
Smoking Status
Never
Ever
Weighted Average
Passive Risk, RR(p)j
Never %
(J-0)
0.67
1.39
1.00
2.07
Format ""
1.07
2.21
1.81
2.071
Ocoasloiial --
2.80
2.80
2.80
LOO1
Regular
'(1-3) ' f
10.0
10.0
10.0
LOO1
1 Assumed.
B-27
05/21/92
-------
DRAFT-DO NOT QUOTE OR CITE
Table B-12. Crude case table - prevalence of cases by smoking status-Correa example
Husband's
Smoking Status
,V, -•, ^ J ,,,. SSf
Occasional
Regular
AH
Never
Ever
All
0.192
0.336
0.528
0.062
0.241
0.303
0.022
0.062
0.084
0.740
2.010
1.016
2.649
2.750 3.665
Table B-13. Normalized case table - prevalence of cases by smoking status-Correa example
Husband's
Smoking Status
Never
Ever
All
., ' --\ ^ Wife '-s Smoking Status ' ., s
* f t s f ^ f tf : r l it'fl : ft
''""' ^ "" *." ','"'' ,
f : Never s " " rosier Occasional' Regular" All
0.052 0.017 0.006 0.202 0.277
0.092 0.066 0.017 0.549 0.723
0.144 0.083 0.023 0.750 1.000
B-28
05/21/92
-------
DRAFT-DO NOT QUOTE OR CITE
Table B-14. Proportionate distribution of observed and true smoking status for wives in Correa example1
Wffe,'s Observed - , Never"
SttkMag Status ' (h'*» Q>
Never (j = 0) 0.500
Ex 0 = 1) 0
Occ'l 0 = 2) 0
Regular 0 = 3) 0
All 0.500
Wife's True i Smolnjig
Pormer - Occasional '
Qi«=l) (h = 2)
0.020 0.006
0.161 0.003
0 0.030
0 0
0.180 0.039
s«,5 - ;- •
-------
DRAFT-DO NOT QUOTE OR CITE
Table B-15. Deletions from the never columns in Tables B-10 and B-13
,'y- - - Wife's &&qj$ng Sfctfus
Husband's
Smoking Status
Table B-10
(pop.)
Table B-13
(cases)
Foqttef ,
Never 0.00678
Ever 0.01274
Never 0.00198
Ever 0.00769
Occl '
0.00157
0.00433
0.00120
0.00331
-Regular'
0.00075
0.00205
0.00206
0.00559
1 IT '
0.00910
0.01892
0.00524
• 0.01659
- Observed
, Never
0.286
0.242
0.05229'
0.09178
Corrected
Never2
0.27690
0.22308
0.04705
0.07519
1 (4) = (1) + (2) + (3)
2 (6) = (5) - (4)
B-30
05/21/92
-------
DRAFT-DO NOT QUOTE OR CITE
Table B-16. Observed ratios of female former smokers to ever-smokers in the U.S.A., U.K., and Sweden:
populations or controls (numbers or %)
-Study '
Hammond (1966)
Buffler
et al. (1984)
Wu et al. (1985)
Lee (1987b)
Brownson
et al. (1987)
Britten (1988)
Humble
et al. (1987)
Svensson
et al. (1989)
Garfinkel and
Stellman (1988)
Time
1960
1978
1980
1980
1980
1982
1982
1984
1982
Never*
381,369,
'41%
92
48.3%
47
767
162
- 120
350,650
Carrent "" Fotmer-
Suwjikef § Smokers, ,
150,017 31,285
38% 21%
73 55
33.6% 18.1%
11 8
558 320
63 48
53 36
132,366 136,909
Ever- ,,
Smqfcera
181,302
59%
128
51.7%
19
878
111
89
269,275
Former/Ever-
0.17,
0.36
0.43
0.35
0.42
0.36
0.43
0.40
0.51
Assumed Ratios bv Years (non-traditional societies)1
Year
Ratio
1960 1965
0.17 0.23
1970 1975 1980
0.28 0.34 0.39
1985
0.45
1 Traditional societies (Japan, Greece, China, Hong Kong) are estimated to lag these ratios by about 20
years, although there are no data in the studies to confirm this. However, because the bias for the
traditional societies is very low, changes in values of this parameter have littte"effect.
B-31
05/21/92
-------
DRAFT-DO NOT QUOTE OR CITE
Table B-17. Observed ratios of current smoker lung cancer risk to ever-smoker risk for females,
Study
Alderson
et al. (1985)
Buffler
et al. (1984)
Garfinkel and
Stellman (1988)
Humble et al.
(1985)
Svensson
et al. (1989)
Wu et al.
(1985)
Overall
s **
•Xs -
Exposed Ca$&s -
Plus Controls, ;
901
701
832
268
261
317
3,280
f
»-^l!l
^Smoker _',
4.5
7.9
12.7
18.0
8.46
6.5
8.05
K&erRR
•",,, ft - "(
Ever-
Smoker
4.75
6.9
8.35
13.0
6.10
4.4
6.52
Ratio , , ,
.. •" •*
"- '^' Cua^ot Smoker RR/
0.95
1.15
1.52-
1.38
1.39
1.48
1.241
1 The summary ratio of 1.24 is the log mean of the individual ratios weighted by the exposed cases
plus controls in that study.
B-32
05/21/92
-------
DRAFT—DO NOT QUOTE OR CITE
APPENDIX C
REVIEW FORMAT FOR CASE-CONTROL STUDIES
PART I GENERAL
Study name_
Location
Time period (data collection).
Study objective(s)
The source of the primary data set is the current study _. or a parent study
(ref) : '
containing CS (current) FS (former) NS (never-smoker)
Study uses term "nonsmoker" or "never-smoker" ^ to mean
nonsmoker
never-smoker
"Exposed" to ETS means (preferably in terms of spousal smoking)
Recall span (how far back in time ETS-exposure was measured).
ETS sources include cigarette
cigar
pipe
other
C-l
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Describe inclusion of non-smoking (never smoking) females not currently married
(number of cases and controls, assumptions exposure)
H DATA COLLECTION (includes NS_
FS
CS
unless noted)
Inclusion/Exclusion criteria
Cases
Controls (include matching variables in PART V)
Main source of subjects
Hospital(s) #
Community
Other
Incident cases Y
Control sampling
Cumulative
Unmatched
Method of collection
Face-to-face
Telephone
Self-admin, ques.
Medical records
Vital stat. records
Other
N
Cases
Controls
Density
Matched
Controls
Collected data verified/corroborated with other sources Y_
N
C-2
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Controls
Sample size
(prior to attrition)
females
males
Attrition
(selection or follow-up)
females
males
Source of response
subject
proxy
Exposure sources NS_
Childhood
Adulthood
Spouse
Parents/in-laws
Other family/
live-ins
Workplace
Other
Age NS_
FS
Distribution
Mean
Standard error
Standard deviation
Range
_FS_
Yes
CS
cs
No
Controls
PART III CLINICAL DATA
Primary lung cancer verified by
Histology
Cytology
Radiology/clinical
NS
FS
CS
C-3
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Death certificate
Tumor registry
Mortality records
Other
Not verified
Airway proximity (no. exp cases/no, cases)
Central
Peripheral
NS
Table
FS
CS
Tumor type (no. exp cases/no, cases)
Squamous cell
Small cell
Adenocarcinoma
NS
Table
FS
CS
Large cell
Others or unspecified
PART IV STATISTICAL ANALYSIS (includes NS FS_
Raw data (for analysis) Cases
females
males
unexp
exp
unexp
exp
CS
unless noted)
Controls
Comments (include measure of exposure) Table
C-4
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Unadjusted (crude) analysis
Estimate OR
Comments
Test of p-value
signif
Test for p-value
trend
Comments
Adjusted analysis
Estimate OR
Test of p-value
signif
Test for
trend p-value
Comments
_% CI (_
Table
Table
Table
PART V DEPENDENT VARIABLES (potential confouders and effects modifiers considered)
In Matching
In Analysis
Age
Gender
Race/ethnicity
Hospital
Residence/
neighborhood
Housing type
House/room sizes
Vital statistics
Smoking status
SES
Medical health
Menstrual/
reproductive
Occupation
Outdoor air
pollution
Cooking habits
Drinking
Diet
Otherwise
C-5
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Education
Family history
of LC
Other indoor
smoke/fumes
Radon
Lifestyle
Climate/
ventilation
C-6
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
APPENDIX D. LUNG CANCER MORTALITY RATES ATTRIBUTABLE TO SPOUSAL
ETS IN INDIVIDUAL EPIDEMIOLOGIC STUDIES
Many of the epidemiologic studies on lung cancer and environmental tobacco smoke (ETS)
were part of larger investigations that included ever-smokers and never-smokers. For those
studies, the lung cancer mortality rate (LCMR) for all causes, appropriate to the location and time
period of the study, has been obtained from other sources. Those values and parameter estimates
from the studies are used to partition the excess LCMR from all causes (i.e., the excess after
allowance for baseline sources) into components attributable to ever-smokers (from current and
former smoking) and never-smokers (from exposure to spousal ETS) and to estimate the LCMR in
the subpopulations of interest—unexposed never-smokers (meaning not exposed to spousal
smoking), exposed never-smokers (exposed to spousal smoking), and ever-smokers ("exposed", is
not used to mean exposure to non-spousal ETS, which applies to the whole target population).
The method is explained in Sections 6.3.1 and 6.3.2.
Lung cancer mortality rates for the case-accrual periods of case-control studies are
displayed in Table D-l. For the studies that collected data on both ever-smokers and never-
smokers, the parameter estimates used are shown in Table D-2. The value for the lung cancer
mortality rate is .from Table D-l, and the remaining estimates are from individual study data. For
HIRA(Coh), the lung cancer mortality rate for the time and location of the nested case-control
study HIRA is used. For GARF(Coh), the rate for GARF in 1971 is assumed, which is the
approximate time of the cohort follow-up. These values may not be very "representative" for lung
cancer mortality in these two cohort studies because they extended over several years, and the
LCMRs changed from year to year, particularly in the United States. This same difficulty arises
in choosing a "representative" year for lung cancer mortality in the case-control studies, although
to a lesser degree. The most extreme examples are KABA, PERS, INOU, and GARF with case-
accrual periods of 10 years or more.
The estimates of prevalence of ever-smokers and the percent of never-smokers exposed to
spousal smoking are the observed proportions in the control group. The extent to which the
control group is representative of the country's population differs between studies, with those
most questionable shown in Table 5-14A. The study reviews in Appendix A provide more
detailed information. The restriction of cell types among cases in some studies is another
consideration. Active smoking is much more strongly associated with occurrence of squamous and
small cell carcinoma than with large cell carcinoma and adenocarcinoma. FONT presents evidence
that passive smoking is more associated with adenocarcinoma than with other cell types. As noted
D-l
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
in Table 5-15, some studies excluded candidate lung cancer cases of specific histopathological
types. This may produce some bias and distortion of comparison between studies. For example,
BROW includes only cases of adenocarcinoma, which should bias the relative risk of ever-smokers
toward unity, thus attributing too little lung cancer mortality to active smoking and too much to
passive smoking and background sources.
Of a more positive nature, there is some advantage to using data from a single study to
assign attributable fractions to different causes. To estimate the yearly number of lung cancers
from each cause, the fraction is multiplied by the LCMR for the location and time of the study;
that figure has to be obtained from sources on vital statistics. As seen in Table D-2, the mortality
rates from lung cancer vary considerably between and within countries. For example, the rates
used for studies in the United States range between 9 and 26. Applying the lung cancer rate
suitable to each individual study should provide better estimates for comparison within a country
than using a single figure for the whole country for some specific year.
Despite the reservations described, partitioning the lung cancer mortality for each study
into components attributable to ever-smoking, spousal ETS, and baseline sources (nontobacco
smoke and nonspousal ETS) provides a broad overview worth noting. The calculated values are
shown in Table D-3. Estimates of relative risk for exposure to spousal ETS (RR2 in notation of
Section 6.3.2) less than 1.0 (see Table 5-8) were replaced by 1.0 to avoid a negative LCMR
attributable to spousal ETS and the consequent inflation of the LCMR attributable to baseline
sources and ever-smoking. Aside from the studies for Hong Kong and China, estimates of lung
cancer mortality due to background sources cluster in the interval 1.5 to 5.5 (excluding BROW,
which is strongly biased), predominantly from 3 to 5. The values for Hong Kong and China,
however, are much higher, ranging from 7 to 14.5. The presence of indoor sources of non-ETS
encountered in some of the studies in China may be a factor, but there is no apparent explanation
for the outcome in Hong Kong. Assuming that the background rate of lung cancer is much higher
in Hong Kong (and possibly China) as it appears, then the question arises as to whether the high
excess rate relative to other countries may be attributable to higher exposure to ETS aside from
spousal smoking or whether it is more likely due to other causes. Summary data from the ten-
country collaborative study of ETS exposure to nonsmoking women conducted by the
International Agency for Research on Cancer (IARC) (Riboli et al., 1991) was kindly submitted to
us for Hong Kong, Japan (Sendai), and the United States (Los Angeles, New Orleans) from Drs.
L.C. Koo, H. Shimizu, A. Wu-Williams, and T.H. Fontham, respectively. The average
cotinine/creatinine (ng/mg) levels for nonsmoking women who are not employed and not married
to a smoker are close for Sendai, Los Angeles, and New Orleans, but they are several times higher
D-2
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
for Hong Kong. Consequently, a high contribution to background lung cancer mortality from
ETS aside from spousal smoking cannot be eliminated as a factor.
The lung cancer attributable to ever-smoking, spousal smoking, and baseline sources
depends on the population proportions for those categories as well as the relative risks. Study
estimates of the LCMR in each category, in units of lung cancer deaths per 100,000 at risk per
year, are shown in Table D-4, The last two columns show the ratios of the LCMR and the excess
LCMR for exposed never-smokers to ever-smokers. As above, relative risk estimates of less than
1.0 were set to 1.0 for the calculations. There is considerable variability across study estimates,
even within the same country, as observed previously in the relative risks for spousal smoking.
To summarize, for studies that included data on ever-smokers, the LCMR for all causes
was partitioned by attributable source (Table D-3). Although there is considerable uncertainty in
the estimates from statistical variability and other sources, the outcomes provide some useful gross
comparisons. For example, the lung cancer mortality rates from all causes differ markedly
between countries and also vary widely between studies within the United States. The proportion
of lung cancers attributable to ever-smoking is very high in the United States, compared to some
more traditional countries (e.g., Japan and Greece).
Individual study estimates of the number of lung cancer deaths per year per 100,000 of
female population from exposure of never-smokers to spousal ETS are predominantly between 0
and about 2.5. Estimates of the LCMR attributable to baseline sources (nonspousal ETS and
nonsmoking causes) are somewhat higher, largely between 2 and 5, except in Hong Kong and
China, where they range between 7+ and 14, (The U. S. study denoted as BROW has a high value,
but that should be upwardly biased because it used only cases of adenocarcinoma, which is not a
common cell type in smokers.) For reasons discussed in Chapter 5, we would be reluctant to draw
conclusions about China on the basis of the epidemiologic studies. The evidence from Hong
Kong, however, is very suggestive that the lung cancer rate in women due to baseline sources is
very high. The extent to which that is attributable to nonsmoking sources of lung cancer and/or
high exposure to nonspousal ETS is not apparent. The cotinine data for Hong Kong from the ten-
country IARC study (Riboli, 1990) is consistent with excessively high ETS exposure, so
nonspousal ETS may be a factor.
D-3
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table D-l. Female lung cancer mortality from all causes in case-control studies'
Study
AKIB
BROW
BUFF
CHAN
CORR2
GAO3
GARF
GENG3
HIRA7
HUMB2
INOU
JANE2
KABA4
KALA4
KATA4
KOO
LAMT4
LAMW
LEE
PERS4
SHIM4
SOBU4
SVEN4
TRIG
WU
WUWIfi
Location
Japan
USA
USA
HK
USA
China
USA
China
Japan
USA
Japan
USA
USA
Greece
Japan
HK
HK
HK
Eng/Wal
Sweden
Japan
Japan
Sweden
Greece
USA
China
Case
Accrual
1971-80
1979-82
1976-80
1976-77
1979-82
1984-86
1971-81
1983
1965-81
1980-84
1973-83
1982-84
1961-80
1987-89
1984-87
1981-83
1983-86
1981-84
1979-82
1961-80
1982-85
1986-88
1983-85
1978-80
1981-82
1985-87
Begin
5.13
15.68
13.94
23.59
26.0
*
9.45
*
4.46
17.7
5.55
23.7
4.69
6.58
*
22.34
22.75
22.34
16.28
3.71
7.46
7.46
7.72
6.88
17.20
*
Average
6.05
17.29
15.29
23.59
26.0
18.0
13.55
27.8
5.70
17.7
6.53
23.7
13.20
6.584
7.464
22.61
23.46
22.88
17.11
5.09
7.464
7.464
7.724
6.40
18.15
11.6
, % End
7.08
19.09
17.20
23.59
26.0
*
17.20
*
7.08
*
7.46
*
17.20
6.58
*
22.75
23.69
23.69
17.89
7.56
7.46
7.46
. 7.72
5.99
19.09
*
Accrual
-*10 yrs
Average
4.57
9.49
7.86
19.05
. 9.49
14.32
6.87
13.82
4.01
10.55
4.93
9.06
6.61
6.75
4.66
19.82
21.33
20.09
12.60
3.954
5.65
6.36
5.78
5.75
10.14
9.22
Accrual
-20 yrs
Average' I
2.30
4.75
4.38
*
4.75
5. 12
*
*
*
5.13
2.95
5.42
4.16
5.834
2.26
*
*
*
8.1
*
4.28
4.93
3.80
5.315
4.96
*
1 Rates are per 100,000 per year. Annual rates for 2-year periods from Kurihara et al. (1989)
were averaged over the years cases were accrued for each study unless otherwise noted. Where
part (or all) of the accrual period fell 1 or 2 years outside the years for which rates were
available, rates from the nearest 2-year period available were assumed to apply to the missing
years. U.S. rates are for white females only.
2 Data for accrual period from 1978-82 rates in IARC (1987), standardized to 1950 world
population from Kurihara et al. (1989). For Correa, weighted average of white and black rates;
for Humble, weighted average of Hispanic and non-Hispanic white rates.
D-4
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table D-l. (continued)
Accrual period data for Gao and Geng derived from IARC (1987) by standardizing to same
1950 world population used by Kurihara et al. (1989), Gao rates are for 1978-82; Geng, 1981-
82. For -10 years, Gao and Geng are 1973-75 rates standardized to the 1960 world population
from China Map Press (1979). Gao -20 years value is nonadjusted 1961 rate from Kaplan and
Tsuchitani (1978).
Where rates for the period were not available in Kurihara et al. (1989), substitutions were made
as follows: Kalandidi from 1984-85 rates; Kabat 1982-83; Katada 1982-83; Lam, T. 1984-85;
Pershagen 1952-53; Shimizu 1982-83; Sobue 1982-83; Svensson 1982-83.
World-standardized rate for 1961-65 from Katsouyanni et al. (1990). [In Greek: translation
provided by Trichopoulos.]
Accrual period value estimated by multiplying LCMR in Shanghai for period 1978-82
(standardized to the 1950 world population) by the ratio of LCMRs in Liaoning and Heilonjiang
to Shanghai, for the period 1973-75 (standardized to the 1960 world population). Data are from
China Map Press (1979). Value for -10 years is the 1973-75 rate.
The nested core-control study of Hirayama.
Data not available.
D-5
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table D-2. Parameter values used to partition female lung cancer mortality into component
sources
Ever-smokers
Case-Control
AKIB
BROW
BUFF
CHAN
CORR
GAO
GARF(Coh)
GENG
HIRA
HIRA(Coh)
HUMS
INOU
KABA
K;ALA
KOO
LAMT
LAMW
LEE
SOBU
SVEN
TRIG
WU
WUWI
Lung Cancer
Mortality
6.05
17.29
15.29
23.59
26.00
18.00
9.45
27.80
5.70
5.70
17.70
6.53
13.20
6.58
22.61
23.46
22.88
17.11
7.46
7.72
6.40
18.15
11.60
Prevalence
(%)
21
29
59
26
47
18
33
41
16
16
41
16
42
17
32
24
22
60
21
43
11
58
37
y "*
Relative
Risk
2.38
4.30
7.06
3.48
12.40
2.54
3.50
2.77
3.20
3.20
16.30
1.66
5.90
3.32
2.77
3.77,
4.12
4.61
2.81
5.97
2.81
4.38
2.24
Never-smokers
Percent
Exposed (%)
70
15
84
47
46
74
72
44
77
77
56
64
60
60
49
45
56
68
54
66
52
60
55
Relative
Risk
1.50
1.50
0.81
0.74
1.90
1.19
1.15
2.16
1.53
1.37
1.98
2.55
OJ4
1.92
1.54
1.64
2.51
1.01
1.13
1.19
2.08
1.31
0.78
1 For studies with data on both ever-smokers and never-smokers. Table entries are drawn from
Tables 5-4, B-8 and D-l, which contain explanatory footnotes.
D-6
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table D-3. Female lung cancer mortality rates by attributable source'
"" ' Baseline
Sources2
Study
AKIB
BROW
BUFF
CHAN
CORR
GAO
GARF (Coh)
GENG
HIRA (Coh)
HUMB
INOU
KABA
KALA
KOO
LAMT
LAMW
LEE
SOBU
SVEN
TRIG
WU
WUWI
Location
Japan
USA
USA
HK
USA
China
USA
China
Japan
USA
Japan
USA
Greece
HK
HK
HK
Eng./Wales
Japan
Sweden
Greece
USA
China
No*
3.47
8.22
3.34
14.34
2.89
12.36
4.67
10.67
3.28
1.57
2.97
4.32
3.04
11.41
10.94
7.35
5.37
5.05
2.19
3.42
5.17
7.95
% '
57
48
22
61
11
69
49
38
58
9
45
33
46
50
47
32
31
68
28
53
28 ,
69
Spousal Smoking
0.96
0.44
0.00
0.00
0.63
1.42
0.33
3.21
0.78
0.51
2.47'
0.00
1.39
2.05
2.39
4.85
0.01
0.28
0.16
1.71
0.40
0.00
16
3
0
0
2
8
4
12
14
3
38
0
21
9
10
21
0
4
2
27
2
0
Ever-smoking
" '~ . No-- '$>
1.61
8.63
11.95
9.25
22.47
4.22
4.44
13.92
1.63
15.62
1.09
8.88
2.15
9.14
10.12
10.68
11.73
2.13
5.37
1.27
12.58
3.65
27
50
78
39
86
23
47
50
29
88
17
67
33
40
43
47
69
29
70
20
69
31
1 Rates are per 100,000 per year. Data not available for GARF, JANE, PERS, SHIM,
BUTL(Coh), and HOLE(Coh).
2 Nonspousal ETS and non-ETS sources.
D-7
05/15/92
-------
DRAFT—DO NOT QUOTE OR CITE
Table D-4. Lung cancer mortality rates of female ever-smokers (ES) and never-smokers (NS) by
exposure status1
Study
AKffi
BROW
BUFF
CHAN
CORR
GAO
GARF (Coh)
GENG
HIRA (Coh)
HUMS
INOU
KABA
KALA
KOO
LAMT
LAMW
LEE
SOBU
SVEN
TRIG
WU
WUWI
Location
Japan
USA
USA
HK
USA
China
USA
China
Japan
USA
Japan
USA
Greece
HK
HK
HK
Eng/Wal
Japan
Sweden
Greece
USA
China
(1)
TJnexposed
NS2
3.47
8.21
3.34
14.34
2.89
12.35
4.67
10.66
3.28
1.57
2.96
4.32
3.04
11.41
10.94
7.35
5.36
5.05
2.18
3.41
5.16
7.95
(2)
Exposed
NS*
5.21
12.32
3.34
14.34
5.49
14.70
5.37
23.03
4.49
3.11
7.56
3.78
5.84
17.57
17.94
18.45
5.42
5.70
2.60
7.10
6.77
7.95
' (3) (2) A& a
Percentage
,E.S. of (3)
11.16
37.99
23.59
49.91
50.70
35.79
18.12
44.62
13.49
39.66
9.80
25.46
15.66
39.98
53.12
55.89
24.91
15.18
14.69
14.99
26.85
17.81
47
32
14
29
11
41
30
52
33
8
77
17
37
44
34
33
22
38
18
47
25
45
(2) ~G)
Asa
Percentage
of (3MI)
23
14
0
0
5
10
5
36
12
4
67
0
22
22
17
23
0
6
3
32
7
0
1 Rates are per 100,000 per year. Data not available for GARF, JANE, PERS, SHIM,
BUTL(Coh), and HOLE(Coh).
2 Exposed to baseline sources-nonspousal ETS and non-ETS sources,
3 Exposed to baseline sources plus spousal ETS.
D-8
05/15/92
-------
DRAFT—NO NOT QUOTE OR CITE
APPENDIX E. STATISTICAL FORMULAE
E.I. CELL FREQUENCIES
The observed outcome of a case-control study or a cohort study may be depicted in a 2 x 2
table, where a, b, c, and d are cell frequencies.
ETS Exposed -..••-•
Yes No
t
Lung Cancer Yes
Present
No
a
c
b
d
E.2. CASE-CONTROL STUDIES
The true (but unknown) odds ratio is estimated by the observed odds ratio (OR), :
OR = ad/bc.
A confidence interval on the (true) odds ratio may be calculated from the normal approximation
to the distribution of log(OR), the natural logarithm of OR (Woolf , 1955). The variance of
log(OR) is estimated by
Var(log(OR)) = I/a + 1/b + 1/c + 1/d
and the standard error by its square root,
SE(log(OR)) = (Var(log(OR)))1'4.
Approximate 90% confidence limits are given by
log(OR) ± 1.645 SE(log(OR)).
The value 1.645 is replaced by 1.96 for 95% confidence limits and, in general, by Za/2 for 100(1 7-
a)% confidence limits. The confidence bounds obtained in this way are sometimes called logit
limits (Breslow and Day, 1980; p. 134). Significance level (P-value) of a test for effect, i.e., H,;
(true) odds ratio = 1 against the alternative Ha: (true) odds ratio > 1 , is the area under the standard
normal curve to the right of the value of the test statistic, given by log(OR)/SE(log(RR)),
E-l
05/15/92
-------
DRAFT—NO NOT QUOTE OR CITE
If the (true) odds ratios are assumed to be equal in k studies, then a pooled estimate is
calculated from
log(OR(P)) =
where the summations are on i, from 1 to k; OR(P) is the pooled estimate; log(OR)i is the
logarithm of OR from the i* study; and W; * (Var(log(OR)j))-I is the weight of the ia study (Breslow
and Day, 1980).
E.3. COHORT STUDIES
The true (but unknown) relative risk is estimated by the observed relative risk (RR),
RR = (a/a+c)/(b/b+d).
A confidence interval on the (true) relative risk may be calculated from the normal approximation
to the distribution of log(RR), using the analogue of Woolfs method referred to above (Katz et
al., 1978). The variance of log(RR) is estimated by,
Var(log(RR)) = c/(a2 + ac) + d/(b2 + bd)
and the standard error by its square root,
SE(log(RR)) = (Var(log(RR)))w.
The remaining calculations follow the description for case-control studies in Section E.2 with
"odds ratio" and "OR" replaced by "relative risk" and "RR," respectively, the pooled estimate of
relative risk from both case-control and cohort studies is calculated by the same methodology for
pooling estimates from case-control studies or from cohort studies separately, i.e., the logarithm of
each individual estimate is weighted inversely proportional to its estimated variance (Kleinbaum
et al., 1982).
E-2
• U.S. GOVERNMENT PRINTING OFFICE: 1992— 6 5 g-
05/15/92
------- |