United States

Environmental Protection Agency

Office of Chemical Safety and
Pollution Prevention

Final Risk Evaluation for

Asbestos
Part 1: Chrysotile Asbestos

Summary of External Peer Review and Public Comments and

Disposition

Supplemental File for Epidemiologic Studies of Automotive

Mechanics

December 2020


-------
Table of Contents

Evaluation of Epidemiologic Studies of Mesothelioma Among Automotive Mechanics	3

Evaluation of Epidemiologic Studies of Lung Cancer Among Automotive Mechanics	11

Page 2 of 15


-------
Evaluation of Epidemiologic Studies of Mesothelioma Among Automotive
Mechanics

The EPA evaluated a paper presenting a review and meta-analysis of research on mesothelioma
risks among motor vehicle mechanics that was cited in public comments at the June, 2020 SACC
meeting (Garabrant et al.,,	965), the original research papers included in that meta-

analysis, and one other paper cited in the Committee comments (Van den Borre & Deboosere.
2015, 3088106). The EPA did not search for additional papers presenting data on cancer risks
among automotive mechanics or other workers performing maintenance and repair on
automotive brakes.

Summary and Conclusions

The EPA conducted a critical review of the meta-analysis by Garabrant et al. (	),

14 of the 16 original epidemiologic studies included in that analysis, and the more recent study
by van den Borre & Deboosere (2 >88106). Two other studies included in the Garabrant
analysis were not available for review. From our evaluation of this evidence, EPA have reached
the following conclusions regarding the three questions posed.

1.	Are the studies of motor vehicle mechanics sufficiently sensitive to detect a cancer effect of
commercial chrysotile asbestos?

The studies reviewed by the EPA have poor sensitivity to detect cancer hazards from commercial
chrysotile asbestos. The low sensitivity of these studies is a consequence of exposure assessment
methods that do not specifically identify asbestos exposure in individual workers, and of
potential bias from the use in most studies of referent groups that include other workers exposed
to asbestos.

2.	Can the studies of motor vehicle mechanics be used to derive a unit risk estimate as was done
in the draft Risk Evaluation, i.e., is there an exposure and response data sufficient to derive lung
cancer and/or mesothelioma risk?

The studies reviewed by the EPA concern only mesothelioma. None of them provide
quantitative exposure-response data that could be used to derive a unit risk for that disease.

3.	Describe the available data to compare fiber size distributions for motor vehicle mechanics
compared to other conditions of use.

The epidemiologic studies reviewed by the EPA do not provide data on associations of cancer
among motor vehicle mechanics with asbestos fiber concentrations, types, or size distributions.
Other studies may be available that provide data on fiber-size distributions without investigating
health effects.

EPA's review of the evidence and the conclusion's basis are detailed below.

Review and meta-analysis of mesothelioma among motor vehicle mechanics

Garabrant et al. (2015. 3019965) searched the Pub Med and Embase electronic databases for
epidemiologic studies of mesothelioma among motor vehicle mechanics published through May

Page 3 of 15


-------
2014; additional searches were conducted for unpublished studies. Eligible studies were scored
for quality using criteria adapted by the authors from standard instruments and were assigned to
3 groups based on their scores. Sixteen studies were ultimately included in the narrative review
and meta-analysis. Relative risks (RRs) for motor mechanics were not reported in four of the
included studies, so the authors estimated crude RRs from other data in the reports or obtained
from the authors. The meta-analysis was carried out with standard methods. The meta-RR for
all 16 studies was 0.80 (95% CI 0.61-1.05) with significant heterogeneity between studies.
Meta-RRs for the three quality groups were 0.76 (95% CI 0.46-1.25), 1.09 (95% C I 0.76-1.58)
and (95%) CI 0.73 0.49-1.08) by descending quality score. For 4 studies that reported data
specifically for workers performing brake work, the meta-RR was 0.64 (95% CI 0.38-1.09). The
authors concluded that their meta-analysis "provides evidence that motor vehicle mechanics,
including workers who were engaged in brake repair, are not at an increased risk of
mesothelioma."

Although the authors did not describe this review as such, the methods used are largely
consistent with standard systematic review methodology. Nevertheless, while the methods
appear to be generally adequate, there are concerns about the quality of the available studies and
the extent to which they are able to support quantitative analysis of the cancer hazard from
exposure to asbestos in automotive brake materials.

Included studies

Fourteen of the 16 studies reviewed by Garabrant et al. (2015, 3019965) and one additional study
mentioned in Committee comments from the July, 2020 SACC meeting were available for
review and are summarized below. Two other studies analyzed by Grabrant et al., including one
registry-based case-control study (Hansen et al. 2003. 6885998) and one registry-based study of
occupational mortality (HSE, 2013), were not available for this review.

The studies included by Garabrant et al. (	5) and available for this review included

10 studies analyzed as case-control studies of mesothelioma (McDonald & McDonal* i l,}N0 I '< '<

; Teta et al. 1983. 205; Hessel et al. 2004. 3530960; Rake et al. 201	!; Teschke et al.

1997. 3081326; Agudo et al. 2000. 733601; Aguilar-Madrid et al. .^ h« . _ I h , I volland et al.
.010 ;0 \'}72; Woitowitz & Rodelsperg^i 30' * , 1 \ ifs et al. _0l'< M I 36), one
occupational cohort study that provided data for multiple cancers (Merlo et al.. 2	),

one register-based descriptive study of mesothelioma (McElvenney et al., 2005, 3089646) and
two death-certificate-based studies of general occupational mortality (Washington DOH. 2011.
6_' • I l , -IQSH. lot l 984620).

None of the available studies reported quantitative exposure-response data for cancer risk in
relation to asbestos exposure, quantitative asbestos exposure levels, or assessments of fiber type
or size distribution. Likewise, no study provided even a qualitative assessment of individual
exposure to asbestos from work with motor vehicle brakes or other automotive maintenance and
repair tasks. The assessment of exposure to asbestos in the available studies depends, instead, on
a variety of indirect approaches. Studies are reviewed below in order of the specificity of the
information they provide about potential exposure to asbestos in automotive brake work.

Page 4 of 15


-------
The most specific information about the risks associated with exposure to asbestos from brake
materials is available from three case-control studies in which information about potential
sources of asbestos exposure was obtained by interview or questionnaire.

Hessel et al. (20	50) re-analyzed a previous case-control study of mesothelioma in three

area of the United States to focus on brake mechanics. Occupational and exposure histories were
obtained through interviews with next of kin in the original study. Analyses compared
individuals who had ever done brake work to those who had not. The OR adjusted for age and 8
other activities assumed to involve potential exposure to asbestos was 0.82 for any (95% CI
0.46-1.41) for any brake work and 0.82 (95% CI 0.36-1.80) for occupational brake work. After
deleting cases and controls with a history of any of the 8 activities involving potential exposure,
the OR for occupational brake work was 0.62 (95% CI 0.01-4.71).

A broader range of potential sources of asbestos exposure was investigated in two other case-
control studies. Teschke et al. (	31326) studied mesothelioma surveillance data to

identify occupational sources of exposure to asbestos in British Columbia, Canada. Information
about occupational history and work tasks was obtained by interview with subjects or next of
kin. In analyses including all cases and controls, the age-adjusted odds ratios (ORs) were 0.8
(95% CI 0.2-2.3) for ever having worked as a vehicle mechanic and 0.3 (95% CI 0.0-1.4) for
ever having done brake lining installation and repair. The OR for vehicle mechanics was
reduced to 0.4 after excluding individuals who had ever worked in occupations with known
exposure to asbestos, but a parallel analysis for brake installation and repair was not conducted.
The unexposed referent group for these ORs was not specifically stated, but calculations from
data in the paper suggest it was all other occupations or tasks.

Rake et al. (2009. 733522) studied the association of mesothelioma with lifetime occupational
and residential exposure to asbestos in Britain. Mesothelioma cases were identified through a
survey of health care providers and matched with controls sampled at random from primary care
rolls. Data on occupational and residential history and other sources of exposure to asbestos
were collected by postal questionnaire and telephone interview. Reported occupations and
activities were ranked by mesothelioma risk according to the results of a previous descriptive
study also included in the Garabrant et al. review (McElvennev et al.. 2005. 3089646). Motor
vehicle mechanics were included in the low-risk group, which constituted the unexposed referent
for most analyses, and no risk data were reported for this group. An OR of 0.4 (95% CI 0.1 -
1.7) for the task of "vehicle maintenance involving work with brakes or gaskets" was reported in
the text; the referent group for this OR was not stated.

A small study by Woitowitz and Rodelsperger (1994. 3081776) also reported on mesothelioma
among motor vehicle mechanics. Data for mechanics were apparently obtained from a larger
case-control study of mesothelioma in Germany, but neither the sources of cases and controls or
of exposure information is reported in this publication. Numbers of cases and hospital and
population controls who had worked as mechanics or done brake work are reported in a table,
but ORs are not provided. For this study, Garabrant et al. (	>) analyzed ORs for

motor vehicle mechanics "definitely engaged in brake service" of 0.89 (0.31-2.47) and 0.87
(95%) 0.43-1.70) compared to population and hospital controls, respectively. These ORs cannot
be reproduced from the data presented by Woitowitz and Rodelsperger (1994. 3081776) and

Page 5 of 15


-------
were apparently calculated by Garabrant et al. using data obtained directly from the original
authors. The unexposed referent for these ORs was not reported.

Two other case-control studies in Spain (Agudo et al. 2000. 733601) and Mexico (Aguilar-
Madrid et al. 2010. 2621487) investigated associations of mesothelioma with potential
occupational exposures to asbestos classified by expert review of occupational histories.
However, neither study reported a measure of association for motor vehicle mechanics. Agudo
et al. (2000. 733601) did report numbers of cases and controls for motor vehicle mechanics in a
footnote, and a crude OR of 0.62 for the risk of mesothelioma in motor vehicle mechanics
relative to subjects never employed in occupations with high risk of asbestos exposure can be
calculated using these data. Aguilar-Madrid et al. Q 621487) classified exposures in a
similar manner, but did not report ORs for specific occupations or industries. Crude ORs for
automobile mechanics can be calculated from data provided in a table, but do not match those
analyzed by Garabrant et al.

As part of a large descriptive study of mesothelioma in North America, McDonald & McDonald
(198( ) conducted a case-control analysis of the occupational exposures. Occupations
elicited in interviews with next of kin were ranked according to potential for exposure to
asbestos using a list obtained from other researchers. The occupation of "garage workers" was
assigned to a presumed low-exposure group for which the OR was not reported. However,
numbers of cases and controls among garage workers and other occupations in the same group
are reported in a table and can be used to compute crude ORs. Garabrant et al. analyzed an OR
of 0.92 based on a comparison of garage workers to other occupations in the same low-exposure
group. The OR for garage workers compared to the group of unexposed occupations would be
more appropriate and can also be calculated (OR 1.6).

In a study of mesothelioma cases from the Connecticut tumor registry matched with deceased
cases, Teta et al. (1983. 205) classified occupations obtained from death certificates and city
directories into classes of known, likely and possible asbestos exposure. The OR for ever
working in the "automobile repair and related service" industry category was 0.65 (95% CI 0.08-
5.63). The unexposed referent for the OR is not stated but appears to be never having worked in
any exposed industry.

In a retrospective cohort study of the impact of air pollution on municipal bus drivers and
maintenance workers, Merlo et al. (2X	) reported a SMR of 1.27 (95% CI 0.66-2.43)

for bus maintenance workers compared to the regional population. No information was reported
about asbestos or work with brakes.

All 5 of the remaining studies (McElvennev et al.. 2005. 3089646; Rolland et al. 2010. 3078972;
Milham et al. 201 I *>84619; NIov U I -i Roetofs et A _0l ' '.Ml 6) reported
ORs or proportionate mortality ratios (PMRs) by occupation or industry category without further
classification of asbestos exposure. Rolland et al. Q	972) coded occupational histories

obtained in interviews with mesothelioma cases and general population controls. ORs were
generated by occupation and industry with all other occupations or industries as the referent
group. The OR for working >6 months versus never as a motor vehicle mechanic was 1.50 (95%
CI 0.76-2.95). Roelofs et al. (	») coded the usual occupation and industry of

Page 6 of 15


-------
mesothelioma cases and controls with other cancer from the Massachusetts cancer registry.
Standardized morbidity odds ratios (SMORs) were generated for each occupation compared to
all others. The SMOR for automobile mechanics was 2.1 (95% CI 1.1-4.0). In a registry-based
descriptive study of mesothelioma in Britain, McElvenney et al. (2005. 3089646) reported PMRs
by occupation using the usual occupation reported on the death certificate. The PMR for "motor
mechanics" was 0.48 (95% CI 0.37-0.62). The references to Milham (Washington DOH, 2011,
6984619) and NIOSH (NIQSH. 2011. 6984620) are links to interactive web sites that allow users
to generate PMRs by cause of death and occupational category, presumably with all other
occupations as the referent. The PMRs from these sources analyzed by Garabrant et al. (2015.
3019965) could not be reproduced, possibly because of additions to the database in ensuing
years.

Additional study

Committee comments from the July, 2020 SACC meeting referenced a study in Belgium by van
den Borre & Deboosere (2015. 3088106). The study authors linked records of 2001 -2009
mesothelioma deaths in 2 Belgian regions with data from the 1991 census. Occupational history
information from both sources was categorized to 3 groups reflecting the potential for asbestos
exposure. SMRs were reported for broad groups and for selected industries. Data were not
reported for the automotive industry or for the subcategory of automotive maintenance & repair
subcategory. However, the authors stated that workers in the automotive industry, among others,
"do not seem to experience significant excess in mortality due to asbestos-related diseases."

Comments

The major limitation of the studies reviewed here is the lack of specific data on asbestos
exposure. None of the 15 available studies provides information on fiber concentrations,
cumulative fiber exposures, fiber type, or fiber-size distributions. Neither do any of the studies
provide individual-level qualitative assessment of work with asbestos brake parts or other
asbestos-containing automotive components.

In the absence of direct assessment of asbestos exposure, potential exposure was assessed in
several studies by inference from self- or proxy-reported tasks (3 studies), by expert review of
occupational histories (2 studies) or by applying matrices or lists of asbestos-exposed jobs to
occupational histories (3 studies). The source of task information in one other study was not
reported. All other studies simply reported associations of mesothelioma morbidity or mortality
with occupation or industry titles.

Subject and proxy reports can identify specific tasks and exposures to the extent that the
respondents had knowledge of them and are able to recall them. However, the four studies that
assessed work with automotive brakes did not report whether workers used asbestos brake
components. Because not all brake pads and shoes contain asbestos, workers may not be able to
report whether they worked with such components. Studies based solely on broad occupational
groups have more limited sensitivity to detect risks arising from work with brake parts or other
components containing asbestos, as many workers will may not have done such work.

A second important limitation is that none of the available studies provides exposure-response
data. In most studies, associations are reported for ever having engaged in a particular occupation

Page 7 of 15


-------
or task or for having done so "usually" or for some minimum amount of time. Five other studies
did not report associations for pertinent tasks or occupations, although some measures of
association can be computed from data provided in these papers.

A third limitation is related to the choice of the unexposed referent group to which automobile
mechanics are compared. Ideally, the referent group should consist of workers with no exposure
to asbestos. However, among the 10 studies in which the authors reported risk data for auto
mechanics or work with brakes, the referent group in 7 studies was reported to be or could be
inferred to be all other occupations or industries. In one study (Teta et al. 1983. 205). the
referent group was explicitly stated to be workers not employed in jobs with known, likely or
possible exposure, and in two studies (Merlo ct al., . ' i '	i; Van den Borre & Deboosere.

2015. 3088106) the general population served as the referent. The selection of all other
occupations as the referent is cause for concern because this would include workers in
occupations known to have high levels of exposure to asbestos and high risks of mesothelioma in
the referent category. This would exert a downward bias on the relative risk downward for
occupations with lower exposure.

Conclusions

The conclusions of this review of the 15 original studies summarized above follow with respect
to the questions posed in EPA's technical directive.

1.	Are the studies of motor vehicle mechanics sufficiently sensitive to detect a cancer effect of
commercial chrysotile asbestos?

Overall, the studies reviewed by the EPA have poor sensitivity to detect a cancer hazards from
commercial chrysotile asbestos. None of the studies directly assessed exposure to asbestos.
Information on fiber type is therefore lacking. The most sensitive studies are four case-control
studies that reported risks associated with a history of working with automotive brakes, which
may contain asbestos. However, none of these studies evaluated whether asbestos-containing
brake components were actually used. The remaining 11 studies report risks based on broader
occupational categories, such as automobile mechanics. These studies have low sensitivity, as
the numbers of workers actually exposed to asbestos was not ascertained. The sensitivity of the
majority of the available studies may also have been reduced by the selection of referent groups
that include other workers exposed to asbestos, which can exert a downward bias on measures of
association.

2.	Can the studies of motor vehicle mechanics be used to derive a unit risk estimate as was done
in the draft Risk Evaluation, i.e., is there an exposure and response data sufficient to derive lung
cancer and/or mesothelioma risk?

The studies reviewed by the EPA concern only mesothelioma. None of them provide exposure-
response data that could be used to derive a unit risk.

3.	Describe the available data to compare fiber size distributions for motor vehicle mechanics
compared to other conditions of use.

Page 8 of 15


-------
The epidemiologic studies reviewed by the EPA do not provide data on associations of cancer
with asbestos fiber concentrations, types, or size distributions. Other studies may be available
that provide data on fiber-size distributions without investigating health effects.

Page 9 of 15


-------
References

Ague onzalez CA. Bleda MI et al. (2000) Occupation and risk of malignant pleural
mesothelioma: a case-control study in Spain. Am J Indust Med; 37: 159-68.

Aguilar-Mato > » ^ bles-Perez E et al. (2010) Case-control study of pleural mesothelioma in
workers with social security in Mexico. Am J Ind Med; 53: 241-51.

Garabrat» U \ H zander DP. Miller PE. Fryzel (-'offetta P. Teta MI. et al. (2015).

Mesothelioma among Motor Vehicle Mechanics: An Updated Review and Meta-analysis.
Ann Occup Hyg. 60(l):8-28.

Hansen J. Meersohn A. (2003) Report on cancer and occupation (1970-97). Kobenhavn,
Penmark: Institut for Epidemiologisk Kraeftforskning. pp. 1-114.

Hessi	i MI. Goodman M et al. (2004) Mesothelioma among brake mechanics: an

expanded analysis of a case- control study. Risk Anal; 24: 547-52.

H ;alth and Safety Executive). (2013) Mesothelioma occupation statistics: male and female
deaths aged 16-74 in Great Britain 2002-2010. London: Health and Safety Executive.

McDonal* i \< * McDoml*! k _ (1980) Malignant mesothelioma in North America. Cancer; 46:

1650-6.

McElvenny DM. Parnt	J et al. (2005) Mesothelioma mortality in Great Britain

from 1968 to 2001. Occup Med; 55: 87.

Merlo DF. Staei E. Fontana V et al. (2010) A historical mortality study among bus drivers and
bus maintenance workers exposed to urban air pollutants in the city of Genoa, Italy. Occup
Environ Med; 67: 611-9.

Washington DOH. (2011) Washington State occupational mortality database, 1950-2010.
https://fortress.wa.gov/doh/occmort/OMQuery.aspx. Accessed 28 August 2020.

NIOSH. (2011) National occupational mortality surveillance (NOMS).

https://www.cdc.gov/niosh/topics/noms/query.html. Accessed 28 August 2020.

Rake C. Gilham C. Hatch J et al. (2009) Occupational, domestic and environmental

mesothelioma risks in the British population: a case-control study. Br J Cancer; 110: 1175—
83.

Roelofs CR. Kern an GJ. Davis LK et al. (2013) Mesothelioma and employment in

Massachusetts: analysis of cancer registry data 1988-2003. Am J Ind Med; 56: 985-92.

Rollaini ^MinoniU < acoun \ i aL (2010) Occupations and industries in France at high risk
for pleural mesothelioma: a population-based case-control study (1998-2002). Am J Ind
Med; 53: 1207-19.

Teschke K, Morgan MS, Checkoway H. et al. (1997) Mesothelioma surveillance to locate sources
of exposure to asbestos. Can J Public Health; 88: 163-8.

Teta MI. Lewinsohn HC. Meigs JW et al. (1983) Mesothelioma in Connecticut, 1955-1977.
Occupational and geographic associations. J Occup Med; 25: 749-56.

van den Borre L. Deboosere P. (2015) Enduring health effects of asbestos use in Belgian
industries: a record-linked cohort study of cause-specific mortality (2001-2009). BMJ
Open 2015; 5: e0073 84. doi: 10.113 6/bmj open-2014-007384

Woitowitz HI. Rodelsperger K. (1994) Mesothelioma among car mechanics? Ann Occup Hyg;
38: 635-8.

Page 10 of 15


-------
Evaluation of Epidemiologic Studies of Lung Cancer Among Automotive
Mechanics

The EPA reviewed epidemiologic studies reporting risk data for lung cancer among motor
vehicle mechanics that were listed in the June, 2020 SACC report prior to Recommendations 48
and 49. This review is based on the studies available to the EPA; although other studies may
present data on asbestos exposures or lung cancer risks among automotive mechanics or other
workers performing maintenance and repair on automotive brakes, the EPA was not aware of
such studies and did not search for them.

Summary and Conclusions

The EPA critically reviewed 11 case-control studies and 1 prospective cohort study. All of the
available studies were broadly focused investigations of associations of lung cancer with
occupation or industry categories based on information obtained by interview or questionnaire.
None of the studies was focused on automotive mechanics, and none included quantitative
exposure-response analyses of lung cancer risk in relation to asbestos fiber exposure, quantitative
measures of asbestos fiber concentrations or size-distributions, or qualitative assessments of
individual asbestos exposure. Two other studies (Williams etai, 1977, 4911; limbec et at.,
1992. 6875651) were not available for review. While the EPA was not able to review these
studies, it is evident from the abstracts and EPA's knowledge of the field that these studies
employed similarly broad approaches and would not provide quantitative exposure data.

The most informative studies were large, population-based case-control studies that adjusted for
smoking and other lung cancer risk factors and classified occupation and industry titles
according to a matrix of potential exposure to occupational carcinogens. Most other case-control
studies and the prospective cohort study contrasted risks among individuals who ever or usually
held various occupations, including automobile mechanics, to those who never did so. The
studies judged to be most informative reported a modest, non-significant increase (Consonni et
al _« h« S0084) or no increase (Richiardi et at.. 2004; 699205. Guida. 2< M I _ * I ) in lung
cancer risk for automobile mechanics compared to occupations with no known or suspected
exposure to lung carcinogens. One them reported a non-significant relative risk of 1.2 (95%
confidence interval (CI) specifically for automobile brake workers (Vineis et at., 1988, 79604).
While these studies are adequate in overall quality, it is likely that they are affected by non-
differential misclassification of asbestos exposure, because not all automobile mechanics work
on brakes and not all brakes contain asbestos. This form of measurement error tends to bias
relative risks toward the null value of 1.0. The less informative studies are likely to suffer from
additional biases related to inadequate control for confounding by tobacco smoking, the use of
controls with other diseases, and comparing auto mechanics to other occupations that could
involve exposure to asbestos.

Based on this review, the EPA conclude that the studies of lung cancer in automotive mechanics
cited in the SACC report do not provide a basis for deriving an IUR for lung cancer and asbestos
exposure.

Page 11 of 15


-------
Similarly, none of the studies provide data on asbestos-fiber size distributions in automotive
repair environments. Such information may be available from exposure studies that do not
assess associations with cancer, but EPA was not aware of such studies and did not search for
them.

Case-Control Studies

Lerchen et al. (198'	) obtained lifetime occupational histories in interviews with lung

cancer cases identified from the New Mexico tumor registry and controls from the general
population. Odds ratios (ORs) were reported for ever versus never having worked in selected
industries and occupations the authors judged to involve increased risk of lung cancer and for
exposure to diesel exhaust fumes, asbestos and wood dust. Smoking-adjusted ORs were 0.9
(95% CI 0.5-1.9) for auto mechanics and 1.1 (95% CI 0.7-1.7) asbestos exposure. An OR for
asbestos exposure among auto mechanics was not reported.

Benhamou et al. (1988. 4868) reported on associations of lung cancer and occupation in French
men who participated in a previous international case-control study of lung cancer and smoking.
Cases and controls were apparently identified from hospital records and lifetime occupational
histories were obtained by questionnaire. ORs were reported for having worked versus never
worked in selected occupations. The smoking-adjusted OR for automobile mechanics was 1.06
(95% CI 0.73-1.54).

In a study aimed at estimating the contribution of occupational exposure to the burden of lung
cancer, Vineis et al. (1988. 79604) combined data from case-control studies in 5 areas of the
United States. Cases and controls were identified from hospitals or death certificates, or from a
cancer registry in one study. Information on jobs held for at least 6 months was collected in
interviews with subjects or next of kin and grouped according to published matrix of occupations
and industries with known or suspected exposure to occupational carcinogens (Simonato &
Saracci, 1983, 6984621). A smoking-adjusted OR of 1.2 (95% CI 0.9-1.7) was reported for
individuals who had ever been automobile brake workers compared to those who never worked
in any exposed occupation.

Morabia et al. (1992, 79634) studied the role of occupational exposure in lung among
participants in an established multicenter hospital-based case-control study in the United States.
Information on usual occupation and exposure to 44 specific agents was collected by interview.
An OR of 0.7 (no CI or p-value reported) for automobile mechanics and repairmen was reported.

Swan son et al. (1993. 6875603) reported on the association of lung cancer risk and duration of
employment in an established case-control study of lung cancer among Black and white men in
Detroit, Michigan. Lung cancer cases and controls with colorectal cancer were identified from
the Detroit cancer registry and lifetime occupational histories were collected by telephone
interview with subjects or proxies. ORs were stratified by race and were reported only for
occupation and industry titles with significantly increased ORs or significant trends with duration
of employment. Control for covariates was not described, so the reported ORs may not be
adjusted for important risk factors such as age or smoking. For automobile mechanics, ORs
were greater than 1.0 for all employment durations and both races, with a significant trend only

Page 12 of 15


-------
for Black men. No significant trend was reported for employment in the automotive services
industry.

Results of a hospital-based case-control study of lung cancer and occupation in Buenos Aires,
Argentina were reported by Matos et al. (	2). Lung cancer cases and controls with

various other diseases unrelated to tobacco smoking were recruited from hospitals and
interviewed to collect information on occupational history, including tasks performed in jobs
held for more than one year. Analyses compared participants never versus never employed in a
given occupation or industry and results were reported for selected "high-risk" jobs. Smoking
adjusted ORs of 2.1 (95% CI 1.0-4.5) for motor vehicle mechanics and 1.2 (95% CI 0.5-2.7) for
the motor vehicle repair industry were reported.

Occupational risk factors for lung cancer were investigated in a population-based case-control
study of lung cancer in Northern Italy (Richiardi et al.. 2004. 699205). Lifetime occupational
histories were collected through interviews with subjects and were grouped using an exposure
matrix similar to that used by Vineis et al. (1988. 79604). Workers ever employed as a motor
vehicle mechanic were classified to the group of occupations and industries with suspected
exposure to carcinogens and were compared to workers with no known or suspected exposure in
the analysis. The adjusted OR for motor vehicle mechanics was 1.1 (95% CI 0.6-2.0) adjusted
for smoking and other risk factors.

MacArthur et al. (2009. 699235) assessed associations of occupation and industry titles with all
lung cancer and lung cancer histologic subtypes in a large case-control study based on the British
Columbia Cancer Registry. All other registered cases with non-lung cancer served as controls.
Occupational histories were collected by self-administered questionnaire, and analyses compared
usual occupation and any occupation ever held to all other occupations. Adjusted ORs for motor
vehicle mechanics and repairers were 0.90 (95% CI 0.68 -1.19) for usual occupation and 1.06
(95%) CI 0.89 -1.26) for ever having held that occupation.

Consonni et al. (2010, 380084) investigated associations of lung cancer and occupation in a large
population-based case-control study in Italy that was originated to explore environmental and
genetic factors in lung cancer. Lifetime occupational histories were collected in personal
interviews and grouped using an approach similar to that of Vineis et al. (1988. 79604) according
to a matrix of jobs with known or suspected exposure to occupational carcinogens (Ahrens &
Merletti. 1998. 737541). The OR for motor vehicle mechanics was 1.29 (95% CI 0.72, 2.33)
compared to occupations with no known or suspected exposure and adjusted for smoking and
other risk factors.

A similar approach was taken in a study of environmental causes of lung cancer in France
(Guida. ,	). Lifetime occupational histories obtained in interviews were grouped

according to a matrix of jobs with known or suspected occupational exposure to carcinogens
( 3ns & Merletti. 1998. 737541; Consonni et al.. 2010. 380084). An OR of 1.06 (95% CI
0.73-1.54) adjusted for smoking and other risk factors was reported for motor vehicle
mechanics.

Page 13 of 15


-------
C orb in et al. (2C 2244) examined associations of occupation and lung cancer in an existing
population-based case-control study in New Zealand. Incident lung cancer cases were identified
from a national cancer registry and controls were recruited from electoral rolls. Occupational
histories were obtained by telephone or personal interview and grouped according to the risk of
lung cancer expected from previous studies. ORs were estimated for ever versus never having
been employed in a given occupation, adjusted for smoking and other risk factors. The
occupation of "motor mechanic" was classified as a priori high-risk: the OR was 1.22 (95% CI
0.59-2.51) and increased to 1.32 (95% CI 0.73-2.39) after semi-Bayes adjustment for multiple
comparisons.

Prospective Cohort Study

Veglia et al. (2007. 3079516) examined associations of occupation and lung cancer in a large,
European prospective cohort study originally designed to investigate associations of diet and
cancer. Information about employment in 52 jobs previously linked to elevated risk was obtained
by questionnaire. Exposure to several agents of particular interest, including asbestos, was
assigned to job titles by expert consensus. Among men and women with any history of exposure
to asbestos the adjusted rate ratio (RR) for incident lung cancer was 1.53 (95% CI 1.2-1.9). In
analyses of selected occupational categories, the RR for "car repair workers" was 1.27 (95% CI
0.8-2.0) adjusted for smoking and other risk factors.

Page 14 of 15


-------
References

Ahrens W. Merloni l: A standard tool for the analysis of occupational lung cancer in
epidemiologic studies. Int J Occup Environ Health. 1998;4(4):236-240.

Benhamou S, Benhamou E, Flam ant R (1988). Occupational risk factors of lung cancer in a
French case-control study. Br J Ind Med 45:231-233.

Consonn : Matteis S. Lubin JH. Wacholder S. Tucker M. Pesatori AC. Caporaso NE.
Bertaz	di MT. (2010). Lung cancer and occupation in a population-based case-

control study. Am J Epidemiol 171(3):323—333.

Corbin M. McLean D. Mannetie A. Dry son E. Walls C. McKenzie F. Maule M. Cheng S.

Cunningham C, Kromhout H, Blair A. Pearce N. (201 1). Lung cancer and occupation: A
New Zealand cancer registry-based case-control study. Am J Ind Med 54:89-101.

Guidn I Pnpadopoulos A. MenvifU^' MatratM l'o\ quo J Coin'-'c. i \t O C.chmaus A.
Carton M. Paget-Bailly S. Radoi L. Tarnaud C. Bara S. Tretau-' 1 ^ o ]> Quicker 1.
(2011). Risk of lung cancer and occupational history: Results of a French population-based
case-control study, the ICARE Study. J Occup Environ Med 53(9): 1068-1077.

limbec Z, Blair AE, Rogot E. Vaught J. (1992). Mortality risks by occupation among US
veterans of known smoking status: 1954-1980. NIH Publication No. 92-3407. US
Department of Health and Human Services (DHHS), Public Health Service, National
Institutes of Health (NIH).

Lerchen ML. Wiggins CL. Samet JM. (1987). Lung cancer and occupation in New Mexico. J
Natl Cancer Inst 79(4):639-645.

MacArth	tng R, Band PR. (2009). Identification of occupational cancer risk in

British Columbia: A population-based case-control study of 2,998 lung cancers by
histopathological subtype. Am J Ind Med 52:221-232.

Matos EL. Vilensky M. Mirabelli D. Boffetta P. (2000). Occupational exposures and lung cancer
in Buenos Aires, Argentina. J Occup Environ Med 42(6):653-659.

Morabia A. Markowitz S. Garibaldi K. Wynder EL. (1992). Lung cancer and occupation: Results
of a multicentre case-control study. Br J Ind Med 49:721-727.

Richiardi L. BoffatU " * imonatoL. Forastiei-' I ambon " I ones C. Gaborieau V. Merleni I
(2004). Occupational risk factors for lung cancer in men and women: A population-based
case-control study in Italy. Cancer Causes Control 15:285-294.

Simonato L. Saracci R. Cancer, occupational. In: International Labor Organization (ILO),

Encyclopedia of Occupational Safety and Health. Geneva, Switzerland: ILO; 1983:369-
375.

Swan son GM. Lin C-S. Burns PB. (1993). Diversity in the association between occupation and
lung cancer among black and white men. Cancer Epidemiol Biomarkers Prev 2:313-320.

Veglia F. Vineit ^ ervad K. Boeing H. Bergmann MM. Trichopoulou \ I'i ichopoulos D.
Gonzalez CA. Marinez C. Dorronsoro M. Barrica "irera L. Quiros JR. Day NE.
Saracci R. Riboli E. (2007). Occupational exposures, environmental tobacco smoke, and
lung cancer. Epidemiol 18(6):769-775.

Vineis P. Thomas T. Hayes R.B. Blot WJ. Mason TJ. Pickle LW. Correa P. Fontham ETH.

Schoenberg J. (1988). Proportion of lung cancers in males, due to occupation, in different
areas of the USA. Int J Cancer 42:851-856.

Williams RR. Stegens ' Idsmith JR. (1977). Associations of cancer site and type with
occupation and industry from the Third National Cancer Survey Interview. J Natl Cancer
Inst 59(4): 1147-1185.

Page 15 of 15


-------