Vlaanderen, Jelle PhD; Taeger, Dirk PhD; Wellman, Jürgen PhD; Keil, Ulrich MD, PhD; Schüz, Joachim PhD; Straif, Kurt MD, PhD
For more than 30 years, there has been evidence that employment in the rubber-manufacturing industry is associated with an increased risk of cancer. During the 1982 International Agency for Research on Cancer (IARC) Monograph evaluation of the rubber industry, the working group concluded that there was sufficient evidence for a causal association between exposures in the industry and urinary bladder cancer and leukemia and sufficient evidence for the excess occurrence of lung and stomach cancers in rubber workers.1 During an update of this evaluation in 2009, it was concluded that for workers in the rubber-manufacturing industry there is now “sufficient evidence” for leukemia and lymphoma, and cancers of the lung, urinary bladder, and stomach, and “limited evidence” for cancers of the prostate, larynx, and esophagus.2
Because of concerns of possible excess risks of laryngeal and other cancers among blue-collar workers in the German rubber industry, a large epidemiological cohort was formed in 1990. The cohort included workers from five tire or general rubber goods producing plants that employed more than 1000 production workers each. Initial publications on this cohort had access to mortality follow-up of individuals in the cohort until December 31, 1991.3,4 In 1996, Weiland et al4 reported on the cancer mortality of male workers in the cohort. Standardized mortality ratios (SMRs) were significantly elevated for lung and pleural cancer.4 Furthermore, SMRs were nonsignificantly elevated for leukemia and for cancers of the pharynx, esophagus, stomach, rectum, larynx, prostate, and bladder.4 The SMR for cancer of the liver or gallbladder was significantly decreased. In 1999, Mundt et al3 reported on the cancer mortality of female workers in this cohort. Although limited by low statistical power because of smaller numbers of workers and therefore cancer cases, nonsignificant excess risks were reported for cancers of the stomach and lung.
More detailed publications on this cohort focused on analysis by workplace5–7 and on specific exposures (nitrosamines, carbon black, and asbestos).8–11
Here, we report cause-specific SMR estimates for the German rubber industry cohort after 9-year additional follow-up. In addition, we present results stratified by work area, year of hire, and duration of employment for the cohort with follow-up until December 31, 2000.
Study Population and Follow-Up
Enumeration and mortality follow-up of the cohort has been described before.3,4 In brief, the cohort of German men comprised blue-collar workers from five rubber plants in Western Germany who were actively employed or retired on January 1, 1981, and had been employed for at least 1 year. Follow-up started on January 1, 1981, and was stopped at the age of 85 years, death, loss to follow-up, or end of follow-up.4 The cohort of German women was restricted to those who were actively employed for at least 1 year on January 1, 1981.3 Retired women were not included because of methodological difficulties with identifying the complete retired workforce.3 The cohort was set up using files from health insurance companies and personnel files of the participating rubber plants.
Vital Status and Assessment of Cause of Death
Former publications on this cohort were based on mortality follow-up from January 1, 1981, until December 31, 1991.3–11 For this publication, mortality follow-up was extended by 9 years to December 31, 2000. For all cohort members alive or with unknown vital status at the end of the first follow-up, a vital status follow-up was initiated using local population registries and starting with the registry of the last known place of residence. For deceased cohort members, a copy of the death certificate was requested from the local community health department. A professional nosologist coded the underlying cause of death according to the ninth revision of the International Classification of Diseases12 to ensure comparability of the coding of causes of death among the deceased cohort members with the coding of the reference population.
Mortality experience of the male and female cohorts was compared with the corresponding German male and female population of Western Germany. Expected mortality was estimated using cause, calendar year (1981 to 2000), sex, and 5-year age group (15 to 85) specific rates from the German Federal Statistical Office. Person-years at risk were calculated for each cohort member from the entry point of the study (January 1, 1981) until the end of follow-up. End of follow-up was defined as December 31, 2000, date of death, or the last date a cohort member was known to be alive, whichever occurred first. Stratified analyses were conducted for “duration of employment” (1 to 9 years, 10 to 19 years, 20 to 29 years, 30 to 39 years, ≥40 years), “year of hire” (<1950, 1950 to 1959, 1960 to 1969, ≥1970), and “work area” (preparation of materials, production of technical rubber goods, production of tires, storage and dispatch, maintenance, and other work area). Workers were assigned to the work area where they spent the majority of their career. “Duration of employment” is a time-dependent variable, and study members may contribute person-years to more than one category. Standardized mortality ratios and 95% confidence intervals (CIs) were calculated in SAS v 9.2 (SAS Institute Inc, Cary, NC) using the stratify macro.13 Heterogeneity across work areas was assessed with the test suggested by Breslow and Day.14
Main cohort characteristics have been described before.3,4 The cohort consists of 11,632 men and 1863 women actively employed in 1981 or retired from (men only) the German rubber industry. Few individuals were excluded from the study: for 47 men and 2 women, we observed inconsistencies between the date of hire and termination point of the study; 4 men and 16 women left the cohort before the start date of follow-up (January 1, 1981). Between 1951 and 1964, 50% of the workers were hired. Median year of hire was 1958. Median working duration of cohort members was 28.6 years for men and 25.2 years for women. With follow-up until December 2000, men accrued a total of 175,458 person-years and women accrued 34,017 person-years. The median age at the end of follow-up was 70 years for men and 69 years for women. Table 1 presents an overview of the assessment of vital status and causes of death. Vital status was unknown for 0.7% of men and 1% of women. Cause of death was unknown for 7% of deceased men and for 10% of deceased women.
In Table 2 (men) and Table 3 (women), we present the observed and expected number of cases and estimated SMR with 95% CI for 46 outcomes (including 12 noncancer outcomes). Among men we observed significantly elevated SMRs for cancer of the lung (1.23, 95% CI = 1.12 to 1.35; 420 cases) and pleura (2.57, 95% CI = 1.59 to 3.93; 21 cases). Other noteworthy (not significantly) elevated SMRs were observed for cancers of the stomach (1.13, 95% CI = 0.93 to 1.35; 113 cases), larynx (1.31, 95% CI = 0.82 to 1.99; 22 cases), bladder (1.22, 95% CI = 0.92 to 1.57; 58 cases), and leukemia (1.18, 95% CI = 0.85 to 1.60; 41 cases).
Among women we observed significantly elevated SMRs for cancer of the lung (1.81, 95% CI = 1.09 to 2.82; 19 cases) and oral cavity (5.75, 95% CI = 1.19 to 16.82; 3 cases). Other noteworthy (not significantly) elevated SMRs were observed for cancers of the pharynx (3.73, 95% CI = 0.45 to 13.46; 2 cases), bladder (2.42, 95% CI = 0.66 to 6.20; 4 cases), and leukemia (2.11, 95% CI = 0.85 to 4.34; 7 cases).
Standardized mortality ratios for noncancer deaths were (significantly) elevated for hypertensive disease among both men (1.88, 95% CI = 1.50 to 2.34; 82 cases) and women (1.89, 95% CI = 0.86 to 3.59; 9 cases), and for chronic liver disease (1.74, 95% CI = 0.98 to 2.88) among women.
To address heterogeneity in the German rubber industry in working conditions across work areas and over time, we conducted analyses stratified by work area and year of hire. We combined the male and female cohorts for these analyses, as stratified analyses in the female cohort were largely uninformative. In Table 4, we report results for a group of cancers of a priori interest (cancers for which IARC concluded in 2009 that sufficient or limited evidence exists for an association with working in the rubber-manufacturing industry.2). Standardized mortality ratios for these cancers were primarily elevated for work areas “preparation of materials” (stomach, lung, leukemia, and larynx), “production of technical rubber goods” (lung, bladder, and leukemia), and “storage and dispatch” (lung, lymphatic system, leukemia, and esophagus). Significant heterogeneity across work areas was observed for the lung (P < 0.0001) and leukemia (P = 0.0145), but not for the other cancers. We report results of stratification by work area in Table S1 (Supplemental Digital Content 1, http://links.lww.com/JOM/A121).
No clear patterns were observed with year of hire among the cancers of a priori interest. Nevertheless, elevated SMRs were observed in several strata. Among workers hired before 1950, significantly elevated SMRs were observed for cancer of the lung and esophagus. Among workers hired in 1950 to 1959, the SMR for leukemia was slightly elevated. Among workers hired in 1960 to 1969, the SMR for cancer of the larynx was elevated. Finally, among workers hired in 1970 to 1981, SMRs for cancers of the stomach, lung, and larynx were elevated. We report results of stratification by year of hire for the other outcomes in Table S2 (Supplemental Digital Content 2, http://links.lww.com/JOM/A122).
In Table 5, we report results for the group of cancers of a priori interest stratified by duration of employment. We did not observe clear trends with the duration of employment for the selected cancers. We report results of stratification by duration of employment for the other outcomes in Table S3 (Supplemental Digital Content 3, http://links.lww.com/JOM/A123).
Our analyses on the basis of 9 years of added follow-up support previously observed elevated cancer risks in a German rubber industry cohort. The added follow-up generally resulted in higher precision of risk estimates because of larger numbers of cancer deaths. The outcome of this study is largely in line with the conclusions that were drawn during the 2009 IARC monograph evaluation with regard to the cancer sites that were evaluated as having “sufficient” (leukemia, lymphoma, lung, urinary bladder, and stomach) or “limited” (prostate, larynx, and esophagus) evidence for an association with working in the rubber-manufacturing industry.2 Nevertheless, we did not observe an elevated risk for prostate cancer and only for cancer of the lung a significant excess cancer risk was observed for the full cohort.
Several aspects should be taken into account with the interpretation of the findings from the updated cohort. Because the cohort is 9 years older, 1852 more deaths in men and 167 in women have occurred and consequently there is an increase in the precision to assess excess risks. Furthermore, for many cancer types, it takes years from the critical exposure to the development of a clinically manifested tumor (latency time). The added 9 years of follow-up therefore allow for a better assessment of excess risks for cancer types with long latency times. Nevertheless, it has also been shown that the effect of exposure on the risk of cancer can diminish with time since exposure. This is particularly known for leukemia mortality with regard to occupational exposures to benzene or ionizing radiation,15,16 but may also apply–-with a somewhat longer time frame–-to solid cancers. For example, Hornung et al17 reported that in a study of uranium miners the relative risk of lung cancer for a given exposure was reduced by 50%, 15 years after cessation of exposure, compared with current miners and those within 5 years of the last exposure.
We observed significantly elevated SMRs for cancer of the lung for workers employed in “preparation of materials” and in “production of technical rubber goods” and for cancer of the stomach for workers in “preparation of materials.” Similar findings were reported for an earlier update of this cohort by Straif et al6 who conducted an analysis on male workers using an internal reference group. Within work area “preparation of materials” tasks typically involve weighing and mixing, giving rise to substantial quantities of airborne dust2,6 and milling and calendering where workers are primarily exposed to rubber fumes.2,6 Within “production of technical rubber goods” tasks involve component assembly and building, curing, and vulcanizing,2,6 and workers are therefore exposed to rubber fumes and potentially considerable exposure to nitrosamines (as a result of the vulcanization process).2,6 Although exposure to rubber dust and fumes has traditionally been reported to be different for workers in “production of technical rubber goods” compared with workers in “production of tires,” more recent studies evaluating exposure measurements have not observed such a difference.18–20 In our analysis, SMRs for workers in “production of technical rubber goods” generally compare well to SMRs for workers in “production of tires” (Table 4).
Among the group of cancers of a priori interest, we did not observe clear declining or increasing trends in SMRs with duration of employment. Nevertheless, because of differences in types and levels of exposures between job titles and downward trends in exposure levels over time in the rubber industry, duration of employment is a less than ideal proxy for the actual total exposure that individuals experienced in the German rubber industry. Furthermore, the definition of the cohort of men as active and retired workers may have resulted in healthy worker survivor effect, which may be even more prominent among workers with longer employment. Yet, SMRs observed for “all cause” and “all cancer” mortality suggest a relatively minor impact of a healthy worker effect. Also, excluding the first 9 years of employment from the person-year calculation resulted in minor changes in SMRs in the main analysis (<5% for cancers of the lung and stomach).
Although the rubber industry made considerable efforts to reduce exposure levels and remove confirmed carcinogens from the production process,18,21,22 we did not observe clear downward trends in SMRs with more recent years of hire among the group of cancers of a priori interest. In fact, for several cancers of a priori interest, the highest risks were observed in the most recent hiring period, 1970 to 1981. Nevertheless, major improvements of occupational hygiene may have been introduced only after 1980 and anticipated trends of downward cancer risks would need to be explored in more recently hired workers.
Our current analyses are based on comparison with an external reference population. Differences in socioeconomic status and smoking behavior or other lifestyle factors between the rubber workers in our cohort and the general population might have introduced bias. Nevertheless, a previous analysis of the men in our cohort using an internal reference population resulted in similar results compared with the results reported here.6 Furthermore, in case of strong confounding by smoking, we would expect to see considerable excess risks for the broad spectrum of smoking-related outcomes. This was not the case. An exception was the increased risk for hypertension in both men and women, which could possibly be also related to obesity. Furthermore, cancers of the oral cavity and pharynx are strongly associated with tobacco smoking and alcohol drinking, which might have confounded our finding of elevated SMRs for these cancers among women.
Although modest in terms of person-years of follow-up and numbers of deaths, this study includes one of the largest cohorts of female rubber workers in the world.3 It was therefore of interest to conduct a separate analysis on women. Many of the findings for the female rubber workers are corroborated by findings for the male rubber workers in our analysis or by other studies. For example, our finding of elevated SMRs for cancers of the esophagus, stomach, lung, and bladder, and leukemia among women parallels the IARC evaluation of 2009, which was primarily based on male cohorts.2 Nevertheless, our findings of elevated SMRs for cancer of the oral cavity and pharynx among women are not corroborated by the current findings for the male cohort. Standardized mortality ratios for these cancers were elevated for men in the previous mortality update of the cohort4 and particularly in association with exposure to nitrosamines.9
Further analysis of the cancer risk of workers in the rubber industry is warranted and should focus on differences between work areas and associated exposures within the rubber industry. We are currently pooling our cohort with other European rubber cohorts to address this issue in more detail and will use the EXASRUB database to harmonize exposure assessment for the pooled cohorts.23
The authors wish to acknowledge the important contributions of the late Professor Stephan Weiland who was one of the initiators of this study. Part of the work reported in this article was undertaken during the tenure of a Postdoctoral Fellowship from the IARC, partially supported by the European Commission FP7 Marie Curie Actions–People–Cofunding of regional, national, and international programmes (COFUND).
1. IARC. The rubber industry. IARC Monogr Eval Carcinog Risks Hum. 1982;28:1–486.
2. IARC. Occupational exposures in the rubber-manufacturing industry. IARC Monogr Eval Carcinog Risks Hum. 2012;100F.
3. Mundt KA, Weiland SK, Bucher AM, et al. An occupational cohort mortality study of women in the German rubber industry: 1976 to 1991. J Occup Environ Med. 1999;41:807–812.
4. Weiland SK, Mundt KA, Keil U, et al. Cancer mortality among workers in the German rubber industry: 1981–91. Occup Environ Med. 1996;53:289–298.
5. Straif K, Weiland SK, Werner B, Chambless L, Mundt KA, Keil U. Workplace risk factors for cancer in the German rubber industry: Part 2. Mortality from non-respiratory cancers. Occup Environ Med. 1998;55:325–332.
6. Straif K, Chambless L, Weiland SK, et al. Occupational risk factors for mortality from stomach and lung cancer among rubber workers: an analysis using internal controls and refined exposure assessment. Int J Epidemiol. 1999;28:1037–1043.
7. Weiland SK, Straif K, Chambless L, et al. Workplace risk factors for cancer in the German rubber industry: Part 1. Mortality from respiratory cancers. Occup Environ Med. 1998;55:317–324.
8. Straif K, Keil U, Taeger D, et al. Exposure to nitrosamines, carbon black, asbestos, and talc and mortality from stomach, lung, and laryngeal cancer in a cohort of rubber workers. Am J Epidemiol. 2000;152:297–306.
9. Straif K, Weiland SK, Bungers M, et al. Exposure to high concentrations of nitrosamines and cancer mortality among a cohort of rubber workers. Occup Environ Med. 2000;57:180–187.
10. Straif K, Weiland SK, Bungers M, Holthenrich D, Keil U. Exposure to nitrosamines and mortality from salivary gland cancer among rubber workers. Epidemiology. 1999;10:786–787.
11. Straif K, Weiland S, Werner B, Wienke A, Keil U. Elevated mortality from nonalcohol-related chronic liver disease among female rubber workers: is it associated with exposure to nitrosamines? Am J Ind Med. 1999;35:264–271.
12. World Health Organization. Manual of the International Statistical Classification of Diseases, Injuries and Causes of Death: Ninth Revision. Geneva, Switzerland: WHO; 1977.
13. Rostgaard K. Methods for stratification of person-time and events—a prerequisite for Poisson regression and SIR estimation. Epidemiol Perspect Innov. 2008;5:7.
14. Breslow NE, Day NE. Statistical Methods in Cancer Research. Lyon, France: IARC; 1987.
15. Richardson DB. Temporal variation in the association between benzene and leukemia mortality. Environ Health Perspect. 2008;116:370–374.
16. Silver SR, Rinsky RA, Cooper SP, Hornung RW, Lai D. Effect of follow-up time on risk estimates: a longitudinal examination of the relative risks of leukemia and multiple myeloma in a rubber hydrochloride cohort. Am J Ind Med. 2002;42:481–489.
17. Hornung RW, Deddens J, Roscoe R. Modifiers of exposure-response estimates for lung cancer among miners exposed to radon progeny. Environ Health Perspect. 1995;103(suppl 2):49–53.
18. Dost AA, Redman D, Cox G. Exposure to rubber fume and rubber process dust in the general rubber goods, tyre manufacturing and retread industries. Ann Occup Hyg. 2000;44:329–342.
19. Kromhout H, Swuste P, Boleij JS. Empirical modelling of chemical exposure in the rubber-manufacturing industry. Ann Occup Hyg. 1994;38:3–22.
20. Vermeulen R, Kromhout H. Historical limitations of determinant based exposure groupings in the rubber manufacturing industry. Occup Environ Med. 2005;62:793–799.
21. Vermeulen R, de Hartog J, Swuste P, Kromhout H. Trends in exposure to inhalable particulate and dermal contamination in the rubber manufacturing industry: effectiveness of control measures implemented over a nine-year period. Ann Occup Hyg. 2000;44:343–354.
22. Wacker CD, Spiegelhalder B, Borzsonyi M, Brune G, Preussmann R. Prevention of exposure to N-nitrosamines in the rubber industry: new vulcanization accelerators based on “safe” amines. IARC Sci Publ. 1987;370–374.
23. de Vocht F, Straif K, Szeszenia-Dabrowska N, et al. A database of exposures in the rubber manufacturing industry: design and quality control. Ann Occup Hyg. 2005;49:691–701.
Copyright © 2013 by the American College of Occupational and Environmental Medicine