American Journal of Epidemiology Copynght © 1991 by The Johns Hopkins University School of Hygiene and Public Health All rights reserved

Vol. 134, No. 6 Printed in U.S.A.

LETTERS TO THE EDITOR RE: "LIFETIME

OCCUPATIONAL PHYSICAL ACTIVITY RISK"

In their recent paper on lifetime occupational physical activity and risk of prostate cancer, Le Marchand et al. (1) reported a negative association between risk of this important malignancy and proportion of life spent in jobs involving only sedentary or light work, among men aged 70 years or older. The subjects were asked whether they had worked with or around several chemicals that may be associated with increased cancer risk (1). In addition, the job-exposure matrix developed by Hoar et al. (2) was used to assess possible chemical or physical exposures using the job held the longest (1). The authors stated that the observed negative association was "unlikely to be due to a relative lack of job-related chemical exposures, since no particular occupation or exposure (either reported by the subjects or imputed from a job-exposure matrix) was associated with prostate cancer risk" (1, p. 108). Although the data from this well-designed study appear to support this conclusion, it is important to recognize the limitations of the methods of chemical occupational exposure assessment that were used. Some of the chemical exposures considered may not have been readily apparent to the subjects, and appeciable underreporting of exposures may have occurred (3). The subjects' recall of exposures that occurred years previously is also likely to have been imperfect (4). In addition, there may be important occupational exposures (that have not been identified by studies of prostatic cancer completed to date) which were not asked about. In studies carried out in a specific geographic locality, errors may occur in linking exposures to job titles if the exposure estimates are based upon national industrial hygiene surveys and general bibliographic sources that are not representative of the local area (3). In a previous validation study of the job-exposure matrix developed by Hoar et al., investigators from Honolulu assigned estimates of exposure levels to known or suspected carcinogens in a population-based casecontrol study of lung cancer in Oahu, Hawaii (5). Statistically significant associations were observed with only three of seven carcinogens associated a priori with lung cancer risk, suggesting that this indirect method of exposure assessment was not sensitive enough to detect the associations of interest (3, 5). The authors noted that the exposure estimates obtained using the matrix ap-

AND PROSTATE

CANCER

proach were based upon different geographic settings and time periods, and that the level of misclassification of exposures may have been high (5). The indirect exposure estimates obtained as part of the case-control study of prostatic cancer (1) may have been similarly biased. In view of these potential limitations, there remains a possibility that the observed negative association between prostate cancer risk and lifetime occupational physical inactivity (1) is explained by reduced chemical exposures among individuals with sedentary or light activity jobs. REFERENCES

1. Le Marchand L, Kolonel LN, Yoshizawa CN. Lifetime occupational physical activity and prostate cancer risk. Am J Epidemiol 1991; 133:103-11. 2. Hoar SK, Morrison AS, Cole P, et al. An occupation and exposure linkage system for the study of occupational carcinogenesis. J Occup Med 1980;22:722-6. 3. Coughlin SS, Chiazze L Jr. Job-exposure matrices in epidemiologic research and medical surveillance. State Art Rev Occup Med 1990;5:633-46. 4. Coughlin SS. Recall bias in epidemiologic studies. J Clin Epidemiol 1990;43:87-91. 5. Hinds MW, Kolonel LN, Lee J. Application of a job-exposure matrix to a case-control study of lung cancer. J Natl Cancer Inst 1985;75:193-7.

Steven S. Coughlin Department of Medicine Divisions of Medical Oncology and Cardiology Georgetown University School of Medicine Washington, DC 20007

THE A UTHORS REPL Y We thank Dr. Coughlin for his comments (1) on our recent paper on lifetime occupational physical activity and risk of prostate cancer (2). We share his concerns about the lack of sensitivity of generic job-exposure matrices in identifying individual exposures to occupational carcinogens from job titles. Indeed, we discussed these issues in our previous validation study (cited by Dr. Coughlin) of one of these matrices (3). Because each method has its own limitations, we used two additional standard approaches to assess occupational exposures in our prostate can672

re: "Lifetime occupational physical activity and prostate cancer risk".

American Journal of Epidemiology Copynght © 1991 by The Johns Hopkins University School of Hygiene and Public Health All rights reserved Vol. 134, No...
96KB Sizes 0 Downloads 0 Views