Abstract

OBJECTIVES--To explore the impression that occupational epidemiologists tend to focus on associations suggestive of increased risk and tend to ignore those associations in which risk is not increased. To examine the risk of colorectal cancer in cohorts exposed to dust, cohorts in which it has been suggested that occupational exposure is a cause of increased risk of stomach cancer. METHODS--A review of the publications in the English language on mortality among hard rock miners, granite, and quarry workers identified from a MEDLINE search and the index of the library of the Ontario Ministry of Labour. RESULTS--When all of the studies were combined, there were significant excesses of lung and stomach cancers, but a significant deficit of colorectal cancer (standardised mortality ratio (SMR) = 83.9; 95% confidence interval (95% CI) 76-91). Overall mortality from gastrointestinal cancer was close to expectation (SMR = 105; 95% CI 99-111). Among those cohorts with increased risk of stomach cancer, rates of colorectal cancer were significantly decreased (SMR = 80; 95% CI 72-88). Among cohorts without increased risk of stomach cancer, the SMR for colorectal cancer was not significantly different from 100 (SMR = 98; 95% CI 81-115). CONCLUSIONS--This review supports the impression that occupational epidemiologists tend to focus on associations suggestive of increased risk and tend to ignore those associations in which risk is not increased. The explanation for the inverse association between risk of stomach and colorectal cancer is uncertain and deserves further study.