Mortality study of Canadian male farm operators: non-Hodgkin's lymphoma mortality and agricultural practices in Saskatchewan.

A cohort study of the mortality experience (1971-1985) of male Saskatchewan farmers has been conducted. This study involved linkage of records of the almost 70,000 male farmers identified on the 1971 Census of Agriculture and the corresponding Census of Population to mortality records. Pesticide exposure indices for individual farm operators for the year 1970 were derived from the 1971 Census of Agriculture records. Although the cohort as a whole had no excess mortality for any specific causes of death, including non-Hodgkin's lymphoma, significant dose-response relationships were noted between risk of non-Hodgkin's lymphoma and acres sprayed in 1970 with herbicides, as well as with dollars spent on fuel and oil for farm purposes in 1970. Using Poisson regression modeling, we found that relative risks for the highest level of herbicide use (greater than or equal to 250 acres sprayed) and fuel purchased in 1970 (greater than or equal to $900) on farms less than 1,000 acres total area were 2.2 (95% confidence interval = 1.0-4.6) and 2.3 (95% confidence interval = 1.1-4.7), respectively.