No abstract

Background An exploratory, case‐control study was used to investigate a new hypothesis about suicide among farm operators. This hypothesis suggested a biologically plausible link between exposures to certain pesticides and the occurrence of suicide among farm operators. These analyses were based on data from the Canadian Farm Operator Cohort. Methods Canadian male farm operators who committed suicide between 1971–1987 (n = 1,457) were compared with a frequency matched (by age and province) sample of control farm operators (n = 11,656) who were alive at the time of death of individual cases. Comparisons focused on past exposures to pesticides reported to the 1971 Canada Census of Agriculture. Results Multivariate logistic regression analyses indicated no associations between suicide and (1) acres sprayed with herbicides, (2) acres sprayed with insecticides, and (3) the costs of agricultural chemicals purchased; after controlling for important covariates. There was, however, a suggestive increase in risk for suicide associated with herbicide and insecticide spraying among a subgroup of farm operators who were most likely to be directly exposed to pesticides: OR = 1.71 (95% CI = 1.08–2.71) for 1–48 vs. 0 acres sprayed. Additional risk factors that were identified included seasonal vs. year‐round farm work (OR = 1.68; 95% CI = 1.15–2.46); and high levels of paid labor on the farm (e.g., OR = 1.61; 95% CI = 1.24–2.10, for >13 vs. 0 weeks per year). Factors that were protective included marriage (odds ratio (OR) = 0.69; 95% confidence interval (CI) = 0.58–0.81), having more than one person resident in the farm house (e.g., two vs. one person; OR = 0.62; 95% CI = 0.42–0.92); and higher levels of education (e.g., postsecondary vs. primary; OR = 0.40; 95% CI = 0.17–0.96). Conclusions This study does not provide strong support for the main hypothesis under study, that exposure to pesticides is an important risk factor for suicide among farmers. Although secondary to the main hypothesis, a number of other risk factors for suicide were suggested. These have implications for the future study and targeting of suicide prevention programs in rural Canada. Am. J. Ind. Med. 34:364–372, 1998. © 1998 Wiley‐Liss, Inc.

(e.g., two vs. one person; OR ϭ 0.62; and higher levels of education (e.g., postsecondary vs. primary; OR ϭ 0.40;