8278855 Occupational pesticide use and colon cancer in the agricultural health study, 1993–2021

<h3>Objectives</h3> Colon cancer is one of the most common cancers worldwide. Besides diet and other lifestyle factors, there are few known etiologic agents. Some studies have suggested links with pesticide exposure, although data are limited on specific chemicals. In addition, recent trends suggest that early onset colon cancer incidence is increasing. We investigated associations between use of individual pesticides and incident colon cancer in the Agricultural Health Study cohort, a prospective study of licensed pesticide applicators in Iowa and North Carolina, USA. <h3>Material and Methods</h3> Among 54,342 male pesticide applicators there were 752 colon cancer cases, including 43 diagnosed prior to age 50, identified through cancer registry linkages between enrollment (1993-1997) and 2021. We calculated rate ratios (RRs) and 95% confidence intervals (CIs) using Poisson regression adjusting for potential confounders to evaluate associations with individual pesticides overall and by age at diagnosis. <h3>Results</h3> Compared to never users, we observed elevated overall colon cancer risk among the highest lifetime users of the herbicides EPTC (RRQuintile5=1.67, 95%CI 1.13-2.48, ptrend=0.03) and chlorimuron ethyl (RRQuintile5=1.88, 95%CI=1.19-2.96, ptrend=0.003), with no differences by age at diagnosis. We observed increased risks of early onset colon cancer for ever use of the insecticides fonofos (RRever=1.98, 95%CI 1.00-3.90) and aldicarb (RRever=3.90, 95%CI 1.69-9.01), and the fungicide chlorothalonil (RRever=2.62, 95%CI 1.06-6.50), although no associations were observed with overall colon cancer for these pesticides. Results were similar when adjusting for correlated pesticides. <h3>Conclusions</h3> In this prospective evaluation, we found evidence of associations between several pesticides and colon cancer. We also found that for some pesticides, elevated risks were only observed for colon cancer diagnosed before age 50. While based on a relatively small number of early onset cases, age-specific associations are intriguing and suggest the need for further study to elucidate potential differences in etiological factors and possible underlying mechanisms.