Introduction Prostate cancer incidence ranks 2nd among men worldwide. Farming and pesticide use have been repeatedly and consistently associated with prostate cancer risk, with suggestion of the role of organochlorine pesticides, but results on individual pesticides remain controversial.

Objectives We assessed associations between prostate cancer and agricultural exposure to organochlorine insecticides (as a group and considering individual pesticides: aldrin, chlordane, HCH, dieldrin, lindane, heptachlore, DDT, DDD, endosulfan, metoxychlore, SPC, toxaphen, chlorfenethol, chlorobenzilate, dicofol, bromopropylate).

Methods Data on pesticide use on 10 crops, including years of beginning and ending, were collected for 81,960 men from the enrolment questionnaire of AGRICAN. Incident prostate cancer cases were identified through linkage with cancer registries. Exposure to organochlorine insecticides was assessed using the crop-exposure matrix PESTIMAT. Hazard Ratios (HR and 95%CI) were estimated using Cox models with attained age as time scale.

Results Until 2015, 3,535 prostate cancers were diagnosed. We found a slight increase in the risk of prostate cancer in men diagnosed until 2009 exposed to organochlorines (HR=1.11[0.96–1.30]), but the trend disappeared with 6 additional years of follow-up (HR=0.95[0.87–1.05]). This result was also globally found when considering individual pesticides. But for substances still in use after 1980 (lindane, dicofol, endosulfan, toxaphen), some increases were still observed with longer follow-up: for lindane used on wheat/barley for ≥40 years (HR=1.27[1.01–1.58]), on rape for ≥30 years (HR=1.52[0.86–2.70]), on sunflower for ≥10 years (HR=1.14[0.87–1.51]) and on vineyard for ≥30 years (HR=1.09[0.90–1.32]), for dicofol used ≥40 years on vineyards (HR=1.17[0.89–1.53]), and for endosulfan used on fruit-growing (HR=1.16[0.95–1.41]).

Conclusions The increased risk of prostate cancer associated with organochlorines initially observed tended to disappear with follow-up duration, except for some substances still in use in the 1980s and 1990s. Explanations could be i) the distance between last exposure for some organochlorines and cancer diagnosis, ii) changes in the exposure intensity due to evolution of practices.