Colorectal cancer among farmers in the AGRICAN cohort study

1

Background

Colorectal cancer (CRC) was the third most common type of cancer with more than 1.9 million new cases and 0.9 million deaths in 2020 estimated worldwide [1] . There has been a gradual shift in cancer incidences from the left to the right colon in the United States and Europe since 1980 s suggesting different risk factors for CRC subsites 234 . Cancer of the right colon is more frequent among women whereas it is opposite for cancer of the left colon [34] .

Associations with CRC have been reported for a number of factors including sex, genetic predisposition, inflammatory bowel disease, cigarette smoking, alcohol, being overweight/obese, high intake of red meat, low intake of whole grains, fruits, vegetables, dairy, low physical activity, hormone-replacement therapy, and aspirin use [5] .

However, associations with occupational and environmental factors is not well-studied. Exposure to asbestos [6] and night-shift work [7] were associated with an increased CRC risk, while physical activity at work was associated with a decreased CRC risk [8] . Exposure to diesel engine exhaust [910] and metalworking fluids [11] was associated with an increased risk of rectal cancer. Moderately increased risk of colon cancer was observed among workers exposed to benzene [1213] .

Farmers are generally at about 10–20% lower risk of CRC compared to general populations 141516171819 , and a deficit is commonly explained by lower rates of smoking, healthier diet and higher rates of physical activity among farmers. Excess mortality due to CRC was observed among Iowa and New-Zealand agricultural workers [2021] . There was no association between farming and CRC risk in an Australian cohort study [22] . An increased CRC risk among farmers was observed in an Italian case-control study [23] and poultry farming was positively associated with colon cancer in the Agricultural Health Study (AHS) [24] .

Farmers are occupationally exposed to various agents (pesticides, dusts, diesel exhaust, disinfectants, endotoxins, zoonotic infections among others) while performing their routine tasks, and some of these agents have been positively associated with CRC risk. Ecological studies found elevated CRC incidence and mortality in regions with higher pesticide use in Brazil [25] and Canada [26] . Pesticide use was associated with an elevated CRC risk in a case-control study from Egypt [27] . Positive associations were also observed for specific pesticide molecules in the AHS cohort studies 2829303132 . Various bacteria and viruses, which is a potential exposure in livestock farming, have also been linked to an increased CRC risk in previous studies [3334] .

There is an evidence that associations with CRC risk factors may vary by anatomical site [2935] . Association with smoking was stronger for rectal cancer than colon cancer in a case-control study of US veterans [35] . Imazethapyr-exposed pesticide applicators in AHS study were at higher risk of colon cancer and excess risk was restricted to proximal colon [29] . In addition, studies demonstrated that right and left colon cancer cases display different characteristics at the molecular, mutational and histological levels [36] . Therefore, CRC subsites represent different disease entities and assessment of associations by CRC anatomical site may provide a better insight into CRC etiology and pathogenesis.

The Agriculture and Cancer (AGRICAN) is a large prospective cohort study conducted in France. It offers an opportunity for detailed evaluation of a large number of agricultural exposures. We aimed to evaluate associations between various farming types and tasks, including pesticide exposure through different tasks, and the risk of CRC, overall and by localization (colon overall, right colon, left colon and rectum) in the AGRICAN cohort.