Residential proximity to animal feeding operations and risk of solid tumors in the Agricultural Health Study cohort

Abstract

Background: Occupational exposure to animals has been associated with several types of cancers, but little is known about risk associated with living near intensive animal farming.Methods: We evaluated associations between residential proximity to animal feeding operations (AFOs) and cancer risk in the Agricultural Health Study, a prospective cohort of pesticide applicators (mostly male farmers) and their spouses. We linked enrollment (1993-1997) addresses for participants in Iowa (33,340 applicators and 20,186 spouses) to a database of permitted AFOs that included animal units (AUs), counts standardized by animal size and manure production. We computed an inverse distance-weighted AU metric that reflects exposure proximity and intensity within 5km of residences. We estimated risk of major solid tumor sites associated with AU metric quintiles using Cox regression (hazard ratio, HR; 95% confidence interval, 95%CI), adjusting for demographics, farming-related factors (occupational pesticide use, direct animal contact), and cancer site-specific potential confounders (e.g., reproductive history). Associations were also estimated separately by animal type (swine, poultry, cattle) and self-reported occupational animal contact.Results: There were 6,137 incident solid tumors diagnosed during follow-up 1993-2015. Among applicators, testicular cancer risk increased with AU quintiles within 5km of the home (HRQ5vsQ1=3.3, 95%CI=1.0-11.0; ptrend=0.03). Prostate cancer risk was significantly elevated in the highest AU quintile (HRQ5vsQ1=1.3; 95%CI=1.1-1.5; ptrend=0.0001). For lung cancer, there was an inverse, but non-significant trend with increasing AUs (HRQ5vsQ1=0.80; 95%CI=0.50-1.2; ptrend=0.09). Among spouses, uterine cancer risk was elevated (HRQ5vsQ1=1.9, 95%CI=1.2-2.9; ptrend=0.03). Risk of breast cancer increased with cattle AUs (ptrend=0.01), but not for other animal types. We observed no effect modification by animal-related work for any cancer site.Conclusions: Our results suggest that residential proximity to AFOs may influence risk of certain solid tumors, even after consideration of farming activities and other risk factors. These findings are novel, but further evaluation is needed to identify etiologic agents.