In order to identify working environments at risk for Parkinson's disease (PD), we investigated the relation between the importance of industry sectors, used as a surrogate for occupational exposures, and PD incidence in French cantons. The number of incident PD cases (2010-2014) in 3689 cantons of metropolitan France was determined using drug claims from French National Health Insurance databases. The proportions of workers in 38 industry sectors in 2006 were calculated for each canton. Associations between the proportions of workers in industry sectors and PD age/sex-standardized incidence ratios were examined using incidence rate ratios (IRR) and 95% confidence intervals (CI) estimated with multilevel negative binomial regressions with a random intercept at the canton-level and adjusted for smoking, deprivation index, and density of neurologists. We then used two-step semi-Bayes hierarchical regression (HR) to include prior information about exposure to pesticides, metals, and solvents in each industry sector. We identified 112,625 incident cases. PD incidence was higher in areas characterized by high proportions of workers in "Agriculture, forestry and fishing" (IRRHR = 1.042; CI 95% = 1.014-1.070; p-TrendHR = 0.004), "Manufacture of textiles, wearing apparel, leather and related products" (IRRHR = 1.024; CI 95% = 1.005-1.044; p-TrendHR = 0.010), and "Manufacture of basic metals and fabricated metal products, except machinery and equipment" (IRRHR = 1.024; CI 95% = 1.003-1.046; p-TrendHR = 0.071). This nationwide study, based on a comprehensive analysis of industry sectors, shows significant associations between high proportions of workers in specific industry sectors (agriculture, metallurgy, textile) and PD incidence that may be targeted in further epidemiological studies to replicate and better understand these associations.
Auteur : Vlaar Tim, Kab Sofiane, Schwaab Yannick, Fréry Nadine, Elbaz Alexis, Moisan Frédéric
European Journal of Epidemiology, 2018, vol. 33, n°. 11, p. 1101-1111
