Green space exposure and risk of amyotrophic lateral sclerosis: a population-based case-control study in Northern Italy

Background: The contribution of environmental determinants in the etiology of amyotrophic lateral sclerosis (ALS) is still unclear. Among the various environmental factors, exposure to green spaces, also known as greenness, is attracting considerable interest as many studies have reported its beneficial associations to health outcomes, particularly to neurodegenerative diseases. Methods: To investigate the relation between greenness and ALS risk, we conducted a population-based case-control study in a Northern Italy population (from Modena, Reggio Emilia and Parma provinces), including 499 cases of ALS newly-diagnosed from 1998 to 2011 and 1,935 sex-, age-, and province-matched controls randomly selected from study provinces residents. We evaluated the association between greenness in the proximity of residence and ALS risk, assessing exposure through multiple satellite-based and land-use derived indices, both conventional and novel devised, for a total of six indices, each providing specific information, including annual and seasonal Normalized Difference Vegetation Index (NDVI), NDVI-weighted to green areas, green cover ratio, accessibility index, and their combined Green Exposure Index (GEI). We used conditional logistic regression models to evaluate disease risk for increasing exposure through both fixed-categories and non-linear restricted cubic splines. Results: We observed a non-linear U-shaped association between greenness and ALS risk with increased odds ratios at both low and high levels. Results were more defined when using NDVI-based indices, while the associations were smoother when considering GEI. The higher risk at low levels may be related to lower accessibility to green spaces with lower physical activity and higher exposure to outdoor air pollutants, whilst elevated greenness may reflect higher exposure to neurotoxic pesticides. These results were confirmed also after adjustment for potential confounders, namely magnetic fields and light at night. Sex stratified analysis yielded similar results, except for more distinct associations in females for GEI. Conclusions: Despite the limitations due to possible unmeasured confounding and exposure misclassification related to the use of residential data, our results provide evidence of an inverse association between intermediate residential greenness and ALS risk, and may have public health implications including disease prevention and urban planning.