Sex-specific individual and joint associations of multiple environmental exposures with diabetes and obesity in the population-based German National Cohort (NAKO)

Recent studies have suggested a potential association of particulate matter (PM) and noise with diabetes and obesity, but studies examining other environmental exposures and their sex-specific and joint associations remain limited. Therefore, we investigated sex-specific individual and joint associations of annual exposure to multiple environmental factors with diabetes and obesity-related measures using cross-sectional data from the population-based multi-center German National Cohort (NAKO). Outcomes included self-reported diabetes mellitus, body mass index (BMI), obesity (BMI ≥30 kg/m(2)), and waist circumference. Annual mean residential exposures included air pollutants, air temperature, day-evening-night road traffic noise (L(den)) and surrounding greenness (normalized difference vegetation index (NDVI)). We used sex-stratified linear and logistic regression models to assess individual associations and quantile g-computation to assess joint associations. Among 174,955 adult participants (50.4% women), 5.6% reported a diabetes diagnosis and 20.9% were obese. An interquartile range increase in PM(2.5) and L(den) was consistently associated with diabetes and obesity-related measures (e.g., PM(2.5)-diabetes for men: odds ratio (OR) [95% confidence interval] = 1.12 [1.02; 1.22]; L(den)-BMI for women: 0.22 kg/m(2) [0.16; 0.27]). Greenness showed non-linear (inverted U-shaped) with all outcomes. An interquartile range increase in multiple exposures simultaneously was associated with higher odds of diabetes, obesity and higher obesity-related measures (e.g., mixture (PM(2.5),L(den), lack of NDVI)-diabetes: OR = 1.20 [1.09; 1.33] for men; mixture (PM(2.5),L(den), lack of NDVI)-BMI: 0.33 kg/m(2) [0.21; 0.44] for women). While longitudinal studies need to confirm these findings, the study highlights that reducing multiple adverse environmental exposures could be potential targets for the prevention of diabetes and obesity.