Fig. 1

Differences in health deficits accumulation as a smooth function of residential traffic density (RTD) at 50 and 100 meters and of residential nearest distance to a petrol station or a major road in the Seniors-ENRICA cohort, 2008–2010 to 2017. Curves represent average differences in deficits accumulation index at each follow-up visit (solid lines) and 95% confidence intervals (dashed lines) based on restricted cubic splines for the exposure variables at baseline. The reference values were set at the 10 th percentile of each exposure variable distribution (8.8 vehicles/day, 32.4 vehicles/day, 254.8 meters and 310 meters, respectively). Average differences were obtained from repeated measures regression models adjusted for age, sex, educational level, baseline levels of the deficit accumulation index (DAI), time-varying lifestyles, social deprivation index (SDI) at the census tract, and presence of natural spaces. Models were weighted by the inverse of the conditional probabilities of censoring given residential exposure to traffic variables, individual time-varying confounders, SDI and presence of natural spaces; and accounted for spatial correlation at the census tract and within-participant correlations induced by repeated measures and weighting. Bars represent the histograms of the exposure variables