Is neighborhood socioeconomic status associated with health behavior in Berlin? Cross-sectional data of the German National Cohort (NAKO)

Abstract Background Neighborhood socioeconomic status (nSES) can complement individual SES to better assess health-behavior inequalities. The aim of this study was to investigate the relationship between the nSES of defined areas in Berlin with healthy lifestyle. Methods This cross-sectional analysis used baseline data from the three Berlin study centers of the German National Cohort (NAKO). We assessed body mass index (BMI), smoking, alcohol consumption, and objectively measured physical activity and combined them to a healthy lifestyle index (HLI; range:0–12 points; 12 = best score). To assess nSES, the Social Index from Berlin’s Social Structure Atlas (1 = best; 7 = worst) was assigned to the participants’ residential locations. We used multivariable regression analyses to examine the association between nSES and the HLI (mean difference with 95% confidence interval, CI) as well as the four individual lifestyle factors (odds ratios (OR) with 95% CI). In sensitivity analyses, nSES was modelled using all seven Social Index categories and as a dichotomy (categories 1–4 vs. 5–7). Results Of 204,801 NAKO participants, 31,075 were recruited in Berlin, of those 11,922 with complete accelerometry data were included (mean ± SD age 50.6 ± 12.9 years; 52.8% women). The mean HLI was 8.3 ± 2.0 points. Worsening of nSES by one point was associated with a 0.08-point lower HLI (-0.08 (95%-CI -0.10; -0.06)), with a reduced odds of normal weight (0.95; 0.93–0.97) and being a never-smoker (0.96; 0.94–0.98), while it was neither associated with alcohol consumption (1.01; 0.99–1.04)) nor physical activity (0.99; 0.97–1.02)). Sensitivity analyses suggested that differences were mainly driven by a contrast between categories 1–4 and the more disadvantaged categories 5–7. However, the overall pattern of results did not change. Conclusions Our analyses suggest a rather small association between Berlin’s nSES and HLI, and slightly stronger associations with BMI and smoking. Future studies using longitudinal data and more neighbourhood measures are needed to better disentangle contextual influences from residential selection and to inform targeted prevention strategies.