Association of neighbourhood unemployment rate with incident Type 2 diabetes mellitus in five German regions

Aim To analyse the association of neighbourhood unemployment with incident self‐reported physician‐diagnosed Type 2 diabetes in a population aged 45–74 years from five German regions. Methods Study participants were linked via their addresses at baseline to particular neighbourhoods. Individual‐level data from five population‐based studies were pooled and combined with contextual data on neighbourhood unemployment. Type 2 diabetes was assessed according to a self‐reported physician diagnosis of diabetes. We estimated proportional hazard models (Weibull distribution) in order to obtain hazard ratios and 95% CIs of Type 2 diabetes mellitus, taking into account interval‐censoring and clustering. Results We included 7250 participants residing in 228 inner city neighbourhoods in five German regions in our analysis. The incidence rate was 12.6 per 1000 person‐years (95% CI 11.4–13.8). The risk of Type 2 diabetes mellitus was higher in men [hazard ratio 1.79 (95% CI 1.47–2.18)] than in women and higher in people with a low education level [hazard ratio 1.55 (95% CI 1.18–2.02)] than in those with a high education level. Independently of individual‐level characteristics, we found a higher risk of Type 2 diabetes mellitus in neighbourhoods with high levels of unemployment [quintile 5; hazard ratio 1.72 (95% CI 1.23–2.42)] than in neighbourhoods with low unemployment (quintile 1). Conclusions Low education level and high neighbourhood unemployment were independently associated with an elevated risk of Type 2 diabetes mellitus. Studies examining the impact of the residential environment on Type 2 diabetes mellitus will provide knowledge that is essential for the identification of high‐risk populations. What's new? Research on the influence of the residential environment on incident Type 2 diabetes in Germany is missing. We adapted an advanced model strategy to take account of the hierarchical data structure and to adjust for the interval‐censored information on incident Type 2 diabetes to avoid a common problem in follow‐up cohort studies. This study shows that neighbourhood unemployment is associated with incident Type 2 diabetes in five German regions.