Sociodemographic, lifestyle, environmental, and neighborhood exposures and incident type 2 diabetes in over 235,000 Dutch adults: an exposome approach

Abstract Introduction Sociodemographic, lifestyle, environmental, and neighborhood exposures are associated with Diabetes Mellitus type 2 (DM-2). Yet, studies that include exposures from all these domains remain limited. In this study, the association between a wide range of exposures across these domains and DM-2 incidence was investigated, stratified by sex. Methods Data from the 2012 Dutch national health survey was used (N = 237 644), enriched with exposure data from multiple sources. In total, 57 sociodemographic, lifestyle, environmental and neighborhood exposures were included. Incidence of DM-2 was based on medication prescription from 2013 to 2022. The most important exposures for DM-2 were identified using Random Forest. Subsequently, the associations between the selected exposures and DM-2 incidence were assessed via a Cox regression, stratified by sex. Results In total, 5328 men and 4298 women developed DM-2 between 2013 and 2022. BMI, age and sex were identified as the most important predictors. The top 15-ranked exposures were included in the Cox regression models. BMI, age and lifestyle exposures (eg, alcohol consumption, physical activity and smoking) were associated with DM-2 in both men and women. Only in men, neighborhood exposures such as property value were associated with DM-2, while education was associated with DM-2 only in women. No associations were found for environmental exposures. Conclusions These results substantiate that sociodemographic and lifestyle exposures are important targets for DM-2 prevention and outperform neighborhood and environmental exposures. We observed sex-specific associations, highlighting the importance of using a sex-stratified approach in future research and clinical practice.