Education and T2DM: the role of differential exposure and susceptibility to overweight and obesity

Abstract

Abstract Background There is a clear educational gradient in type 2 diabetes and differential exposure and susceptibility to overweight and obesity across educational groups may explain part of this gradient. Previous studies have mainly addressed differential exposure to overweight/obesity, even though the pathways are highly intertwined. We take advantage of novel methodological developments to assess the simultaneous contribution of both pathways. We further compare these analyses to more traditional approaches. Methods A prospective cohort of 53,159 Danish men and women, aged 50-64 years at enrollment in 1993-1997, were followed for a mean of 14.7 years. We used national register data to ascertain education and incident type 2 diabetes. Overweight/obesity was determined by BMI. Using a marginal structural approach and a three-way effect decomposition, we estimated rate differences of type 2 diabetes by education and decomposed these into direct, indirect and mediated interactive proportions. We compared this approach to traditional and separate approaches to mediation and interaction analysis. Results Compared with high education, medium education was associated with 316 (95% CI: 268-363) extra cases of type 2 diabetes per 100,000 person-years, and low education with 454 (95% CI: 398-510) extra cases. Differential exposure to overweight/obesity accounted for 29% (95% CI: 24-36) and 37% (95% CI: 31-45) of the cases in the medium and low educational levels, respectively, while differential susceptibility accounted for 6% (95% CI: 3-10) and 9% (95%CI 4-14). Traditional approaches suggested stronger effects of both pathways. Conclusions Differential exposure and susceptibility to overweight/obesity are both important pathways in the education-type 2 diabetes association, and the results emphasize the importance of addressing them simultaneously. Key messages Lower educated groups are more exposed and susceptible to the adverse effects of overweight/obesity with regard to type 2 diabetes, suggesting a large preventive potential. Future studies could benefit from estimating both differential exposure and differential susceptibility to better understand social inequality in type 2 diabetes and other health outcomes.