Using genetics to understand the role of lifestyle and socioeconomic factors in the relationship between Major Depressive Disorder and Type 2 Diabetes

Background Major depressive disorder(MDD) and type 2 diabetes(T2D) are global health problems that co-occur frequently. The bidirectional causal relationship between them is well-established, though the potential mediators remain understudied. Here, we investigated the role of lifestyle and socioeconomic traits as mediators of the MDD-T2D relationship using genetic data. Methods Using genome-wide association study(GWAS) summary statistics, we tested if the genetic correlation(rg) between MDD(N=2,000,702) and T2D(N=1,812,017) was influenced by lifestyle {physical activity and sedentary time(N=95,776), cigarettes per day(N=618,489), drinks per week(N=941,280)} and socio-economic {education(N=293,723), poverty factor(N=453,689)} traits. Mendelian randomization(MR) analyses, using both GWAS and UK-Biobank data, were conducted to investigate potential causal relationships. Finally, we applied mediation analysis to quantify the indirect effect of potential mediators. Results The strong genetic correlation between MDD and T2D(rg:0.184, p=3.5E-37) was attenuated when accounting for cigarettes per day (partial_rg:0.098), physical activity (partial_rg:0.141), educational attainment(partial_rg:0.115) and poverty(partial_rg:0.118). MR demonstrated that higher genetic liability to MDD and T2D was associated with higher cigarettes per day, poverty, and lower education. MR showed that higher education was associated with lower MDD and T2D odds, whilst poverty associated with higher odds of both conditions. Mediation analysis demonstrated that a genetic liability to education and poverty traits substantially mediated the causal relationship in both directions: 28.67% and 24.98% for MDD->T2D, and 34.3% and 29.87% for T2D->MDD, respectively. Conclusion Our findings highlight the role of socioeconomic traits in the MDD-T2D relationship and demonstrate both the strength and limitation of using MR for investigating complex traits. Further research incorporating gene-environment interactions, policy implications and diverse populations is needed to better understand this relationship.