Mendelian randomization evidence for the causal effects of socio-economic inequality on human longevity among...

Human longevity correlates with socio-economic status, and there is evidence that educational attainment increases human lifespan. However, to inform meaningful health policies, we need fine-grained causal evidence on which dimensions of socio-economic status affect longevity and the mediating roles of modifiable factors such as lifestyle and disease. Here we performed two-sample Mendelian randomization analyses applying genetic instruments of education, income and occupation (n = 248,847 to 1,131,881) to estimate their causal effects and consequences on parental lifespan and self-longevity (n = 28,967 to 1,012,240) from the largest available genome-wide association studies in populations of European ancestry. Each 4.20 years of additional educational attainment were causally associated with a 3.23-year-longer parental lifespan independently of income and occupation and were causally associated with 30-59% higher odds of self-longevity, suggesting that education was the primary determinant. By contrast, each one-standard-deviation-higher income and one-point-higher occupation was causally associated with 3.06-year-longer and 1.29-year-longer parental lifespans, respectively, but not independently of the other socio-economic indicators. We found no evidence for causal effects of income or occupation on self-longevity. Mediation analyses conducted in predominantly European-descent individuals through two-step Mendelian randomization suggested that among 59 candidates, cigarettes per day, body mass index, waist-to-hip ratio, hypertension, coronary heart disease, myocardial infarction, stroke, Alzheimer's disease, type 2 diabetes, heart failure and lung cancer individually played substantial mediating roles (proportion mediated, >10%) in the effect of education on specific longevity outcomes. These findings inform interventions for remediating longevity disparities attributable to socio-economic inequality.