Educational attainment and drinking behaviors: Mendelian randomization study in UK Biobank

Educational attainment has been associated with drinking behaviors in observation studies. We performed Mendelian randomization analysis to determine whether educational attainment causally affected drinking behaviors, including amount of alcohol intakes (in total and various types), drinking frequency, and drinking with or without meals among 334 507 White British participants from the UK Biobank cohort. We found that genetically instrumented higher education (one additional year) was significantly related to higher total amount of alcohol intake (inverse variance weighted method (IVW): beta=0.44, 95% confidence interval (CI) 0.40 to 0.49, P=1.57E-93). The causal relations with total amount and frequency of alcohol drinking were more evident among women. In analyses of different types of alcohol, higher educational attainment showed the strongest causal relation with more consumption of red wine (IVW beta=0.34, 95% CI 0.32 to 0.36, P=2.65E-247), followed by white wine/champagne, in a gender-specific manner. An inverse association was found for beer/cider and spirits. In addition, we found that one additional year of educational attainment was causally related to higher drinking frequency (IVW beta=0.54, 95% CI 0.51 to 0.57, P=4.87E-230) and a higher likelihood to take alcohol with meals (IVW: odds ratio (OR)=3.10, 95% CI 2.93 to 3.29, P=0.00E+00). The results indicate causal relations of higher education with intake of more total alcohol especially red wine, and less beer/cider and spirits, more frequent drinking, and drinking with meals, suggesting the importance of improving drinking behaviors, especially among people with higher education.