Abstract
Alcohol is widely used but recognized as a risk factor for several adverse health outcomes based on observational studies. How alcohol affects lifespan remains controversial, with no trial to make such an assessment available or likely. We conducted a Mendelian randomization (MR) to assess the effect of alcohol on lifespan in men and women, including a possible role of smoking and education. Strong (p < 5e(- 8)), independent (r(2) < 0.001) genetic predictors of alcohol consumption in 2,428,851 participants of European ancestry from the Sequencing Consortium of Alcohol and Nicotine use (GSCAN) consortium genome wide association study (GWAS) were applied to sex-specific GWAS of lifespan (paternal and maternal attained age) and age at recruitment to the UK Biobank. We used multivariable MR to allow for smoking and education, with systolic and diastolic blood pressure as control outcomes. Inverse variance weighted was the primary analysis with sensitivity analysis. Alcohol consumption decreased lifespan overall (- 1.09 years (logged alcoholic drinks per week), - 1.89 to - 0.3) and in men (- 1.47 years, - 2.55 to - 0.38), which remained evident after adjusting for smoking (- 1.81 years, - 3.3 to - 0.32) and education (- 1.85 years, - 3.12 to - 0.58). Estimates from sensitivity analysis were similar, and when using the genetic variant physiologically associated with alcohol use. Alcohol consumption was associated with higher blood pressure as expected. Our study indicates that alcohol does not provide any advantages for men or women but could shorten lifespan. Appropriate interventions should be implemented.