of long-term alcohol consumption, calculated on the basis of ADH1B genotype (instrumental variable), on blood pressure, body mass index (BMI), and triglyceride levels [Lawlor and others 2013]. A larger Mendelian-randomization study (N = 261,991) replicated these results, reporting that the carriers of the protective allele had a more favorable cardiovascular profile and a reduced risk of coronary heart disease than those without the genetic variant [Holmes and others 2014]. Since there is a U-shaped relationship between alcohol use and cardiovascular events – that is, moderate alcohol consumption leads to better outcome than very low or very high – the Alcohol-ADH1B Consortium tested this hypothesis using ADH1B rs1229984 as a genetic instrument and they confirmed the presence non-linear causal effects of alcohol intake [Silverwood and others 2014]. This non-linear relationship was also confirmed by the ADH1B genotype-differential effect of initiating moderate red wine consumption on 24-h blood pressure observed in a randomized trial of patients with type-2 diabetes [Gepner and others 2016]. In Asian alcoholic individuals, both ADH1B and ALDH2 protective alleles showed association with high serum triglyceride levels and low serum cholesterol levels, leukocyte, granulocyte, and monocyte counts [Yokoyama and others 2016; Yokoyama and others 2015]. Our GWAS of BMI
