Individuals were removed from the present study based on non-British ancestry (within those who self-identified as being British, principal component analysis was used to remove outliers, n=32 484), high missingness (n=0), relatedness (n=7948), quality control failure in UK BiLEVE (n=187) and gender mismatch (n=0). Relatedness was defined as having a KING-estimated kinship coefficient of >0.0442. For the GWAS analysis we used hard-called genotypes with an imputation info score ⩾0.9, MAF ⩾0.1% and Hardy–Weinberg equilibrium (HWE) P-value ⩽1 × 10−6 (no. of SNPs=12 489 782). SNPs were then filtered to only include those where >80% of the sample had a hard-called genotype.
