To validate our merged GWAS datasets, we conducted two proof-of principle GWAS of one quantitative trait (BMI) and one binary trait (VTE). We defined BMI as weight (kg)/height2 (cm) and obtained it by extracting information on weight from the accompanying questionnaire collected at time of blood draw. If weight information was missing, we extracted it from the questionnaire closest in time to time of blood draw. Height was extracted from the baseline questionnaire. We obtained data on BMI for 20,283 participants. VTE is a spectrum of disease that includes pulmonary embolism (PE) and deep vein thromboembolism (DVT). Physician-diagnosed PE has been asked on every biennial NHS questionnaire since 1982, and every NHSII and HPFS questionnaire since cohort inception. In the NHS, DVT without PE is captured when a nurse answers that she has had phlebitis or thrombophlebitis (ICD-9 = 453.x). In NHS, NHSII and HPFS cohorts through 2010 (we did not have VTE data for PHS), we identified 6,041 individuals who reported VTE. Self-reported PE was verified through medical records review by a trained physician (CK). DVT cases are based
