Extended Data Fig. 8: Incident disease and PGS validity.

a) Incident disease events over the 7.7 year of follow-up in the n = 3,087 INTERVAL participants. Endpoint: incident disease definition available in INTERVAL for the relevant PGS, as defined by CALIBER phenotyping algorithms. Age of onset: median age of first hospitalisation with the respective endpoint. Numbers in brackets gives the interquartile range. b) Hazard ratio (HR) (points) and 95% confidence interval (95% CI) (horizontal bar) for 7.7 year risk of hospitalisation with the respective endpoint per standard deviation increase in the respective PGS in cox proportional hazards models using follow-up as time scale and adjusting for age, sex, 10 genotype PCs, sample measurement batch, and time between blood draw and sample processing in n = 3,087 INTERVAL participants. P-values are two-sided. c) Association between CKD PGS with estimated glomerular filtration rate (eGFR), a marker of renal function used in chronic kidney disease diagnosis: decreased eGFR is indicative of reduced renal function⁹⁸. EGFR was computed from serum creatinine in n = 3,307 participants using the CKD-EPI equation⁹⁹. Association was fit with linear regression adjusting for age and sex, and 10 genotype PCs. The point corresponds to the change in eGFR per standard deviation increase in CKD PGS, and the horizontal bar corresponds to the 95% CI. P-values are two-sided.