A weighted genetic risk score based on 279 signals of association with lung function predicts Chronic Obstructive Pulmonary Disease

Lung function measures are used in diagnosis and grading of COPD. We have previously shown that the genetic determinants of lung function, analysed in large general population cohorts, are informative about risk of COPD and a powerful alternative to case-control studies. We previously reported a significant association (P=5.65x10-36) with COPD for an unweighted genetic risk score based on the 95 variants reported as associated with lung function at that time. We undertook genome-wide association analyses of spirometry in 404,165 individuals from UK Biobank and the SpiroMeta consortium to identify novel signals (P We identified 139 novel signals which we combined with the 140 previously reported signals to construct a weighted risk score that was then tested for association with COPD in a combined analysis of 5991 COPD cases and 3378 controls. The mean (SD) number of risk alleles per individual across the 279 signals was 295 (10.4). The weighted risk score was significantly associated with risk of COPD (P=6.64x10-63), with an OR of 1.55 (95% confidence intervals 1.47, 1.63) for each standard deviation of the risk score (approximately 12 risk alleles). Our findings show that increasing the number of signals in a weighted risk score increases predictive power for COPD, and provides new biological insight.