Genetic predisposition to longer lifespan, lifestyle factors, and all-cause mortality: a 17-year prospective cohort study

Background We used a genome-wide polygenic lifespan score (PLS) to investigate whether genetic predisposition to a longer lifespan is associated with lower mortality risk and how this association compares with associations between long-term lifestyle factors and mortality. Methods A PLS was computed for the Older Finnish Twin Cohort (mean age 57·4 years; 45·2% males; N=5575). Cox regression models were used to analyse the effects of the PLS on all-cause mortality risk before and after adding covariates (sex, physical activity, BMI, alcohol consumption, smoking, and education level). Differences between concordance indices (C-indices) were used to evaluate each predictor's contribution to the model's discriminatory performance. Findings Over a mean follow-up of 17·5±8·3 years, 1405 deaths (25·2%) were recorded. A one standard deviation increase in the PLS was statistically significantly associated with a lower all-cause mortality risk (hazard ratio [HR]=0·838, 95% confidence interval [CI]=0·792–0·887) and remained relatively unchanged after adding all covariates (HR=0·863, 95% CI=0·816–0·912). Smoking 20 or more cigarettes per day showed the strongest association with increased mortality risk (HR=3·341, 95% CI=2·751–4·056), whereas female sex was linked to the greatest risk reduction (HR=0·678, 95% CI=0·597–0·770). Smoking had the largest impact on model discrimination (ΔC-index=0·027), whereas all other covariates increased the C-index by less than 0·006. Interpretation Genetic predisposition to a longer lifespan was associated with a small decrease in all-cause mortality risk. This association remained after adjusting for lifestyle factors and covariates. Smoking behaviour and female sex were stronger predictors of mortality than genetic scores for lifespan. Funding This study was funded by the Research Council of Finland (341750, 346509, and 361981 for ES) and the Juho Vainio, and Päivikki and Sakari Sohlberg Foundation (ES). Finnish Twin Cohort data collection was supported by the Research Council of Finland (264146, 308248, 336823, and 352792 for JK), the Wellcome Trust Sanger Institute, the Broad Institute, ENGAGE, and FP7-HEALTH-F4-2007 (201413).