Long-term risk of cardiac arrest according to a cardiovascular polygenic risk score in a population-based cohort

Background

Polygenic risk scores predict atherosclerotic cardiovascular disease, but their association with cardiac arrest and utility for absolute risk stratification in the general population are uncertain.

Methods

We included 29,004 participants from a large population-based Swedish cohort (1991–1996) with genotyping and baseline risk factors. A genome-wide cardiovascular polygenic risk score, GPSMult, was standardised per SD and categorised as low (lowest 20%), intermediate (middle 60%) and high (highest 20%). Incident cardiac arrest was ascertained by registry linkage with death before cardiac arrest treated as a competing event. Associations were examined using Cox proportional hazards models with adjustment for age, sex, smoking, hypertension, diabetes, and the ApoB/ApoA1 ratio. Absolute risks at 10 and 15 years were estimated using cumulative incidence functions accounting for competing mortality.

Results

Over a median follow-up of 27.3 years, 326 participants experienced cardiac arrest (mean age 58.1±7.6 years; 60.4% women).

Per 1-SD higher GPSMult, cardiac arrest risk increased in unadjusted (HR 1.40, 95%CI 1.25–1.56), age/sex-adjusted (HR 1.45, 1.30–1.62) and multivariable-adjusted models (HR 1.35, 1.21–1.52).

Participants in the highest genetic risk group had a higher risk of cardiac arrest compared with those in the low and intermediate risk groups combined after adjustment for age and sex (HR 1.80, 95% CI 1.42–2.28), and this association persisted after additional multivariable adjustment (HR 1.49, 95% CI 1.16–1.92).

The proportional hazards assumption was met for the polygenic risk score, while age showed evidence of non-proportionality; results were robust in sensitivity analyses allowing time-varying effects of age. In age-stratified analyses, separation in 15-year cumulative incidence by genetic risk was most pronounced among participants aged 55–65 years at baseline, whereas absolute risks were uniformly low in younger individuals and attenuated by competing mortality in older individuals.

In competing-risk analyses, the 10-year cumulative incidence of cardiac arrest was 0.16% in the low genetic risk group, 0.29% in the intermediate group, and 0.31% in the high-risk group. At 15 years, corresponding cumulative incidences were 0.38%, 0.56%, and 0.83%, respectively, corresponding to an excess of 4.5 cardiac arrests per 1,000 individuals in the high compared with the low genetic risk group.

Conclusion

In a large population-based cohort with long follow-up, a cardiovascular polygenic risk score was associated with future cardiac arrest independent of measured baseline risk factors. Absolute risks were low, but high genetic risk identified higher long-term cumulative incidence, particularly in midlife, and highlighted the influence of age and competing mortality on absolute risk.Fig 1.Incidence of cardiac arrest