Using data from the UK Biobank, the well-validated CHARGE-AF score was used to estimate risk of AF based on clinical risk factors (including age, height, weight, blood pressure, use of antihypertensive medication, diabetes, heart failure and history of myocardial infarction at baseline). A polygenic risk score (PRS) was constructed based on 142 independent genetic variants previously associated with AF in a meta-analysis from the AFGen Consortium.
Amongst the 270,254 individuals contributing to the analyses, with a median of over 8 years of follow-up, 12,407 incident AF cases were identified using hospital episode statistics and death registry data. The CHARGE-AF risk score strongly predicted incident AF and was associated with a ~3-fold higher risk of AF per standard deviation (SD) (hazard ratio [HR] 2.88; 95% confidence interval [CI] 2.82–2.94). Independently of the clinical score, the PRS was associated with a ~50% higher risk of AF per SD (HR 1.57; 95% CI 1.54–1.59) and a 2.5-fold higher risk (HR 2.59; 95% CI 2.47–2.71) when comparing those in the top vs. the bottom third of the PRS. Furthermore, the additional information from the PRS over the CHARGE-AF score resulted in correct reclassification of 8.7% of AF cases and 2.6% of non-cases at 5 years. The findings from this study indicate that an AF PRS can materially improve risk prediction over conventional risk factors.
The authors concluded that incorporating genetic information into AF risk prediction tools may help to identify higher risk individuals who may benefit from earlier monitoring and personalised risk-management strategies.