Polygenic risk score for CAD especially useful in younger adults

Predictive Utility of a Coronary Artery Disease Polygenic Risk Score in Primary Prevention

Literature - Marston NA, Pirruccello JP, Melloni GEM, et al. - JAMA Cardiol. 2022 Dec 28. doi: 10.1001/jamacardio.2022.4466

Introduction and methods


Risk factors for coronary artery disease (CAD) can be captured in a single polygenic risk score (PRS). However, clinical implementation of such a CAD PRS is hindered by several barriers, for example, because it is not yet clear in which populations targeted genetic testing will impact management. Furthermore, it is not evident how the CAD PRS can be integrated with established clinical risk factors to help guide therapies.

Aim of the study

The study aim was to investigate the predictive power of the CAD PRS across several age groups and clinical risk and assess the degree to which polygenic risk contributes to the development of MI in younger adults. In addition, the CAD PRS was integrated with an established clinical ASCVD risk score to help guide in which population statin therapy should be initiated for primary prevention.


This was a longitudinal cohort study using data from the UK Biobank. The analysis included 330,201 participants of European ancestries without a history of ASCVD who were not taking lipid-lowering therapy. The authors also performed a replication analysis with data from 104,439 individuals obtained from the BioBank Japan. The CAD PRS was based on 241 conditionally independent genome-wide significant single-nucleotide variants that were recently identified in a recent CAD genome-wide association study [1].


The first endpoint was incident MI, which is a well-validated and the most specific outcome associated with the CAD PRS. This endpoint was used to assess the predictive utility of the CAD PRS across the age spectrum.

The second endpoint was incident ASCVD events, which was defined as a new diagnosis of clinical CAD (including MI), stroke, or CV death. This endpoint was used to analyze the effect of integration of the CAD PRS with a calculated 10-year ASCVD risk score (with the following risk categories: low (<5.0%), borderline (5.0–7.4%), intermediate (7.5–19.0%), and high (≥20.0%)).

Main results

CAD PRS by age

  • During a median follow-up time of 11.2 years (IQR: 9.8–12.6), 4454 individuals had an MI and 19,914 ASCVD events occurred.
  • The CAD PRS was associated with the risk of MI in all age groups but had a significantly stronger association at younger ages (P for interaction<0.001). The HR per 1-SD increase in the CAD PRS was 1.72 (95%CI: 1.56–1.89) for individuals aged <50 years, 1.46 (95%CI: 1.38–1.53) for those aged 50–60 years, and 1.42 (95%CI: 1.37–1.48) for those aged >60 years.
  • Similar results were seen in the BioBank Japan analysis.
  • In the group aged <50 years, individuals with a high PRS (top 20%) had a 3- to 4-fold increased associated risk of MI compared with those in the low PRS category (bottom 20%) (HR at age 44 years: 3.61; 95%CI: 2.71–4.59; P<0.001).
  • A similar pattern was found when the attributable fraction of MI risk due to high CAD PRS was analyzed as a spline function of age at baseline: The proportion of MI risk associated with the CAD PRS decreased from 30% (age ~42 years) to <10% (age ~72 years).

Integration of CAD PRS with 10-year ASCVD risk score

  • When CAD PRS testing was added to the ASCVD risk score, 1154/4817 individuals aged <50 years (24% ) with borderline ASCVD risk and a high CAD PRS had an observed 10-year rate of an ASCVD event of 8%. This reclassified them as having intermediate ASCVD risk, which warranted initiation of statin therapy according to the American Heart Association/American College of Cardiology guideline s.
  • In addition, 20% of the individuals aged<50 years with borderline or intermediate ASCVD risk and a low CAD PRS had an observed 10-year rate of an ASCVD event of <5%, thereby stratifying them as “low risk,” for which statin therapy is not warranted.
  • In the age groups 50–60 years and >60 years, similar trends were seen among participants with low, borderline, or intermediate ASCVD risk.
  • In the entire population, addition of the CAD PRS to the ASCVD risk score resulted in a modest improvement in the Harrell’s C index (from 0.75 (95%CI: 0.75–0.76) to 0.76 (95%CI: 0.75–0.77; P=0.001)). However, among participants aged <50 years with borderline to intermediate ASCVD risk, the addition of the CAD PRS improved the C index more strongly (from 0.61 (95%CI: 0.56–0.66) to 0.71 (95%CI: 0.67–0.75; P<0.001)).


This longitudinal, population-based cohort study showed that using a recently developed PRS for CAD had greatest predictive power in younger adults. Furthermore, individuals with borderline or intermediate clinical ASCVD risk could be risk stratified to help determine who was eligible for statin therapy. The authors believe that their “results suggest that a targeted strategy for CAD PRS testing based on age and clinical risk has the potential for a much higher yield in terms of clinically actionable results and requires significantly less testing compared with an approach that tests everyone.”


1. Aragam KG, Jiang T, Goel A, et al. Discovery and systematic characterization of risk variants and genes for coronary artery disease in over a million participants. Nat Genet. Published online December 6, 2022. doi:10.1038/s41588-022-01233-6

Find this article online at JAMA Cardiol.

Facebook Comments


We’re glad to see you’re enjoying PACE-CME…
but how about a more personalized experience?

Register for free