NEW YORK – Genome-wide polygenic risk scores are more predictive of coronary heart disease in three US racial or ethnic groups than restricted polygenic risk scores, a new study has found.
Scores of genetic loci have been linked to the risk of developing coronary heart disease (CHD), and researchers have used these to generate polygenic risk scores to predict coronary events. But as these scores largely have been developed in cohorts of European ancestry, their utility among other ancestry groups has not been clear.
A Mayo Clinic-led team of researchers sought to examine the associations between different types of polygenic risk scores — restricted PRSs that only include variants that reached genome-wide significance and genome-wide PRSs that have lower thresholds for variant inclusion — and coronary events in individuals of African ancestry, European ancestry, and Hispanic ethnicity. As they reported in the American Journal of Human Genetics Thursday, the researchers found the association of PRSs with coronary events were similar for individuals of European ancestry and Hispanic ethnicity, but less strong among individuals of African ancestry.In addition, genome-wide PRSs performed the best across the three groups.
"These results highlight the potential clinical utility of PRSs for CHD as well as the need to assemble diverse cohorts to generate ancestry-/ethnic-specific PRSs," Mayo's Iftikhar Kullo and his colleagues wrote in their paper.
They assessed how well two restricted and two genome-wide PRSs predicted coronary events using genotyping and electronic health record data gleaned from the EMERGE consortium. This dataset included 45,645 individuals of European ancestry, 7,597 individuals of African ancestry, and 2,493 individuals of Hispanic ethnicity. The PRSs analyzed were all developed in an European ancestry cohort. The restricted PRSs — PRSTikkanen and PRSTada — included 28 and 50 variants, respectively, while the genome-wide PRSs — PRSmetaGRS and PRSLDPred — included 1.7 million and 6.6 million variants, respectively.
Among individuals of European ancestry, the restricted PRSs and genome-wide PRSs were associated with up to 1.2-fold and 1.5-fold increased risk, respectively, of a coronary event per increase in standard deviation. This risk was lower, though, in African-ancestry individuals, as the restricted PRSs and genome-wide PRSs were associated with up to 1.1-fold and 1.3-fold increased risk of a coronary event, while they were associated with up to 1.1-fold and 1.5-fold increased risk, respectively, among Hispanic individuals.
Across all three groups, PRSmetaGRS was the most strongly associated with coronary heart disease.
For the European and African ancestry groups, the researchers estimated individuals' 10-year absolute risk of having a heart attack. When they added the PRSmetaGRS to the 10-year absolute risk for having myocardial infaction, it shifted a number of European and Afrian ancestry individuals who otherwise fell in the intermediate category to other risk categories. About 40 percent of European ancestry individuals were reclassified to lower risk, while nearly a quarter were reclassified to higher risk. Meanwhile, about a quarter of African ancestry individuals were reclassified to a lower risk group and about 20 percent to a higher risk group.
"Our results highlight the potential utility of PRSs for CHD in the clinical setting and suggest that until ancestry-/ethnic-specific PRSs become available, a genome-wide PRS could be considered in [African-ancestry] individuals," the researchers wrote.