A team at Penn Medicine looked at data from nearly 48,000 adults to see how well genomic risk scores predict health problems. They used a tool called GIRA that looks at both single gene changes and many small genetic variations together. This approach helps spot people who might get sick before symptoms appear. The study found that 30.4 percent of the participants were labeled as high-risk for at least one of nine conditions. That means more than three out of every ten adults in this group carried a genetic signal for future illness. The researchers also checked if these scores worked equally well for everyone. They found that people of African or African American ancestry had higher rates of being labeled high-risk compared to East Asian or South Asian groups. This difference suggests the tool might need adjustments to work fairly for all communities. The scores also predicted who already had the disease and who would develop it later. However, the prediction was not perfect for every single condition. Some groups had lower accuracy rates. The study did not report any safety issues because it used existing health records rather than giving new drugs or treatments. If health systems use these scores widely, they could help doctors focus care on those most likely to need it. But the team warns that the tool must be tested carefully across different populations to ensure it does not miss or overestimate risk for specific groups.
Review evaluates utility of GIRA high-risk genomic criteria across 9 adult conditions in 48279 patientsGenomic risk scores identify 30 percent of adults as high-risk for nine common conditions
AI-generated summary of the cited source, checked by automated accuracy review. How we work
This retrospective evaluation assesses the utility of GIRA high-risk criteria, which incorporate monogenic and polygenic genomic components, for assessing health risk across 9 adult conditions. The study utilized data from the Penn Medicine Biobank, a health system independent of eMERGE, involving a sample size of 48279 adults. The primary outcome measured the utility of these criteria, while secondary outcomes included stratifying by ancestry, enrichments of high-risk individuals in social deprivation index, and prediction of prevalent and incident disease.
The analysis found that 30.4% of patients were labeled as high-risk, representing 14676 absolute numbers. High-risk classification rates were higher in African or African American populations at 56.6% versus 50.1%, and lower in East and South Asian ancestries at 42.0% and 40.0% respectively. A p-value of 7.43x10-36 was reported for these differences. Enrichments of high-risk individuals were observed in the highest quartile of social deprivation index.
The polygenic component served as a good predictor of prevalent cases. However, the GIRA prediction of incident disease showed lower accuracies for some conditions. Demographic composition of high-risk atrial fibrillation individuals was enriched for European ancestries, whereas incident AFIB cases were enriched for AFR ancestries. Safety data, including adverse events and tolerability, were not reported. Funding or conflicts were not reported.
The authors suggest potential impact on the health system if these criteria are implemented at scale. The study does not establish causality. Limitations regarding the retrospective nature and the specific setting are inherent to the design. Practice relevance is tempered by the lower accuracies for incident disease prediction in certain contexts.
View Original Abstract ↓
Genotype-Guided SSRI Prescribing Shows No Short-Term Benefit but Higher Remission at 6 Months
Genetic Testing Helps Some Depression Patients Feel Better Long Term
Related in Genetics & Precision Medicine
From Other Specialties
