Skip to main content
The 5th ELSI Congress - ELSIcon2022

ELSIcon2022 Late-Breaking: Laypeople overestimate the predictive power of polygenic scores but do not view them as any more anxiety-producing than other scores for the same trait


ELSIcon2022 • Networking Session • Late-breaking Abstracts • May 31, 2022

Michelle Meyer

Several companies return polygenic scores (PGSs) for medical and nonmedical traits. There is no consensus about how best to present PGS results and almost no research on how well laypeople understand them. In Study 1, we conducted a survey experiment with 999 online panel participants who were asked about hypothetical, but typical, PGS reports. Participants were randomized to answer questions about PGS reports: for educational attainment, intelligence, subjective well-being, height, or coronary artery disease; that reported that an individual was at the 20th or 80th percentile of the distribution; and that were marked with a “speculative,” “fair,” or “high” confidence level. Seventy percent of participants misinterpreted the percentile as reflecting the individual’s place in the general population rather than in a selected sample of platform users, as the reports explicitly noted, and 48% mistakenly confused actual for predicted phenotype. In estimating a PGS’s predictive power, although 70% participants were appropriately sensitive to the confidence label, across all conditions, participants substantially overestimated the predictive power of PGSs. Nonetheless, participants did not consider PGSs any more anxiety-inducing or accurate than other standard or hypothetical scores (e.g., standardized tests) to predict the same traits, and estimated that any negative affect would significantly diminish by six months. Study 2 will test reactions to 5th and 95th percentiles; estimated predictive power of a PGS without confidence label; and whether participants appreciate the reduced cross-ancestry predictive power of PGSs. Our results can inform decisions about whether and how to present PGSs in direct-to-consumer and clinical settings.


Networking Session
Late-breaking Abstracts
Screening for polygenic conditions and risk scores
Return of results and incidental findings

Videos in Series

Related Videos