Abstract
Polygenic risk scores (PRS) for different diseases are expected to become more widely available to the public in the coming decades. In addition to the investigation of the clinical relevance of polygenic risk scores, an assessment of the health behavioral impact is needed. The present study used data from a personalized medicine project that combined genomic and traditional health data to evaluate respondents' risk for common diseases. Specifically, we investigated if supplementing traditional risk estimates of type 2 diabetes and coronary heart disease with PRS influenced respondents' self-reported physical activity, alcohol consumption, fruit/vegetable consumption or prompted the respondents to seek medical treatment/examination. As an exploratory hypothesis, we also tested if there was an interaction between the disease risk level and the experimental/control group for any of the outcomes. A randomized controlled trial was conducted, where the experimental group (n = 216 for seeking treatment and 523-459 for other outcomes) received risk estimates based on traditional risk and PRS, and the control group (n = 216 and 526-498) based solely on traditional risk factors. On average, approximately 80 days elapsed between the risk disclosure and outcome measurements. We found no significant difference between the groups regarding health behavior (ps > .28, ds < 0.07) or likelihood of seeking medical treatment/examination (p = .86, OR = 1.06). Likewise, no significant interactions were detected (ps > .08, ds < .11, ORs < 1.2). We conclude that we did not find support for either a beneficial or detrimental effect of supplementing traditional risk estimates with PRSs. However, several limitations should be noted when generalizing the results.