Polygenic risk scores (PGSs) are used to predict an individual’s risk for particular conditions using data from multiple genes. Although the scores may be able to provide valuable insight into disorders, such as type 2 diabetes risk, without a health care professional, they can be misused. In a new article, researchers from the Carl R. Woese Institute for Genomic Biology and Harvard Medical School argue for increased regulation of genetic tests by the US Food and Drug Administration, specifically in regard to PGSs.

In more traditional direct-to-consumer genetic tests, like 23andMe, consumers provide their genetic samples, which are sequenced by a company. The information returned to consumers are based on data from individual genes. “These tests tell you about a range of traits from ancestry to whether you have mutations in a specific gene that is related to breast cancer, for example,” said Jacob S. Sherkow (GSP), a professor of law. “There is a clear line from which specific gene is being looked at and, in some cases, what you can do about it when you get a result. It is similar to going to an actual health care provider.”

Historically, such tests have been subject to oversight by the FDA. The agency has not, however, begun to regulate direct-to-consumer PGSs. “Instead of focusing on a single gene, these tests focus on a number of them, and as a result the line from the outcome of the tests to the clinical condition is unclear,” Sherkow said.

For most diseases, a variety of genes contribute to disease risk. PGSs combine the different variations of genes a person has and, using an algorithm, calculates a “risk score” of developing a particular condition, such as alcoholism. Because these algorithms are often opaque, consumers receive little information about how their scores might be calculated.

“The first problem is that it is not always transparent how companies are calculating these scores. They claim to have a proprietary algorithm, which, in reality, is a total black box,” Sherkow said. “If they are not completely accurate, consumers may make adverse health choices on the basis of misinformation.”

An example of such a problem, as outlined by Sherkow, is if a consumer is told that their body reacts to medication differently from the general population. Based on that PGS, some may choose to adjust their dosage without asking a healthcare provider, which could be dangerous. Another example is if consumers are told that they are at diminished risk for heart disease. Such a claim could be misleading without the intervention of a healthcare provider who is likely to know more about a patient’s lifestyle choices and environmental factors that may contribute to disease risk.

“The second problem with PGSs is that it is unclear which agency should regulate it. There has been some back-and-forth within the federal government about whether the FDA has the power to regulate all laboratory-developed tests,” Sherkow said. “While the FDA can regulate some tests that are run by clinical labs, many PGSs are different. They give you a risk score, which many providers claim is not a diagnosis of any disorder.”

To make matters worse, some companies that provide PGSs directly to consumers bypass the most stringent regulatory requirements by marketing their services as general wellness products, an area the FDA does not typically police. “We believe that it is wrong because they are actually providing consumers with health information, and the risk of misusing it is harmful,” Sherkow said.

The authors have two recommendations for how to regulate PGSs. First, the FDA needs to reevaluate their regulatory guidelines and police PGSs the same way they regulate direct-to-consumer tests. Second, Congress needs to clearly delineate what aspects of genetic testing should be regulated by the FDA.

“Every time there is a new administration in the White House, the guidelines shift. We should not have to rely on the FDA to continuously update its guidelines in order to stay ahead of the technology,” Sherkow said.

PGSs are neither the first nor the last to challenge the FDA’s authority. However, as genomic sequencing and its applications improves and diversifies, the regulatory approaches need to keep pace in order to safeguard public health.

The paper “Regulating Direct-to-Consumer Polygenic Risk Scores” was published in the Journal of the American Medical Association and can be found at 10.1001/jama.2023.12262. The work was carried out with contributions from Jin K. Park, an MD student at Harvard Medical School and Christine Y. Lu, a professor of pharmacy at the University of Sydney and a visiting professor of population medicine at Harvard Medical School.