Polygenic risk scores combine information from many genetic variants to estimate an individual’s inherited predisposition for conditions such as coronary artery disease or type 2 diabetes. Research by Sanjay V. Khera at Brigham and Women's Hospital and Harvard Medical School showed that polygenic scores can stratify risk in European-ancestry cohorts, but the clinical meaning and stability of those scores vary widely. The question for insurance underwriting is not just technical performance but whether use of these scores aligns with justice, accuracy, and respect for persons.
Predictive validity and equity
Empirical work by Alicia R. Martin at the Broad Institute and colleagues demonstrated that polygenic risk scores developed in European populations have markedly reduced accuracy in people of African, Latin American, and other ancestries. This systematic bias creates a real risk that underwriting driven by these scores would produce disparate impact, worsening existing inequities across racial, ethnic, and territorial lines. Even when predictive in a research cohort, scores often explain only a fraction of disease risk, and environmental and social determinants may be larger drivers for many outcomes.
Ethical, legal, and social consequences
Use of genetic information in insurance raises well-documented concerns about discrimination, privacy, and trust. The Genetic Information Nondiscrimination Act protects against genetic discrimination in health insurance and employment in the United States, but does not cover life, disability, or long-term care insurance, leaving regulatory gaps. The Association of British Insurers negotiated a moratorium limiting use of predictive genetic tests in the United Kingdom, reflecting policy caution. The World Health Organization and bioethicists including Maya Sabatello at Columbia University argue that governance must account for social harms and community perspectives, especially for groups historically marginalized by medical systems.
Applying polygenic risk scores to underwriting could lead to higher premiums, denial of coverage, and deterrence from clinical or research testing that benefits public health. Insurers might claim actuarial justification, yet actuarial use that amplifies social disadvantage is ethically fraught. In territories where genetic databases are sparse or where consent norms differ, implementation could also erode collective trust and exacerbate cross-border inequities.
Given current science and social context, routine use of polygenic risk scores in insurance underwriting is ethically problematic. Responsible pathways require stronger evidence of predictive value across populations, robust anti-discrimination law, transparent governance, and meaningful involvement of affected communities before any limited, closely regulated application would be defensible.