Overview #

Polygenic risk scores (PRS), which are also called (PGS) or (GPS), are statistical tools that estimate an individual’s genetic susceptibility to complex diseases or traits by aggregating the effects of numerous genetic variants across the genome. Unlike monogenic diseases, which are caused by variants in a single gene, most common diseases—including cardiovascular disease, diabetes, and many cancers—result from the interaction of multiple genetic and environmental factors. PRS aim to quantify the cumulative contribution of genetic variation to disease risk. 

The development of PRS generally begins with Genome-Wide Association Studies (GWAS), which identify genetic variants associated with specific diseases, traits, or health outcomes across large populations. These associations are then weighted and combined into a score that estimates an individual’s relative genetic risk. As the predictive performance of PRS depends heavily on the size and diversity of the datasets used to develop them, large-scale biobanks have become essential resources. By integrating genomic data with clinical, phenotypic, imaging, environmental, and lifestyle information, biobanks enable the development, validation, and refinement of increasingly sophisticated PRS. 

This data-intensive approach enables expanded PRS scores as the sample sizes and available data increase, but also has the potential to embed structural biases stemming from especially the earlier steps in the production of data, such as recruitment bias and imbalances, population underrepresentation and variation in healthcare access. The growing attention toward AI-enhanced PRS may leverage AI for countering these issues with the promise of improved prediction, diagnosis and prognosis for the global human population thus bearing hopes and opportunities; however, such approaches also bring their own challenges and risks. 

Key ethical, legal and societal implications #

From an ELSI standpoint, PRS development and use intersect directly with debates about fairness, representativity, clinical validity and utility, communication of findings that are probabilistic and relative, and potential for exacerbating health inequities. While laws, regulations and ethical principles specifically on PRS generation is lacking, biobanks are central actors as infrastructures, especially for PRS research, and thus it is important to consider named challenges, especially regarding topics such as recruitment bias, data governance, public trust, consent, return of findings and benefit sharing. Topics such as clinical applications, population-level screening, embryo selection/preimplantation use, commercial direct-to-consumer context, risks of misinterpretation, genetic determinism, genomic privacy as well as potential and current of use AI have been some of the topics that are discussed in the literature (see for more: Fritzsche et al., 2023). 

More in detail, the relevant ELSI issues are the following: 

Population bias and health inequalities  #

Most existing PRS have been developed using datasets derived predominantly from individuals of European ancestry. As a result, their predictive accuracy often decreases when applied to underrepresented populations, potentially exacerbating existing health inequalities. Structural biases arising from recruitment practices, unequal healthcare access, and limited population diversity may therefore be embedded into PRS models and perpetuated in clinical practice. 

Data governance, privacy and biobank stewardship  #

The development of robust PRS requires access to large volumes of genomic and health-related data, making biobanks central to their implementation. The integration of genetic, clinical, environmental, and lifestyle information increases both scientific value and privacy risks. Strong governance frameworks are therefore necessary to ensure secure data management, responsible data sharing, pseudonymisation, and compliance with data protection legislation. 

Artificial Intelligence and algorithmic fairness  #

Artificial intelligence is increasingly used to improve the construction and interpretation of PRS by integrating heterogeneous datasets and identifying complex relationships that conventional statistical methods may overlook. While AI may help mitigate certain biases through improved modelling techniques, it may also introduce new sources of algorithmic bias, reduce transparency, and make clinical decision-making less explainable. Ensuring fairness, accountability, and human oversight remains essential when AI-assisted PRS are used in healthcare. 

Clinical validity and utility  #

Although PRS show considerable promise, their predictive value varies across diseases, populations, and clinical settings. Ethical implementation requires careful evaluation of clinical validity, analytical validity, and clinical utility before PRS are incorporated into healthcare decision-making. Overestimating predictive accuracy may lead to inappropriate interventions, unnecessary anxiety, or false reassurance. 

Informed consent and secondary use of data  #

PRS development often relies on the secondary use of genomic and health data collected through biobanks. Participants may not have anticipated that their data would contribute to predictive algorithms or AI development. Transparent consent procedures and robust governance mechanisms are therefore necessary to ensure that secondary data use remains consistent with participants’ expectations and applicable legal requirements. 

Communication of genetic risk  #

Polygenic risk represents a probabilistic estimate rather than a definitive diagnosis. Communicating PRS results to patients requires careful consideration to avoid misunderstanding, stigmatisation, or psychological harm. Appropriate genetic counselling and clear communication strategies are essential to support informed decision-making and responsible clinical use. 

Equity in precision medicine  #

If implemented responsibly, PRS may improve disease prevention and personalise healthcare. However, unequal access to genomic testing, digital infrastructure, and specialised healthcare services could result in unequal distribution of the benefits of PRS. Promoting diversity in genomic research and ensuring equitable implementation remain important societal priorities. 

BBMRI-ERIC resources  #

• Fritzsche, M.-C., Akyüz, K., Cano Abadía, M., McLennan, S., Marttinen, P., Mayrhofer, M. T., & Buyx, A. M. (2023). Ethical layering in AI-driven polygenic risk scores—New complexities, new challenges. Frontiers in Genetics, 14. https://doi.org/10.3389/fgene.2023.1098439. 

External resources   #

Considering the lack of laws, regulations and specific ethics guidelines regarding PRS generation and use in biobanking, we have compiled below relevant statements and proposals by major professional societies and groups that are relevant for PRS. 

• European Society of Human Genetics (ESHG),Viewpoint: The use of polygenic risk scores in pre-implantation genetic testing: an unproven, unethical practice . Forzano, F., Antonova, O., Clarke, A., de Wert, G., Hentze, S., Jamshidi, Y., Moreau, Y., Perola, M., Prokopenko, I., Read, A., Reymond, A., Stefansdottir, V., van El, C., Genuardi, M., Peterlin, B., Oliveira, C., Writzl, K., Houge, G. D., Cordier, C., . . . Professional Policy Committee of the European Society of Human Genetics (2022). The use of polygenic risk scores in pre-implantation genetic testing: an unproven, unethical practice. European Journal of Human Genetics, 30(5), 493-495. https://doi.org/10.1038/s41431-021-01000-x (see also supporting statement by European Society of Human Reproduction and Embryology (ESHRE): https://www.eshre.eu/Europe/Position-statements/PRS) 

• American College of Medical Genetics and Genomics (ACMG), The clinical application of polygenic risk scores: A points to consider statement . Abu-El-Haija, A., Reddi, H. V., Wand, H., Rose, N. C., Mori, M., Qian, E., & Murray, M. F. (2023). The clinical application of polygenic risk scores: A points to consider statement of the American College of Medical Genetics and Genomics (ACMG). Genetics in Medicine, 25(5). https://doi.org/10.1016/j.gim.2023.100803  

• Human Genetics Society of Australasia, Position Statement: Use of Polygenic Scores in Clinical Practice and Population Health . Young, M.-A., Yanes, T., Cust, A. E., Dunlop, K., Limb, S., Newson, A. J., Purvis, R., Thiyagarajan, L., Scott, R. J., Verma, K., James, P. A., & Steinberg, J. (2023). Human Genetics Society of Australasia Position Statement: Use of Polygenic Scores in Clinical Practice and Population Health. Twin Research and Human Genetics, 26(1), 40-48. https://doi.org/10.1017/thg.2023.10 

• Polygenic Risk Score Task Force of the International Common Disease Alliance, Perspective: Responsible use of polygenic risk scores in the clinic: potential benefits, risks and gaps . Polygenic Risk Score Task Force of the International Common Disease Alliance. (2021). Responsible use of polygenic risk scores in the clinic: potential benefits, risks and gaps. Nature Medicine, 27(11), 1876-1884. https://doi.org/10.1038/s41591-021-01549-6  

• Clinical Genome Resource (ClinGen) Complex Disease Working Group and the Polygenic Score (PGS) Catalog, Polygenic Risk Score Reporting Standards (PRS-RS). Wand, H., Lambert, S. A., Tamburro, C., Iacocca, M. A., O’Sullivan, J. W., Sillari, C., Kullo, I. J., Rowley, R., Dron, J. S., Brockman, D., Venner, E., McCarthy, M. I., Antoniou, A. C., Easton, D. F., Hegele, R. A., Khera, A. V., Chatterjee, N., Kooperberg, C., Edwards, K., . . . Wojcik, G. L. (2021). Improving reporting standards for polygenic scores in risk prediction studies. Nature, 591(7849), 211-219. https://doi.org/10.1038/s41586-021-03243-6  

• National Society of Genetic Counselors , Clinical genetic counseling and translation considerations for polygenic scores in personalized risk assessments: A Practice Resource . Wand, H., Kalia, S. S., Helm, B. M., Suckiel, S. A., Brockman, D., Vriesen, N., Goudar, R. K., Austin, J., & Yanes, T. (2023). Clinical genetic counseling and translation considerations for polygenic scores in personalized risk assessments: A Practice Resource from the National Society of Genetic Counselors. Journal of Genetic Counseling, 32(3), 558-575. https://doi.org/https://doi.org/10.1002/jgc4.1668 

Relevant literature:  #

• Lewis CM, Vassos E. Polygenic risk scores: from research tools to clinical instruments. Genome Med. 2020 May 18;12(1):44. doi: 10.1186/s13073-020-00742-5. PMID: 32423490; PMCID: PMC7236300. 
• Torkamani A, Wineinger NE, Topol EJ. The personal and clinical utility of polygenic risk scores. Nat Rev Genet. 2018 Sep;19(9):581-590. doi: 10.1038/s41576-018-0018-x. PMID: 29789686. 
• Wand H, Lambert SA, Tamburro C, Iacocca MA, O’Sullivan JW, Sillari C, Kullo IJ, Rowley R, Dron JS, Brockman D, Venner E, McCarthy MI, Antoniou AC, Easton DF, Hegele RA, Khera AV, Chatterjee N, Kooperberg C, Edwards K, Vlessis K, Kinnear K, Danesh JN, Parkinson H, Ramos EM, Roberts MC, Ormond KE, Khoury MJ, Janssens ACJW, Goddard KAB, Kraft P, MacArthur JAL, Inouye M, Wojcik GL. Improving reporting standards for polygenic scores in risk prediction studies. Nature. 2021 Mar;591(7849):211-219. doi: 10.1038/s41586-021-03243-6. Epub 2021 Mar 10. PMID: 33692554; PMCID: PMC8609771. 

Acknowledgements:  #

This Knowledge Base entry benefited from funding from INTERVENE (INTERnational consortium for integratiVE geNomics prEdiction), a project that has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No. 101016775.