Problems with Using Polygenic Scores to Select Embryos

This article critically examines the use of polygenic scores for embryo selection (ESPS), a service increasingly marketed by companies like Genomic Prediction and Orchid Health. While polygenic scores can predict traits such as educational attainment or disease risk in adults, their application to embryos is fraught with limitations:

• Reduced Predictive Power: Shared parental genetics and gene-environment correlations diminish the accuracy of polygenic scores within families. For example, selecting embryos for higher educational attainment yields an expected gain of only 0.23–0.53 years of schooling, far lower than between-family estimates.

• Ancestry Bias: Polygenic scores derived from European-ancestry populations perform poorly for non-European groups, exacerbating health disparities.

• Unintended Consequences: Pleiotropy may inadvertently select for undesirable traits (e.g., a 16% increased risk of bipolar disorder when selecting for educational attainment).

• Ethical Risks: ESPS could reinforce eugenic practices, devalue certain traits, and widen societal inequalities.

The authors urge transparent communication of absolute (not relative) risk reductions, ancestry specific limitations, and societal implications. They call for FTC oversight, professional guidelines, and public dialogue to address the ethical and regulatory challenges posed by ESPS.

This article critically examines the use of polygenic scores for embryo selection (ESPS), a service increasingly marketed by companies like Genomic Prediction and Orchid Health. While polygenic scores can predict traits such as educational attainment or disease risk in adults, their application to embryos is fraught with limitations:

• Reduced Predictive Power: Shared parental genetics and gene-environment correlations diminish the accuracy of polygenic scores within families. For example, selecting embryos for higher educational attainment yields an expected gain of only 0.23–0.53 years of schooling, far lower than between-family estimates.

• Ancestry Bias: Polygenic scores derived from European-ancestry populations perform poorly for non-European groups, exacerbating health disparities.

• Unintended Consequences: Pleiotropy may inadvertently select for undesirable traits (e.g., a 16% increased risk of bipolar disorder when selecting for educational attainment).

• Ethical Risks: ESPS could reinforce eugenic practices, devalue certain traits, and widen societal inequalities.

The authors urge transparent communication of absolute (not relative) risk reductions, ancestry specific limitations, and societal implications. They call for FTC oversight, professional guidelines, and public dialogue to address the ethical and regulatory challenges posed by ESPS.