Genetic selection, polygenic scores, and the parent's choice

The traits most often marketed for polygenic selection—cognitive ability, depression risk, ADHD susceptibility—reflect neurodevelopmental cascades involving thousands of genes acting through synaptic density, myelination timing, neurotransmitter receptor distribution, and prefrontal-limbic connectivity. No single common variant moves any of these mechanisms appreciably; the effect sizes detected in GWAS rarely exceed 0.1 percent of variance per SNP. The biological reality is that a polygenic score is an aggregation of vanishingly small individual signals across loci whose mechanistic role is, in most cases, unknown. The brain that emerges from a given genome is also shaped by maternal nutrition, prenatal stress hormone exposure, the microbiome, and postnatal environment, each interacting with the genome in ways that the linear additive model of polygenic scoring does not capture. Selecting an embryo for a higher cognitive score is, in neurobiological terms, selecting for a marginal shift in the probability distribution of a developmental trajectory that has not yet been run. The substrate is real; the precision claimed is theatrical.

Parents engaging with embryo reports report a recognizable pattern: initial discomfort, followed by a sense of obligation. Once a number is on the page, the cognitive load of ignoring it exceeds the load of acting on it. This is the well-documented anchoring effect operating in a context where the anchor carries the weight of a child's future. The psychological cost shows up later. Parents who selected on cognitive scores show elevated rates of disappointment when their child's school performance is ordinary, because they have implicitly internalized the score as a promise. Parents who declined selection but were offered it sometimes describe a different burden: the suspicion that they should have. Neither group is comfortable, which suggests the technology is producing a new category of parental anxiety rather than relieving an old one. The clinic's role in framing—whether the report is presented as medical, as informational, or as a recommendation—substantially determines which psychological track the parent enters.

A child selected on polygenic scores develops the same way any child does: through the interaction of genome, environment, and chance. The selected variants are present from conception, but their expression unfolds across decades and is heavily mediated by experience. A child with a high educational attainment polygenic score raised in a household with no books will not become the predicted scholar. A child with an average score raised in a verbally rich, stable, well-resourced environment will often outperform the prediction. The developmental literature is consistent on this: heritability estimates rise across the lifespan as people increasingly select environments matching their dispositions, but the variance explained by any current polygenic score remains modest at every age. What does change developmentally is the parent's interpretation. The score, once known, becomes a lens through which ordinary developmental variation is read as confirmation or betrayal.

The cultural reception of embryo selection splits along familiar lines. In Silicon Valley discourse, it is framed as the rational extension of parental investment—a natural step from prenatal vitamins to genetic optimization. In bioethics journals, it is framed as a renewed eugenic temptation requiring institutional restraint. In disability communities, it is read as the latest statement that certain lives are pre-judged not worth living. In evangelical and Catholic frameworks, it is rejected at the prior step of IVF itself. These framings are not equally weighted in the policy conversation; the optimization frame has the venture capital, the regulatory frame has the journals, and the disability and religious frames have the smallest seat at the table despite representing the largest constituencies affected. Cultural acceptance is therefore being manufactured in the gap between what is technically offered and what is publicly debated.

The defensible practical use of polygenic screening is in single-gene and chromosomal conditions where the predictive accuracy approaches certainty—cystic fibrosis, Tay-Sachs, Huntington's, aneuploidies. Here the technology extends existing preimplantation genetic diagnosis and the ethical conversation is decades old. The contested application is in common, complex, polygenic conditions—type 2 diabetes, coronary disease, schizophrenia—where a high-risk embryo may still never develop the condition and a low-risk embryo may still develop it. The most contested application is in non-medical traits. Practically, families considering selection should ask three questions: what is the absolute risk reduction (usually small), what is the predictive accuracy in their ancestry group (often poor), and what trait substitution is happening when one embryo is chosen over another (each embryo is a different potential person, not an upgraded version of the same one).

Selection alters the relational fabric of family before the family exists. The unchosen embryos—stored, discarded, or donated—are not nothing in the moral imagination of the parents who chose among them. The chosen child carries, often unspoken, the knowledge of having been picked. Siblings conceived through different cycles may have different selection profiles, introducing a new kind of hierarchy into family life. Extended family members react variably: grandparents often resist, framing the practice as an insult to the family's natural inheritance; in-laws may push for it as a marker of seriousness about the next generation. The reproductive partner relationship is itself stressed by the choice, particularly when the partners weight medical and non-medical traits differently. Clinics rarely offer counseling that addresses these relational layers because the appointment is structured around the embryo report, not the family system receiving it.

The philosophical question underneath selection is what parents owe to children who do not yet exist. Procreative beneficence, as Julian Savulescu has formulated it, holds that parents have a moral reason to select the child expected to have the best life. Michael Sandel has argued the opposite: that the disposition to optimize children corrodes the unconditional love that makes parenthood a moral good rather than a project. Adrienne Asch and Erik Parens have developed the expressivist objection—that selection against a trait communicates a judgment about people living with that trait. None of these positions is fully consistent with everyday parental practice, which combines selection (of schools, partners for children to meet, neighborhoods) with acceptance (of the child who arrives). Polygenic screening forces the contradiction to the surface at the moment of conception, which is why it feels different even when its consequences may be similar to choices made later.

The eugenic movements of the early twentieth century are the inescapable backdrop. State-mandated sterilization in the United States, Nazi racial hygiene laws, and decades of forced sterilization of indigenous, Black, and disabled women established that population-level genetic intervention is not a hypothetical evil. Polygenic embryo selection is structurally different—it is voluntary, individual, and operates on embryos rather than living people—but the surface differences should not obscure the continuity in logic: that some genomes are preferable to others and that the preference can be acted on. Dorothy Roberts has documented how reproductive technologies have consistently been deployed in ways that track existing racial hierarchies. The historical record suggests that the question is not whether the technology will be misused but who will be insulated from its misuse and who will not.

Polygenic selection sits at the intersection of IVF access, insurance coverage, ancestry-specific predictive validity, and national regulatory posture. IVF in the U.S. costs roughly fifteen to thirty thousand dollars per cycle, largely out of pocket, which restricts the practice to higher-income families. Polygenic scores trained predominantly on European-ancestry data perform worse in other groups, meaning the technology is both economically and statistically stratified. National regulation varies: Germany prohibits most embryo selection; the U.K. permits it only for approved conditions; the U.S. has no federal framework; some jurisdictions in Asia move faster than their public conversations. These contextual variables are not background—they are the actual mechanism by which a globally available technology produces locally divergent outcomes.

The technology integrates with several existing systems: the IVF clinic infrastructure, the consumer genomics market, the insurance reimbursement architecture, and the medical research enterprise that produces GWAS data in the first place. Each integration creates feedback. Selection generates new cohorts whose outcomes will, in principle, refine the scores; consumer demand drives clinic adoption; clinic adoption drives data accumulation; data accumulation drives further marketing. Without intentional friction, the system runs toward expansion regardless of whether the science supports it. The friction points available—professional society guidelines, insurance coverage decisions, mandated counseling, ancestry-specific accuracy disclosures—exist but are inconsistently applied. The default trajectory is integration without deliberation.

Synthesizing across the substrate, the psychology, the development, the culture, the practice, and the system, what emerges is a technology whose predictive power is modest, whose social consequences are significant, whose distribution is unequal, and whose regulatory framework is underdeveloped. The collective question is not whether the technology will continue to advance—it will—but whether the public conversation will catch up to the choices being offered in clinics today. The sixth law is the operative one: the framework must revise faster than the practice, or the practice will write the framework by default. Parents making individual choices are not the appropriate site of resolution. The site is professional standards, public funding decisions, and the substantive debate about what a society wants the threshold of a human life to be.

Within ten years, polygenic scores will improve, ancestry coverage will expand, and the menu of traits offered will grow. Within twenty, the combination of in vitro gametogenesis and polygenic selection may allow hundreds of embryos to be ranked per cycle, dramatically increasing the marginal gain from selection. Within thirty, the children of the current cohort will be adults, and the empirical question of whether selection actually delivered on its promises will have an answer. The collective preparation work that is not happening now—public deliberation, equitable access, regulatory architecture, support for the children who arrive with or without selection—will determine whether the answer is a story of careful expansion or another chapter in the history of reproductive technologies that ran ahead of the societies deploying them.

1. Greely, Henry T. The End of Sex and the Future of Human Reproduction. Cambridge, MA: Harvard University Press, 2016. 2. Harden, Kathryn Paige. The Genetic Lottery: Why DNA Matters for Social Equality. Princeton: Princeton University Press, 2021. 3. Plomin, Robert. Blueprint: How DNA Makes Us Who We Are. Cambridge, MA: MIT Press, 2018. 4. Sandel, Michael J. The Case Against Perfection: Ethics in the Age of Genetic Engineering. Cambridge, MA: Harvard University Press, 2007. 5. Savulescu, Julian. "Procreative Beneficence: Why We Should Select the Best Children." Bioethics 15, no. 5-6 (2001): 413-426. 6. Parens, Erik, and Adrienne Asch, eds. Prenatal Testing and Disability Rights. Washington, DC: Georgetown University Press, 2000. 7. Roberts, Dorothy. Killing the Black Body: Race, Reproduction, and the Meaning of Liberty. New York: Pantheon, 1997. 8. Glannon, Walter. Genes and Future People: Philosophical Issues in Human Genetics. Boulder: Westview Press, 2001. 9. Turley, Patrick, et al. "Problems with Using Polygenic Scores to Select Embryos." New England Journal of Medicine 385, no. 1 (2021): 78-86. 10. Lencz, Todd, et al. "Utility of Polygenic Embryo Screening for Disease Depends on the Selection Strategy." eLife 10 (2021): e64716. 11. Karavani, Ehud, et al. "Screening Human Embryos for Polygenic Traits Has Limited Utility." Cell 179, no. 6 (2019): 1424-1435. 12. Martin, Alicia R., et al. "Clinical Use of Current Polygenic Risk Scores May Exacerbate Health Disparities." Nature Genetics 51, no. 4 (2019): 584-591.