Sociogenomics and Polygenic Scores

^{This article was originally published as “Soziogenomik und polygene Scores” in issue 272 (February 2025) of the German-language journal Gen-ethischer Informationsdienst (GID); translated by the authors.}

In mid-November 2024, the British organization Hope not Hate published its investigative research ‘Inside the Eugenics Revival’. In addition to documentating an active international “race research” network, the investigation also brought to light the existence of a US start-up that offers eugenic embryo selection. Heliospect Genomics aims to enable wealthy couples to select the “most intelligent” embryos for in-vitro fertilization (IVF) using the latest genomic technology: so-called polygenic (risk) scores (PGS). However, experts strongly doubt the validity of predictions based on these scores and criticize them as ethically questionable.

Belief in the power of genes

Heliospect Genomics is operating in a legal and ethical grey area with its dubious offer. While probably legal in the U.S., embryo selection based on an expected IQ would currently be prohibited In many countries, including Germany and Great Britain. Furthermore, it is extremely doubtful that the IQ predictions made for embryos would have any validity at all. They are based on dubious statistical correlations and cannot yield reliable statements about the future development of individuals.

Nevertheless, the claims in this and similar offers build on a new model of genetic and genomic research that has gone relatively unnoticed by the critical public. Called sociogenomics, social science genetics, and social or behavioral genomics, this field recycles a well-worn claim with a new scientific-technological basis: namely, to be able to precisely calculate the heredity of all possible social characteristics. This includes behaviors such as intelligence, educational attainment, addiction behavior, divorce, church attendance and more.

The sociogenomics model has been enabled by new technological developments, in particular the ever faster and more cost-effective sequencing of complete genomes paired with the capacity to process ever larger amounts of data. It follows on the failure of the ‘one gene-one trait’ model, which searched for one or very few genes that were thought to be the direct cause of certain human characteristics (e.g., ‘the gene for homosexuality’). Sociogenomics researchers are trying to take advantage of the fact that individual human genomes—although 99.9% identical—contain an average of around three million minimal variations of individual DNA building blocks. Researchers now seeks statistical correlations between these variations and differences in the aforementioned social characteristics and behaviors. However, relevant statistical correlations can only emerge when hundreds of thousands, or even millions, of such genetic variants are aggregated into polygenic scores.

Yet, these scores usually correlate with only a very small part of the observed variation in the examined traits: in the case of educational attainment, for example, no more than 10% of the differences in characteristics can be “explained” by genetics, while for risk tolerance, this figure drops to just 2%. It is important to note that these are purely statistical correlations and by no means causal biological relationships, even if sociogenomics researchers repeatedly try to blur this distinction. It is also unclear whether the correlations found might indicate structures of social inequality rather than genetic influences. Despite the modest results so far, many in the field suggest that genes are the most important determining factor in people’s lives. Sociogenomics sounds a lot like genetic determinism, although proponents of the field try to distance themselves from this history.

Prediction instead of scientific explanation

PGS are statistical tools that cannot reveal anything about causal relationships or interactions between genes and the environment. For this reason, sociogenomicists emphasize that PGS provides a prediction of supposedly genetically determined individual or social risks. A programmatic essay by the US psychologist and geneticist Robert Plomin and the psychologist Sophie von Stumm shows how this condenses into ideas of risk prevention. They argue that the “predictive power” of PGS does not depend on understanding the genetic processes that influence the behavior of interest. Rather, it is crucial to improve the predictive power of polygenic scores so they can be used as an “early warning system” to prevent problems. According to the authors, it is essential to be able to use PGS to identify ”individuals at risk,” i.e., people with an increased risk of problematic characteristics and behaviors.

Not all proponents of sociogenomics would express this as openly as Plomin and von Stumm do. But as the central analytical tool of sociogenomics, polygenic scores are treated primarily as a forecasting instrument that uses genomic data to predict individual differences, for example in educational attainment. This vision of risk avoidance and prevention, however, remains illusory, especially since PGS do not provide any knowledge about the cause of, for example, low school achievements.

It is therefore hardly a coincidence that the promise of risk avoidance is shifting to focus on prenatal prevention and selection, as in the example of Heliospect Genomics. Predicting the future IQ of an embryo is completely unfounded in scientific terms; but it does suggest tangible routes for action and intervention, and a market for polygenic embryo screening is apparently already emerging.

Containing the dangers

The promise of being able to predict the intelligence and other characteristics of embryos based on genomic data—and thus select the “best” possible future child—dramatically highlights the dangers of using PGS. Even if such promises are unlikely ever to be fully realized, they can still have profound social effects that are difficult to control. Katie Hasson, associate director of the Center for Genetics and Society in California, warns that selling PGS-based tests “normalizes this idea of ‘superior’ and ‘inferior’ genetics.” The sociogenomic claims of predicting social and psychological characteristics of babies, children, and adults using genome analyses not only revive a form of genetic determinism but also risk reinforcing the stigmatization of those deemed “genetically inferior.”

The resulting pressure for commercialization and technological “optimization” of reproduction, as well as the heightened expectations of parents and institutions in the social and educational sectors, massively call into question the fundamental values and legal norms of democratic societies, such as the equality of all people and the recognition of human diversity. Given these concerns, an urgent social and scientific debate is needed on the regulation or ban of commercial services like those offered by Heliospect Genomics. Under certain circumstances, such as when it concerns sensitive social characteristics like intelligence and educational attainment, a moratorium may be necessary even on purportedly neutral basic PGS research.

Peter Wehling is a sociologist at Goethe University Frankfurt. Tino Plümecke is a sociologist and science studies scholar at the University of Freiburg. Isabelle Bartram is a molecular biologist and program director of Gen-ethisches Netzwerk.