«««< HEAD In the 2023 January issue of Nature Human Behavior, we have a new paper on indirect genetic effects. The main result is that we cannot see much of an effect of within-nuclear-family “genetic nurture”. Instead, indirect effects arise for other, dynastic reasons: effects of genetics in the wider extended family network, sorting of families or assortative mating. ======= In the 2023 January issue of Nature Human Behavior, we have a new paper on indirect genetic effects. The main result is that we cannot see much of an effect of within-nuclear-family “genetic nurture”. Instead, indirect effects arise for other, dynastic reasons: effects of genetics in the wider extended family network, sorting of families or assortative mating.

bec2a50e6742af61c0ada8e70b19665febac5609

In the review process, we were asked to provide a FAQ in order to avoid misunderstandings of the results The FAQ is included in the paper’s supplement, but also posted here.

Frequently Asked Questions

«««< HEAD

Contents

What did this study do? What is a polygenic index and how is it calculated? What are “indirect” genetic effects? Who are the people in this study? What do the findings mean for the effects of parents’ nurturing behavior on children? What do the findings mean for the research on education and educational performance?

=======

bec2a50e6742af61c0ada8e70b19665febac5609

What did this study do?

This study examined how genes and environments were related to educational outcomes in Norwegian children, who completed standardized academic tests (“Nasjonale prøver”) of reading comprehension, mathematics and English (as a second language) at ages 10,15 and 16. For every children we calculated something known as a polygenic index, which is a measure of genes thought to be correlated with going further in formal education. In addition, we also calculated the polygenic index of the children’s parents and of any of their parents’ siblings. In our work, the correlations between genes and outcome, education in this case, aren’t thought of as immutable or fixed, but rather a function of society, environment and population. If a society shifts its emphasis on what is thought of as important to attain education, then that will subtly shift which genes correlate to educational outcomes.

Our study tested three things. First, does a child’s own genes predict their own academic test performance? Second, does a child’s parents’ genes predict their test performance, above and beyond the child’s own genes? Such an association (termed indirect genetic effects) cannot be due to genetic inheritance and is evidence for an environmental process. Third, does a child’s parents’ genes predict their test performance, even after controlling for the parents’ siblings’ genes? This type of analysis is unusual because most studies don’t have data on extended pedigrees of relatives. This analysis is important because it tests whether the environmental processes that are “tagged” by polygenic indices, and are associated with children’s educational outcomes, are operating within nuclear families (parents and children) or are operating multi-generationally.

What is a polygenic index and how is it calculated?

As a genetic effect the study considers a polygenic index, which is a weighted sum of your genotypes, weighted by their suspected effect on education, based on previous genetic studies. Since these previous genetic studies are large, a polygenic index of education explains about ±7-13% of variance in educational outcomes, though that includes both contributions from direct and indirect genetic effects.

What are “indirect” genetic effects?

Indirect genetic effects – sometimes called genetic nurture – is a term for the presence of an association between your parents’ genotypes, and your outcome, over and above the effect the genotype has through its presence in you, as you have inherited genes from your parents. If ignored in the analysis of genetic effect, or environmental effects for that matter, indirect genetic effects might appear like direct genetic effects and inflate our estimate of those. In particular when it comes to educational outcomes, genetic effects and environmental effects seem steeply correlated, which has inflated the estimates of genetic effects on educational outcomes, but also raises the question what causes these indirect genetic effects to occur? There are a few processes researchers think might give rise to indirect genetic effects on educational outcomes. For example genetic variants that every so slightly improve you parents socio-economic position (for example, trough effects on their education), will become correlated to your education if their socio-economic advantage puts you in a better school/neighborhood, or if it means they have more time to be available to help you learn, resulting in better educational outcomes. However, there are also competing explanations where the indirect genetic effects don’t really act through your parents in some way shaping your environment. If for example people with similar education levels consistently (for a number of generations) marry and have kids, this could also give rise to indirect genetic effects, without any effect through nurture or socio-economic advantage on part of your parents. We would like to find out whether these indirect genetic effects, of your parents’ polygenic index for education on your school outcomes, reflect processes that plays out within the nuclear family, or are consequences of some of these multi-generational mate choice processes.

Who are the people in this study?

The research used data from Norwegian Mother, Father and Child Cohort Study, a large cohort data set collected in Norway. MoBa is a unique study where over 90,000 pregnant women were recruited from 1998 to 2008. More than 70,000 fathers have participated. It provides data on children, mothers and fathers from questionnaires, biobanks and administrative registers. For more information about MoBa, see its homepage.

Norway, the country in which data were collected, is a Scandinavian welfare state with a universal, free education system. The Norwegian state also provides a range of social insurance and benefits to the population. The results of the research must be interpreted in this context. Results might not be similar in other countries that have different economies and educational systems.

What do the findings mean for the effects of parents’ nurturing behavior on children?

Our findings related parents’ polygenic indices for education to their child’s educational achievement. Obviously, your nurturing behavior is more than your propensity for having a long education. Any nurture processes that do not correlate to the education PGI, of which there are probably many, can obviously influence kids’ success, and other outcomes in their lives. As so often, our findings are narrow and specific, and don’t easily translate to broad lay concepts like “nurture” or “parenting”. So our findings are not a test of the importance of nurture, rather they are a test of the particular indirect genetic effect researchers have previously observed.

What do the findings mean for the research on education and educational performance?

Our work, and that of many others, highlights that there are many sociological and historical reasons why your genotype can be correlated with any outcome. In particular, when that outcome is as socially determined as education, this inevitably leads to gene-environment correlation. Our work suggests the source of gene-environment correlation is likely predominantly due to a longer term process. One such process is assortative mating, the phenomenon where partners – and thus parents – are matched on their education level or other traits that are partially shaped by genetics. It is then also important to understand that assortment itself is a very complex sociological and psychological process, profoundly shaped by social and cultural factors. We can’t fully rule out a modest effect of your parents’ educational polygenic index on your success in school, nor can we rule out other multi-generation processes that would slowly introduce gene-environment correlation.

---