Neither nature nor nurture: Using extended pedigree data to understand indirect genetic effects on offspring educational outcomes

Abstract

Background: More highly educated parents tend to raise children who go on to complete more education themselves. Strong evidence for environmental transmission arises from the fact that offspring outcomes correlate to parental genotypes even when controlling for the offspring genotype. The process that gives rise to an environmentally mediated correlation between parental genotype and offspring education (“indirect genetic effects”) remains poorly understood. A key question is whether intergenerational transmission reflects within-family processes, such as parenting behaviors (i.e. “Genetic nurture”), or social inheritance that could reflect either genetic ancestry or social stratification (i.e. “Dynastic effects”). Methods: We analyzed data from N = 25,215 genotyped parent-offspring trios participating in the Norwegian Mother, Father, and Child Cohort Study (MoBa), where many of the participants in the parental generation are siblings (N = 3,500 sibling pairs). We correlate genetic differences within the parental family, conditional on offspring genetics to offspring’s educational outcomes. The strategy isolates indirect genetic effects that play out in the nuclear family from indirect genetic effects that arise through social or ancestral genetic stratification. We additionally partition genetic variants associated with EA into those associated with cognitive versus non-cognitive skills. Results: We find that children’s genetics, measured using polygenic indices (PGIs) for cognitive and non-cognitive components of EA, are associated with their educational performance. Parents’ genetics are also associated with their children’s educational performance, over and above the child’s own genetics. However, we find no evidence that parents’ PGI are specifically related to offspring academic achievement over and above the average PGI of the siblings in the parental generation. Conclusions: Our result suggests that the effects of the environmental processes captured by genome-wide association studies of EA and characterized as “nurture” are explained less by parents’ specific behaviors and more by dynastic stratification in environments relevant to success in school.

Conventional twin studies overestimate the environmental differences between families relevant to educational attainment

“…49,63 Likewise, the absence of passive rGE in our PA models appears to contradict recent molecular genetic evidence that suggests EA polygenic scores partly capture the effects of passive rGE. 29,30,[50][51][52][53] Here we stress that, while our bestfitting model was a PA model, our SH models did not fit the data significantly worse than our PA models. It's also possible that a mixed homogamy scenario, in which phenotypic assortment and social homogamy both play a role, might explain the data better than the PA and SH models compared in this study.…”

“…However, we also note that phenotypic assortment is expected to produce some of the molecular genetic effects that have been interpreted as passive rGE or "genetic nurture". 29,50,64 We briefly describe how phenotypic assortment can induce these effects in Supplementary Note 3.…”

“…26 However, a large Australian study found that spousal convergence played a negligible role in partner similarity for EA, 27 and recent molecular genetic studies have found strong evidence for phenotypic assortment on EA and/or associated traits. [28][29][30][31][32][33] A recent Norwegian study estimated the genetic correlation between spouses for EA at 0.37. 32 It also estimated the genetic correlation between siblings at 0.67 -a figure which suggests that CTD estimates of EA have been doubling the difference between MZ and DZ twin correlations to estimate heritability when tripling the difference might be more appropriate.…”

The next 10 years of behavioural genomic research

“…These effects have been labelled 'indirect effects' or 'genetic nurture', but again only passive rGE is controlled in these analyses. Such indirect effects have so far only been found for cognitive traits, not behaviour problems (Willoughby et al, 2021) and it has been suggested that these effects of 'genetic nurture' are actually between-family stratification effects (Nivard et al, 2022). These analyses do not specify which environmental factors are responsible, but measured environments can be incorporated in this approach.…”

Luck, lottery, or legacy? The problem of confounding. A reply to Harden

“…Thus, the latest GWAS for educational attainment reports that, although 12‐16% of the variance in educational attainment among people of European ancestry is predicted by polygenic scores, less than 5% of the variance is due to direct genetic effects (Okbay et al., 2022 ). A recent paper, which Harden co‐authored, attempts to tease apart parental indirect effects and confounders by analyzing siblings of the parents as well as nuclear families; it concludes that much of the non‐direct genetic effects may be due to “dynastic stratification in environments relevant to success in school” (Nivard et al., 2022 ), a form of confounding.…”