Preprint: RAVEL framework suggests assortative mating drives most apparent genetic nurture effects on child cognition

Researchers have introduced RAVEL, a structural equation modelling framework designed to separate genuine genetic nurture effects — whereby parental genotypes shape the environments they provide to offspring — from statistical artefacts introduced by assortative mating, measurement error, and unobserved genetic variation. The preprint, posted to bioRxiv from Cold Spring Harbor Laboratory, applies the method to parent-offspring trio data using childhood cognitive performance as the outcome trait.

The central finding is that the majority of the apparent association between non-transmitted parental polygenic scores and offspring cognitive outcomes — a signal that has been interpreted in recent literature as evidence of genetic nurture — can instead be accounted for by assortative mating. Partners who mate non-randomly on cognitive ability share correlated genomes, which inflates the apparent influence of parental genotype on offspring environments when standard polygenic score methods are used.

RAVEL addresses this by using two independently constructed polygenic scores for the same trait as cross-validators, allowing the framework to estimate latent genetic liability associations while correcting for the biases identified. The authors report analytic derivations alongside simulation results.

This work is directly relevant to statistical geneticists, quantitative geneticists, and behavioural geneticists who work with polygenic scores in family-based study designs. It has implications for interpreting the broader genetic nurture literature, including studies of educational attainment, cognitive ability, and developmental outcomes. The preprint has not yet undergone peer review.