Exploring Boundaries for the Genetic Consequences of Assortative Mating for Psychiatric Traits

Considerable partner resemblances have been found for a wide range of psychiatric disorders, meaning that partners of affected individuals have an increased risk of being affected compared with partners of unaffected individuals. If this resemblance is reflected in genetic similarity between partners, genetic risk is anticipated to accumulate in offspring, but these potential consequences have not been quantified and have been left implicit.

The anticipated consequences of partner resemblance on prevalence and heritability of psychiatric traits in the offspring generation were modeled for disorders with varying heritabilities, population prevalence (lifetime risk), and magnitudes of partner resemblance. These models facilitate interpretation for a wide range of psychiatric disorders, such as autism, schizophrenia, and depression. The genetic consequences of partner resemblance are most pronounced when attributable to phenotypic assortment (driven by the psychiatric trait). Phenotypic assortment results in increased genetic variance in the offspring generation, which may result in increased heritability and population prevalence. These consequences add generation after generation to a limit, but assortative mating is unlikely to balance the impact of reduced fecundity of patients with psychiatric disorders in the long term. This modeling suggests that the heritabilities of psychiatric disorders are unlikely to increase by more than 5% from 1 generation of assortative mating (maximally 13% across multiple generations). The population prevalence will increase most for less common disorders with high heritability; for example, the prevalence of autism might increase by 1.5-fold after 1 generation of assortative mating (≥2.4-fold in the long term) depending on several assumptions.

The considerable partner resemblances found for psychiatric disorders deserve more detailed interpretation than has been provided thus far. Although the limitations of modeling are emphasized, the anticipated consequences are at most modest for the heritability but may be considerable for the population prevalence of rare disorders with a high heritability.