- Substantial partner similarity for psychiatric disorders (overall r=0.28; schizophrenia r=0.38), including both within-disorder and cross-disorder correlations.
- Observed nonrandom mating inflates prevalence of any psychiatric disorder by 1.8% and family-based heritability by 12.5%.
- Weaker somatic correlations were largely explained by psychiatric comorbidity, producing familial clustering of disease risk with significant societal and public health implications.
JAMA Psychiatry. 2026 Sep 9. doi: 10.1001/jamapsychiatry.2026.2712. Online ahead of print.
ABSTRACT
IMPORTANCE: Partner similarity for psychiatric traits, often attributed to assortative mating, has been repeatedly documented, but its theoretical consequences for the prevalence and heritability of psychiatric disorders have been examined only under strong and often unrealistic assumptions and have never been empirically estimated using population-based data.
OBJECTIVE: To quantify the extent of partner similarity for psychiatric and somatic disorders and to estimate its empirical intergenerational implications for the prevalence and heritability of psychiatric disorders and their corresponding somatic comorbidities.
DESIGN, SETTINGS, AND PARTICIPANTS: This population-based cohort study used linked nationwide Danish registry data on residents born between 1969 and 2017. Parental pairs and offspring were identified through the Danish Civil Registration System, and psychiatric and somatic diagnoses were obtained from national health registers. Data were analyzed from 2021 to 2026.
MAIN OUTCOMES AND MEASURES: Parental correlations for psychiatric and somatic disorders were calculated. Prevalence and heritability in the observed data were then compared with corresponding estimates under a random mating scenario.
RESULTS: Among 4 244 585 Danish residents born between 1969 and 2017 (2 200 795 male [51.9%]; mean [SD] age, 26.3 [13.5] years for males and 25.9 [13.5] years for females), we identified 549 541 parental couples and 822 209 offspring. Parental correlations were substantial for psychiatric disorders overall (r = 0.28; SE, 0.003), with the highest within-disorder correlation observed for schizophrenia (r = 0.38; SE, 0.015) and the lowest for anorexia (r = 0.11; SE, 0.029). Cross-disorder correlations ranged from 0.01 to 0.26. Parental correlations were also observed for somatic illnesses but were weaker (r = 0.03-0.14) and were almost entirely accounted for by comorbid psychiatric disorders. Simulations further showed that this pattern of nonrandom mating inflated the prevalence and family-based heritability of any psychiatric disorder by 1.8% and 12.5%, respectively.
CONCLUSIONS AND RELEVANCE: These findings suggest that partner similarity for psychiatric disorders is substantial and includes both within- and cross-disorder patterns, suggesting assortative mating alongside other mechanisms that may generate similarity between coparents. Regardless of its underlying mechanisms, this structure contributes meaningfully to observed family-based heritability and prevalence and leads to clustering of familial disease risk with important societal and public health implications.
PMID:42714899 | DOI:10.1001/jamapsychiatry.2026.2712
Share Evidence Blueprint
Save to Google Notes
Search Google Scholar
Save as PDF
⭐ My Revision List
