Mental disorders (MDs) significantly increase the risk of cardiometabolic disorders (CMDs) and mortality, reducing life expectancy by 7-10 years. This comorbidity is bidirectional, meaning individuals with CMDs also have a higher risk of developing MDs. The risk varies considerably across specific MD-CMD pairs. Understanding the underlying causes—genetic and environmental factors—is crucial for developing effective prevention and treatment strategies. Previous studies using genome-wide association studies (GWAS) summary statistics have shown genetic overlap between MDs and CMDs, but these studies have limitations, including biased sampling and sensitivity to statistical model assumptions. This study leverages nationwide health registers and near-complete genealogies from Denmark and Sweden, representing a large, population-representative sample, to overcome these limitations and quantify the relative contributions of genetic and environmental factors to MD-CMD comorbidity.

Existing literature demonstrates a strong association between mental and cardiometabolic disorders, highlighting a significantly increased risk of premature mortality among individuals with mental health diagnoses due to the elevated likelihood of developing somatic comorbidities such as cardiovascular diseases. While previous research suggested substantial genetic overlap between these conditions, methodologies employed often relied on GWAS summary statistics, inherently susceptible to biases stemming from sample selection and statistical modeling. Consequently, a systematic and comprehensive analysis of the relative importance of genetic and environmental factors in a large, truly representative population was lacking, necessitating this study.

This study utilized nationwide health registers and near-complete genealogical data from Denmark and Sweden encompassing 17 million individuals across four generations. The researchers analyzed the cumulative incidence of six mental disorders (attention-deficit/hyperactivity disorder (ADHD), anorexia nervosa (AN), autism spectrum disorder (ASD), affective disorders (AFF), bipolar disorder (BD), and schizophrenia (SCZ)) and 15 cardiometabolic disorders. Heritability (h²) and genetic correlations (rg) were estimated using the liability threshold model, incorporating the cumulative incidence functions for the general population and individuals with affected relatives. The relative contributions of genetic (G) and environmental (E) factors to the observed phenotypic correlation (rp) were quantified using a quantitative genetic model combining heritability and genetic correlation estimates. For comparison, SNP-based heritabilities (h²SNP) and genetic correlations (rg SNP) were calculated using LDSC and summary statistics from well-powered GWAS. The study included a detailed description of the data sources, including the Danish Civil Registration System, National Patient Register, Psychiatric Central Research Register, and the Swedish Total Population Register, Multi-Generation Register, and Inpatient Register.

The study found that heritability estimates ranged from 25% to 75% across the disorders and were highly correlated between Denmark and Sweden. Thirty-two significant genetic correlations (Bonferroni p<5.95×10⁻⁶) were observed between MD-CMD pairs. However, the analysis of the relative contributions of genetic and environmental factors to comorbidity revealed that environmental factors were predominantly responsible for the observed comorbidity in most MD-CMD pairs. The genetic component was substantial (around 50%) for the comorbidity between affective disorders and most CMDs, intermediate (around 30%) for CMDs with ADHD or schizophrenia, and negligible for anorexia nervosa with most CMDs. The genetic correlations observed using SNP-based analyses were highly concordant with those obtained using register data, demonstrating a substantial shared genetic contribution between many MDs and CMDs. However, the SNP-based analysis underestimated the genetic component of comorbidity due to the lower heritability estimates obtained from this method.

The findings highlight that the comorbidity between mental and cardiometabolic disorders is predominantly driven by shared environmental factors, rather than shared genetic factors alone, though genetics plays a notable role in some specific disease pairings. This challenges the assumption that genetic correlation alone is sufficient to explain comorbidity. The study emphasizes the importance of considering each MD-CMD pair individually, as the relative contributions of genetic and environmental factors vary substantially. The observed environmental effects may be attributed to both individual behaviors associated with MDs (e.g., lifestyle factors) and external factors like medication side effects. The high concordance between register-based and SNP-based genetic correlations validates the findings but underscores the limitations of SNP-based heritability estimates in fully capturing the genetic component of comorbidity.

This study provides the first systematic quantification of genetic and environmental contributions to MD-CMD comorbidity using large-scale population data. It demonstrates that shared environmental factors are major contributors to comorbidity, but genetic factors play a substantial role in some MD-CMD pairs. This emphasizes the need for individual-level analysis of each MD-CMD pair when investigating the origins of comorbidity. Future research should focus on identifying specific environmental exposures and genetic determinants involved in these complex relationships to inform precision medicine approaches.

The study's limitations include the use of secondary care hospital diagnoses, potentially leading to underrepresentation of some recently introduced diagnoses in older individuals. The potential influence of shared environments in close relatives on heritability and genetic correlation estimates is not quantified and could slightly inflate these estimates. The study assumes no interaction between genetic and non-genetic factors, and the use of hospital records may represent a more extreme phenotype compared to primary care data. The generalizability of findings to non-Scandinavian populations may be limited.