The etiological link between clinical psychiatric diagnoses and continuous variation in related traits: evidence from a twin study

Psychiatric disorders are often argued to represent the extreme manifestation of continuous traits, which present to varying degrees of severity throughout the general population. As such, one hypothesis is that genetic risk... [ view full abstract ]

Psychiatric disorders are often argued to represent the extreme manifestation of continuous traits, which present to varying degrees of severity throughout the general population. As such, one hypothesis is that genetic risk factors for such disorders are also associated with traits characteristic of these disorders. In this study, we utilized a large twin study in which participants have been comprehensively screened for psychiatric disorders using continuous trait measures, and which has been linked with Swedish patient registries. Parents of approximately 14,000 twin pairs taking part in the Child and Adolescent Twin Study in Sweden (CATSS) completed scales assessing traits characteristic of autism spectrum disorders (ASD), attention-deficit/hyperactivity disorder (ADHD), intellectual disability, anxiety disorders, depression, obsessive-compulsive disorder (OCD), and tic disorder when twins were aged 9 or 12. Information on clinical diagnoses of these seven conditions was extracted from the Swedish National Patient Register. We then applied an approach to twin analysis that involved jointly modeling liability to each dichotomous diagnosis and continuous variation in each trait, which estimated the genetic correlation between clinical diagnoses and continuous traits. Each dichotomous disorder and continuous scale was under significant genetic influence (disorder liability heritability=36-94%; trait heritability=23-69%). For six phenotypes, there was a significant genetic correlation between the dichotomous diagnosis and traits characteristic of the same disorder: genetic correlations=.55 (ADHD), .50 (intellectual disability), .40 (tic disorder), .38 (ASD), and .24 (anxiety disorders and depression). The genetic correlation between OCD and characteristic traits was weak and non-significant (.13). Thus, our findings support the hypothesis that psychiatric disorders represent the extreme end of a continuous distribution of population traits.