Convergent etiology among schizophrenia, eating disorders, and other medical conditions

Abstract

Schizophrenia and eating disorders (EDs) are complex traits with considerable somatic and psychiatric morbidity. Despite distinct diagnostic criteria and few symptoms in common, significant genetic correlations between schizophrenia and anorexia nervosa (AN) have been illustrated in several recent genomic studies. However, the genetic association between schizophrenia and EDs other than AN is still unknown, and whether these disorders have common familial risk is unclear. This thesis extends the understanding of the underlying shared etiological pathway between schizophrenia and EDs, investigates its influences on clinical outcomes of EDs, and explores the broad impact of schizophrenia genetic risk on a wide range of medical conditions.

In Study I, we applied a family design to estimate the familial co-occurrence and coaggregation of schizophrenia and EDs in population-based cohorts from Sweden and Denmark. Significantly increased risk of schizophrenia was found in individuals with EDs (6- 7 times the risk) as well as their relatives (highest in first-degree relatives and diminished with decreasing genetic relatedness) compared with individuals without EDs and their relatives, indicating that familial liabilities including common genetic and/or shared environmental factors contribute to the associations between schizophrenia and EDs.

In Study II, we investigated the effect of a family history of schizophrenia on clinical outcomes including ED clinical features, psychiatric comorbidities, and somatic and mental health burden in individuals with EDs from a population-based ED case cohort identified from Swedish national registers. Family history of schizophrenia was associated with increased somatic and mental health burden, but less severe ED-related symptoms in individuals with AN. It was also associated with increased risk of psychiatric comorbidities, cumulative health burden, and suicide attempts in individuals with other eating disorders (OED). These findings suggest that family history of schizophrenia might be a potential factor for clinically meaningful ED case stratification.

In Study III, we evaluated the impact of schizophrenia genetic predisposition conferred by polygenic risk scores (PRS) on clinical outcomes of EDs in cases from two ED studies: the Anorexia Nervosa Genetics Initiative-Sweden (ANGI-SE) and the Binge Eating Genetics Initiative-Sweden (BEGIN-SE). Schizophrenia PRS was positively associated with comorbid major depressive disorders (MDD), but negatively associated with clinical impairment assessment (CIA) scores in ANGI cases. BEGIN cases who had higher schizophrenia PRS had increased risk of MDD and substance abuse disorders (SUD), greater cumulative health burden, and higher ED pathology scores. These findings suggest a different pattern of influence of schizophrenia PRS for AN compared with OED with reference to clinical outcomes.

In Study IV, we explored the pleiotropic effect of schizophrenia PRS in over 400,000 individuals without schizophrenia from the UK Biobank. We found poorer self-reported overall health condition, more hospital inpatient contacts, more total diseases diagnosed in individuals with higher schizophrenia PRS. Specifically, we found schizophrenia PRS were statistically significantly associated with multiple illnesses including mental disorders, musculoskeletal diseases (negatively associated), respiratory diseases, digestive diseases, pituitary hyperfunction, etc., demonstrating that genetic risk for schizophrenia also impacts risk for numerous other health conditions.

In conclusion, this thesis enhances the understanding of the convergent etiology among schizophrenia, EDs, and other medical conditions; highlights the influences of schizophrenia genetic predisposition on clinical outcomes of EDs; and demonstrates the broad impact of schizophrenia genetic risk on a wide range of medical conditions in individuals without schizophrenia.