Polygenic risk for type 2 diabetes mellitus among individuals with psychosis and their relatives

Background: An elevated prevalence of Type 2 diabetes (T2D) has been observed in people with psychotic disorders and their relatives compared to the general population. It is not known whether this population also has increased genetic risk for T2D. Methods: Subjects included probands with schizophrenia, schizoaffective disorder, or psychotic bipolar I disorder, their first-degree relatives without psychotic disorders, and healthy controls, who participated in the Bipolar Schizophrenia Network for Intermediate Phenotypes study. We constructed sets of polygenic risk scores for T2D (PGRS_T2D) and schizophrenia (PGRS_SCHIZ) using publicly available data from genome-wide association studies. We then explored the correlation of PGRS_T2D with psychiatric proband or relative status, and with self-reported diabetes. Caucasians and African-Americans were analyzed separately. We also evaluated correlations between PGRS_SCHIZ and diabetes mellitus among Caucasian probands and their relatives. Results: In Caucasians, PGRS_T2D was correlated with self-reported diabetes mellitus within probands, but was not correlated with proband or relative status in the whole sample. In African-Americans, a PGRS_T2D based on selected risk alleles for T2D in this population did not correlate with proband or relative status. PGRS_SCHIZ was not correlated with self-reported diabetes within Caucasian probands. Conclusion: Differences in polygenic risk for T2D do not explain the increased prevalence of diabetes mellitus observed in psychosis probands and their relatives.