RESUMO
OBJECTIVE: To systematically determine the prevalence of diabetes mellitus (DM), impaired fasting glucose (IFG), and impaired glucose tolerance (IGT) in psychiatric inpatients and explore the impact of patient and study variables on prevalence estimates. METHOD: We searched EMBASE, PsychINFO, Medline and CENTRAL from database inception until 1st December 2015. We included studies of any design reporting prevalence of abnormal glucose metabolism in any adult psychiatric inpatients. We conducted a random effects meta-analysis to generate pooled prevalence estimates. Chi-square tests compared differences within categorical variables (inpatient setting, continent of study and patient diagnostic category) and Spearman's correlation analyses assessed the impact of linear variables (age, year of data collection and study quality). Study quality was assessed using an adapted Newcastle-Ottawa Scale. RESULTS: 36 study reports representing 42 unique cohorts were included. Across all studies prevalence of unspecified type DM was 10% (95%CI: 9-12), of T1DM was 1% (0-1), of T2DM was 9% (6-13), of IFG 18% (8-28), and of IGT was 22% (16-28). These estimates were not affected by study quality. CONCLUSIONS: All estimates are higher compared to the general population. Mental health professionals should be aware of this elevated prevalence to improve screening and management of abnormal glucose metabolism.