National trends in metabolic risk of psychiatric inpatients in the United States during the atypical antipsychotic era

Abstract

Although the cardiometabolic effects of atypical antipsychotics have been well-described in clinical samples, less is known about the longer-term impacts of these treatments. We report rates of metabolic syndrome in a nationally representative sample of U.S. adult inpatients 1993-2018 admitted for schizophrenia-spectrum disorders (n = 1,785,314), any mental health disorder (n = 8,378,773), or neither (n = 14,458,616) during a period of widespread atypical antipsychotic use. Metabolic syndrome, derived from additional diagnoses, was defined as three or more of hypertension, dyslipidemia, type II diabetes, hyperglycemia, and overweight or obese. Using an ecological age and period design, a 4-level period variable was constructed to proxy for atypical antipsychotic exposure as the minimum of age minus 20 years or the calendar year minus 1997 in accord with the disease course for schizophrenia-spectrum illness and the market share of atypical antipsychotics in the U.S. Logistic regression models, adjusted for age, year, and exposure main effects, estimated odds ratios (ORs) of metabolic syndrome. Relative to other mental health or other discharges, schizophrenia-spectrum discharges had an elevated risk for metabolic syndrome regardless of potential atypical antipsychotic exposure (OR = 1.46; 95 % CI, 1.30-1.64). For schizophrenia-spectrum discharges, periods of potential atypical antipsychotic exposure conferred additional metabolic syndrome risk OR = 1.21; 95 % CI, 1.04-1.41 for exposures of 1-2 years, OR = 1.29; 95 % CI, 1.13-1.46 for 3-7 years, OR = 1.27; 95 % CI, 1.12-1.44 for 8-12 years, and OR = 1.10; 95 % CI 0.98-1.24 for >12 years. In summary, cardiometabolic disease and related risks were elevated among a nationally representative sample of adult inpatients with schizophrenia-spectrum disorders during a period of pervasive atypical antipsychotic use.

Keywords: Metabolic syndrome; Schizophrenia; Side effects-medication; Type 2 diabetes.

Figure 1.

Difference-in-differences for Metabolic Syndrome Diagnosis Rate Relative to Non-Mental Health Discharges for Inpatients Aged 20-89 Years, Admitted for Schizophrenia-Spectrum Illness and Total Mental Health, 1993-2018. National Inpatient Sample from 1993-2018. Analysis limited to patients aged 20-89, N=22,898,472. Discharge types are defined by primary diagnosis. Schizophrenia-spectrum discharges are a subset of mental health discharges. Metabolic syndrome is defined as having three or more of diagnoses hypertension, dyslipidemia, type II diabetes, hyperglycemia, overweight or obese included as associated diagnoses with the discharge. Quantities displayed in the solid line reflect the rate of metabolic syndrome among schizophrenia-spectrum discharges in year t minus the rate of metabolic syndrome for psychosis discharges in year 1993, all minus the rate of metabolic syndrome among non-mental health discharges in year t minus the rate of metabolic syndrome for non-mental health discharges in year 1993. Quantities displayed in the hollow line reflect the rate of metabolic syndrome among mental health discharges in year t minus the rate of metabolic syndrome for mental health discharges in year 1988, all minus the rate of metabolic syndrome among non-mental health discharges in year t minus the rate of metabolic syndrome for non-mental health discharges in year 1993.