Abstract

To examine the associations between atypical antipsychotic (AAP) exposure and the development of type 2 diabetes mellitus (T2DM) in Korean pediatric patients with psychiatric disorders, we conducted a nested case–control study using the claims data of the National Health Insurance system of Korea between 2010 and 2014. A cohort of patients with psychiatric disorders was identified, and enrollment was taken as the date of the first psychiatric diagnosis. Cases involved patients with a diagnosis of T2DM or prescriptions for glucose lowering drugs after enrollment, and the identification of T2DM was defined as the index date. We performed a conditional logistic regression analysis for matched case–control data to assess associations between AAP exposure and T2DM, and adjusted odds ratios (aORs) with 95% confidence intervals (CIs) are presented. From 1,092,019 patients aged 2–19 years, we identified 20,263 cases with T2DM and 80,043 controls, matched by sex, age, enrollment date, and primary psychiatric diagnosis. After adjusting for comorbidities, psychotropic medication history, and the healthcare institution characteristics, the aOR of having T2DM was significantly higher in multi-AAP users compared with non-users (aOR 1.89; 95% CI 1.63–2.20). Particularly high ORs for T2DM were observed in clozapine users compared with non-users (aOR 3.47; 95% CI 1.88–6.41). We observed a linear relationship between the increase in risperidone dose and the increase in the risk of developing T2DM. Our findings suggest a significantly increased risk of developing T2DM in child or adolescent patients with psychiatric disorders exposed to AAPs compared with those not exposed to AAPs.