Bradley Ng, MBChB, FRANZCP
Are antipsychotic drugs associated with hyperglycemia in older patients with diabetes mellitus?
Nested case–control study with mean follow-up of 2 years.
15 874 patients ≥ 66 years (mean age 78 y, 50% women) with diabetes who began antipsychotic therapy. Exclusion criteria were antipsychotic prescription in the preceding year, dialysis or palliative care, and residence in a long-term care facility. Cases were patients hospitalized for hyperglycemia (n = 1504). Controls (≤ 10 per case) were patients who had not been hospitalized for hyperglycemia but were at risk for an event, matched to cases for age, sex, year of cohort entry, and duration of follow-up (n = 14 370). Each patient could serve as a control more than once and could subsequently become a case.
Prescribed use of atypical (i.e., olanzapine, quetiapine, and risperidone; excluding clozapine) or typical antipsychotics by time of use: current (first-time and prevalent), recent past (discontinued 15 to 180 d before index date), and remote (discontinued > 180 d before index date).
Hyperglycemia (hospitalization or emergency department visit for hyperglycemia, diabetic ketoacidosis, hyperosmolar nonketotic coma, or diabetes).
Current antipsychotic use was associated with an increased risk for hyperglycemia for all diabetes treatment groups (insulin, oral hypoglycemic, or no treatment) (overall adjusted rate ratio 1.50, 95% CI 1.29 to 1.74). Results by time of use and type of diabetes treatment are presented in the Table. The annual estimated numbers needed to harm for insulin-treated, oral agent–treated, and untreated patients with any current antipsychotic use were 21 (95% CI 10 to 138), 42 (CI 23 to 127), and 37 (CI 20 to 86), respectively.
In older patients with diabetes, current antipsychotic use was associated with increased risk for hyperglycemia regardless of type of diabetes treatment. Risk was highest in first-time antipsychotic users.
Associations between use of antipsychotics and hyperglycemia in older patients with diabetes*
*CI defined in Glossary.
†Compared with remote antipsychotic use (discontinued > 180 d before index date). Adjusted for multiple potential confounding variables.
‡Not statistically significant.
Ng B. Use of antipsychotic drugs was associated with increased risk for hyperglycemia in older patients with diabetes. Ann Intern Med. 2009;151:JC5–12. doi: 10.7326/0003-4819-151-10-200911170-02012
Download citation file:
Published: Ann Intern Med. 2009;151(10):JC5-12.
Cardiology, Coronary Risk Factors, Diabetes, Endocrine and Metabolism, Geriatric Medicine.
Results provided by:
Copyright © 2018 American College of Physicians. All Rights Reserved.
Print ISSN: 0003-4819 | Online ISSN: 1539-3704
Conditions of Use