Risk of New-onset Type 2 Diabetes in Antidepressant Users: A Population-based Cohort Study in Denmark

Background: The association between antidepressants and type 2 diabetes has been controversial due to the bidirectional association of depression and type 2 diabetes. Recently large population-based studies from France and Japan and meta-analyses have reported significantly increased risk despite adjusting for depression.

Objectives: To investigate the association between antidepressant initiation and incident type 2 diabetes in different age-groups.

Methods: We performed a cohort study in 1 million random sample of the Danish population, based on national registries. New-users of antidepressants were identified from 2001-2015 and 1:4 matched to controls on age, sex, and index date. The outcome was incident type 2 diabetes or oral glucose-lowering drugs. All analyses were age-stratified as young-middle aged (18-64 years) and older adults (≥65 years). Intention-to treat analysis with a follow-up of 5 years, and an analysis with time-dependent exposure status (current: 0-120, recent: 121-365 or past: >365 days) was conducted. Cox proportional hazards regression models were used to estimate hazard ratios (HRs) with 95% confidence intervals (CIs), adjusted for potential confounders, including somatic and psychiatric comorbidities and comedications.

Results: The age-sex matched cohort comprised of 117,691 antidepressant users and 470,764 controls. In the < 65 age-group, the incidence rate (IR) of type 2 diabetes in antidepressant users and controls was 5 per 1000 person-years (PY) and 3.2 per 1000 PY, respectively, corresponding to an unadjusted HR of 1.57 (1.50-1.64) and adjusted HR (aHR) of 1.29 (1.22-1.36) in 5-year follow-up. Whereas, among the ≥65 age-group, IR was 9.2 and 8.6 per 1000 PY corresponding to an unadjusted HR of 1.07 (1.01-1.14) and aHR 0.97 (0.90-1.04). Finally, in the time-dependent exposure model, the aHR in < 65 age-group for current, recent and past use was 1.43 (1.30-1.56), 1.38 (1.20-1.60) and 1.17 (1.07-1.29). However, the aHR in the ≥65 age-group, was 0.95 (0.86-1.06), 0.93 (0.76-1.14), and 1.00 (0.87-1.15).

Conclusions: Antidepressant use was associated with significantly increased risk of type 2 diabetes in the < 65 age-group and was increased among all exposure categories of current, recent and past use. Contrastingly, there was no increased risk among the older adults.