Aim To determine the risk of type 2 diabetes onset associated with accumulated inhaled corticosteroids (ICS) dose during the previous year in patients with chronic obstructive pulmonary disease (COPD).

Materials and methods We conducted a nationwide observational cohort study based on data from patients with COPD between 1 January 2010 and 31 December 2017 extracted from Danish health databases. Patients were followed for 7 years, until death or a type 2 diabetes event. A propensity-matched Cox model and an adjusted Cox proportional hazards model (stratified on body mass index [BMI]) were used to estimate the hazard ratio (HR) for new-onset type 2 diabetes.

Results A total of 50 148 patients with COPD were included, 3566 (7.1%) of whom had a type 2 diabetes event. During the previous year before study entry, 35 368 patients (70.5%) used ICS. The propensity-matched Cox model (N = 33 466) showed an increased risk of type 2 diabetes, which progressed with increasing accumulated ICS dose (low-ICS: HR 1.076, confidence interval [CI] 1.075-1.077, P <.0001; medium-ICS: HR 1.106, CI 1.105-1.108, P <.0001; high-ICS: HR 1.150, CI 1.148-1.151, P <.0001), compared with no ICS use. Results were confirmed in the adjusted Cox analysis on the entire study population, but only for patients with BMI

Conclusions In patients with COPD, ICS use was associated with a moderate dose-dependent increase in the occurrence of type 2 diabetes.