This study aimed to investigate whether metformin can affect risk of chronic obstructive pulmonary disease (COPD) in type 2 diabetes (T2D) patients.

T2D patients newly diagnosed during 1999–2005 were enrolled from the reimbursement database of Taiwan's National Health Insurance system and followed up to 31 December 2011. Analyses were conducted in an unmatched cohort (92,272 ever-users and 10,697 never-users of metformin) and a propensity score (PS) matched pair cohort (10,697 ever-users and 10,697 never-users). Cox regression incorporated into the inverse probability of treatment weighting using the PS was used to estimate hazard ratios (HRs).

In the unmatched cohort, 2573 never-users and 13,840 ever-users developed COPD with respective incidences of 5994.64 and 3393.19 per 100,000 person-years. The overall HR was 0.560 (95% confidence interval [CI]: 0.537–0.584). HRs for the first (<25.27months), second (25.27–55.97months) and third (>55.97months) tertiles of cumulative duration were 1.021 (0.975–1.070), 0.575 (0.548–0.603) and 0.265 (0.252–0.280), respectively. Analyses of the matched cohort showed an overall HR of 0.643 (0.605–0.682), with HRs of 1.212 (1.122–1.309), 0.631 (0.578–0.689) and 0.305 (0.273–0.340) for the respective tertiles.

A reduced risk of COPD is observed in metformin users with T2D.