BackgroundSevere exacerbations of COPD are commonly associated with hyperglycaemia, which predicts adverse outcomes. Metformin is a well-established anti-hyperglycaemic agent in diabetes mellitus, possibly augmented with anti-inflammatory effects, but its effects in COPD are unknown. We investigated accelerated metformin therapy in severe COPD exacerbations, primarily to confirm or refute an anti-hyperglycaemic effect, and secondarily to explore its effects on inflammation and clinical outcome.MethodsThis was a multicentre, randomised, double-blind, placebo-controlled trial testing accelerated metformin therapy in non-diabetic patients, aged ≥35 years, hospitalised for COPD exacerbations. Participants were assigned in a 2:1 ratio to 1 month of metformin therapy, escalated rapidly to 2 g/day, or matched placebo. The primary end point was mean in-hospital blood glucose concentration. Secondary end points included the concentrations of fructosamine and C reactive protein (CRP), and scores on the COPD Assessment Test and Exacerbations of Chronic Pulmonary Disease Tool.Results52 participants (mean (±SD) age 67±9 years) were randomised (34 to metformin, 18 to placebo). All were included in the primary end point analysis. The mean blood glucose concentrations in the metformin and placebo groups were 7.1±0.9 and 8.0±3.3 mmol/L, respectively (difference −0.9 mmol/L, 95% CI −2.1 to +0.3; p=0.273). No significant between-group differences were observed on any of the secondary end points. Adverse reactions, particularly gastrointestinal effects, were more common in metformin-treated participants.ConclusionMetformin did not ameliorate elevations in blood glucose concentration among non-diabetic patients admitted to hospital for COPD exacerbations, and had no detectable effect on CRP or clinical outcomes.Trial registration numberISRCTN66148745 and NCT01247870.