Background Whether inhaled corticosteroids (ICSs) reduce major adverse cardiovascular events (MACEs) in people with chronic obstructive pulmonary disease (COPD) is debated.

Objectives To establish, within people with COPD, (1) whether ICS reduced MACE rates (acute coronary syndrome (ACS), heart failure (HF), ischaemic strokes or cardiovascular-specific death) compared with long-acting bronchodilators; and (2) whether drug class, incident usership or patient cardiovascular history influenced the ICS-MACE relationship.

Methods We conducted a cohort study including patients with COPD in England, using Clinical Practice Research Datalink Aurum data, linked with Hospital Episode Statistics and Office of National Statistics death data, between 1 January 2010 and 31 December 2019. We implemented Cox proportional hazard regressions, adjusting for time interactions or using propensity score-adjusted models, as necessary. Our exposures included prescriptions of any ICS (vs any long-acting bronchodilators) and triple therapy (vs combination long-acting bronchodilators), determined during the year prior to follow-up. The outcomes of interest were MACE collectively and individual MACE subtypes.

Measurements and main results Among 113 353 people with COPD (mean age 67.9 years old, 53.3% male), ICS prescription was not associated with MACE (adjusted HR (95% CI)=0.98 (0.95, 1.02), p=0.41) but was associated with reduced HF, specifically, until year 6 of follow-up (average adjusted HR (95% CI)=0.91 (0.86, 0.96), p<0.001). HF reduction was driven by the ICS group containing mometasone furoate, beclomethasone, budesonide or ciclesonide (HR (95% CI)=0.89 (0.84, 0.94), p<0.001). Incident ICS use was associated with increased ACS (HR (95% CI)=1.27 (1.09, 1.47), p<0.001) but was not sustained beyond incident use. There was no association between triple therapy and MACE. Results did not differ by cardiovascular history.

Conclusions ICS did not reduce MACE, except HF, likely by reducing misclassified COPD exacerbations.