Twenty-three variable quality RCTs were included (n=23,396 participants, range 186 to 6,184): nine were classified as having a low risk of bias (clear descriptions of blinding, method of randomisation, allocation concealment and loss to follow-up) and the others had an unclear risk of bias. Twelve observational studies were included: five were at risk of confounding bias, six at risk of immortal time bias and in one study it was unclear whether patients with asthma had been excluded reliably.
Based on the RCTs, there was no statistically significant difference between inhaled corticosteroids and control in risk of myocardial infarction (RR 0.95, 95% CI 0.73 to 1.23), cardiovascular death (RR 1.02, 95% CI 0.81 to 1.27) and death (RR 0.96, 95% CI 0.86 to 1.07). There was no evidence of statistical heterogeneity among trials for any of the outcomes. The results were robust to all the sensitivity analyses.
Based on a pooled analysis of the observational studies, inhaled corticosteroids were associated with a statistically significant reduced risk of cardiovascular death (RR 0.79, 95% CI 0.72 to 0.86; two studies) and a reduced risk of death (RR 0.78, 95% CI 0.75 to 0.80; 11 studies) but not myocardial infarction (RR 0.83, 95% CI 0.63 to 1.08; one study). There was moderate statistical heterogeneity in both mortality analyses. There was evidence of publication bias in the observational studies.