Adverse events of inhaled corticosteroids in adult patients with asthma or chronic obstructive pulmonary disease: pairwise, network and dose–response meta-analyses

Objectives This study aimed to evaluate the associations between inhaled corticosteroids (ICSs) and adverse events (AEs) in adults with asthma or chronic obstructive pulmonary disease (COPD) and to explore variations by ICS type and dosage. Design and setting Medline/PubMed, Embase, the Cochrane Central Register of Controlled Trials and ClinicalTrials.gov were searched from inception to 20 February 2025. Data extraction and synthesis were performed, estimating risk ratios (RRs) with 95% confidence intervals (CI) or credible intervals (CrI) through pairwise and network meta-analysis (NMA), and an NMA using the Emax model was employed to find the dose–response relationships of the AEs of each ICS. Risk of bias was evaluated across both pairwise and NMAs. The certainty of evidence, based on the GRADE approach, was assessed exclusively for the NMA estimates. Participants Adults diagnosed with asthma or COPD treated with ICSs compared with non-ICS controls or those receiving different types or doses of ICSs. Main outcome measures The primary outcomes were AEs, including pneumonia, oral candidiasis, upper respiratory tract infection (URTI), fracture, diabetes, cataract and plasma cortisol abnormalities. Secondary outcomes included efficacy-related endpoints: asthma exacerbations, COPD exacerbations and all-cause mortality. Results 129 trials (120 900 participants) were included. According to pairwise meta-analysis, ICSs were associated with a higher risk of pneumonia (RR 1.49, 95% CI 1.38 to 1.61) and oral candidiasis (RR 2.29, 95% CI 2.01 to 2.60) and with a slightly higher risk of URTI (RR 1.17, 95% CI 1.10 to 1.25), compared with control. Besides, ICSs were linked to a reduced risk of asthma exacerbations (RR 0.30, 95% CI 0.20 to 0.45) and COPD exacerbations (RR 0.90, 95% CI 0.86 to 0.94) vs control. The results of the NMA that explored differences by ICS type showed that beclomethasone and fluticasone increased pneumonia risk compared with control, with absolute risk increases of up to 2.3%. Fluticasone also showed a slightly higher pneumonia risk than budesonide. Besides, mometasone, beclomethasone, budesonide and fluticasone were associated with increased risk of oral candidiasis, with the highest increase observed for mometasone (4.3%). In contrast, ciclesonide consistently showed lower risk across comparisons, reducing oral candidiasis by up to 4.5% with other ICSs. Fluticasone modestly increased URTI, while ciclesonide reduced it. Budesonide and mometasone were also associated with a small increase in cataract risk. Most of the outcomes reported here were supported by moderate- to high-certainty evidence. Furthermore, the dose–response relationship estimated from the Emax model indicated significant associations between fluticasone dose and the risks of pneumonia and URTI, between multiple ICS doses and oral candidiasis and between budesonide dose and cataract. Conclusions ICSs are correlated with the escalated risk of pneumonia, oral candidiasis and URTI in adult patients with asthma or COPD, but ICSs could reduce asthma exacerbations and COPD exacerbations. Fluticasone, beclomethasone, budesonide and mometasone are independently linked to certain AEs, and some Emax dose–response relationships are detected. Our findings may help guide individualised ICS use by informing benefit-risk considerations across various formulations and dosing regimens. PROSPERO registration number CRD42024527797.