Adrenal insufficiency following systemic glucocorticoid therapy: a systematic review and meta-analysis

BACKGROUND: Adrenal insufficiency (AI) is a common complication following glucocorticoid (GC) therapy. However, reliable predictors of AI and GC regimens that mitigate this risk remain unclear. Previous systematic reviews (SRs) have included many observational studies with imprecise reporting of GC dose and duration. We conducted a systematic review focusing exclusively on studies with well-defined GC exposure. METHODS: We searched Embase, Web of Science, and MEDLINE for original studies published since 2000 that investigated adults receiving protocolised systemic GC regimens and assessed all participants for AI using the adrenocorticotropic hormone (ACTH) stimulation test. Data extracted included AI prevalence, diagnosis, administration route, GC dose and tapering, treatment duration and timing of outcome assessment. Risk of bias was assessed using an adapted Newcastle-Ottawa Scale. Meta-analysis was performed using random-effects models in R. RESULTS: Of 1340 unique records identified, 12 studies (1170 participants) met inclusion criteria. AI was present in 423 individuals, with a pooled prevalence of 30% (95% CI: 19-45%) at the earliest time point tested. Reported prevalence ranged from 0 to 70%, and was not consistently associated with diagnosis, administration route, treatment duration, or study quality. AI was significantly less common in studies that included tapering (p < 0.01) and in those assessing AI ≥15 days after stopping GCs (15%, 95% CI: 8-26%) compared to testing at cessation (48%, 95% CI: 35-62%). CONCLUSION: AI is frequent following systemic GC use, with considerable variability. Tapering regimens and delayed testing post-cessation appear to reduce the observed prevalence. No specific GC regimen reliably prevents AI.