Clinical indicators of adrenal insufficiency following discontinuation of oral glucocorticoid therapy: A Danish population-based self-controlled case series analysis

BACKGROUND: Biochemical adrenal insufficiency induced by glucocorticoid treatment is prevalent, but data on the clinical implications are sparse. We investigated clinical consequences of glucocorticoid-induced adrenal insufficiency after oral glucocorticoid cessation. METHODS: We conducted a Danish population-based self-controlled case series utilizing medical registries. In this design each individual serves as their own control allowing event rates to be compared as a function of time and treatment. Clinical indicators of adrenal insufficiency were defined as diagnoses of gastrointestinal symptoms, hypotension, cardiovascular collapse, syncope, hyponatremia, and hypoglycaemia. We included 286,680 persons who discontinued long-term (≥ 3 months) oral glucocorticoid treatment. We defined five risk periods and a reference period (before treatment): period 0 (on treatment), withdrawal period (1 month before and after cessation), followed by three consecutive 2 month-risk periods after withdrawal (periods 2–4). RESULTS: Median age at cessation was 69 years and 57% were female. Median treatment duration was 297 days and median cumulative dose was 3000 mg prednisolone equivalents. The incidence rates of hypotension, gastrointestinal symptoms, hypoglycemia and hyponatremia were increased in the withdrawal period compared to before treatment started (reference period). Incidence rate ratios comparing the withdrawal period with the reference period were 2.5 [95% confidence interval (CI): 1.4–4.3] for hypotension, 1.7 (95% CI: 1.6–1.9) for gastrointestinal symptoms, 2.2 (95% CI: 0.7–7.3) for hypoglycemia, and 1.5 (95% CI: 1.1–2.0) for hyponatremia. During 7 months of follow up, the rates of hypotension and gastrointestinal symptoms remained elevated compared to the reference period. Risk factors included use of antibiotics, increasing average daily dose of glucocorticoids, cumulative dose, and age. CONCLUSION: Oral glucocorticoid withdrawal was associated with adverse outcomes attributable to adrenal insufficiency. Our study underscores the need for future research to establish evidence-based clinical guidance on management of patients who discontinue oral glucocorticoids.