228 Glucocorticoid-associated adrenal insufficiency in patients with rheumatoid arthritis: a study combining direct data collection with electronic health records

Abstract

Background: Glucocorticoids (GCs) remain widely used in inflammatory rheumatic disease, including 50% patients with RA and higher proportions in other diseases like SLE and PMR. It is known that exposure to GCs can lead to suppression of cortisol production by the adrenal glands which can persist after withdrawal of GCs. The resulting adrenal insufficiency (AI) can impact quality of life and is potentially dangerous in times of stress. Questions remain about the prevalence of and risk factors for developing GC-induced AI. The aims of this study were to determine the prevalence of AI in a sample of patients with RA currently and formerly exposed to GCs, and to identify predictors of GC-induced AI. Methods: This cross-sectional study was nested within UK primary care electronic health records (EHR) provided by the Clinical Practice Research Datalink. Participants were adult patients with RA who had been prescribed oral GCs in the past two years. AI was defined by measuring morning salivary cortisol levels: salivary cortisol <5mcg/dl was taken to represent AI. Information about current use and dose of oral GCs was captured using patient diaries. Cumulative duration of exposure (total time with an active prescription over the past two years) and other potential predictors, including age, gender, body mass index, socio-economic status, and duration of RA, were determined from the EHR. The proportion of current and former users of oral glucocorticoids with AI was determined. Univariate logistic regression was used to investigate potential predictors of AI; this was stratified according to current use of oral GCs (only results for current users are presented). Odds ratios (OR) and 95% confidence intervals are presented. Results: In total, 526 patients were invited to take part and 76 completed the study. 29 participants (38%) had AI. Among current GC users, the prevalence was 66%. Among non-current users the prevalence was 11%, with some participants having low cortisol levels at least 6 months after their last GC exposure. For current GC users the daily dose ranged from 2-30mg prednisolone (median 5mg); 30% of patients taking low dose GCs (<5mg) had AI. In current users, the risk of AI increased according to increasing current daily dose (mg) (OR 1.83, 1.11-3.03) but not cumulative duration (weeks) (OR 1.00, 0.98-1.02). Increasing age and RA duration (years) were significantly associated with increased risk of AI (ORs 1.92, 1.04-3.56; and 1.42, 1.10-1.81). Conclusion: Two thirds of patients currently taking oral GCs had suppressed cortisol levels, including some patients taking low-dose GCs (<5mg), and the risk of adrenal suppression appeared to increase with increasing current daily dose. The prevalence of AI among current GC users is high, and the risk of AI should be considered whenever GCs are withdrawn.