Performance of Dehydroepiandrosterone Sulfate and Baseline Cortisol in Assessing Adrenal Insufficiency

CONTEXT: Diagnosing adrenal insufficiency (AI) often requires complex testing, which can be time-consuming and expensive. Dehydroepiandrosterone sulfate (DHEAS) is a promising marker of hypothalamic-pituitary-adrenal axis function; however, its diagnostic performance has not yet been evaluated in a large-scale study. OBJECTIVE: Evaluate the performance of DHEAS and baseline cortisol in assessing AI. DESIGN: Single-center retrospective cohort study. SETTING: Referral center. PATIENTS: Adults who underwent cosyntropin stimulation testing (CST) between 2005 and 2023 and had DHEAS measured within 3 months prior to CST. MAIN OUTCOME MEASURES: Area under receiver operating characteristic curve (AUROC) for DHEAS and baseline cortisol. Prevalence of AI based on various DHEAS and baseline cortisol concentrations. RESULTS: Among 1135 patients, 195 (17.2%) had AI. Both baseline cortisol and DHEAS independently had good diagnostic performance with AUROC 0.81 95% confidence interval (CI) 0.77-0.84 and 0.81 (95% CI 0.78-0.85), respectively. Time of CST performance had no significant effect on diagnostic accuracy of baseline cortisol while recent glucocorticoid use decreased diagnostic performance of DHEAS (AUROC 0.72 vs 0.83). Only 1.2% of patients with baseline cortisol ≥10 mcg/dL had AI based on CST. Among patients with baseline cortisol between 5 and 9.9 mcg/dL, only 1.3% had AI if DHEAS was ≥60 mcg/dL. Conversely, the majority (72.2%) of patients with both baseline cortisol <5 mcg/dL and DHEAS <25 mcg/dL were found to have AI. CONCLUSION: DHEAS has good diagnostic performance in assessing AI. Measuring both baseline cortisol and DHEAS concentrations may eliminate the need for further dynamic testing in many patients.