An accurate, non-invasive approach to diagnose Cushing’s syndrome in at-risk populations

Abstract

Background: The prevalence of Cushing’s syndrome (CS) in at-risk populations in developing countries remains uncertain. Evening urinary cortisol (UFC22–23) and salivary cortisol after treatment with 1-mg DST (SAFdex) have seldom been used as diagnostic tools in these populations. Objectives: (1) To establish the prevalence of CS in adults with cortisol-related morbidities using UFC22–23 and SAFdex as markers along with all first-line diagnostic tests recommended for CS; and (2) to assess the performance of each test and define a non-invasive diagnostic approach for CS in at-risk outpatient subjects. Methods: A total of 128 outpatients were evaluated, including type 1 and 2 diabetic patients with poor metabolic control (DM1 and DM2), hypertensive subjects with central obesity (HBP) and premenopausal women with osteoporosis (OS). Controls included 100 healthy volunteers and 23 patients with CS. Total urinary cortisol (UFC), UFC22–23, late-night salivary cortisol (SAF23) and suppression of cortisol levels in saliva (SAFdex) and serum (Fdex) after treatment with 1-mg DST were assessed. Results: CS was diagnosed in one DM2 and one HBP patient; both women exhibited central obesity. Among CS patients, UFC showed more within-person variability than UFC22–23 or SAF23. UFC22–23 and SAF23 were positively and significantly correlated in all groups (r ⩾ 0.70; p ⩽ 0.0001). UFC22–23 > 44.0 ng/mg creatinine or SAF23 > 3.8 nM were 100% sensitive (S) and specific (E) for CS. Furthermore, SAFdex > 2.0 nM or Fdex > 50.0 nM were 100% S and 97.3% E for CS. Conclusion: CS was diagnosed in 1.5% of at-risk patients. The combination of UFC22–23 or SAF23 with SAFdex offers a non-invasive diagnostic tool to assess cortisol nadir and feed-back status in outpatients.