Performance of Free Versus Total Cortisol Following Cosyntropin Stimulation Testing in an Outpatient Setting

Abstract

Free cortisol (FC) is potentially superior to total cortisol (TC) measurements in selected clinical settings; however, the advantages of uniform use of FC in outpatient settings are unclear. The objectives of this study were to describe the dynamic response of FC during cosyntropin stimulation testing (CST) compared to TC and to determine the rates of discordance. This is a cross-sectional study of 295 stable patients who underwent CST in an outpatient Endocrine Testing Center. The main outcome measures were TC and FC measurements during CST. The mean age of the 295 subjects was 49.1 (16.9) years. Of 218 females, 43 were taking estrogen therapy (ET) at the time of testing. Adrenal insufficiency (AI) was diagnosed in 41/295 (14%) patients. The FC concentrations were associated with TC concentrations at baseline (R(2) = 0.77, P<.001), 30 minutes (R(2) = 0.87, P<.001), and 60 minutes (R(2) = 0.90, P<.001). The FC cutoffs for AI were 873 and 1,170 ng/dL at 30 and 60 minutes, respectively. The FC had a more pronounced fold change from baseline to peak than TC (median 3.2 vs. 1.7, P<.001). Both TC and FC at baseline were higher in females on ET compared to those who were not and to males; however, peak TC and FC values were similar. In 3/43 females on ET, FC, and TC results were discordant (P = .003). We report 99% concordance of TC and FC measurements in a large outpatient cohort. The discordant rates were high in females treated with ET (7%). The FC measurements during CST in females on ET may provide a more rapid and accurate diagnosis of AI.