Primary adrenal insufficiency (PAI) is a condition in which dysfunction or destruction of the adrenal glands leads to cortisol deficiency. To confirm a diagnosis of PAI, adrenal response is typically assessed by the gold standard adrenocorticotropic hormone (ACTH) stimulation test.Traditionally, cut-offs of 18 or 20μg/dL have been cited as the standard peak serum cortisol thresholdsabovewhich PAI could be excluded following stimulation testing. Thesecut-offs wereincluded in the Endocrine Society’s 2016 clinical practice guidelines for the diagnosis of PAI, although the guidelines emphasized that these values are assay-dependent.A number of factors contribute to wide variability in cortisol measurements across platforms, including lack of standardization, variable strategies to release cortisol from binding proteins, and the use of different detection antibodies with varying specificities. Thus, establishing an appropriate serum cortisol cut-off for each assay is critical to support clinical decision-making and to prevent misdiagnosis.

Around the same time the Endocrine Society guidelines were released, a second generation of cortisol immunoassays was developed, further challenging the historical 18-20 μg/dL thresholds which were defined by older, non-standardized immunoassays utilizing polyclonal antibodies to bind cortisol. Polyclonal antibodies against cortisolexhibit a high degree of cross-reactivity with a number of structurally-similar compounds, such as endogenous steroids or steroid drugs, leading to an overestimation of cortisol levels. The new generation of cortisol immunoassays implements monoclonal antibodies, decreasing the amount of cross-reactivity and consequently increasing the specificity of cortisol measurements. As a result, cortisol measurements were found to be 20 to 30% lower using the monoclonal antibody immunoassays according to several studies. Moreover, these studies showed that a substantial number of healthy subjectsdid not achieve the traditionalpeak level of 18 µg/dL, highlighting the need for new reference ranges to prevent over-diagnosis of PAI.

Although the majority of clinical laboratories measure serum cortisol by automated immunoassays, there has been a continually-growing interest in liquid chromatography-tandem mass spectrometry (LC-MS/MS) for steroid measurements in routine clinical laboratories. Because of its superior analytical specificity, LC-MS/MS is considered the gold standard for cortisol measurements. While several studies, including one multi-center evaluation, describe excellent agreement between serum cortisol measurements obtained by the monoclonal antibody immunoassays and LC-MS/MS, cortisol values obtained by new immunoassays are slightly higher, on average(Table). This suggests that even the best-calibrated immunoassays are still subject to some degree of cross-reactivity, and therefore, separate immunoassay-specific and LC-MS/MS-specific decision-making points are needed.

Table: A comparison of serum cortisol in healthy control subjects measured by LC-MS/MS and two second generation cortisol immunoassays.Adapted from Ueland et al., 2018

  Median cortisol, μg/dL (range)
Immunoassay A
Immunoassay B
Baseline 10.1 (3.9-23.7)  10.4 (6.4-22.0)  11.0 (6.0-21.2) 
30 minutes
19.5 (14.3-30.6)  21.1 (14.9-27.2)  21.4 (15.7-32.6) 
     30 minΔ
60 minutes
22.4 (16.2-35.9)  23.4 (18.2-30.9)  24.4 (19.8-37.1)
     60 minΔ

With the implementation of highly specific cortisol methodologies, establishing appropriate peak serum cortisol thresholds to determine adrenal insufficiency is more critical than ever. Because there is considerable inter-assay variation of cortisol measurements as evidenced by proficiency test results released by the College of American Pathologists, each manufacturer should determine their own clinical cut-off. Furthermore, laboratorians and clinicians should be familiar with which cortisol assay is being used at their institution. Although more work needs to be done to evaluate the new cut-offs proposed in the literature, these studies demonstrate how failure to recognize that new assays result in lower peak serum cortisol levels could lead to over-diagnosis of PAI and inappropriate therapeutic intervention.


  3. Raverot V, Richet C, Morel Y, Raverot G, Borson-Chazot F. “Establishment of revised diagnostic cut-offs for adrenal laboratory investigation using the new Roche Diagnostics Elecsys® Cortisol II assay” Ann Endocrinol (Paris). 2016; 77(5) 620-622.
  4. Kline GA, Buse J, Krause RD. “Clinical implications for biochemical diagnostic thresholds of adrenal sufficiency using a highly specific cortisol immunoassay.” Clin Biochem 2017; 50(9):475–480
  5. Grassi G, Morelli V, Ceriotti F, Polledri E, Fustinoni S, D’Agostino S, Mantovani G, Chiodini I, Arosio M. “Minding the gap between cortisol levels measured with second-generation assays and current diagnostic thresholds for the diagnosis of adrenal insufficiency: a single-center experience.” Hormones 2020; DOI: 10.1007/s42000-020-00185-y
  6. Vogeser M, Kratzsch J, Ju Bae Y, Bruegel M, Ceglarek U, Fiers T, Gaudl A, Kurka H, Milczynski C, Prat Knoll C, Suhr AC, Teupser D, Zahn I, Ostlund RE. “Multicenter performance evaluation of a second generation cortisol assay.” Clin Chem Lab Med 2017; 55(6): 826–835 15.
  7. Ueland G, Methlie P, Øksnes M, Thordarson HB, Sagen J, Kellmann R, Mellgren G, Ræder M, Dahlqvist P, Dahl SR, Thorsby PM, Løvås K, Husebye ES. “The short cosyntropin test revisited: new normal reference range using LC-MS/MS.” J Clin Endocrinol Metab 2018; 103(4):1696–1703