Measuring cortisol in serum, urine and saliva - are our assays good enough?
Level 5 - mechanism / opinion, no new human data
Narrative review of laboratory methods without systematic search or primary empirical data
PubMed 28068807 · doi:10.1177/0004563216687335
What was done
This narrative review synthesized the analytical methods (primarily automated immunoassays and liquid chromatography–tandem mass spectrometry [LC-MS/MS]) used to measure cortisol in serum, urine, and saliva across various clinical scenarios, including cortisol excess, adrenal insufficiency, acute illness, and neonates.
What was found
The abstract reports that ~10% of total serum cortisol is biologically active and unbound. Standard automated immunoassays measure total cortisol, which can give misleading results when cortisol-binding proteins vary, and suffer from poor specificity and high inter-assay variability. LC-MS/MS improves specificity and sensitivity for serum and saliva, but diagnostic cut-offs for the short Synacthen and Dexamethasone suppression tests remain unvalidated for LC-MS/MS. Salivary cortisol tracks free serum cortisol reliably but lacks a single validated reference range and standardized assays. 24-hour urine free cortisol correlates well with mean serum-free cortisol in hypercortisolemic states. Beyond the ~10% free fraction estimate, no specific quantitative comparative metrics were reported in the abstract.
Why it matters
Understanding the limitations of immunoassays and the matrix-specific nuances of cortisol testing is critical for avoiding diagnostic errors in adrenal disorders, particularly in patients with altered binding proteins or acute illness.
Limits
This is a narrative review that does not employ systematic review methods, meta-analytic pooling, or formal risk-of-bias assessments. The abstract reports no empirical sample sizes or quantitative performance comparisons (such as analytical sensitivity or cross-reactivity percentages).
Cited by
- supports Salivary cortisol measurements over multiple daily time points reflect stress response and cortisol dynamics more accurately than a single transient serum cortisol draw.