Inter-method comparison of salivary cortisol measurement

Background: Salivary cortisol is increasingly used in the diagnostic work-up of suspected Cushing's disease as well as in stress research. In this study, the agreement of different methods for salivary cortisol measurement was assessed. Methods: Saliva samples from five healthy volunteers were distributed to three routine clinical chemistry laboratories. As a reference, all samples were also analyzed using liquid chromatography-tandem mass spectrometry involving stable isotope labeled cortisol for internal standardization in our laboratory. Results: All immunometric routine methods substantially overestimated salivary cortisol concentrations by approximately 50%. The agreement between these immunoassays was moderate with between-method coefficients of variation of up to 36% for individual samples. Conclusions: Standardization of salivary cortisol measurement requires substantial improvement, in particular by implementation of proficiency testing schemes, a defined reference method, and reference materials. Interpretation of salivary cortisol data in psycho-neuroendocrinological research has to consider the inappropriate level of standardization realized for salivary cortisol measurement to date.