Timing and Frequency of Serum Cortisol and Salivary Cortisone Sampling Required to Estimate 24-Hour Cortisol Exposure

Presentation Number: MON 380
Date of Presentation: April 3rd, 2017

Richard J Ross*1, Robert F Harrison2, Martin J Whitaker3, Miguel Debono4 and Brian G. Keevil5
1University of Sheffield, United Kingdom, 2University of Sheffield, 3Diurnal Limited, 4University of Sheffield, Sheffield, United Kingdom, 5University Hospital of South Manchester, Manchester, United Kingdom


Context: Measuring cortisol exposure is important as inadequate levels are associated with adrenal crisis and excess with all the complications of Cushing’s syndrome. Salivary cortisone now provides a non-invasive method for measuring cortisol(1), and we have further investigated the use of salivary cortisone to estimate 24-hour cortisol exposure. Objective: To examine the timing and frequency of serum cortisol and salivary cortisone measurement required to estimate the AUC over 24 hours of serum cortisol. Participants & method: cortisol exposure was assessed in 14 male volunteers by the sampling of hourly serum cortisol throughout 24 hours and salivary cortisone hourly between 15:00h and 22:00h and between 07:00h and 15:00h. All cortisol measurements were assayed by LCMSMS. The relationship between the estimated AUC (eAUC) for serum cortisol and the absolute AUC (AUC24), generated from the full 24 hour measurements, was investigated by deriving the relative standard deviation (coefficient of variation) for the use of 1 to 4 sampling points in the 24 hours. For salivary cortisone, values were converted to serum cortisol using a previously published fixed effects model(1). Results: If a single measurement of serum cortisol was used to estimate the AUC the average CV of the different time points was 20% (range 7.4 - 39%); highest at 06:00h and lowest at 24:00h. When two serum samples at 12 hour intervals were used the average CV was 14% (range 8.7 - 23%). When 3 serum samples, at 8 hourly intervals, were used the average CV was 10% (range 5.2 - 14%). When 4 serum samples, at 6 hourly intervals, were used the average CV was 8% (range 5.4 - 10%). Single salivary cortisone measurement showed similar high CVs ranging from 15 - 40% and was again highest in the early morning. For twice daily salivary cortisone CV range was between 10 - 20% and as sampling was interrupted for saliva it was only possible to compute approximate 8 hourly sampling of 15:00, 22:00 & 07:00h which had a CV of 17%. Conclusion: A single measurement of serum cortisol or salivary cortisone is a poor measure of 24-hour cortisol exposure especially when taken first thing in the morning. Three samples taken at approximately 8 hourly intervals give an eAUC which is within 80-120% of the AUC24. This could represent a simple, cheap method of estimating the physiological cortisol exposure in an out-patient or research setting.


Disclosure: RJR: Director, Diurnal. MJW: Director, Diurnal. Nothing to Disclose: RFH, MD, BGK