Time specific reference intervals for cortisol — ASN Events

Time specific reference intervals for cortisol (#344)

Shalini Mohan 1 , Narelle Hadlow 1 , Suzanne Brown 1 , Wardtrop Robert 1 , David Henley 1
  1. Sir Charles Gairdner Hospital, WA, Perth, WA, Australia

Background: Cortisol varies during the 24 hour period with peak levels at early morning.1 There is a diurnal drop in cortisol over the morning, which is not adjusted for in the current conventional morning reference range. As the decreasing levels are not taken into account, this may lead to indeterminate cortisol levels in diagnosing adrenal insufficiency and unnecessary short Synacthen tests. The prevalence of adrenal insufficiency is low at 0.03%.2
Objective: To quantify the diurnal drop in morning cortisol in a large population, determine time specific reference intervals and assess the utility of multiple of medians (MoMs) for classifying patients.
Methods: We undertook a retrospective analysis of 19320 cortisol measurements collected between 7 am and 12 pm from community pathology from January 2000 to December 2012. We excluded results from daylight savings periods and from subjects with various disease states, extreme cortisol values and specialist referrals. The values were divided into 15 minute intervals and converted to MoMs.
Results: Figure 1 shows the gradual drop in cortisol throughout the morning in subjects. Upper and lower dashed lines denote 97.5 and 2.5 centiles and the moving line represents these values at each time. Figure 2 shows MoM transformed data. Using the time referenced lower limits, 126 (26%) of the patients identified below the conventional lower limit would be classified as normal. Median cortisol decreased by 31 nmol/L each hour from 8-11 am. The MoMs for each 15 minute interval were plotted in Figure 2 and the most notable feature was that the lower limit of the MoMs appeared steady.

1059-Cortisol%20Figure%201%20jpg.jpgFigure 1

1058-Cortisol%20Figure%202%20jpg.jpg

Figure 2

Conclusion: We have quantified the diurnal drop in morning cortisol and developed time specific reference intervals. Classification using MoMs at the lower limit for cortisol more consistently classifies patients and would be a more accurate assessment of adrenal insufficiency.  

  1. Weitzman ED, Gallagher TF, Hellman L et al. Twenty-four hour pattern of the episodic secretion of cortisol in normal subjects. J Clin Endoc 1971;33:14-22
  2. Wiebke Arlt, Bruno Allolio. Adrenal Insufficiency. Lancet 2003;361:1881-93
@ESASRB