Background: Serum total cortisol responses to the short Synacthen® test (SST) are used to confirm or refute a diagnosis of adrenal insufficiency. Salivary cortisol measurement offers a non-invasive alternative and has the potential advantage of evaluating free, bioavailable cortisol. However, reference ranges and its performance in the investigation of suspected hypoadrenalism (SH) have not been firmly established.
Aims: 1. To establish the salivary cortisol response to SST in healthy volunteers (HV), patients with low protein concentration (secondary to nephrotic syndrome or liver cirrhosis; NS-C) and women taking an oestrogen-containing oral contraceptive pill (OCP).
2. To evaluate the performance of a lower reference limit (LRL) determined in HVs in patients with SH.
Methods: An SST was undertaken in 139 HVs, 29 patients with SH, 24 OCP-females and 10 NS-C patients. Salivary cortisol was measured using an in-house liquid chromatography-tandem mass spectrometry assay. The mean and LRL (mean-1.96SD) of the salivary cortisol response to Synacthen® were derived from log-transformed concentrations; the LRL was applied as a diagnostic cut-off in SH patients.
Results: The 30-minute salivary cortisol response to SST in HVs showed a non-Gaussian distribution, with no significant gender difference. Concentrations ranged between 10.1 and 39.7 nmol/L, with mean concentration 19.3 nmol/L and LRL 10.3 nmol/L. Mean concentrations in OCP-females (19.7 nmol/L) and NS-C patients (19.0 nmol/L) were not different from HVs (P=0.69 & P=0.75, respectively). The LRL correctly identified all patients with hypoadrenalism (7/7) and incorrectly diagnosed hypoadrenalism in 4/22 unaffected patients.
Conclusions: This study provides a reference range and estimated LRL for the salivary cortisol response to Synacthen® in healthy volunteers. The LRL has 100% sensitivity and 82% specificity when applied as a cut-off in patients with suspected hypoadrenalism. Salivary cortisol responses to the SST may offer an alternative to serum measurement and may be especially useful in patients with altered protein states, although assay-specific reference ranges are needed.