Göran Brattsand

Context Adrenal insufficiency (AI) is usually diagnosed by low plasma cortisol levels following a short Synacthen test (SST). Most plasma cortisol is bound to corticosteroid-binding globulin, which is increased by estrogen in combined estrogen-progestin oral contraceptives (COCs). Women with AI using COCs are therefore at risk of having an apparently normal plasma cortisol level during SST, which would not adequately reflect AI. Objective This work aimed to test whether salivary cortisol or cortisone during SST is more robust against the COC effect and to calculate the lower reference limits (LRLs) for these to be used as tentative diagnostic cutoffs to exclude AI. Methods Forty-one healthy women on COCs and 46 healthy women without exogenous estrogens underwent an SST with collection of plasma and salivary samples at 0, 30, and 60 minutes after Synacthen injection. The groups were compared using regression analysis with age as covariate and the LRLs were calculated parametrically. Results SST-stimulated plasma cortisol levels were significantly higher in the COC group vs controls, while mean salivary cortisol and cortisone levels were slightly lower in the COC group. Importantly, COC use did not significantly alter LRLs for salivary cortisol or cortisone. The smallest LRL difference between groups was seen for salivary cortisone. Conclusion Salivary cortisol and especially salivary cortisone are considerably less affected by COC use than plasma cortisol during SST. Due to similar LRLs, a common cutoff for salivary cortisol and cortisone during SST can be used to exclude AI in premenopausal women irrespective of COC use.