Serum free cortisol as an ancillary tool in the interpretation of the low-dose 1-μg ACTH test

Objective Serum free cortisol, rather than serum total cortisol (TC), determines glucocorticoid activity in vivo, but how the considerable inter-subject variation in ambient serum free cortisol affects the outcome of dynamic hypothalamic-pituitary-adrenal (HPA) assessment in noncritically ill subjects is unknown. Design, patients and measurements We performed the low-dose 1-μg ACTH test in 75 subjects referred for HPA evaluation. Serum TC was determined by a chemiluminescence method, and serum free cortisol was measured by the same method following equilibrium dialysis. In a subset of these patients, salivary cortisol was also measured. Results Mean fraction of free cortisol was 5·07 ± 4·08% (±SD; range 1·77-10·1%). Although no correlation was seen between TC and the fraction (%) of free serum cortisol, a positive correlation existed between baseline total and free cortisol (R = 0·539 P = 0·01), as well as between peak ACTH-stimulated total and free cortisol (R = 0·619; P = 0·01). There was no correlation between baseline salivary cortisol and serum free cortisol and between peak ACTH-stimulated salivary and serum free cortisol. Using the lowest attained peak serum free cortisol in subjects whose TC response to ACTH was normal (≥500 nm), the minimal 'pass' level for normal serum free cortisol response to 1 μg ACTH was set at 25·0 nm. Five of the 19 subjects showing subnormal TC response to 1 μg ACTH had normal serum free cortisol response. Conclusions Discrepancies between the peak free and TC were noted mostly for subjects whose ACTH-stimulated TC peaked between 440 and 580 nm. At this range, the measurement of serum free cortisol allows further refinement of the assessment of borderline responses to 1-μg ACTH.