Plasma free cortisol in depressive illness--a review of findings and clinical implications

Measurements of a variety of parameters of biologically active cortisol indicate that a small number of patients with depressive illness have significantly elevated levels of unbound plasma cortisol. Abnormalities in corticosteroid binding globulin do not account for the hypercortisolemia, and elevated urinary levels of free cortisol confirm the data obtained for plasma free cortisol. Direct measurements of free cortisol suggest that the absence of physical effects of cortisol in patients with depression is related to the very mild elevation in plasma levels of free cortisol in these patients and that high levels are not sustained throughout the day as they are in patients with pathologic glucocorticoid excess. The apparent discrepancy between elevated total cortisol levels and the mild elevation of unbound plasma cortisol is best explained by the generous binding capacity of corticosteroid binding globulin. Total cortisol levels of greater than 25 micrograms/dl are necessary to saturate the binding sites of corticosteroid binding globulin. Only then can free cortisol be detected in plasma. Although these studies suggest that unbound plasma cortisol is elevated, whether these levels represent biologically active cortisol in other tissue sites as well, particularly the central nervous system, is still unclear. It is well accepted that discrete binding sites for cortisol and dexamethasone occur in hypothalamic and pituitary tissue, but the mechanism of glucocorticoid regulation of these binding sites has not been elucidated. Cortisol levels in cerebrospinal fluid correspond closely to plasma levels of unbound cortisol. Ruf and Steiner have shown that corticosteroid responsive neurons are present in the periventricular gray matter of the third ventricle, where other investigators have identified active ependymal transport mechanisms for cortisol.(ABSTRACT TRUNCATED AT 250 WORDS)