Leptin inhibits appetite by activating several neuroendocrine systems, including the hypothalamo-pituitary-adrenal cortical (HPA) axis. In turn, chronically elevated glucocorticoids increase circulating leptin. HPA axis hyperactivity occurs in 30-50% of patients with major depression, but the few prior reports of leptin measurements in this illness have shown inconsistent results. We, therefore, measured plasma leptin in 12 female and 8 male unipolar major depressives and 12 female and 8 male individually matched normal controls administered low-dose physostigmine (PHYSO) and arginine vasopressin (AVP) to stimulate the HPA axis. The subjects underwent four test sessions 5-7 days apart: PHYSO (8 mug/kg IV); AVP (0-08 U/kg IM); PHYSO+AVP; and saline control. Serial blood samples were taken before and after pharmacologic challenge and analyzed for leptin, ACTH(1-39), cortisol and AVP. Estradiol and testosterone also were measured at each test session. PHYSO and AVP produced no side effects in approximately half the subjects and predominantly mild side effects in the other half, with no significant patient-control differences. Correlations between side effects (absent or present) after PHYSO or AVP and the corresponding leptin responses were non-significant in all groups. Baseline plasma leptin concentrations (mean+/-S.D.) were significantly higher in the female patients compared to the female controls (22.5 +/- 13.9 ng/ml vs. 12.3 +/- 9.7 ng/ml), whereas they were similar in the male patients and the male controls (3.9 +/- 1.4 ng/ml vs. 3.6 +/- 2.0 ng/ml). Leptin concentrations following PHYSO remained unchanged from baseline, indicating that the short-lived ACTH and cortisol increases produced by PHYSO did not affect leptin secretion. In contrast, AVP administration, while also increasing ACTH and cortisol, significantly suppressed leptin, more so in the women than in the men. Baseline leptin and the leptin decrease after AVP were moderately positively correlated with the Hamilton Depression Scale 'somatization' factor in the female patients (r=0.50) and more strongly correlated with the 'mood-depression' factor in the male patients (r=0.81). These findings indicate a sexual diergism (functional sex difference) in plasma leptin measures between major depressives and matched normal controls. (C) 2002 Elsevier Science Ireland Ltd. All rights reserved.