OBJECTIVES Obesity with abdominal body fat distribution (A‐BFD) and hypothalamic‐pituitary‐adrenal (HPA) axis activity are somehow linked, but the exact interactions still need clarification. Obese subjects display normal circulating plasma cortisol concentrations with normal circadian rhythms. However, when the HPA axis is pharmacologically challenged, body fat distribution matters and then A‐BFD obese women differ from those with subcutaneous body fat distribution (P‐BFD). We hypothesized that lower dose provocative and suppressive tests than those used to diagnose hypercortisolism of tumour origin or adrenal insufficiency would shed some light on the characteristics of the HPA axis activity in relation with body fat distribution.

PATIENTS AND METHODS Fifty premenopausal obese women were grouped according to their body fat mass distribution. Their plasma cortisol responses to (i) two low doses of dexamethasone (0·25 and 0·5 mg) with (ii) low dose of the ACTH analogue tetracosactrin (1 µg) were assessed. Salivary cortisol was also determined during the ACTH test.

RESULTS A‐BFD differed from P‐BFD women in terms of HPA axis responsiveness. They had comparatively: (i) increased nocturnal cortisol excretion (9·38 ± 2·2 vs. 6·82 ± 0·91 nmol/µmol creatinine, A‐BFD vs. P‐BFD, respectively, P = 0·03); (ii) increased salivary cortisol response to ACTH stimulation (1 µg) [salivary cortisol peak: 33·4 (14·1–129) vs. 28·5 (13·2–42·8) nmol/l; salivary AUC: 825 (235–44738) vs. 537 (69–1420) nmol/min/l; A‐BFD vs. P‐BFD, P = 0·04 for both]; and (iii) increased pituitary sensitivity to dexamethasone testing [postdexamethasone (0·25 mg) plasma cortisol levels: 163 (26–472) vs. 318 (26–652) nmol/l and postdexamethasone (0·5 mg) plasma cortisol levels: 26 (26–79) vs. 33 (26–402) nmol/l; A‐BFD vs. P‐BFD, P = 0·01 for both).

CONCLUSIONS These data demonstrate differences in the HPA axis activity and sensitivity to glucocorticoids between obese women differing in their body fat distribution, with both enhanced negative and positive feedback in those with abdominal obesity. Several mechanisms may explain these differences: central vs. peripheral hypotheses. Thus, abdominal obesity does not appear to be linked solely to one pathophysiological hypothesis.