Patients with malaria can have features of adrenal insufficiency. Because of the pathophysiological and clinical implications of an Addisonian state, the hypothalamic-pituitary-adrenocortical axis was assessed in nine Vietnamese adults with complicated malaria. A CRH test was performed on admission (in convalescence in five cases) and in six healthy controls. Basal plasma ACTH concentrations in the patients and controls were similar [median (range): 2.9 (0.2-9.7) vs. 3.5 (1.9-13.4) pmol/L, respectively; P > 0.1]. Serum cortisol levels were greater in the patients [882 (294-1682) vs. 190 (110-676) nmol/L; P < 0.01], but three (33%) had values within the control range. Basal serum corticosteroid-binding globulin concentrations were similar in patients and controls (P = 0.23). The post-CRH rise in plasma ACTH was attenuated in the patients [peak: 6.1 (0.9-23.2) vs. 14.5 (6.2-21.5) pmol/L in controls; P < 0.05]; basal and peak plasma ACTH correlated with plasma interleukin-6 in this group (rs > or = 0.60; P < or = 0.04). Serum cortisol responses to CRH were depressed in acute illness [peak 990 (394-1, 805) nmol/L or 10 (0-50%) above baseline vs. 500 (429-703) nmol/L or 160 (10-380%) in controls; P < 0.05]. The median estimated serum cortisol t1/2 was 4.6 h in the patients and 1.6 h in the controls. These data suggest that, relative to a normal stress response, primary and secondary adrenal insufficiency can occur in severe malaria but may be attenuated by increased circulating interleukin-6 concentrations and impaired cortisol metabolism. The benefits of stress-dose corticosteroid replacement are unknown but could be considered in hypoglycemic patients or those with a serum cortisol within or below the reference range.