Evidence for a hypothalamic-pituitary versus adrenal cortical effect of glycemic control on counterregulatory hormone responses to hypoglycemia in insulin-dependent diabetes mellitus

The epinephrine and cortisol responses to hypoglycemia are reduced in insulin-dependent diabetes mellitus (IDDM) patients in strict glycemic control. However, it is not known whether these abnormalities are mediated at a central (hypothalamic-pituitary) or peripheral (adrenal) level. To examine this question, we measured counterregulatory hormone secretion during a 3-h hypoglycemic hyperinsulinemic clamp (12 pmol/kg.min) that lowered glucose from 5.0 to 2.2 mmol/L in steps of 0.55 mmol/L every 30 min in 13 well controlled IDDM subjects (hemoglobin A1, 7.8 +/- 0.2%), 14 poorly controlled IDDM subjects (hemoglobin A1, 12.3 +/- 1.5%), and 20 healthy volunteers. Basal levels of ACTH, cortisol, and epinephrine were similar in all 3 groups before hypoglycemia. At the nadir glucose level (2.2 mmol/L), ACTH, cortisol, and epinephrine levels were significantly lower in well controlled IDDM compared to healthy controls, and the glucose levels required for significant secretion of ACTH, cortisol, and epinephrine also were lower in well controlled IDDM compared to those in both poorly controlled IDDM and healthy volunteers (P < 0.05). During hypoglycemia, ACTH levels were significantly correlated with cortisol levels (r = 0.43; P < 0.05). Because adrenomedullary epinephrine synthesis is partially dependent on adequate adrenocortical function, we also determined whether the blunted epinephrine response might result from the reduced cortisol secretion. Eleven of the control subjects underwent a second identical insulin clamp study during which metyrapone was administered to produce adrenal cortical blockade. Despite higher basal ACTH levels after metyrapone and sustained elevations in ACTH during hypoglycemia, the cortisol response was abolished during metyrapone treatment, indicating effective blockade. However, epinephrine responses did not differ during hypoglycemia with or without metyrapone treatment. We conclude that 1) ACTH, cortisol, and epinephrine responses during hypoglycemia are reduced in IDDM patients in strict glycemic control; 2) the lower cortisol response is correlated with reduced ACTH levels; and 3) in healthy subjects, the cortisol response to hypoglycemia is abolished by adrenocortical blockade with metyrapone, whereas the epinephrine response to hypoglycemia remains intact. These data suggest that central adaptations in hypothalamic-pituitary responses to hypoglycemia rather than alterations in adrenal gland function per se underlie the reduced co