Although the secretion of growth hormone (GH) is widely accepted as being increased in diabetes mellitus, this conclusion appeared to us to be the result of bias in case selection. Newly diagnosed and hospitalized diabetics cannot be assumed as representative of the clinically stable diabetic state. In view of the postulated importance of growth hormone to the vascular complications of diabetes, we re-examined this problem in 58 clinically stable diabetics and 38 age-matched controls. GH was measured in serum samples collected hourly during a 24-h period of normal activity.
Control and clinically stable diabetic subjects of normal weight had similar mean serum GH concentrations for 24 h. In subjects of normal weight, values were 4.4 ± 0.6 ng/ml for women of the control group and 4.0 ± 0.8 ng/ml for diabetic women (P = NS) and values in men were 2.4 ± 0.3 ng/ml in controls and 2.1 ± 0.3 ng/ml in diabetic men (P = NS). Obese controls and obese diabetics had similar, but significantly lower values than nonobese subjects.
When hyperglycemia (125 mg to 175 mg/dl) was induced in 12 control subjects for 24 h, glycemia was obtained comparable to that seen in a select group of nine diabetics. Mean serum concentrations of GH were significantly less in the glucose-infused controls than in both euglycemic controls and diabetics.
These data do not support an absolute increase of circulating GH in stable diabetics; GH levels appear to be normal in this group but inappropriate for the level Of glycemia.