Is growth hormone hypersecretion in diabetic adolescent girls also a daytime problem?

Glycaemic control often deteriorates during puberty in girls with insulin dependent diabetes mellitus (IDDM). This may be due in part to the normal psychosocial changes associated with adolescence. Puberty is, however, also characterized by rapid somatic development, orchestrated by hormonal changes. Some of these hormones play a major role in glucose homeostasis. We have examined the insulin-GH-IGF-I axis in 11 adolescent girls with poorly controlled insulin dependent diabetes and compared the data with those of 10 non-diabetic girls matched for age, pubertal stage and body mass index (BMI).

Serum profiles of glucose, insulin, GH and IGF binding protein 1 (IGFBP1) were analysed in addition to IGF-I in serum and nocturnal urinary excretion of GH.

Serum glucose, insulin and IGFBP1 were measured every hour for 24 h, whereas GH in serum was measured every 30 minutes during the same period. Nocturnal urinary GH was analysed as a mean of three consecutive nights.

The insulin profiles of the IDDM patients were flat with low post-prandial peaks, corresponding to only one-third of the peaks of the non-diabetic girls. The integrated insulin levels, both during 24-h sampling and during daytime, were significantly lower in the diabetic group. There were no differences during night-time. The diabetic patients had elevated mean baseline levels of serum GH (IDDM 2.8 +/- 0.5 mU/l, controls 0.7 +/- 0.2; P < 0.001), a higher 24-h mean serum GH level (9.8 +/- 1.7 mU/l vs. 4.4 +/- 0.7; P < 0.001), significantly more peaks and a urinary GH excretion twice as high as in the non-diabetic group. An interesting observation was the finding of marked differences in daytime GH concentrations between the groups, both regarding overall integrated levels (GH AUC 103 +/- 15.8 and 35.9 +/- 7.1 mU/l x 12 h, respectively; P < 0.005) as well as baseline levels (3.8 +/- 0.6 mU/l vs. 0.7 +/- 0.2; P < 0.001). In contrast, during night-time only the mean basal levels of GH differed. The level of IGF-I was reduced in the diabetic group compared with the healthy controls (IDDM 233 +/- 19 micrograms/l vs. controls 327 +/- 21; P < 0.005). In addition, the IDDM patients had significantly increased concentrations of IGFBP 1, but kept a normal diurnal rhythm with a pronounced night peak.

Hypoinsulinaemia in adolescent IDDM patients, particularly in the portal hepatic circulation, results in decreased IGF-I and increased IGFBP 1 production in the liver. High levels of IGFBP 1 may, in turn, reduce the bioactivity of IGF-I even further. Low levels of IGF-I will lead to increased GH secretion. Earlier studies on the relationship between GH and diabetic control have focused on elevated GH levels during the night. In this study we have observed markedly elevated levels of GH also during daytime in adolescent IDDM patients. This indicates increased insulin resistance and insulin demand also during the day in diabetic subjects. The increased insulin resistance may result in hyperglycaemia leading to additional insulin resistance. A vicious circle may thus be induced, accelerating metabolic impairment in poorly controlled adolescent IDDM girls.