The activity of GH/IGF-I axis in anorexia nervosa and in obesity: a comparison with normal subjects and patients with hypopituitarism or critical illness.

GH/IGF-I axis activity changes have been reported both in anorexia nervosa (AN) and in obesity (OB). AN is characterized by GH hypersecretion and very low IGF-I levels as a result of undernutrition and acquired peripheral GH resistance. On the other hand OB is a GH hyposecretory state but IGF-I levels are generally preserved. The activity of GH/IGF-I axis in AN and OB has never been directly compared with that of other pathophysiological conditions such as hypopituitarism and critical illness in which a reduction of both GH and IGF-I secretion has been demonstrated. To this aim, we evaluated IGF-I levels and both basal and GHRH (1 microgram/kg) IV-induced GH secretion in 20 female patients with anorexia nervosa (mean age: 19.1 +/- 0.8 years) and in 15 female and 5 male patients with simple obesity (mean age: 39.0 +/- 3.0 years). We then compared the results with those of hypopituitaric patients with severe GH deficiency (GHD), including 10 female and 10 patients (mean age: 32.0 +/- 2.1 years), and with 4 female and 7 male patients with critical illness (CRI) following multiple trauma 72 hours after ICU admission (mean age: 59.2 +/- 1.2 years). Twenty-six normal subjects (NS) including 14 female and 12 male patients (mean age: 37.8 +/- 3.7 years) were studied as controls. Basal IGF-I levels in AN patients (93.5 +/- 11 micrograms/L) were lower (p < 0.001) than in the NS (201.7 +/- 13.5 micrograms/L) and OB (194.5 +/- 28.6 micrograms/L), which, in turn, were similar. IGF-I levels in AN patients were lower than in CRI patients (162.8 +/- 17.4 micrograms/L) and higher than in GHD patients (76.7 +/- 13.5 micrograms/L) but these differences did not attain statistical significance. Basal GH levels in AN patients (7.6 +/- 2.5 micrograms/L) were higher (p < 0.001) than in NS (1.8 +/- 0.3 micrograms/L), OB patients (1.1 +/- 0.5 micrograms/L), CRI patients (1.8 +/- 0.5 micrograms/L) and GHD patients (0.3 +/- 0.1 microgram/L), which were the lowest (p < 0.01). The GHRH-induced GH rise in AN patients (AUC: 2032.9 +/- 253.5 micrograms/L/h) was three fold higher (p < 0.001) than in NS (662.1 +/- 80.3 micrograms/L). On the other hand in OB (332.4 +/- 74.7 micrograms/L/h) the GH response to GHRH was similar to that in CRI (199.6 +/- 98.8 micrograms/L/h); both were clearly higher (p < 0.01) than in GHD patients (25.1 +/- 5.2 micrograms/L/h) but lower (p < 0.01) than in NS. These findings demonstrate that in AN patients, in spite of a clear increase of both basal and GHRH-induced GH secretion, IGF-I synthesis and release are as markedly impaired as in patients with panhypopituitarism and severe GHD. On the other hand in OB and in CRI, IGF-I synthesis and release are preserved despite marked impairment to GHRH-induced GH secretion. These results reinforce the major role of nutrition in conditioning the activity of GH/IGF-I axis in different patho-physiological states.