Growth hormone and ghrelin responses to an oral glucose load in adolescent girls with anorexia nervosa and controls

Abstract

Anorexia nervosa (AN) is associated with high levels of GH and low levels of IGF-I suggestive of a nutritionally acquired lack of GH action or GH resistance. The suppression of GH levels after administration of inhibitors of GH secretion such as oral glucose is the definitive test to distinguish normal from pathological states of GH excess, such as acromegaly. However, suppression of GH by glucose has not been well characterized in states of adaptive GH excess, such as AN, especially in a younger adolescent population with relatively higher GH levels, compared with adults. In this study, we investigated GH suppression after a 100-g oral glucose load over a 1-h period in 19 adolescent girls with AN and 20 healthy controls of similar chronologic and bone age. We also compared nocturnal GH secretion characteristics by deconvolutional analysis in both groups to determine differences in secretory patterns between adolescents whose GH values suppressed vs. those whose values did not after oral glucose. Fasting levels of ghrelin, a GH secretagogue, and suppression of ghrelin with oral glucose were also determined to assess whether GH suppression or nonsuppression could be related to ghrelin values at respective time points. At 0 min (0′) of the oral glucose tolerance test, girls with AN had significantly lower levels of glucose (P = 0.009) and higher levels of GH (P = 0.04) than controls. Nadir GH values were higher in AN than in controls (2.0 ± 1.8 vs. 0.5 ± 0.5 ng/ml, P = 0.001). Only 31.6% of girls with AN suppressed their GH values to 1 ng/ml or less vs. 85.0% of healthy adolescents (P = 0.0005). All healthy controls had nadir postglucose GH values of 2 ng/ml or less. Nadir GH concentrations during the oral glucose tolerance test correlated directly with all measures of GH secretion [basal (r = 0.37, P = 0.02), pulsatile (r = 0.56, P = 0.0002), and total (r = 0.57, P = 0.0002)]. Adolescent girls who did not suppress their GH values to 1 ng/ml or less had significantly higher levels of ghrelin at 0′, 30′, and 60′ (P = 0.02, 0.004, and 0.008), significantly higher GH at 0′ (P = 0.001), and higher nocturnal basal (P = 0.002), pulsatile (P = 0.05), and total GH secretion (P = 0.03) than those who did suppress below this level. Ghrelin values were higher in AN than in controls at each time point (P = 0.02, 0.0002, and 0.01 at 0′, 30′, and 60′) but did not predict GH values at these time points. Adolescent girls with AN fail to adequately suppress their GH values after a 100-g oral glucose load. This lack of suppression may be related to the higher GH secretion seen in adolescents with this disorder. In contrast, all healthy adolescents suppress their GH values to 2 ng/ml or less but not 1 ng/ml or less after a glucose load. Although ghrelin values are higher in AN than in controls, we could not demonstrate a relationship between ghrelin and GH values. The inability of healthy girls to uniformly suppress GH levels to 1 ng/ml or less, a normal level defined for adults, may be related to higher GH secretion in the pubertal years, compared with adult life. Further studies are needed to define GH suppression in an adolescent population.