Growth hormone secretory pattern and response to treatment in children with short stature followed to adult height

OBJECTIVE To compare the relative utility of GH stimulation tests and assays of spontaneous GH secretion as predictors of change in height standard deviation score at the end of GH treatment in children with short stature. PATIENTS AND METHODS We retrospectively studied 116 children (67 boys and 49 girls) with subnormal growth rates and short stature, defined as a height of more than 2SD below the mean for age and sex. The patients were classified according to their pattern of findings on baseline pharmacological GH stimulation tests and a 12-h assay of nocturnal spontaneous GH secretion. Twenty-eight patients (24%) had normal hormone levels by both methods (group I); 14 (12%) had normal levels by stimulation tests but subnormal levels by the physiological assay (group II); 48 (41%) had subnormal levels on pharmacological stimulation, with normal physiologic levels (group III); and 26 (22%) had subnormal levels by both methods (group IV). All children in groups II and IV, and 27 in group III, designated IIIb, were treated with recombinant GH at 0.7 U (0.23 mg/kg) of body weight per week. GH secretory patterns were related to final height SD scores and other growth parameters, after the patients had attained their adult stature 6.7 +/- 2.2 years (SD) after GH evaluation. RESULTS The five groups were similar with respect to mean baseline height SD scores for chronological as well as bone age. Whether assessed as absolute or parentally adjusted (relative) values, mean gains in height SD scores were significantly greater in treated patients with physiologicial hormone deficiency (groups II and IV) than in those with normal hormone levels (group I, untreated controls). Relative height gains were 1.03 +/- 1.45 cm (6.6 +/- 9.28 cm) and 1.85 +/- 1.21 cm (SDS; 11.8 +/- 7.74 cm) in groups II and IV respectively, compared with only 0.11 +/- 0.42 cm (0.7 +/- 2.68 cm) in group I (P < 0.01 and P < 0.001). GH treatment failed to improve either the absolute or parentally adjusted final height of patients with GH deficiency by stimulation tests but normal levels by physiological assay. CONCLUSION Long-term administration of GH to short children with normal spontaneous GH secretion is not associated with an appreciable increase in adult height.