Objective: We wished to evaluate the use of urinary GH measurements when compared to conventional GH provocation tests in the assessment of short children.
Design: Children presenting for the first time to a regional growth clinic were assessed clinically by one observer. Investigations comprising standard GH provocation tests and measurement of urinary GH were undertaken to exclude GH deficiency.
Patients: Fifty-eight children aged 5.8-16 years were enrolled. Ten were diagnosed on clinical assessment as GH deficient, 43 had delayed growth and/or familial short stature, and five had idiopathic short stature; the 48 children in the last two groups were defined as short normal.
Measurements: GH secretion was evaluated by two standard provocation tests and by the measurement of GH in five overnight urine collections. A normal peak GH concentration was defined as > 16 mU/l. The values for urinary GH excretion were compared to normal ranges (+/- 2 standard deviations from the mean), established in healthy schoolchildren of normal stature.
Results: All children considered GH deficient on clinical grounds had low peak GH concentrations on provocation tests, while 8/10 had low values of urinary GH excretion. All short normal children with normal peak GH concentrations (n = 36) on provocation tests and 11/12 children with low peak GH concentrations had urinary GH excretion within the normal range. There was therefore a significant difference in the classification of 'normal' GH secretion in the two tests. If the clinical diagnosis was used as the standard by which GH tests were judged, the predictive value of a positive urinary GH test in the diagnosis of GH deficiency was 89% compared with 45% for GH provocation tests.