Reliability of stimulated and spontaneous growth hormone (GH) levels for identifying the child with low GH secretion.

The reliability of stimulated and spontaneous GH levels for identifying the child with low GH secretion has been the subject of debate. We compared the ability of GH concentrations after pharmacological stimulation with levodopa and clonidine and of spontaneous peak and 12-h pooled GH concentrations during sleep on a single night to estimate the maximum spontaneous GH secretion from 2 nights in 55 children, aged 5-16 yr, with heights below the 3rd percentile and/or height velocities below the 25th percentile for age, who had two consecutive overnight GH secretory profiles. Maximum stimulated GH concentrations correctly categorized 80% of children who had maximum spontaneous GH concentrations above and below 4 micrograms/L using a double monoclonal immunoradiometric assay for GH (Tandem-R HGH, Hybritech). The remaining 20% of children had stimulated GH concentrations below but spontaneous GH concentrations above 4 micrograms/L. Using this cut-off, the maximum GH concentrations from the first and second nights correctly categorized 98% and 95% of the children, respectively. Night to night variation in GH secretion was low in children who had low spontaneous GH secretion (maximum spontaneous peak and pool GH concentrations, less than 4 and less than or equal to 0.7 micrograms/L, respectively), and pooled GH concentrations from the 2 nights were concordant in 98% of the cases. We conclude that it is not uncommon for stimulated GH concentrations to underestimate spontaneous GH secretion. Even without acclimatization to the hospital setting, measurement of spontaneous GH secretion on a single night was more reliable for identifying the child with low endogenous GH secretion than was GH stimulation testing alone.