Overnight sampling for growth hormone (GH) is a research tool for quantifying characteristics of spontaneous GH secretion. However, the study is costly in assays and blood volume, particularly that required from a small child.
Design and patients: Existing overnight GH data from 126 normal children and from 227 children with GH deficiency or short stature were reanalyzed, examining 6-h and 4-h segments of this data for accuracy in representing each child's 12-h GH secretion. The goal was to see whether the test could be made shorter and more practical without losing accuracy.
Results: The 6-h segment 2200-0400 h consistently contained the majority of GH peaks. Correlation was high between GH values from 2200-0400 h and from the 12-h period. Normal 95% confidence limits (CL) for GH during 2200-0400 h were derived from data in normal children for gender and each pubertal stage. Data from short children were compared with the normal 6-h 95% CL. In short children, GH values low for 12-h were also low for 6-h. Only a few children with normal 12-h values (1.5% of normals, 0.5% with short stature) had GH values outside 95% CL for 6-h.
Conclusions: Six-hour GH sampling (2200-0400 h) is accurate and cost-efficient compared to the 12-h overnight GH study. These studies are primarily useful in research settings.