Background: The diagnosis of GH deficiency (GHD) in children and adolescents is classically established when GH concentrations fail to reach an arbitrary cut-off level (usually 710 μg/l) after at least two provocative tests (PT).
Objective and hypotheses: Aim of the study was to define optimal GH cut-offs to different PT in children and adolescent with short stature.
Method: This was a retrospective study in 437 subjects who underwent GH secretion studies for short stature, after exclusion of other causes for their shortness. GH and IGF1 were measured by the same chemiluminescence assay in all samples (Immulite, Siemens). Patient group (P) consisted of 126 patients (66 boys, 60 girls, aged 11.4±3.7): 121 organic or genetic GHD and five subjects with peak GH <8.5 μg/l in two PT, H-SDS <−3 SDS and IGF1 SDS <−2 SDS. Control group (C) consisted of 311 subjects (196 boys, 115 girls, aged 9.8±4.5) with normal GH response to at least one PT. The PT used were Arginine (154 C, 90 P), insulin tolerance test (ITT; 74 C, 86 P) and Clonidine (173 C, 26 P). All PT were performed between 08.00 and 09.00 h after fasting overnight. Receiver operating characteristic (ROC) analysis and likelihood ratio (LR) were used to evaluate the optimal GH cut-offs and the diagnostic accuracy of the PT.
Results: ROC analysis showed that optimal GH cut-off for Arginine test is 8.2 μg/l (Sens=83.3%, Spec=81.2%, LR=4.43), for ITT is 5.3 μg/l (Sens=79.1%, Spec=87.8%, LR=6.5) and for Clonidine test is 8.6 μg/l (Sens=88.5%, Spec=90.2%, LR=9). ROC analysis showed that IGF1 SDS has low accuracy in diagnosing GHD (AUC=0.69).
Conclusion: The results of ROC analysis showed that the cut-off limits which discriminate between normal and GHD are lower than those commonly employed, and differ according to the stimulation test. IGF1 is characterized by low diagnostic accuracy.
20 - 22 Sep 2014
European Society for Paediatric Endocrinology