ESPE Abstracts (2015) 84 P-2-446

BMI Negatively Correlates with GH Response to GH Provocation Testing

Ioannis-Anargyros Vasilakisa, Ruth Gauschec, Christoph Begerc, Juergen Kratzschb, Wieland Kiessa, Antje Koernera & Roland Pfaefflea,c


aPaediatric Department, Paediatric Endocrinology and Diabetology Unit, University Hospital Leipzig, Leipzig, Germany; bInstitute for Laboratory Medicine Clinical Chemistry and Molecular Diagnostics, University of Leipzig, Leipzig, Germany; cCrescNet, University of Leipzig, Leipzig, Germany


Background: In adults it has been shown, that GHmax values after provocation testing are negatively correlated to BMI. Preliminary studies in children have found a similar correlation. Consequently children with elevated BMI would be overdiagnosed with GHD. However, studies so far were too small to define this correlation exactly. This would be a condition to judge whether and to what extend adjustments of GH cut-off levels should be considered also in children with elevated BMIs.

Objective and hypotheses: To establish a more precise correlation between BMI and GH peak levels in GH provocation tests in a larger cohort of children with short stature.

Method: Using the local CrescNet database we retrospectively investigated data from 804 children with short stature who were tested between 2004 and 2014. A total of 1109 GH stimulation tests were performed (812 arginine initially, and 297 glucagon as confirmatory test). Children with known syndromes (i.e. UTS), severe chronic illness, or under antipsychotic or sex steroid medication were excluded from study. A linear model was used to assess the correlation between BMI-(SDS) and GHmax. We divided the GH stimulation tests in two groups (prepubertal: 905 and pubertal: 204) to account for possible effects of puberty on GH secretion. Also the correlations of GHmax to BMI-SDS was compared in respect to whether the fixed cut-off level for GHD (7.09 ng/ml) was reached or not.

Results: There was a significant (negative) correlation of BMI-SDS and the GH max levels reached during stimulation tests. This correlation was neither altered when studied according to gender of the probands nor by the substance used for testing. Also, this (negative) correlation was not found to be significantly changed by puberty, although GHmax levels in average were significantly higher in pubertal children. Finally, in prepubertal children no significant difference of this correlations (r2: 0.458, P-value: <2.2e-16) was observed between children with and without GHD (cut-off GH max >7.09 ng/ml).

Conclusion: We found that increasing BMI significantly and negatively influences GH max values in both arginine and glucagon GH stimulation tests. As this effect is similarly observed on ‘both sides’ of established cut-off levels for GHD, BMI should be taken into account when interpreting GH stimulation tests by development of a correction factor. BMI must be taken into account when interpreting GH stimulation tests and the development of a correction factor, could be very helpful in order not to overdiagnose GHD.

Article tools

My recent searches

No recent searches.