Validation of the growth hormone (GH)-dependent marker method of detecting GH abuse in sport through the use of independent data sets

Context: The detection of exogenously administered growth hormone (GH) poses a formidable challenge but a detection method based on the measurement of two GH-dependent markers, IGF-I and type 3 pro-collagen (P-III-P) has been proposed. The measurement of multiple markers in conjunction with discriminant functions can improve the sensitivity and specificity of detection compared with single marker analysis. Objective: To provide further validation of the GH-dependent marker approach. Design: Analysis of discriminant function scores for GH detection on independent datasets. Setting: Two independent (GH-2000 and Kreischa) double blind, placebo controlled, hGH administration studies. Subjects: Healthy active male volunteers. Intervention: GH-2000 proposed a discriminant function involving IGF-I and P-III- P while the Kreischa function involved IGF-I, P-III-P and IGFBP-3. After adjustment for assay differences the formulae were applied to the other dataset. Outcome measures: Ability to detect GH use in independent datasets using a predefined specificity of approximately 1 in 10000. Results: The GH-2000 formula was able to detect 90% of those receiving GH in the Kreischa study at one or more time points during the study period. This sensitivity was similar to that obtained on the original GH-2000 dataset. The Kreischa formula correctly identified 41% of individuals receiving GH in the GH-2000 study. Conclusions: The study provides further validation that the test proposed by GH-2000 based on IGF-I and P-III-P concentrations can be used to detect subjects receiving exogenous GH. (c) 2007 Elsevier Ltd. All rights reserved.