The use of insulin-like growth factor 1 reference values for the diagnosis of growth hormone deficiency in prepubertal children

OBJECTIVE: This study was done to determine whether the use of reference values obtained in children with idiopathic short stature (ISS) improved the clinical value of serum insulin-like growth factor I (IGF-1) as a tool for diagnosing GH deficiency (GHD) in prepubertal children. PATIENTS AND METHODS: Serum IGF-1 was measured with a new IRMA kit (IGFI-RIA CT, Cis Bio, Gif sur Yvette, France) in 168 prepubertal normal children and in prepubertal children with ISS (n = 68), organic GHD due to a craniopharyngioma (oGHD, n = 15) and permanent idiopathic GHD (iGHD, n = 28). RESULTS: IGF-1 was lower (P < 0.001) in iGHD than in either ISS or oGHD and was below the fifth percentile of the normal range in 29/68 ISS (43%), 8/15 oGHD (53%) and 28/28 (100%) iGHD patients. Three oGHD (20%) and two iGHD (7%) patients had a serum IGF-1 below the fifth percentile of the normal group but above the fifth percentile of the ISS group. Thus, a serum IGF-1 below the fifth percentile of the normal group distinguished between normal children and iGHD with 100% sensitivity, between normal and oGHD with 53% sensitivity and between normal and all GHD (idiopathic + organic) with 84% sensitivity; the overall specificity was only 57%. Conversely, a serum IGF-1 below the fifth percentile of the ISS population distinguished between ISS and iGHD with 93% sensitivity, between ISS and oGHD with 33% sensitivity and between ISS and all GHD with 72% sensitivity; the overall specificity was then 95%. CONCLUSIONS: A serum IGF-1 within the normal range virtually excludes idiopathic GHD but does not rule out organic GHD, whereas an IGF-1 below the ISS range is strongly in favour of GHD, after exclusion of poor nutritional status and/or liver disease. An IGF-1 below the normal range but in the idiopathic short stature range gives no definitive conclusion even when it is associated with a low GH peak. Thus, whereas reference values obtained in normal children must be used to interpret serum IGF-1 in short prepubertal children, reference data obtained in idiopathic short stature children should also be taken into account.