To evaluate the usefulness of IGF-1 and IGFBP-3 for the diagnosis of growth hormone deficiency (GHD), we studied a group of GHD, short non GHD (NGHD) and normal prepuberal boys between the ages of to 5-9 years. The GHD group (n=14, 10 boys) was defined as height <2SD for age, growth velocity <10th percentile for age, and a maximum GH response to 2 stimulation tests ≤7 ng/ml (mean±SD: 1.2±0.6), one of which was insulin. NGHD(n=28, 22 boys), was defined as height <2SD for age, growth velocity <25th percentile for age, and a maximum GH response to 2 stimulation tests >7 ng/ml (16.3±6.9). 26 (17 boys) normal school children with average heights were studied as controls (C). IGF-1 was measured with a polyclonal RIA, and IGFBP-3 with DSL IRMA. Other underlying diseases were excluded. Results are shown in the Table as mean ±SD.

Table 1
Full size table

IGF-1 was significantly lower in GHD and NGHD than in controls. IGFBP-3 was significantly lower in GHD only. We defined the normal range as ±2SD of the control group. Using this criteria, IGF-1 was low in 11/14 GHD (79% sensitivity), and in 16/28 NGHD (42% specificity). IGFBP-3 was low in 11/14 GHD (79% sensitivity), and in 2/28 NGHD (93% specificity). Both tests were normal in 2 GHD, suggesting that these patients may be NGHD, and abnormal in 1 NGHD, suggesting that this patient may be GHD. We conclude that: 1.- Although we have used an arbitrary cut off point to diagnose GH deficiency after the GH stimulation tests, IGFBP-3 appears to be more specific than IGF-1 to make this diagnosis. When both IGF-1 and IGFBP-3 are considered together they improve their diagnostic accuracy. 2.- According to these results, we may have erroneously classified 3 patients based upon GH stimulation tests. 3.- IGF-1 and IGFBP-3 are important tools for the diagnosis of GHD.