[Provocative tests in the diagnosis of childhood onset growth hormone insufficiency].

INTRODUCTION The incidence of short stature associated with growth hormone deficiency has been estimated to be about 1:4000 to 1:10000. It is the main indication for treatment with recombinant growth hormone. OBJECTIVES The aims of the study were to evaluate the results of growth hormone stimulation tests and identify the growth hormone deficiency predictors. MATERIAL AND METHODS A cross-sectional, analytical and observational study was conducted. We studied all the children and adolescents submitted to growth hormone pharmacological stimulation tests between January 2008 and May 2012. Growth hormone deficiency diagnosis was confirmed by two negatives growth hormone stimulation tests (growth hormone peak < 7 ng/ml). The statistical analysis was performed using student t-test, chi-square, Pearson correlation and logistic regression. Statistical significance determined at the 5% level (p ≤ 0.05). RESULTS Pharmacological stimulation tests were performed in 89 patients, with a median age of 10 [3-17] years. Clonidine (n = 85) and insulin tolerance test (n = 4) were the first growth hormone stimulation tests performed. Growth hormone deficiency was confirmed in 22 cases. In cases with two growth hormone stimulation tests, the growth hormone peak showed a moderate correlation (r = 0.593, p = 0.01). In logistic regression model height (z-score) and the growth hormone peak in first stimulation test were predictors of growth hormone deficiency diagnosis (each one unit increase in z-score decrease the growth hormone deficiency probability). DISCUSSION Measurement of IGF-1 cannot be used in diagnosing growth hormone deficiency. CONCLUSION Auxological criteria associated with a positive test seems to be a reliable diagnostic tool for growth hormone deficiency.