Effect of growth hormone therapy on height in children with idiopathic short stature: a meta-analysis

CONTEXT: Use of growth hormone (GH) therapy to promote growth in children with idiopathic short stature is controversial. A fundamental issue underlying the controversy is uncertainty about the magnitude of effectiveness of GH for this condition. OBJECTIVE: To determine the effect of GH on short- and long-term growth in idiopathic short stature. STUDY DESIGN: Systematic review of controlled and uncontrolled studies. DATA SOURCES: MEDLINE (1985-2000), key journals, cross-referencing of bibliographies, abstract booklets, and experts. STUDY SELECTION AND DATA EXTRACTION: We performed a meta-analysis of all studies satisfying the inclusion criteria for idiopathic short stature: initial height below the 10th percentile, normal stimulated GH levels (>10 microg/L), absence of comorbid conditions, no previous GH therapy, treatment with biosynthetic GH, and inclusion of major outcome measures. PRIMARY OUTCOME MEASURES: Growth velocity and height SD score (number of SDs from mean height for age and sex) at baseline and after 1 year to evaluate the short-term effect of GH. Adult height was analyzed to evaluate the long-term effect of GH. DATA SYNTHESIS: Ten controlled trials (434 patients) and 28 uncontrolled trials (655 patients) met the inclusion criteria. While baseline growth velocities were equivalent at baseline, 1-year growth velocity of the GH-treated group significantly exceeded that of controls by 2.86 cm/y. Similarly, in uncontrolled trials, growth velocity increased after 1 year, and height SD score increased from -2.72 at baseline to -2.19. In controlled studies, the adult height of the GH-treated group significantly exceeded controls by 0.84 SD, and in uncontrolled trials the adult height attained after GH treatment (-1.62 SDs) exceeded that predicted at baseline (-2.18 SDs). These results suggest an average gain in adult height of approximately 4 to 6 cm (range, 2.3-8.7 cm) with GH therapy. Given current treatment costs, this corresponds to more than $35 000 per inch (2.54 cm) gained in adult height in idiopathic short stature. CONCLUSIONS: Treatment with GH results in short-term increases in growth for children with idiopathic short stature, and long-term GH can increase adult height. These results are fundamental to decisions about GH use and raise questions about the goals of treatment. Use of GH for idiopathic short stature in clinical practice will depend on its efficacy in promoting growth and the value of this effect to families, physicians, and third-party payers.