OBJECTIVE We studied growth rate, bone density, and bone metabolism in patients affected by type I osteogenesis imperfecta (OI) with quantitative defect in type I collagen synthesis during treatment with human growth hormone (hGH), being aware of its collagen-stimulating synthesis activity in vitro. METHODS Fourteen patients (6 boys; ages 4.8 to 10.8 years) were studied. Any structural alteration in the collagen chains was excluded, and reduced production of structurally normal type I collagen (increase in type III/type I collagen; reduction in the messenger ribonucleic acid alpha 1 (I)/ alpha 2 (I) ratio) was demonstrated. The patients were divided into two groups comparable in sex, age, height, and clinical severity of OI; seven patients (three boys) were treated for 12 months with hGH at a dosage of 0.2 mg/kg per week (0.6 IU/kg per week), in six injections subcutaneously, and seven were followed as control subjects. Auxologic data were measured every 3 months, and bone age was determined at the start, after 1 year of treatment, and 1 year after its completion. Every 3 months, serum insulin-like growth factor type I, osteocalcin, carboxyterminal propeptide of type I procollagen, alkaline phosphatase, calcium, and phosphorus levels and urinary hydroxyproline and calcium levels were determined. Bone mass measurements were carried out at the start of the study in all patients and repeated after 12 months in treated patients at the lumbar spine by dual-energy x-ray absorptiometry and by anteroposterior (second, third, and fourth lumbar vertebrae) and lateral (third lumbar vertebra) scan. Results were expressed as areal (anteroposterior and lateral) bone density (in milligrams per square centimeter) and as calculated true density (in milligrams per cubic centimeter). RESULTS After 12 months, linear growth velocity in treated patients increased significantly in comparison with the pretreatment period (from 3.57 +/- 0.55 to 6.04 +/- 0.69 cm/yr; p < 0.05) and with the untreated group (p < 0.05). Bone age did not advance faster than chronologic age. The fracture index per year was low before treatment, and during therapy no patient had any fractures. Serum osteocalcin levels were statistically lower than in control subjects before treatment and increased significantly after 12 months (3.3 +/- 1.0 vs 2.1 +/- 0.9 nmol/L; p < 0.05). Serum levels of carboxyterminal propeptide of type I procollagen were significantly lower than normal values before treatment (164.6 +/- 46.7 vs 310.3 +/- 97.6 ng/ml; p < 0.05) and rose, but not significantly, during and after treatment. Before therapy, patients with OI had significantly lower lumbar anteroposterior, lateral, and calculated true bone density than the normal population of the same sex compared for both age and height. After hGH treatment, bone density increased significantly in the lumbar spine, in anteroposterior and lateral scans (+2.6 +/- 2.5% and +9.8% +/- 14.0%, respectively; p < 0.05). CONCLUSIONS From our results, we conclude that hGH treatment in moderate OI does not increase the fracture risk in treated patients in the short term, significantly increases the rate of linear growth velocity, and increases bone turnover and mineral content in trabecular bone at the lumber spine.