Context: Long-term safety of growth hormone (GH) treatment is an area of much debate. Studies including children treated with GH not only due to GHD, but also due to non-GHD causes like idiopathic short stature or like short stature in children born small for gestational age have suggested that GH treatment is associated with increased mortality or stroke.
Objective: To study the impact of GH replacement on overall and cause-specific mortality in childhood-onset GHD (CO GHD) patients.
Design: A nationwide population-based registry study on patients with CO GHD and general population controls matched on age and gender. Mortality hazard ratios (HRs) were computed comparing patients and controls, and comparing GH-replaced patients and non-GH-replaced patients, using Cox regression. Comparing GH- and non-GH-replaced patients HRs were adjusted for birth year, year of diagnosis, gender, irradiation, ACTH insufficiency and primary disease.
Patients and controls: A total of 494 patients with CO GHD each matched with 100 general population controls were included.
Results: Mortality was substantially increased comparing patients with CO GHD and general population controls, HR = 7·51 (95% CI = 6·06-9·31). Comparing GH-replaced patients with non-GH-replaced patients mortality was significantly decreased in total (HR = 0·27, CI = 0·17-0·43) and due to malignancy (HR = 0·14, CI = 0·07-0·28) in GH-replaced patients. Adjusting for relevant confounders, this decrease remained significant both in total (HR = 0·56, CI = 0·32-0·96) and due to malignancy (HR = 0·33, CI = 0·16-0·69). Overall and cause-specific mortality was increased in both GH-replaced and non-GH-replaced patients compared to general population controls, but mortality was generally highest in non-GH-replaced patients.
Conclusion: The present data from a national cohort of patients with CO GHD do not support the suggestion that GH replacement is associated with increased mortality.
© 2015 John Wiley & Sons Ltd.