Growth hormone treatment for childhood short stature and risk of stroke in early adulthood

OBJECTIVES: We investigated the incidence of stroke and stroke subtypes in a population-based cohort of patients in France treated with growth hormone (GH) for short stature in childhood. METHODS: Adult morbidity data were obtained in 2008-2010 for 6,874 children with idiopathic isolated GH deficiency or short stature who started GH treatment between 1985 and 1996. Cerebrovascular events were validated using medical reports and imaging data and classified according to standard definitions of subarachnoid hemorrhage, intracerebral hemorrhage, and ischemic stroke. Case ascertainment completeness was estimated with capture-recapture methods. The incidence of stroke and of stroke subtypes was calculated and compared with population values extracted from registries in Dijon and Oxford, between 2000 and 2012. RESULTS: Using both Dijon and Oxford population-based registries as references, there was a significantly higher risk of stroke among patients treated with GH in childhood. The excess risk of stroke was mainly attributable to a very substantially and significantly higher risk of hemorrhagic stroke (standardized incidence ratio from 3.5 to 7.0 according to the registry rates considered, and accounting or not accounting for missed cases), and particularly subarachnoid hemorrhage (standardized incidence ratio from 5.7 to 9.3). CONCLUSIONS: We report a strong relationship between hemorrhagic stroke and GH treatment in childhood for isolated growth hormone deficiency or childhood short stature. Patients treated with GH worldwide should be advised about this association and further studies should evaluate the potentially causal role of GH treatment in these findings.