Fractures in children with cerebral palsy

Introduction: We studied the fracture history in a large population of patients with cerebral palsy to determine which children were at the highest risk for fracture. Methods: The International Classification of Diseases (Ninth Revision) coding identified 763 children with cerebral palsy. Patients and caregivers were contacted for information about fracture history and risk factors for low bone density. Of the 763 children identified, 418 children (54.8%) were available for this study; 243 (58%) had quadriplegia, 120 (29%) diplegia, and 55 (13%) hemiplegia. Three hundred sixty-six children were spastic, 23 mixed tone, 13 athetoid, and 16 classified as others. We identified 50 children (12%) who fractured; 15 of these same children had, over time, multiple fractures. Results: The number of fractures showed a normal distribution by age, with a mean of 8.6 (SD, 4.0). Children with cerebral palsy with mixed tone had a higher rate of fracture (chi(2) = 14.7, p < 0.01); chi(2) analysis indicated that the children who fractured were, as a group, more likely to use a feeding tube, have a seizure disorder, take valproic acid (VPA), and use standing equipment in therapy. Multiple regression analysis demonstrated older age and VPA use as predictive of fracture and gave the following equation: fracture = -0.01 + (VPA x 0.17) +(age x 0.15). The subgroup that sustained multiple fractures were older at the time of first fracture than the children who had only one reported fracture (t = -2.3, P < 0.05). Conclusions: The main finding of our article is that older age at first fracture and use of VPA are predictive of fractures and define a group of children with cerebral palsy who may benefit from treatment interventions to increase bone density.