Risperidone versus aripiprazole fracture risk in children and adolescents with autism spectrum disorders.

Abstract

Risperidone and aripiprazole, commonly used antipsychotics in children with autism spectrum disorder (ASD), have previously been associated with elevated fracture risk in other populations. The aim of this study was to evaluate and compare the risk of fracture among children with ASD using risperidone or aripiprazole. This was a retrospective, propensity-score matched cohort study, set between January 2013 and December 2018. We used the MarketScan Medicaid insurance data, which covers multiple states of the United States. We included ASD children aged 2-18 years, who were new users of aripiprazole or risperidone and with no prior history of antipsychotic use or fractures. The main exposure was the continued use of aripiprazole or risperidone. The incidence rates of any fracture during follow-up were evaluated, and the risk between aripiprazole and risperidone was compared via Cox-proportional hazard models. Results were stratified by age, sex, duration of exposure and fracture site. In total, 3312 patients (78% male; mean [SD] age 11.0 [3.7] years) were identified for each cohort. Over the full duration of follow-up, fracture incidence rates per 1000 patient-years were 23.2 for risperidone and 38.4 for aripiprazole (hazard ratio and 95% confidence interval: 0.60 [0.44-0.83]). Risks were similar between cohorts throughout the first 180 days on treatment, but significantly higher in the aripiprazole group thereafter. Extremity fractures drove most of the increased risk, with the biggest differences in lower leg and ankle fractures. Differences widened for children aged 10 years or younger (HR [95% CI]: 0.47 [0.30-0.74]). In conclusion, compared to aripiprazole, risperidone was associated with 40% lower risk of fracture. Further analysis on the mechanism and long-term bone health of antipsychotic-treated children with ASD is warranted. LAY SUMMARY: We compared the risk of bone fractures among 6624 children with autism spectrum disorder (ASD), half of whom used risperidone and half of whom used aripiprazole. Taking other factors into account, risks were similar between the two groups throughout the first 180 days on treatment, but significantly higher in the aripiprazole group thereafter. The biggest differences were in lower leg and ankle fractures. Overall, compared with aripiprazole, risperidone was associated with 40% lower risk of fracture.