The Effect of Early Intervention on Long-Term Developmental Outcomes of Children with Autism Spectrum Disorders: A Systematic Review and Meta-Analysis

Background: Early intervention has been shown to improve short-term developmental outcomes in children with autism spectrum disorders (ASD), but little is known about its effect on long-term outcomes.

Objectives: This meta-analysis aimed to evaluate the long-term effects of early intervention on multiple developmental outcomes of children with ASD.

Methods: We searched PubMed, EMBASE, PsycINFO, Scopus, the Cochrane Library, OVID, ERIC and Web of Science from inception through April 24, 2018 for relevant full-text articles in English. Studies were selected if they investigated the long-term developmental outcomes of early intervention in 6-to-18-year-old children with ASD; they were excluded if they targeted subjects with a medical complications or drug treatment. The quality of included studies was examined by the Cochrane risk of bias tool and the Risk Of Bias In Non-randomized Studies-of Interventions assessment tool. Meta-analyses were performed on studies judged sufficiently similar and appropriate to pool using fixed-effect or random-effect models. The standard mean difference (SMD) was used as the effect measure: up to 0.5=small effect, 0.51 to 0.79=moderate effect, and larger than 0.8= large effect. Additionally, we conducted stratified analysis to explore the potential differences in effects of the various intervention methods on children with ASD.

Results: A total of 8 pre-post studies and 9 between-group studies were selected, with aggregate data on 634 non-overlapping children with ASD. Among these studies, 11 conducted the Application of Behavioral Analysis therapy (ABA), 2 performed the Early Start Denver Model (ESDM), and the others used comprehensive interventions. The quality assessment showed that 6 of included studies were low risk of bias. Meta-analysis demonstrated that early intervention can prominently enhance the IQ level (SMD= 0.94, 95% CI: 0.57-1.32), expressive language (SMD= 1.12, 95% CI: 0.70-1.53), and receptive language (SMD= 1.07, 95% CI: 0.73-1.42) in children with ASD. In addition, early intervention can improve certain domains of adaptive behavior, including socialization (SMD= 0.59, 95% CI: 0.02-1.16), social (SMD= 0.75, 95% CI: 0.35-1.16) and communication (SMD= 0.85, 95% CI: 0.62-1.08). However, the results indicated that early intervention has no significant long-term effect on adaptive behavior composite (SMD= 0.19, 95% CI: -0.32-0.71), as well as on ASD symptoms (SMD= -0.61, 95% CI: -1.22-0.01). Moreover, the stratified analysis indicated that ABA was more effective than other methods in reducing symptoms (pooled SMD= -1.33, 95% CI: -2.28 to -0.39). In contrast, ESDM displayed the largest effects in improving the IQ level of affected children (pooled SMD= 1.37, 95% CI: 0.95-1.80).

Conclusions: Our study demonstrated that early intervention improves the long-term prognosis of children with ASD, especially for the cognitive function, language and communication. Furthermore, ABA and ESDM show different effects in improving certain long-term outcomes. These findings may have implications for developing customized interventions that are targeted at improving specific developmental outcomes in children with ASD.