Comparative Effectiveness of Non-pharmacological Interventions for Autism Spectrum Disorder: A Network Meta-analysis Using the Social Responsiveness Scale
Felipe Gouhie1, Daniela Fagundes1, Lucas Figueira Vieira1, Caio Nazar Cunha1, Arthur Gheno1
1Federal University of Uberlândia
Objective:
To compare the effectiveness of non-drug interventions for autism through a network meta-analysis.
Background:
 Autism Spectrum Disorder (ASD) is a neurological condition marked by distinct patterns of social interaction and communication. The heterogeneity and complexity of its symptoms make diagnosis and treatment strategy development challenging. Although several non-drug therapies have been proposed, their effectiveness remains uncertain, warranting comparative evaluation.
Design/Methods:

A comprehensive search of PubMed, Embase, and Cochrane databases identified randomized clinical trials evaluating non-drug interventions — defined as therapies not involving prescription psychotropic medications — including behavioral, digital, and nutritional approaches aimed at improving Social Responsiveness Scale (SRS) scores. Studies not meeting these criteria were excluded. Data were extracted, and analyses were performed using R Studio 4.3.1. Appropriate models accounted for heterogeneity, and pooled estimates were generated. Results were interpreted considering potential biases and limitations using the Risk of Bias 2 (RoB2) tool.


Results:

Fifty-six studies, comprising 117 treatment arms and 3,261 participants, were included. The network meta-analysis showed that, compared with placebo, the most effective interventions for reducing SRS scores were metaverse-based skills training (MD: –21.74; 95% CrI [–37.46, –6.28]), vitamin B12 supplementation (MD: –8.97; 95% CrI [–21.01, 3.04]), and high-dose vitamin A (MD: –4.13; 95% CrI [–17.00, 9.35]). Only metaverse-based skills training demonstrated a statistically significant improvement over placebo.


Conclusions:

Metaverse-based skills training was the only non-drug therapy to significantly improve SRS scores compared to placebo, highlighting its potential as a promising intervention for ASD. However, current evidence on efficacy and safety is preliminary; thus, routine use of these approaches, including nutritional supplementation, cannot yet be recommended.

10.1212/WNL.0000000000217026
Disclaimer: Abstracts were not reviewed by Neurology® and do not reflect the views of Neurology® editors or staff.