Summary: A comprehensive umbrella review of 248 meta-analyses examining complementary, alternative and integrative medicines (CAIMs) for autism found no robust evidence to support their effectiveness. The review covered 19 intervention types — from music therapy and animal-assisted interventions to probiotics and vitamin D — but most results were unreliable and evaluations of safety were sparse.
Given that up to 90% of autistic people report trying at least one CAIM in their lifetime, the authors warn that families and clinicians should not rely on single, low-quality studies. The research team also developed a publicly accessible online platform to help users explore and interpret the accumulated evidence.
Key Facts
- Scope: The review reanalyzed 248 meta-analyses derived from 53 meta-analytic reports, which collectively included 200 clinical trials with more than 10,000 participants.
- Quality of Evidence: Most CAIMs yielded weak or very low-quality evidence; none had high-quality support for improving core or associated symptoms of autism.
- Safety Reporting: Fewer than half of the treatments had any reported data on acceptability, tolerability or adverse events.
Source: University of Southampton
Overview
Researchers from Paris Nanterre University, Paris Cité University and the University of Southampton published an umbrella review in Nature Human Behaviour that systematically evaluated the clinical evidence for CAIMs in autism. Their search covered five databases up to 31 December 2023 and aimed to provide a clear, consolidated picture of what the body of research actually shows.
The team focused on 19 categories of intervention, including animal-assisted interventions, acupuncture, herbal medicine, music therapy, probiotics and vitamin D, among others. Instead of assessing individual trials in isolation, they reestimated the meta-analyses using consistent statistical methods and rated each meta-analysis for methodological quality using AMSTAR-2 and for certainty of evidence using an algorithmic adaptation of the GRADE framework.
To make the results accessible to clinicians, autistic people and families, the researchers also launched an online platform that consolidates their findings and presents them in an easy-to-navigate format.
Autism commonly affects social communication and interaction, sensory processing, anxiety, and repetitive behaviors, all of which can reduce quality of life. Because many people seek CAIMs hoping for benefits with fewer side effects than conventional treatments, the authors stress the importance of relying on high-quality evidence before adopting these approaches.
Professor Richard Delorme, Head of the Child and Adolescent Psychiatry Unit at Robert Debré Hospital in Paris, notes that while CAIMs are often pursued with good intentions, careful assessment of rigorous randomized trials is necessary before recommending them.
Dr. Corentin Gosling, Associate Professor at Paris Nanterre University and first author, emphasizes that reviewing meta-analyses rather than individual trials allowed the team to evaluate the comprehensive evidence base across multiple interventions. He adds that the online tool is intended to support shared decision-making between autistic people and practitioners.
Although some interventions showed preliminary promise, the effects were generally supported by low- or very low-quality evidence, undermining confidence in the findings. Alarmingly, safety and tolerability were rarely reported: less than half of the reviewed CAIMs included assessments of adverse events or acceptability, a critical gap that limits clinical guidance.
Professor Samuele Cortese, NIHR Research Professor at the University of Southampton and co-senior author, cautions against drawing conclusions from single, small, or poorly conducted studies: robust decision-making requires considering the full body of evidence and the quality of that evidence.
The authors recommend that future research place a stronger emphasis on rigorous trial design, consistent outcome measures, and systematic reporting of harms to build a trustworthy evidence base for any CAIMs being considered for autism.
The interactive platform presenting the study results is available at: https://ebiact-database.com
Funding: This research was funded by the Agence Nationale de la Recherche (ANR).
COI Statement
S.C. has declared reimbursement for travel and accommodation from the Association for Child and Adolescent Mental Health (ACAMH) in relation to lectures, and has received honoraria from Medice. M.S. has received honoraria or consultancy fees from Angelini, AbbVie, Boehringer Ingelheim, Lundbeck and Otsuka. All other authors report no competing interests.
About this Autism research news
Author: Steven Williams
Source: University of Southampton
Contact: Steven Williams – University of Southampton
Image: The image is credited to Neuroscience News
Original Research: Open access. “Complementary, alternative and integrative medicine for autism: an umbrella review and online platform” by Richard Delorme et al., Nature Human Behaviour.
Abstract
Complementary, alternative and integrative medicine for autism: an umbrella review and online platform
Complementary, alternative and integrative medicine (CAIM) is commonly used by autistic individuals, with surveys indicating up to 90% have tried at least one CAIM. Despite this prevalence, the overall evidence supporting CAIMs for autism has remained unclear.
To address this uncertainty, the authors conducted an umbrella review of systematic reviews with meta-analyses that evaluated CAIM effects across the lifespan in autism, and developed a web platform to share the findings. Five databases were searched through 31 December 2023; eligible papers were identified and data were independently extracted by pairs of investigators.
Each included meta-analysis was reestimated with a consistent statistical approach, its methodological quality was assessed with AMSTAR-2, and the certainty of its findings was appraised using an algorithmic adaptation of GRADE.
The review identified 53 meta-analytic reports, enabling 248 separate meta-analyses covering 19 CAIMs. The overall conclusion is that there is no high-quality evidence supporting the efficacy of any CAIM for core or associated symptoms of autism. While some CAIMs showed potentially promising results, these were based on very low-quality evidence. Safety and adverse event reporting were infrequent, highlighting a key priority for future trials.
To facilitate evidence-based consideration of CAIM interventions, the authors provide an interactive platform that makes the results accessible and easier to interpret: https://ebiact-database.com.