The Views of Autistic Adults on Early Autism Interventions: A Mixed-Methods Systematic Literature Review.
Elizaveta Nosova, Alexandra Sturrock, Neil Humphrey and 1 others
Across 19 studies, autistic adults consistently said early support should respect the child's autonomy, be play-based and enjoyable, accept autism rather than target traits, and change the environment rather than the child.
They also reported significant harm from compliance-focused behavioural interventions.
Key findings
1The qualitative synthesis produced five themes: Hear the child's voice (autonomy, assent, tailoring to the child), Let the child be a child (developmentally appropriate, enjoyable, play-based), Celebrate autistic lives (acceptance over normalisation, authentic communication, involving autistic adults), Understand autistic challenges (hidden difficulties, accessible environment, supporting the community around the child), and Damage Done (a cross-cutting theme capturing harm from past interventions, including power imbalance, bodily autonomy violations, and compliance training).
2Behavioural interventions were strongly discouraged in qualitative experiential studies, with accounts of being physically restrained, having objects removed, and being trained to comply; in quantitative surveys, ratings were mixed, with some strategies (planned ignoring, food rewards) rated unacceptable and others (token boards, behavioural momentum) rated neutral or positive.
3Goals rated high priority across studies included safety, well-being, and adult support; goals rated low priority or deemed unacceptable by more than 10% of participants included autism traits and play.
Still to come
How it was doneWhat they foundWhat it means for OTsWhat it means for SLPs
Read the rest of this summary
You get three full summaries a month, free, and we do not ask for a card. Search, the TL;DRs and your library stay unlimited either way.
What it does not show
Wide focus and generous inclusion criteria produced substantial heterogeneity across the 19 studies, limiting the depth of synthesis (the authors' own stated limitation). Participants in the larger quantitative studies were typically women, white or European, with higher education and an average diagnosis age of 24–25 years; people who actually receive early interventions are more likely to be boys, diagnosed earlier, and have higher support needs or learning disability, so their views are underrepresented. Online recruitment limits access to those who use the internet, can read, and can spend 20–25 minutes completing a survey. ABA was defined and interpreted differently across studies; only one study (Jonkman et al., 2025) collected specific information about intervention type, outcomes, and practices, making it hard to know exactly what participants experienced. Not all autistic adults who participated have direct experience with young children or current early support systems, so some views may reflect general principles rather than practical knowledge of current services. The review could not draw conclusions about any particular educational context, country, or cultural setting due to international sampling.
Declared interests
The research is self-funded. Two authors (K.L. and N.H.) conduct research into the development and evaluation of early autism interventions, and one author (A.S.) conducts research into social communication support for autistic students. The author team is theoretically aligned with neurodiversity-affirming approaches to autism interventions, and some are engaged in primary research to develop interventions that are more acceptable to autistic people.
The easy way to misread this
Do not read this as evidence that all behavioural interventions are harmful or that ABA does not work. The review found strongly negative accounts in qualitative experiential studies but mixed ratings in quantitative surveys, with some specific strategies (token boards, behavioural momentum) rated neutral or positive by autistic adults. This is a review of views and past experiences, not a trial of intervention effectiveness, and the negative accounts come from a subset of participants who had direct experience with specific practices. Conversely, do not read the mixed quantitative ratings as endorsement: the qualitative data consistently describe harm, and the two data sources are not in agreement.
Summarised by AI from the full paper, without a clinician reviewing it. Check it against the source before it changes what you do. Read it on PubMed →