Summary
A medRxiv systematic review of 39 studies found that mental-health treatment outcomes for autistic people varied by intervention. Where significant differences were reported, they generally involved smaller improvements rather than no benefit.
A systematic review posted as a medRxiv preprint found that mental-health treatment outcomes for autistic people differed across interventions and services. Across 39 studies covering 4,787,730 participants, significant differences between autistic and non-autistic groups generally meant smaller improvements for autistic people—not an absence of benefit.
The review compared outcomes for autistic and non-autistic people receiving the same mental-health interventions or services. Its findings point to variation between types of care, rather than a single pattern that applies across mental-health treatment as a whole.
What the review examined
The researchers included 17 studies of psychological therapies and 22 of multicomponent or service-level approaches. They searched academic databases, grey literature, trial registries and preprint servers for work published from 1 January 1992 to 30 April 2026. Results were synthesised narratively, with random-effects meta-analyses used where appropriate.
The review found that some larger, higher-quality studies reported less favourable outcomes for autistic participants, including less improvement in symptoms or functioning, and higher rates of restrictive practices or acute service use. Other interventions showed comparable or mixed outcomes across studies. In some comparisons, small samples made findings harder to interpret.
These differences matter because a treatment or service designed around non-autistic people may not produce the same outcomes for autistic people. The review’s findings support examining whether care is accessible and appropriate, while avoiding the assumption that all interventions have the same results.
How strong and inclusive was the evidence?
The authors rated 10 studies (25.6%) as low quality, 22 (56.4%) as medium quality and seven (17.9%) as high quality. Nearly half of the studies were based in the UK, and fewer than a quarter included people with intellectual disabilities.
The review’s Autism-Inclusive Research Assessment also raised concerns about how autistic people were represented in the research. These included limited involvement of autistic people and a lack of validated or adapted outcome measures. Such measures are important for judging whether changes reported in studies accurately reflect outcomes for the people receiving care.
The authors conclude that mental-health outcomes vary by intervention and outcome type, and call for more accessible care and higher-quality, co-produced research across a wider range of countries and service settings. Because this is a preprint systematic review and the included studies varied in quality and size, its results are best read as evidence of differences across particular comparisons—not as a single verdict on mental-health treatment for autistic people.