Attachment-Based Play Therapy Shows Promise for Autistic Children With Developmental Trauma

Autistic children in the care system face a distinctive double burden that researchers and clinicians have long struggled to address. They carry the organic interpersonal and sensory differences that define autism, and many have also endured the caregiver abuse, neglect, and separation that constitute developmental trauma and attachment disruption. A new retrospective service evaluation published in the Journal of Child & Adolescent Trauma suggests that a structured, play-based psychotherapeutic intervention may help this vulnerable cohort, with outcomes that appear broadly comparable to those seen in non-autistic children with similar trauma histories.

The study, conducted at Chrysalis Associates, a specialist developmental trauma and attachment service in the north of England, evaluated a 30-hour neuro-collaborative Theraplay-informed intervention delivered between April 2020 and April 2024. The therapeutic package comprised three equal components: ten hours of Theraplay-informed family therapy involving parents and children, ten hours of therapeutic parenting sessions guided by the principles of dyadic developmental parenting, and ten hours of sensory regulation work delivered directly to children or psycho-educationally to parents. This integrative design deliberately expands on Perry’s neurosequential model, which proposes treating trauma by targeting the brainstem first, then the limbic system, and finally the cortex. Instead of moving sequentially, the neuro-collaborative approach addresses multiple brain systems concurrently, pairing body-based regulation with attachment-focused play and reflective parenting support.

The researchers compared 35 autistic care-experienced children with 172 non-autistic children who received the same intervention. The autistic group had an average age of just over nine years, ranged from four to seventeen, and attended on average more than 96 percent of their scheduled sessions. Outcomes were measured before and after therapy using four well-validated caregiver-report instruments: the Trauma Symptom Checklist for Young Children, the Child Behaviour Checklist, the Behaviour Rating Inventory of Executive Function, and the Assessment Checklist for Children. Paired t-tests and analyses of covariance controlling for age and gender were used to assess change within and between the groups.

The results told a nuanced story. Across the full cohort, every outcome measure improved significantly, and after applying a Bonferroni correction for multiple comparisons, total post-traumatic stress, post-traumatic stress arousal, total behavioural problems, and oppositional defiant problems remained significantly reduced, with effect sizes ranging from small to approaching medium. The larger non-autistic group alone also showed statistically significant improvements on most measures. The smaller autistic group did not reach statistical significance, but scores moved consistently in the direction of improvement with encouraging effect sizes, including a moderate effect for reduced rule-breaking behaviour. Crucially, the analyses of covariance found no significant difference in treatment response between autistic and non-autistic children once pre-treatment scores, age, and gender were accounted for.

The theoretical rationale for the intervention rests on how early relationships shape developing regulatory systems. Drawing on Schore’s account of right-hemisphere development and Porges’ polyvagal theory, the authors argue that attachment experiences sculpt the autonomic and nonverbal systems underlying emotional attunement, bodily awareness, and social engagement. Theraplay, a structured, playful, and largely preverbal approach organised around the domains of structure, engagement, nurture, and challenge, is designed to intervene precisely at this nonverbal, embodied level. Because autistic children often process sensory input and social cues differently, the therapy was adapted flexibly, honouring the child’s developmental stage and sensory profile rather than demanding eye contact or touch that could prove dysregulating.

The authors also invoke monotropism theory, which describes the intensely focused attentional tunnel characteristic of autism, to explain how trauma and autistic hypersensitivity can interact. A child whose nervous system is hypervigilant for danger after trauma, and whose attention is already narrowly channelled by autistic monotropism, may become overwhelmed by stimuli inside that channel while missing cues outside it. The intervention’s sensory regulation component, drawing on the Just Right State program and the Alert Program, helps parents create predictability and reduce overload, while Theraplay-informed sessions teach carers to synchronise with the child’s attention rather than fight against it, building experiences of being seen, understood, and co-regulated.

Not every measure responded identically across groups. On the Assessment Checklist for Children, which probes attachment-related relational behaviours, the non-autistic group improved significantly while the autistic group’s scores on indiscriminate and non-reciprocal scales remained stable or rose slightly. The researchers argue this divergence is expected and even reassuring: non-reciprocal and indiscriminate friendliness in trauma-exposed children typically reflects damaged internal working models, which attachment-rich therapy can repair, whereas analogous presentations in autistic children stem from innate neurobiological differences in social reciprocity. Attempting to modify those differences therapeutically would be neither appropriate nor neurodiversity-affirming, and the authors caution that the checklist itself may simply be insufficiently sensitive to capture relational change in autistic children.

The study’s significance lies partly in dismantling a persistent clinical misconception. Some practitioners have assumed that Theraplay’s emphasis on touch and eye contact makes it unsuitable, or even harmful, for autistic children. This evaluation found no evidence of deterioration in functioning and instead suggests the model can be adapted in a neurodiversity-affirming way to achieve trauma-related outcomes comparable to those of non-autistic peers. Prior scoping reviews had already found that interventions relying less on verbal skills and emphasising caregiver involvement, both hallmarks of Theraplay, suit neurodivergent children who have experienced interpersonal trauma. The present study extends that literature to the specific and understudied intersection of autism, developmental trauma, and attachment disruption in care-experienced children.

Nevertheless, the authors are candid about the limitations inherent in practice-based evidence. The evaluation was retrospective rather than randomised, the autistic subgroup was small, the integrative design makes it impossible to isolate which component drove change, and all outcome data relied on caregiver report vulnerable to expectancy effects. Formal treatment fidelity checks were not routinely conducted, dropout information was not collected, and the generic psychometric measures were not adapted to autistic symptomatology. Changes in executive function and attachment-related outcomes, while directionally positive and at times showing moderate effect sizes, did not reach significance, likely reflecting insufficient statistical power.

The research team concludes that the 30-hour neuro-collaborative Theraplay-informed intervention is effective for children with developmental trauma and attachment disruption, and may also benefit autistic children, though larger, adequately powered trials are needed to confirm this. They call for qualitative research capturing the lived experiences of autistic children, their carers, and their therapists, alongside the development of autism-sensitive assessment tools. For now, the message for services supporting care-experienced autistic children is cautiously optimistic: attachment-based, sensory-informed, neurodiversity-affirming therapy deserves a place in the treatment conversation, and services are encouraged to evaluate and publish their own anonymised data to build the collective evidence base.

Subject of Research: Evaluation of a neurocollaborative Theraplay-informed psychotherapeutic intervention for developmental trauma and attachment difficulties in autistic looked after children

Article Title: The Impact of a Neurocollaborative Theraplay®-Informed Intervention on the Presentations of Developmental Trauma and Attachment Difficulties in a Sub-Group of Autistic Looked after Children: a Retrospective Service Evaluation

Article References: Purrington, J., Hartley, G., Price, A., Bass, E., Godfrey, C., Lynch, J., & Mukherjee, R. (2026). The Impact of a Neurocollaborative Theraplay®-Informed Intervention on the Presentations of Developmental Trauma and Attachment Difficulties in a Sub-Group of Autistic Looked after Children: a Retrospective Service Evaluation. Journal of Child & Adolescent Trauma. https://doi.org/10.1007/s40653-026-00964-6

Image Credits: AI Generated

DOI: 10.1007/s40653-026-00964-6

Keywords: autism, developmental trauma, attachment disruption, Theraplay, looked after children, sensory regulation, therapeutic parenting, neurodiversity, psychotherapy outcomes, service evaluation, child and adolescent trauma, care system