This self-talk exercise may help reduce emotional dysregulation in autistic children

A new study published in Autism Research suggests that helping autistic children strengthen their inner speech may improve their ability to regulate emotions. The pilot trial tested a novel therapy called Thinking in Speech and found promising early evidence that it may help reduce emotional distress in some autistic children.

Autism spectrum disorder is a developmental condition that can affect how people communicate, interact socially, and manage their emotions. Many autistic children experience emotional dysregulation, which means they may have trouble recognizing their feelings, calming themselves down, or expressing emotions in socially accepted ways. These challenges can lead to frustration, anxiety, and difficulty coping with everyday stress.

The researchers behind this study wanted to test whether a therapy focused on building inner speech—the ability to talk silently to oneself—could help autistic children better manage their emotions. Inner speech plays an important role in self-regulation and problem-solving. In non-autistic individuals, it often supports thinking through difficult tasks, calming down in stressful situations, and planning actions. However, prior research has found that inner speech may be less developed in some autistic children, which could make it harder for them to use language as a tool for emotional regulation.

To explore this possibility, the researchers tested an intervention called Thinking in Speech (TiS), which was developed by an autistic speech-language pathologist. Rather than instructing children on what to think or feel, TiS helps children become more self-aware by modeling how to reflect on feelings and use words to handle challenges.

For example, if a child appears frustrated, the therapist might say something like, “This is what hard feels like,” and then help the child find and say a strategy like, “I need help.” Over time, the goal is for the child to internalize this kind of self-talk and apply it independently.

“My wife, Janice Nathan, is a speech-language pathologist who specializes in autism and related neurological disorders,” explained corresponding author Barry R. Nathan. “She had a successful private practice from about 2003 to 2020, when she retired. The mom’s of her autistic clients said they had never seen a therapy like hers, and said we should do a study of it and write a book, and we did both. (It turns out Janice, my wife, is autistic; she received her diagnosis in 2020 during the pandemic. Her brother and mother were autistic as well.)”

The study included 22 verbal children with a formal autism diagnosis, all between the ages of 7 and 11. They were randomly assigned to either begin therapy immediately or to wait ten weeks before starting. This design allowed the researchers to compare emotional changes between those receiving the therapy and those not yet treated. All children eventually received the full course of therapy, which consisted of sixteen 30-minute sessions delivered remotely over 8 to 10 weeks by trained speech-language pathologists.

The participating therapists—nine in total—were certified professionals who underwent an 11-hour training program. The training included presentations, practice sessions, role-playing, and feedback. This training was conducted remotely over a five-week period and totaled approximately 11 hours. It was led by Janice Nathan and followed a structured format that combined theoretical instruction with practical application.

To be approved to work with study participants, therapists had to demonstrate mastery of the material, the ability to apply TiS techniques with flexibility and sensitivity, and a strong commitment to preserving the child’s self-worth throughout the intervention. During therapy, a caregiver was present, and each session began by letting the child choose from various activities, helping ensure the sessions were child-centered and collaborative.

To measure changes in emotional dysregulation, the researchers relied on caregiver-completed questionnaires at three points: before therapy, after 10 weeks, and again at 20 weeks. The main measure was the Emotion Dysregulation Inventory, which includes two scales. One tracks dysphoria—feelings of unease or low mood—while the other measures reactivity, or how quickly and intensely a child becomes emotionally upset. A third scale, from a separate inventory, measured how well children were able to control their emotional responses in daily life.

The results showed that children who received TiS therapy experienced a reduction in dysphoria compared to those who were still waiting for treatment. This was seen in both groups once they had completed therapy, suggesting that the timing of the intervention mattered less than receiving it at all.

Reductions in emotional reactivity were also observed, though these changes were marginally significant in the full sample and more robust in the group that received therapy later. Notably, younger children seemed to benefit less from the therapy in terms of emotional reactivity, highlighting the possible role of developmental stage in how children respond to inner speech interventions.

While TiS did not significantly improve children’s ability to control emotions in the broader behavioral sense measured by the executive functioning scale, it did appear to reduce the emotional distress children felt. These findings suggest that the therapy may help children feel better even if it doesn’t immediately change all aspects of emotional behavior.

Nathan highlighted three main takeaways: 1.) Autism is neurological, not ‘behavioral’ and there’s a lot of research that backs this up. 2.) Janice’s therapy reduces emotional dysregulation because it focuses on problem-solving, not problematic behaviors. 3.) Her therapy can be taught remotely and delivered remotely.”

Importantly, no adverse events were reported during the study, and families generally completed all sessions. Therapist fidelity to the TiS method was also high, with most completing over 75% of the key tasks across sessions. However, one unanticipated finding was that therapists often did not consistently communicate the therapy strategies to caregivers.

“The one thing that we did not expect is that trained speech-language pathologist did not communicate about the strategies they were using with caregivers as we had expected,” Nathan said. “We will revise the training to emphasize this more, and provide some role-playing exercises. Janice also believes that maintaining the child’s self-esteem is a critical part of her therapy, but we did not measure this.”

There are some other limitations to keep in mind. The sample was small, consisting of only 22 children and nine therapists, all of whom were non-Hispanic white females. Because all participants were verbal and had internet access, the findings may not apply to nonverbal autistic children or to families without access to telehealth services. Also, the study did not directly measure inner speech, so it’s unclear how much the children internalized the modeled strategies. Future studies should include measures of inner speech development to better understand how these changes relate to improvements in emotional regulation.

The researchers acknowledge that these findings are preliminary and that a larger, more definitive trial is needed to confirm the therapy’s effectiveness and they hope to expand the work. “We’d like to continue this research with more therapist, more children, and more sessions with the autistic children,” Nathan explained. “We also would like to add self-esteem measures.”

“Doing research outside academia is extremely difficult,” he added. “We were only able to do this because we found academics who believed in what we were doing, and made this a part of their research program. But this opportunity is also drying up with the new Secretary of the Department of Health & Human Services. Hopefully, under a new administration we can continue our research. But Janice and I are both old, and we’re not sure if there still time.”

The study, “Developing Inner Speech to Help Autistic Individuals Improve Their Self-Regulation Ability: A Pilot Randomized-Controlled Trial,” was authored by Barbara L. Baumann, Janice Nathan, Barry R. Nathan, Miriam Sheynblyum, Valire Carr Copeland, Carla A. Mazefsky, and Shaun M. Eack.