Drexel Study Finds Early Intervention Helps Most Autistic Children Acquire Spoken Language

After receiving evidence-based early interventions, roughly two-thirds of non-speaking kids with autism speak single words, and approximately half develop more complex language, according to a new study led by researchers at Drexel University’s A.J. Drexel Autism Institute. The findings, which offer insights that might help improve success rates for the kids who remain non-speaking or minimally speaking (e.g. not combining words to form short phrases) after therapy, were recently published in the Journal of Clinical Child and Adolescent Psychology.

About one in 31 kids in the United States are autistic, according to a 2025 CDC report, a number that has risen in recent years due to improved screening, more understanding and awareness. About one in three kids with autism are estimated to be nonspeaking. Although spoken language can vary considerably among kids with autism, those who do not gain spoken language beyond preschool years are at increased risk of disabilities later in life. Therefore, language skills are often a major target for early child development.

In the first large scale study to look at how many kids do not experience these gains, and the factors influencing success, the team used data across multiple studies on 707 autistic preschoolers who completed different types of evidence-based early interventions from six months to two years for at least 10 hours per week designed to teach spoken language, among other skills.
The participants who did not see success from intervention started with lower scores in cognitive, social, adaptive and motor imitation measures compared to their peers who developed spoken language. The team also found an association between development of spoken language and modifiable intervention factors, such as age at intervention start and intervention duration. The researchers suggest that these interventions are not one size fits all and individuals should be monitored for progress with adjustments made periodically to help language development.

These interventions remain evidence-supported ways of teaching language to nonspeaking kids with autism, authors say, noting that about half of the kids who had single or no words at the start were combining words by the conclusion of the intervention.

“When parents ask me if their child should do these interventions to gain spoken language, the answer after doing this study is still yes,” said Giacomo Vivanti, PhD, an associate professor and leader of Early Detection and Intervention in the Autism Institute at Drexel’s Dornsife School of Public Health. “What our study is telling us is that even when we’re implementing practices that are evidence-based, some children remain behind. So, we should carefully monitor the response of each child and see what to add or change to tailor therapy for the individual as needed.”

The researchers found that duration of the intervention, rather than intensity of the intervention, was associated with outcomes in the children who are nonverbal. Therefore, it is possible that, instead of packing in as many hours as possible working with a child 20-40 hours each week to facilitate language in a therapy room, more success might be achieved by doing the therapy for a longer term for more opportunities to practice skills, but with fewer hours per week.

“It’s imperative to track and monitor progress, even when the therapy meets your evidence-based practice criteria,” Vivanti said. “You need to monitor and be ready to take action, and adjust the strategies that you are using, the goals that you are targeting and the duration of the intervention to meet the needs of each child. For some children the ‘standard’
version of an intervention might be sufficient to support the acquisition of spoken language. For others, maybe a longer duration or more focus on preverbal skills that are foundational to communication is appropriate.”

“For young autistic children, often the highest priority is figuring out ways of supporting their communication, both their understanding and use of spoken language as well as other important behaviors, such as gestures,” said co-author Catherine Lord, PhD, distinguished professor at UCLA. “Many well-established interventions have reported improvements in test scores and parent report measures. However, it is important to remember that not all children learn at the same rate, and to be sure that we know when and how to shift strategies when progress isn't happening or is occurring in such small steps that it is hard to see.” 

As suggested in earlier studies, the researchers found that children who imitate more through motor imitation, such as encouraging children to join when people are clapping hands, nodding or imitating other gestures, were more likely to obtain spoken language.

“Those nonspeaking prerequisites of communication may help create infrastructure for spoken language,” Vivanti said. “Imitating what others are doing may help may them later to imitate what people are saying, and from there using language to express their thoughts.”

The team observed that the type of intervention used made little difference in language learning outcomes, despite underlying differences between these options in orientation, reasons why they should work and theories about what causes children to develop spoken language.

As this was a retrospective study, methodology varied among interventions, such as using different types of sites, such as a home, in a clinical practice, etc. Additionally, children were followed only for the duration of the intervention.

Despite these limitations, this study fills gaps in the understanding of spoken language status and stands as a rare large-scale commitment to share insights among many autism research institutions on a single study.

“Often scholars are weary about sharing intervention data and examine children who are not showing an optimal response to their interventions, especially for interventions that are already established as ‘evidence-based,’” said Vivanti. “This paper shows a willingness in the early intervention community to collaborate on data and learn more about how to help all children.”

The interventions in the study included the “Early Start Denver Model,” which is designed to help participants grow skills in many areas (including language) by engaging the child in joint play routines; other “Naturalistic Developmental Behavioral Interventions,” which similarly embed teaching within the child’s preferred activities and routines; “Early Intensive Behavioral Interventions” that use structured, adult-led teaching practices like discrete trial training, which breaks up skills into smaller parts for better learning; and TEACCH which focuses on changing an environment to adapt to learning preferences.

This work was supported by funding from the National Institute on Deafness and Other Communication Disorders.

In addition to Vivanti and Lord, additional authors on the research include Michael V. Lombardo from Istituto Italiano di Tecnologia, Ashley Zitter from Drexel, Brian Boyd from University of North Carolina at Chapel Hill, Cheryl Dissanayake from La Trobe University, Sarah Dufek from University of California Davis Health, Helen E. Flanagan from IWK Health Centre, Suzannah Iadarola from University of Rochester Medical Center, Ann Kaiser from Vanderbilt University, So Hyun Kim from Korea University, Lynne Levato from University of Rochester Medical Center, Joshua Plavnick from Michigan State University, Diana L. Robins from Drexel, Sally J. Rogers from University of California Davis Health, Isabel M. Smith from Dalhousie University, Tristram Smith of University of Rochester Medical Center, Aubyn Stahmer from University of California Davis Health and Linda Watson of University of North Carolina at Chapel Hill.