Six Meta-Analyses, Same Direction: Children in Comprehensive Early ABA Scored Higher on IQ, Other Skills

Young autistic children who received comprehensive ABA scored higher on average on IQ tests and on measures of everyday skills than children in comparison groups, according to an umbrella review published October 1 in the Journal of Autism and Developmental Disorders. All six meta-analyses the review examined pointed in the same direction; together, they drew on 36 distinct studies.

An umbrella review gathers and compares the findings of existing meta-analyses. It runs no new trial and, in this case, computed no new combined estimate. This one covered comprehensive programs, which the National Academies distinguish from narrower, focused ABA practices, and searched six databases through December 2025.

The comparison groups were not untreated: most received usual care or eclectic programs, and some received lower-intensity ABA, according to the earlier version of the review that the Academies commissioned. The findings cover comprehensive programs for young children, not focused ABA or older children.

The estimated effects were larger for IQ than for adaptive behavior. Three of the six meta-analyses reported the IQ gap in points, at about 12 points or more on the standard scale. For adaptive behavior, the everyday communication, self-care, and social skills that standardized scales measure, three reviews reported gaps of about 6 points or more. These are average differences between groups, not a gain any one child can expect. The Academies committee that commissioned the earlier version described the effects as “primarily in the moderate to large range, especially for IQ.”

The authors, Brian Reichow and Jinwei Song, both of UConn Health, and Erin E. Barton, based in Nashville, Tennessee, call the pattern “consistent, replicated evidence.” Reichow is an associate editor of the journal and was not involved in the editorial process for the paper, the disclosure says.

Where the Evidence Is Less Certain

The consistency weakens beyond those two outcomes. Communication, social skills, daily living skills, and autism symptoms were reported less often across the six reviews, and where they were reported the findings varied more, the authors write. In its assessment of the earlier version, the Academies committee described effects on social skills and autism characteristics as positive but not clinically significant.

The six reviews also drew in part on the same trials. The authors’ overlap index came in at about 21%; a 2014 scale published in the Journal of Clinical Epidemiology labels anything above 15% very high. The index counts how often the same studies recur across reviews, so the six are not six independent replications. The 2,241 participants the abstract reports are summed across the reviews, which means children in studies that appear in more than one review are counted more than once. One of the six is Reichow and Barton’s own 2018 Cochrane review. The review contains no new trials. Its contribution is that all six meta-analyses, on average, favored comprehensive early ABA for IQ and adaptive behavior, though their estimates of the size differed.

The underlying trials limit how much weight the numbers can carry. The committee warned that the individual studies carry risks of bias, including small sample sizes, lack of randomization, and single-blind designs. Reichow and Barton have said as much themselves. Their 2018 Cochrane review of early intensive behavioral intervention (EIBI) covered five studies, one of them randomized, and 219 children. Its IQ and adaptive-behavior analyses, each drawing on 202 of those children, found scores about 15 and 10 points higher than in comparison groups; the review rated the evidence low quality and concluded there was “weak evidence that EIBI may be an effective behavioral treatment for some children with ASD.”

The TRICARE Coverage Fight

The National Academies, funded by Congress through the fiscal 2022 defense authorization law, conducted part of the review to examine TRICARE’s Autism Care Demonstration. The demonstration covered ABA for about 17,600 beneficiaries with an autism diagnosis and cost $519 million in fiscal 2024, according to the Defense Health Agency’s April 2026 report to Congress. The agency sponsored the Academies’ analysis.

The committee, chaired by George W. Rutherford of the University of California, San Francisco, commissioned Reichow and Barton to produce an earlier version of the umbrella review, and its report, released September 9, 2025, weighed that review alongside other evidence to reach a conclusion the agency has disputed: that ABA meets TRICARE’s own standard for proven medical care. The committee’s first recommendation was to replace the demonstration with ABA coverage as a basic TRICARE benefit.

“For these reasons, the committee concludes that ABA-based interventions do meet TRICARE’s hierarchy of reliable evidence standard for proven medical care.” – National Academies of Sciences, Engineering, and Medicine, The Comprehensive Autism Care Demonstration: Solutions for Military Families (2025)

The Defense Health Agency has told Congress the opposite in its annual reports for fiscal 2022, 2023, and 2024, most recently in April 2026, seven months after the Academies published: “As of now, ABA services do not meet the TRICARE hierarchy of reliable evidence standard for proven medical care.” The April report described the Academies’ work as providing “valuable insights into the delivery of ABA services.”

On June 26, 2026, Senators Kirsten Gillibrand, a New York Democrat, and Eric Schmitt, a Missouri Republican, wrote to Defense Secretary Pete Hegseth that more than nine months had passed since the recommendations and the department had not adopted them, and urged it to cover ABA as a basic TRICARE benefit. “It is unacceptable that military families are being denied essential care for their autistic children,” Gillibrand said in a statement.

The senators set no deadline for a reply. TRICARE’s current guidance still covers ABA through the demonstration, which is authorized through December 31, 2028.

AT A GLANCE

SOURCES & REFERENCES