Stem Cell Therapy for Autism: What the Evidence Actually Shows

Parents researching options for a child with autism will come across a wide range of claims about stem cell therapy. This article focuses on what serious, peer-reviewed science has actually found — most notably a large randomized trial from Duke University that identified a specific, real signal worth continued study — while being clear about where the evidence stands today: no stem cell product currently holds FDA approval for autism, and the trial's primary measure did not show a benefit over placebo.

Duke's Randomized Trial: A Real Signal, and a Result to Take Seriously

The most scientifically important work on this topic is a randomized, double-blind, placebo-controlled Phase II trial led by researchers at Duke University (Geraldine Dawson and Joanne Kurtzberg), published in the Journal of Pediatrics in 2020. The trial enrolled 180 children with autism spectrum disorder, ages 2 to 7, across three arms: autologous cord blood, allogeneic (donor) cord blood, and placebo. This is exactly the kind of rigorous design the field needs — randomized, blinded, placebo-controlled — and it's worth taking both of its findings seriously. On the trial's primary endpoint, change in socialization skills measured by the Vineland Adaptive Behavior Scales, a single cord blood infusion did not show a statistically significant benefit over placebo across the full study population.

The Genuine, Specific Signal Worth Watching

Within that trial, the Duke team identified something real: children ages 4 to 7 with a nonverbal IQ above 70 who received cord blood showed measurable improvements in communication, sustained attention (measured via eye tracking), and EEG markers of brain connectivity compared with placebo. This is a genuine, specific result in about half the enrolled children (101 of 180 met the IQ threshold) — the kind of finding that justifies continued, carefully controlled research into which children might benefit and why. It's a real scientific lead, not yet broad evidence of benefit.

The Earlier Phase I Trial: Establishing Safety

Before the Phase II study, the same Duke group ran a smaller, open-label Phase I trial (Stem Cells Translational Medicine, 2017) testing autologous cord blood infusions in young children with autism, designed to assess safety and feasibility. It had no placebo arm, which is exactly why the placebo-controlled Phase II trial that followed carries more scientific weight.

Where the Regulatory Approval Stands Today

Beyond the blood-forming stem cell products approved for certain cancers and blood disorders, no stem cell, mesenchymal stem cell, exosome, or regenerative biologic product currently holds FDA approval for autism. Duke's own research has operated under FDA-regulated investigational protocols throughout, the appropriate pathway for studying a promising but still-unproven approach.

FDA Oversight in This Space

Because the science here is still early, it's worth knowing how to check whether a specific provider is operating within that FDA-regulated research framework. In January 2025, for example, the FDA issued a warning letter to one company, Chara Biologics, Inc., after it marketed unapproved stem cell and exosome products for a range of conditions including autism without the required biologics license — a useful reminder that clinical-sounding language on a website isn't the same as FDA authorization. Asking whether a specific treatment is offered through a registered clinical trial or as a commercial product is the most useful single question a family can ask.

What's Known About Safety

Because unapproved cell products aren't manufactured or reviewed the same way FDA-approved products are, they carry real safety variability. A 2021 Pew Charitable Trusts analysis of adverse events tied to unapproved stem cell products (spanning many indications, not autism specifically) documented cases including infections, tumors or abnormal growths, and vision loss — a useful reminder of why treatment inside a regulated trial framework carries a different risk profile than treatment outside one.

Bottom Line

For a condition where legitimate research is still ongoing, the most useful thing a family can do is separate two things that often get blurred: the actual science, and specific commercial offers. Duke's trial is real and rigorous, and it found both a null primary result and a genuine, narrow signal worth more study — that's an honest, current picture of where things stand, not a verdict on whether cell-based approaches will ultimately help. Before considering any stem cell, cord blood, or exosome treatment for autism, ask directly: Is this part of a registered, FDA-authorized clinical trial with IRB oversight, or is it a commercial product? What does the trial's own primary data actually show? A provider who can answer those questions with real citations is engaging with the same evidence this article does.