Stem cells for autism: what the science says

If your child was recently diagnosed, or exhausting years of therapy haven't brought the results you hoped for, the idea of trying stem cells for autism can sound appealing. Instead of slow, everyday work on behavior, it promises something faster, as if the cells could "clean up" or "repair" something in the body. Clinics in different countries offer this therapy, usually for a lot of money. Before paying for it, it helps to understand in plain terms what the research actually shows.

What the research found

In 2026, scientists pulled together the results of 22 different studies where children with autism were given stem cells, and worked out an overall result (Fekrat et al., 2026). At first glance the numbers look encouraging: on average, children improved noticeably, noticeably enough that it's unlikely to be chance (statisticians call this a "statistically significant result"). And where a comparison group got a placebo (a fake treatment), the difference was even bigger.

But there's an important catch. When researchers looked closer, the 22 studies turned out to be very different from each other: different doses, different types of cells, different ways of measuring results. So different that the averaged number says almost nothing about what would happen to a specific child, a bit like calculating the average temperature across an entire hospital. The authors themselves call this spread "extremely high" and put a number on it, 95 out of 100 (scientists write this as I² = 95%, and the closer it is to 100, the less you can trust the averaged result). On top of that, some of these studies didn't even compare treated children to untreated ones, they just watched one group and wrote down what changed. In cases like that, parents and doctors who know a child got an expensive treatment often see improvement even when nothing has actually changed. That's called an expectation effect.

Why "seems to help" isn't the same as "it works"

Autism isn't damage that extra cells can patch up. It's a difference in how the nervous system is built and works from birth, not an injury that needs repairing or replacing. No one can yet explain, in plain terms and with evidence, how cells injected into the bloodstream (usually taken from cord blood or bone marrow) are supposed to change the core features of autism, like how a child communicates or behaves. Until that explanation exists, calling the method effective is premature.

What the FDA, the US drug regulator, says

In the US, there's a government agency that checks medicines and medical treatments: the FDA. It issued a specific warning for consumers about stem cells and similar products, and it says plainly that no such product is approved to treat autism. The only thing that is approved is cord blood cells, and only for treating blood disorders, which has nothing to do with autism.

The FDA also warns that unproven cell injections aren't a harmless "just in case" procedure. There have been real cases of serious complications, including cells growing uncontrollably right at the injection site. We're deliberately not naming specific clinics here: the issue isn't one bad provider, it's that the method as a whole isn't proven yet.

What to do if you're offered this therapy

If you are offered stem cell therapy, here are a few simple questions worth asking first:

• Ask exactly how the method was tested. If it's just stories of "it helped a few kids," that's not the same as a real experiment comparing treated children to untreated ones.
• Ask whether it's registered as an actual clinical trial. Real trials are usually free for the family, or partly funded by the organizers, rather than sold as a paid service.
• Ask about safety separately. Just because the marketing doesn't mention complications doesn't mean they don't happen.
• Don't drop methods that already work for the sake of an experiment. Even if you decide to try it, the time and money taken from applied behavior analysis (ABA) or speech support are also part of the cost.

What we offer instead

Unlike stem cells, applied behavior analysis (ABA) has been tested in dozens of serious studies over many years. The specialists who work with it, called BCBAs, don't rely on a general sense that things got better, they measure specific, visible changes in a child's behavior. That doesn't mean ABA solves everything: every method has its limits, and we talk about them honestly in a consultation. Stem cells aren't the only method sold on promises alone: if you'd like the bigger picture on other similar methods, from special diets to acupuncture, we've covered a large scientific review of 19 of them in a separate article: Myths about autism treatment.

You can start with a free thirty-minute consultation, where we'll talk honestly about which of the methods offered to you are backed by research and which aren't.