Do Rhythmic Movement Training and reflex exercises help autistic children?
Rocking-type movement programs, and the idea behind them. Here’s what 8 sources can and can’t tell you.
A care note before you read. “Not enough research” is not the same as “proven not to work.” It means nobody has tested it well yet in autistic children. Your child’s doctor, physical therapist or occupational therapist can help you think it through.
Podcast·Two-voice deep dive
Listen to the discussion
Chapters & transcript
This is an AI-generated discussion of this article, transcribed automatically. The written summary below is the reviewed source — if the two differ, go by the summary.
0:00What is this discussion about?
Host A You know, usually when we talk about a medical diagnosis, there is this kind of comforting expectation of precision.
Host B Right, like it feels like engineering or something.
Host A Yeah, exactly. Like if you fall and break your arm, the x-ray shows that jagged white line, the doctor points to it, and well, that's that.
Host B It's visible. It's categorized.
Host A Right. And the path forward is obvious. But then you step into the world of neurodevelopment and suddenly that x-ray machine is just broken.
Host B Oh, completely broken. It is a landscape filled with theories, overlapping symptoms, and honestly, a whole lot of parents just trying to find the best way to support their kids without a clear roadmap.
Host A And that is exactly the territory we are exploring today. Welcome to the Deep Dive.
Host B Glad to be here.
Host A So today we are unpacking a very specific, heavily marketed intervention, rhythmic movement training, or RMT, and reflex integration programs for autistic children.
Host B It's a big topic.
Host A It really is. And today's deep dive is built on a pretty unique stack of sources. We have a set of internal documents from Spectrum Connect.
Host B Specifically, a demo-grade findings report and a really detailed audit document.
Host A Right. And these files trace the evidence, or in some cases, the surprising lack of evidence surrounding these movement programs.
Host B Which is so important to look at.
Host A Yeah. But because this landscape is so murky, I want to be crystal clear with you up front. We are acting as research translators today, not doctors.
Host B Exactly. This is not medical advice.
Host A Strict summary of research findings here. We are looking at what the data actually says, how it was gathered, where the gaps are. But absolutely nothing we discuss here today is a recommendation for any individual child.
Host B Our mission is just to pull back the curtain on the evidence base, you know, breaking down the clinical research so you can understand the science behind all those marketing claims you see online.
1:50What is Rhythmic Movement Training?
Host A So let's just get right into it. If a family is looking into rhythmic movement training, they are probably being told it's a physical intervention.
Host B That's right.
Host A Based on the findings report, this approach, which was originally developed by Harald Blomberg, involves a child doing gentle, repeated physical movements.
Host B Yeah. So you're looking at activities like rhythmic rocking while lying on the floor or sitting up or getting on their hands and knees.
Host A Just very repetitive, specific motions.
Host B That is the core of it. And to understand why they are doing these specific rocking motions, you really have to look at the underlying theory the program relies on.
2:26What are primitive reflexes, and what is the idea behind these programs?
Host A Which is all about primitive reflexes, right?
Host B Exactly. When babies are born, they have these automatic hardwired movements. Like if you've ever seen a newborn suddenly throw their arms out when they are startled.
Host A Oh, yeah, the Moro reflex.
Host B Right, the Moro reflex. Or if they turn their head to one side and the arm on that side stretches out while the other arm bends, that is the asymmetric tonic neck reflex, often called the ATNR.
Host A Okay, and in typical development, those reflexes don't just stick around forever, right?
Host B No, they don't. The nervous system matures, and those automatic responses just naturally fade away within like the first year of life.
Host A I think the clinical term is that they integrate?
Host B Yes, integration.
Host A But the central premise of RMT and reflex integration programs is that some children retain these primitive reflexes for far too long. Okay, so they stick around when they shouldn't.
Host B Right. Okay. And the proponents of these programs theorize that having these leftover reflexes acts as a roadblock. They claim it causes a massive cascade of other issues.
Host A Like what kind of issues?
Host B Well, they blame retained reflexes for problems with physical coordination, attention, learning delays, and behavioral challenges.
Host A Wow. So they're linking a lot to this.
Host B Yeah. So the claim is that by performing these highly specific repeated rhythmic exercises for weeks or months, you can target those leftover reflexes and physically force the nervous system to integrate them.
Host A Okay, I'm trying to visualize this mechanical barrier they are describing. It feels a lot like leaving training wheels on a bicycle for way too long.
Host B Oh, that's a really good way to put it.
Host A Right? Because when a kid first learns to ride, training wheels are essential. They keep the bike upright, just like primitive reflexes keep a vulnerable newborn functioning.
Host B Exactly.
Host A But if you leave those training wheels on when the kid is five years old and trying to, you know, lean into a sharp turn or pedal over uneven ground, the wheels actually get in the way. They prevent advanced riding.
Host B They become a barrier.
Host A Right. So this intervention is basically claiming to be the wrench that finally removes those training wheels so the child can navigate the world smoothly.
Host B That analogy perfectly captures the mechanical claim being sold here. The retained reflex is viewed as the physical barrier to higher-level brain development, and the rocking movement is the specific wrench designed to remove it.
4:49Has Rhythmic Movement Training been tested in autistic children?
Host A Okay, but a claim is just a claim until you back it up with data.
Host B Right.
Host A And the Spectrum Connect audit document tracks exactly how the evidence base for this wrench actually developed. And honestly, the methodology of their search is fascinating.
Host B The scoping search.
Host A Yeah. They ran a scoping search, which basically means they cast a massive net across the Internet and clinical databases to see what exists.
Host B Right. Checking everywhere.
Host A And they yielded 19 web results and ultimately cited 12 records. But when I was reading through the audit, one detail just stopped me in my tracks.
Host B I think I know which one you're talking about.
Host A They found zero studies, literally not a single one, testing the actual named branded Blomberg rhythmic movement training in autistic children.
Host B Zero.
Host A The official RMT International site didn't even surface in the audit in a way that provided an autism specific clinical study.
Host B Which is wild because when you have an intervention that is actively marketed to families, you kind of expect to find at least a baseline of branded clinical trials. You'd think so. But instead, the evidence base that the audit had to pull from relies entirely on generic related reflex integration programs.
Host A Not the branded one.
Host B Right. And the types of studies making up this base are far from gold standard randomized controlled trials.
Host A What are they then?
Host B They are primarily uncontrolled pre-post studies.
Host A Okay, meaning they measure a single group before an exercise and then after without comparing them to a control group.
Host B Exactly. That along with cross-sectional association studies.
Host A Okay. And the audit also zeroes in on where this information actually lives online. The retrievable material on the web isn't really coming from major pediatric journals.
Host B No, it's heavily dominated by proponent and vendor websites.
Host A Right. The audit specifically names a site called MovePlayThrive, and it flags something even more critical. Developer authorship is inferred for several of these foundational studies.
Host B Which is a huge flag.
Host A Yeah. Like, for example, there's a 2016 study on the MNRI Reflex program by a researcher named Masgutova. The auditor infers developer authorship right from the surname, though it wasn't independently verified in this run.
Host B And there is also a group led by Robert Melillo that appears across multiple narrative reviews and premise studies.
Host A Which feels like a pretty big conflict of interest.
Host B Yeah. And the Spectrum Connect audit engine actually has a built-in protocol flag for this exact scenario.
Host A Oh, really?
Host B Yeah. It flags when an evidence base is dominated by people who teach, trademark, or sell the program, and lacks independent, high-quality reviews from outside researchers.
Host A So it's just marking that bias.
Host B Right. It doesn't mean the data is automatically false, but it forces evaluators to view the claims through a lens of potential bias and, you know, contested premises.
Host A That makes sense.
7:36What did the closest autism study measure?
Host A So if the overarching evidence is mostly self-published or uncontrolled, the auditors must have had to dig really deep into whatever independent data did exist.
Host B They did.
Host A The audit report spends a lot of time breaking down a recent study from 2025 by Hirose, Toshiro, and Takasaki, published in the journal Children.
Host B But we aren't talking about a massive, rigorous clinical trial here.
Host A No, not at all.
Host B They only looked at 15 autistic children.
Host A Mostly boys between the ages of 6 and 12.
Host B Right, who went through a 12-week program of generic rhythmic balance and coordination exercises.
Host A And this is where we have to look incredibly closely at what the researchers actually measured.
Host B Because in any clinical study, the outcome measures tell you what the scientists value.
Host A So what did they value in the Hirose study?
Host B They measured three main things. First, they measured reflex retention scores, specifically that left-standing ATNR we talked about earlier.
Host A Okay, the mechanical barrier.
Host B Right. Second, they used a right-hand finger movement test, abbreviated as FOT. And third, and arguably most importantly for families, they used a parent-rated scale called the SRS-2.
Host A Which is designed to measure real-world social skills, right?
Host B Exactly.
Host A Now, the audit also includes a second separate study by Alsaman and Elghidi from 2026, but this one didn't test the rocking exercises at all, right?
Host B No, it just tested the core premise. It took a cross-sectional snapshot, basically comparing 30 autistic children with 30 neurotypical children.
Host A Just to see if the autistic kids actually had higher levels of those retained primitive reflexes.
Host B Right. It was purely observational. They just wanted to see if the two things existed in the same room, so to speak.
Host A Okay, let's go back to the Hirose study for a second because I have to bring up a protocol defect the audit system flagged because it honestly blew my mind.
Host B Oh, the double counting?
Host A Yes. The flag is called mediator as outcome double counting. If I am understanding the logic here, the whole theory is that retained reflexes cause autism related difficulties.
Host B Yes.
Host A So the reflex is the mechanism or the mediator. But the Hirose study tries to use a drop in the reflex score as proof of the program's overall clinical benefit.
Host B Which is deeply problematic.
Host A Right. If a parent puts their kid in a 12-week therapy program, they want to see better social skills or easier transitions or better communication. They don't care about a clinical reflex score on a piece of paper.
Host B Absolutely not.
Host A So why are researchers using a drop in a reflex score as proof of success? Isn't that just proving the kid got better at the exercise rather than proving an actual real-world benefit?
Host B You have just articulated one of the most critical foundational flaws in how this type of research is designed. Wow. Yeah. Reflex retention scores are not an autism-relevant outcome. If a child's ATNR score goes down on a clinical test, but their social skills or everyday behaviors remain exactly the same, the intervention has not delivered a meaningful functional benefit to that family.
Host A It's just a number on a page.
Host B Right. The reflex is merely the proposed mechanism. You cannot double count a change in the mechanism as a change in the real world outcome.
Host A To use our bike analogy again, it's like a mechanic taking those training wheels off the bike, pointing at the unattached wheels, lying on the driveway and shouting, look, a successful bike ride.
Host B While the bike is just sitting there.
Host A Exactly.
Host B Without ever bothering to check if the kid actually rode the bike down the street without falling over.
Host A The removal of the wheels is the mechanism. The riding of the bike is the outcome. You have to measure the riding.
11:07Did reflexes, movement or social skills improve?
Host A Okay, so did they measure the riding? Let's look at the actual findings here.
Host B Okay, let's break it down.
Host A So the Alsaman Observational Study found that the 30 autistic children did, in fact, have higher levels of retained reflexes compared to the neurotypical kids.
Host B Right, and those leftover reflexes were linked to several emotional and behavioral traits.
Host A So the premise that they exist together holds up in that snapshot.
Host B It does.
Host A And in the 12-week Hirose intervention study, the researchers reported that the left-standing ATNR reflex score went down and the right-hand finger movement test improved.
Host B But I'm guessing you were probably wondering about that third measure.
Host A The SRS-2.
Host B The social skills test.
Host A Yes. Because if they specifically set out to measure social skills and they had parents fill out the rating scales, I am guessing they wanted to prove that dropping the reflex score bumped up the social skills.
Host B That would be the logical assumption.
Host A Okay. So did they actually prove that?
Host B This is perhaps the biggest revelation in the entire audit.
Host A Okay.
Host B The Hirose abstract, which is what the retrieval engine pulled, did not report a social result for the autistic children.
Host A Wait, what? They listed the SRS-2 as a primary outcome measure, but the actual results for the autistic kids are just missing from the summary?
Host B Entirely missing.
Host A You're kidding.
Host B No, and it gets even more convoluted. The Hirose study was actually a mixed diagnosis study. It included two different arms.
Host A Okay, so who was the other arm?
Host B There were the 15 children with autism and a separate group of 12 children with ADHD.
Host A Ah, I see where this is going.
Host B Yeah. The abstract reports that behavioral gains, specifically measured on a different scale called the Conners 3 Index, were seen in the ADHD group.
Host A But not the autism group.
Host B Right.
Host A Okay. If a proponent website is willing to take that data and run with it, we have a massive problem.
Host B And the audit flagged exactly that happening on the MovePlay Thrive website.
Host A Oh, wow.
Host B Yeah. The source checker uncovered a profound misattribution. The Move, Play, Thrive evidence page summarized the Hirose study by stating that the combination of rhythmic movements improved behavioral outcomes.
Host A Let me guess. They listed this under autism.
Host B They listed it explicitly under their autism section. They took the behavioral gains measured exclusively in the ADHD arm and credited them to the autism group to market the program.
Host A That is staggering. I mean, a parent researching therapies at midnight reads that promotional page and thinks, wow, this study proved these exercises improve behavior in autistic kids.
Host B Exactly.
Host A But the actual abstract of the study shows the behavior gains were only measured in the ADHD group and the autism social scores just vanished from the results.
Host B It's incredibly misleading. The misattribution was so blatant that the Spectrum Connect audit document had to create a brand new protocol flag just for this run. What did they call it? They labeled it DRMT2, Mixed Diagnosis Arm Attribution. And that was simply to force the system to ensure results are only credited to the specific diagnosis in which they were actually measured.
Host A Honestly, if a proponent website is willing to pass off ADHD behavioral gains as autism results, it really makes you question the foundation of this entire industry.
Host B It really does.
14:28Why is the evidence rated very low?
Host A When the auditors stepped back and graded this evidence, how did it actually hold up?
Host B The grading is brutal, but necessary.
Host A Let's hear it.
Host B The sources grade the evidence for related reflex programs as very low, which equates to insufficient. And the overall status for the headline question, does named Blomberg RMT help autistic children, is categorized firmly as not supported, not enough research yet.
Host A And the findings report is really careful to note that not enough research yet doesn't mean science has definitively proven it doesn't work.
Host B Right. It means nobody is bothered to test it rigorously enough to know either way.
Host A Exactly. And aside from the missing social data, the audit cites a massive design weakness in the Hirose study, which is the complete lack of a control group.
Host B Ah, yes. They call it a maturation confound.
Host A A maturation confound. Okay, what does that actually mean for the listener?
Host B It is simply the reality of human biology and development. Think about a 12-week period in the life of a child between 6 and 12 years old.
Host A They change a lot.
Host B Right. They are constantly growing. Their nervous systems are actively developing. They are learning new concepts in school. They are interacting with peers. And they might be receiving other therapy simultaneously. Right. If you run a study without a control group of similar children who are not doing the rhythmic exercises, you have absolutely no baseline.
Host A So when you see a finger-tapping test improve at the end of 12 weeks, you can't possibly know if the rocking exercises caused the change or if it was just the passage of time and the natural, beautiful process of a child growing up.
Host B Precisely. Without a control group, any claims of causation are scientifically hollow. And that is compounded by another limitation, the audit flags, which is unblinded assessments.
Host A Okay. If I understand unblinded bias correctly, it means the people measuring the finger tapping or the reflex angles at the end of the study actually knew which kids had just spent 12 weeks doing the therapy.
Host B They did. And human bias is incredibly powerful, even in clinical settings.
Host A Oh, for sure.
Host B If a researcher expects to see an improvement because they know the child just completed the intervention, they might subconsciously score a reflex angle a few degrees better or offer slightly more encouragement during a finger-tapping test.
Host A And they kind of just see what they want to see.
Host B Exactly. That is exactly why blinding, where the evaluator has no idea who got the therapy and who didn't, is a non-negotiable requirement in rigorous science.
Host A I also noticed the audit engine ran into a very modern, very frustrating measurement problem with this.
Host B Oh, the paywall.
Host A Yes. The retrieval system found a 2026 autism study by Bijle and Malawad that was supposed to look at reflex integration exercises. But the source checker hit an HTTP 429 rate limit.
Host B Right. An access block.
Host A A paywall. so they couldn't even read it.
Host B They could only pull the cross-referenced metadata. So there is a piece of quasi-experimental evidence sitting out there in the ether, but it couldn't be factored into the findings because it was locked away.
Host A Which means this very low evidence grade relies heavily on just dissecting abstracts rather than being able to fully interrogate the methodology in the full texts.
Host B Exactly.
Host A So if we look at the big picture here, we have an intervention heavily marketed to families, built on zero branded studies, relying on uncontrolled trials with unblinded assessors and paywall data.
Host B It's a lot of red flags.
Host A The gaps in this science aren't just small cracks. They are massive craters.
17:58Is it safe, and are there long-term results?
Host A So what exactly does this evidence not answer?
Host B The open questions are stark, and they represent pretty much everything a family would actually need to know.
Host A Okay, like what?
Host B First, as we established, does the actual branded Blomberg RMT work for autistic children? We have zero data to answer that.
Host A None.
Host B Second, are there any long-term results? None exist in the captured literature. We only have that 12-week snapshot. Right. Third, what about teenagers or adults? The data entirely ignores anyone outside the 6 to 12 age bracket.
Host A Oh, interesting. So it's very narrow.
Host B Very. And fourth, what about functional everyday life skills? They measured finger tapping, but they completely failed to report on the practical skills that families actually care about.
Host A But I think the most glaring omission and the one the audit points out strictly is the complete lack of safety data.
Host B Yes.
Host A Like no study reported on side effects, adverse events or the burden of the intervention. It's not just that they forgot to measure safety. It means we have no idea if forcing a child through weeks of these repetitive movements causes physical discomfort or, you know, spikes their anxiety.
Host B And harm isn't just physical pain either.
Host A What do you mean?
Host B The audit emphasizes that the burden and cost of time were unmeasured. Placing an undue financial drain on a family, burning through their limited free time, and offering false hope based on mixed diagnosis marketing, that is a profound type of harm.
Host A Wow. Yeah, I didn't think about it like that.
Host B As the auditors note, harm isn't shown to be absent in these studies. It simply wasn't studied at all.
Host A With so many gaps and an online landscape that's perfectly designed to blur the lines between marketing and science, how does a clinician actually evaluate this? I mean, the findings report gives us a window into how professionals weigh this kind of evidence, right?
Host B It does. A rigorous evaluator looks at study designs to see if they test cause rather than just association.
19:56Do leftover reflexes cause autism-related difficulties?
Host B Okay. Let's look back at that Alsaman study, which found that autistic children had higher levels of retained reflexes. That is an association. But think of it like noticing that people who carry umbrellas often get wet.
Host A The umbrella doesn't cause the rain.
Host B Exactly.
Host A They just happen to exist in the same environment on the same day.
Host B Right. Just because retained primitive reflexes and autism traits are in the same room, or the same nervous system, does not mean one is causing the other.
Host A So what would proof cause?
Host B To strengthen confidence, an evaluator would need to see robust, randomized controlled trials. They would need a design that proves changing the reflex directly and independently changes the developmental outcome.
Host A Which we don't have.
Host B Because that is entirely lacking. The clinical premise remains graded as unestablished, not disproven.
Host A Unestablished, not disproven. The findings report actually translates that clinical reasoning into incredibly practical guidance for families, which I love.
Host B Oh, the questions to ask.
Host A Yes. If a provider offers you a reflex integration program, you shouldn't just ask, does this work?
Host B Because they'll just say yes.
Host A Exactly. The report suggests asking highly specific, targeted questions. Things like, what exact functional goal will we measure? Or how will we know if it's working outside of a clinical reflex score? And what is the exact cost in time, energy, and money?
Host B Which is brilliant.
Host A Right. By forcing the conversation away from abstract theories about primitive reflexes and grounding it in measurable reality, you protect your family's resources.
Host B It completely separates the glossy marketing claims from clinical reality. You are demanding that the intervention prove its worth in the real world.
Host A Well, we have covered a massive amount of ground today, navigating from theoretical training wheels to paywalled studies and mixed diagnosis marketing.
Host B It's been quite the journey.
Host A It has.
21:50What are the key takeaways?
Host A So let's synthesize this deep dive into the core findings you should actually remember from these Spectrum Connect documents.
Host B Sounds good.
Host A First, we started by looking for the science behind a heavily marketed fix. But what the auditors found was an industry built on zero studies testing the named, branded Blomberg rhythmic movement training in autistic children.
Host B Zero branded studies. Right.
Host A Second, we found that the core premise, that leftover reflexes actually cause autism-related problems, is completely untested for causation. The observational research only shows that they sometimes exist in the same child.
Host B The umbrella in the rain.
Host A Exactly. Third, we uncovered massive limitations in the existing research, specifically the lack of control groups in studies like the 2025 Hirose paper. That makes it impossible to separate the therapy's effects from a child's natural growth and maturation.
Host B The maturation confound.
Host A Right. And finally, a massive open question remains regarding safety, long-term effects, and the actual financial and emotional burden on families, because those vital factors simply haven't been measured.
Host B They just aren't in the data.
Host A Now, before we wrap up, I want to shift gears and speak directly to you, the listener, with a very important reminder. Decisions about any individual child belong in a conversation with that child's pediatrician or care team. And this deep dive cannot make that call. Research on this topic continues, and Spectrum Connect will update its coverage as new evidence appears.
Host B Which is great.
Host A Just as a reminder, Spectrum Connect translates autism research into plain language for parents and checks findings for accuracy against misinformation. If you are a credentialed expert, an MD, PhD, PsyD, or DO, and would like to help validate our findings, email research@spectrumconnect.us.
Host B And I want to leave you with a final thought to mull over today.
Host A Let's hear it.
Host B Think deeply about that maturation confound we unpacked today. If an intervention with no control group takes credit for the natural growth and development of a child over several months, how do we ever truly measure success? That's a great question. It is a vital reminder to always look closely at the data and ask, you know, is this program actually working? Or is this intervention just taking credit for the beautiful, natural process of a child growing up?
Host A I think that's a perfect note to end on. We'll see you next time on The Deep Dive.
Key Takeaways
We found no study of Rhythmic Movement Training itself in autistic children. The research can’t say whether it helps, and it can’t say it doesn’t.
The closest study was small and had no comparison group. In 2025, researchers gave 15 autistic children aged 6 to 12 a 12-week program of rhythmic, balance and coordination exercises. One reflex score went down and a finger movement test got better on the right hand. With no comparison group, we can’t tell whether the exercises caused those changes, since children grow and change over 12 weeks anyway.
No social or behavior gain was reported for autistic children. That study also measured social skills in the autistic children, but its summary did not report a social result. The behavior gains it reported were in a separate group of children with ADHD.
The idea behind these programs has not been shown. The idea is that baby reflexes that stick around cause problems with movement, attention or behavior, and that exercises can make them fade. One study found that 30 autistic children had more leftover reflexes than 30 children without autism, but going together is not the same as one causing the other.
If you’re thinking about it, a few questions help. Ask the provider which studies in autistic children support it, what goal they will measure, how you will know if it is working, and what it costs in time and money.
What this means for you
Rhythmic Movement Training (RMT) is a set of gentle, repeated movements, such as rocking, that a child does lying down, sitting or on hands and knees. It was developed by Harald Blomberg. Other programs, often called “reflex integration” programs, use similar exercises. Babies are born with automatic movements called primitive reflexes, and most fade in the first year. These programs say that some children keep these reflexes for too long, that this causes problems with movement, attention, learning or behavior, and that the exercises make the reflexes fade (“integrate”). We looked at two things separately: does the program help autistic children, and is the idea behind it true?
On the first question, we found no published study of Blomberg’s RMT in autistic children. We found one small study of a similar program, described in the takeaways above, and a second study of reflex exercises in autistic children published in 2026. We could not open that second study, so we don’t know what it found. On the second question, one study found more leftover reflexes in autistic children and some links with emotional and behavior problems, but that kind of study can’t show cause. A 2026 review of 14 studies in children without autism or other developmental conditions found that 5 tested reflex programs. Most found some improvement, but the review called the evidence “limited and heterogeneous,” and it tells us nothing about autistic children.
We are not at all sure about any of this. The only autism study of a program that we could read was small and had no comparison group, and people who expect change may see it. Much of what is easy to find online comes from people who teach or sell these programs. We found nothing on side effects: no study we read reported on them, so harm has not been studied, which is different from being shown to be absent.
This page does not cover sensory integration therapy, physical therapy, water-based therapy or general exercise. Those are covered in other reports (sensory integration, physical therapy and aquatic therapy). The 2025 autism study used a general “rhythmic, balance and coordination” program rather than branded RMT, and it could be read as general exercise. We show it here because it is the only autism study of its kind we could read. A 2024 trial of sensory-motor integration exercises also turned up in our search, and it belongs with the sensory integration report, not here. This page does not tell you to use or avoid Rhythmic Movement Training. The choice stays with your family and your child’s care team.
Who was studied. The one autism program study enrolled 15 autistic children aged 6 to 12 (13 boys) for 12 weeks; we did not capture the country. The reflex comparison included 30 autistic and 30 non-autistic children, with ages not captured. We did not find long-term results of a movement program, results for teenagers or adults, or results on everyday life skills. For most studies we read only the short summary, not the full paper. Whether any of this applies to a particular child is not established.
Where the studies landed
Not enough research yetThis is a very small field — all 8 sources we could read are shown below. None shows that Rhythmic Movement Training or a reflex exercise program helps autistic children.
3 Autism studies: small, and none can show the exercises work
5 Nothing we can use for autistic children yet
Tap any tile to read it. Three more records were found by title only and are not counted: a 2022 review of retained reflexes in autism, a study of parents’ views on RMT, and a 2024 paper on attention. A web page from a program promoter is not counted either, because it is not a study.
Autism studies: small, and none can show the exercises work3
Nothing we can use for autistic children yet5
Spectrum Connect reviews published research on interventions parents are exploring for their autistic children — so you can see where the evidence actually stands. No agenda, no selling, no cherry-picking. Just the studies, our method, and what it means for you.
See the research behind this
For anyone who wants to check our work · 8 sources
Where we looked
This was a web-based scoping search, not a full multi-database search: 2 web searches plus 9 page-read attempts. Of those 9, 7 returned content and 2 failed. One failed because the address was too long and was re-run as a shorter query that was read; the other was refused by a website’s rate limit and was not retried, and page reading stopped there. No database was searched directly, and PubMed and PMC pages were not used because they showed a CAPTCHA, so there are no per-database counts. The search of record for a full run covers PubMed/MEDLINE, Embase, APA PsycINFO, CINAHL, ERIC, Cochrane CENTRAL, Epistemonikos, trial registers (ClinicalTrials.gov and ICTRP) and evidence pages from RMT International and MNRI. Here is the PubMed search of record.
("rhythmic movement training"[tiab] OR "rhythmic movement therapy"[tiab]
OR "reflex integration"[tiab] OR "primitive reflex*"[tiab]
OR "retained reflex*"[tiab] OR "persistent primitive reflex*"[tiab]
OR "Blomberg"[tiab] OR MNRI[tiab] OR Masgutova[tiab])
AND ("Autism Spectrum Disorder"[MeSH] OR autis*[tiab] OR "ASD"[tiab])
Run on PubMed →
Counts for every database, and the dates searched, are pending. No protocol was registered in advance. This search is not reproducible from the record we kept.
What we did with what we found
This is a scoping count, not a formal screening flow. No full text was assessed; works were judged by title and abstract or record, with no second reviewer. The 22 and the 4 duplicates are approximate in our record.
What the strongest evidence says
No study found. The official Rhythmic Movement Training website could not be opened.
One small study of a similar program, 15 autistic children, no comparison group. The summary reports no social result for the autistic children.
A right-hand finger movement test improved after 12 weeks in that one study. No comparison group, so growth over 12 weeks, practice on the test and expecting change could all explain it.
One reflex score went down in that study. A reflex score is the idea’s own measure, not a result for daily life, so a drop in it does not by itself show that a child is better off.
One small study says yes (30 autistic vs 30 non-autistic children). How reliable the reflex tests were was not reported in what we read.
Not tested. No study in what we read was designed to test cause.
A 2026 review found most of 5 program studies reported some improvement, but called the evidence limited and varied. It left autism out on purpose, so it does not carry over.
None reported in what we read, but side effects were not studied, and neither were time, money or opportunity cost.
Keep reading
What would you like to do?
Want to talk to a parent who’s been there?
We match you with families whose kid is like yours. Free, takes two minutes.
Heard of something we haven’t covered?Suggest an article
If another parent, a clinic or a video mentioned a therapy you can’t find here, tell us. We’ll read the research and review it next.
Suggest an articleThink we got something wrong?Report a factual error
We publish the whole record so it can be checked, and that only counts if we act on what you find. If a number looks wrong, a study is missing or has been retracted, or we’ve read a finding in a way the evidence doesn’t support, tell us.
You don’t need a research background to file one. “This doesn’t match what our doctor told us” is a useful report. Every one reaches a person: we reply within seven days, and within thirty we have either corrected the page or told you when we will. Substantive reports send the affected steps back through the protocol and need fresh sign-off before anything here changes.
Report a factual errorAre you a clinician or researcher?Review this article
If you have read this and the evidence holds up, you can sign off on it. We send you a personal review link where you confirm your credentials, and our team checks every review before anything is published. Once approved, your name and a link to your profile appear on this article.
Ask to review this