Search for therapeutic movement and you will be told it is a family of gentle practices that build body awareness and retrain how you move. The phrase sounds like a category. It is not one.
No evidence body treats these practices as a unit. When a government review set out to assess them, it did not produce one report on therapeutic movement. It produced a separate report on the Alexander Technique, another on Feldenkrais and another on Pilates, with conclusions ranging from moderate certainty of benefit to effects unknown. Its list is not this article’s list either: among the sixteen therapies it assessed is Rolfing, a hands-on bodywork method, and dance therapy is not there at all. It pooled nothing and named no parent category, because there was nothing to pool.
Their theories do not match. Their delivery does not match. Their training does not match, and the gap at the extremes is enormous: one of these is a master’s-level psychotherapy requiring a 700-hour clinical internship, and another is a name a federal court ruled anybody may use.
The closest thing to a legal definition of the group is a New Hampshire statute, and what it defines them as is the set of practices a license does not reach.
A state law defines the group by what it leaves out
New Hampshire’s chapter 328-B governs massage therapists and massage establishments. Section 10 is an exemption provision, and it does something unusual: it names specific practices in order to carve them out.
Paragraph II covers practitioners who use touch, words and directed movement to build awareness of movement patterns, and exempts them when they work within a profession with established standards and ethics, provided their services are not designated as or implied to be massage or massage therapy. The practices it names are the Feldenkrais method of somatic education, the Trager approach to movement education, the Alexander technique, and body-mind centering.
Paragraph III does the same for practitioners who use energy or superficial touch to affect the body’s energy systems, naming polarity therapy, therapeutic touch, Bowen therapy and reiki.
Two things follow. The legislature itself drew a line between movement-awareness practices and energy practices, which is more discrimination than most consumer writing manages. And the provision creates no protected title and prohibits nothing. It is the opposite of a license: it is a statement that the massage licensing chapter does not apply to these people.
That is the best single fact available about this group, and it is worth sitting with. A legislature looked at the field, found four movement practices distinct enough to name, and the only legal effect of naming them was to place them outside regulation.
The Alexander technique has the best evidence, and it is for function
The Alexander Technique is one-to-one teaching, usually hands-on and verbal, about habitual patterns of posture and effort. Lessons are the unit, and the training behind the teacher is long: professional societies typically require a three-year full-time course amounting to 1,500 to 1,600 hours of instruction.
In August 2008 the BMJ published the trial that anchors this subject. Researchers recruited 579 patients with chronic or recurrent low back pain from 64 general practices in England and randomized them four ways: 144 to normal care, 147 to massage, 144 to six Alexander lessons and 144 to twenty-four. Half of each group was additionally randomized to an exercise prescription from a doctor with behavioral counseling from a nurse. The primary outcomes were a disability score counting activities impaired by pain, and days in pain.
At one year, against a control disability score of 8.1, twenty-four lessons produced a reduction of 3.4 points, with a confidence interval from 4.76 to 2.03. Six lessons produced 1.40, interval 2.77 to 0.03. Exercise prescription produced 1.29. Massage produced 0.58, with an interval running from 1.94 in favor to 0.77 against, meaning it crossed zero and cannot be distinguished from normal care.
The most communicable number in this entire subject is the second outcome. Control patients reported a median of 21 days in pain in the previous four weeks. Twenty-four lessons cut that by 18 days. Six lessons cut it by 10. Massage cut it by 7. No significant harms were reported in 579 patients.
Two details keep this honest. The comparator was normal care, not a dummy treatment, so the 24-lesson effect contains everything that comes with 24 hours of individual attention from a trained teacher. But the trial contains its own test of that objection: massage delivered comparable hands-on contact time and did not beat control at one year, while the lessons did. The trialists also found that exercise after six lessons achieved 72 percent of the effect of 24 lessons alone, which is a dose-response pattern rather than a flat result.
A companion economic analysis published in the same journal that December priced it. Intervention costs ran from 30 pounds for exercise prescription to 596 pounds for 24 lessons plus exercise. Six lessons followed by exercise worked out at an extra 43 pounds per additional pain-free day and 5,332 pounds per quality-adjusted life year, which the health economists judged the most effective and cost-effective option tested. It is the only practice in this library with a published price tag attached to its own effect size.
A second large trial, published in Annals of Internal Medicine in November 2015, tested 20 Alexander lessons against 12 acupuncture sessions in 517 people with neck pain of at least three months, median duration six years. Both arms delivered exactly 600 minutes of one-to-one contact on top of usual care. At twelve months both beat usual care by a similar margin, 3.79 percentage points for lessons and 3.92 for acupuncture on the Northwick Park questionnaire, each statistically significant. No serious adverse events were judged related to either.
What the authors proposed as the mechanism is the interesting part, and it is not technique-specific. Self-efficacy, meaning a patient’s confidence in managing their own pain, improved in both arms and was strongly associated with the twelve-month result. The trialists suggested that improved self-efficacy may partly explain why the benefit lasted. They also noted that their practitioners came from the two main British professional associations, which limits how far the findings travel.
A smaller 2002 trial in Clinical Rehabilitation randomized 93 people with confirmed idiopathic Parkinson’s disease to 24 Alexander lessons, 24 massage sessions or no additional intervention, and found the lessons group improved on a self-assessed disability scale against the no-intervention group, with gains maintained at six months. Read it carefully: four of the six reported confidence intervals have an upper bound of exactly zero, which makes them borderline, and the first author was affiliated with an Alexander teaching society.
Pooling it all, the Australian government review’s Alexander evaluation, prepared by Cochrane Australia and dated 4 November 2024, found seven randomized trials comparing the technique with an inactive control. Its summary of findings table, for chronic low back or neck pain at three to six months, rates physical function and disability at moderate certainty from three trials and 655 participants, with a standardized difference of 0.39 and a tight interval of 0.32 to 0.47. Pain is rated low certainty from four trials and 611 participants, and its interval crosses zero. Quality of life and emotional functioning are both low certainty with intervals crossing zero.
One caution for anyone citing that document. Its narrative text gives different trial and participant counts from its own table for two outcomes, listing 611 participants for disability where the table says 655, and four trials for emotional wellbeing where the table says two. The certainty ratings and the disability estimate are consistent in both places and are the figures to use. Those two participant counts are not settled.
Feldenkrais: ten small trials and nothing certain
The Feldenkrais method combines group classes in guided attention to movement with one-to-one hands-on sessions. The same Australian review, also prepared by Cochrane Australia and dated 4 November 2024, assessed it separately and reported that there is very little evidence on the effects of Feldenkrais, including as an add-on therapy.
Ten randomized trials contributed to the synthesis against inactive controls. The included ones are small, running from about 12 to 65 participants each. Four covered chronic musculoskeletal conditions, and their per-outcome numbers tell the story: pain from four trials and 154 participants, disability from one trial and 51, emotional functioning from one trial and 26, breathing patterns from one trial and 34. Six covered mobility and falls risk, with falls from three trials and 114 participants, mobility from five trials and 205, quality of life from three trials and 133.
Every single outcome in both groups was rated very low certainty.
That is the whole finding and it should not be padded. This is not a hostile verdict; it is a government evidence review reporting that nobody has done enough research to support a claim in either direction. Feldenkrais is the clearest case in this group of a practice whose evidence base is simply too small to carry weight.
Pilates: moderate certainty for disability, and a name anyone can use
Pilates is the outlier in volume. The Australian review’s Pilates evaluation, prepared not by Cochrane Australia but by a separate contractor, Health Technology Analysts, and dated August 2022, found 138 eligible studies, assessed 105 across 26 conditions, and synthesised 66 covering 22 conditions against inactive control. Evidence on a critical or important outcome existed for 15 conditions. Breast cancer, hypertension, chronic widespread pain and shoulder pain could not be assessed because no study reported an outcome the review counted.
For chronic low back pain, disability was rated moderate certainty from 12 studies and 937 participants, described by the reviewers as a moderate reduction. Pain was rated low certainty from 12 studies and 1,062 participants, described as a large decrease. Those descriptors are the review’s own qualitative labels, and the certainty attached to the pain label is the lower of the two. The overall verdict was that the evidence provides moderate to low certainty that practicing Pilates beats not practicing Pilates for some of the conditions assessed, with the effect unknown for others.
Then the part that tells you more about this field than any effect size. In 2000, in the US District Court for the Southern District of New York, Judge Miriam Cedarbaum held that “Pilates” had become a generic term for a method of exercise, incapable of identifying a single source, and directed the Patent and Trademark Office to cancel the marks. The genericness finding rested partly on evidence that the plaintiff and its predecessors, beginning with Joseph Pilates himself, had used the word generically. He had obtained patents on several pieces of equipment he invented, encouraged others to teach his methods, and never acted to stop third parties using his name for the exercises they taught. The equipment was protected. The name was not.
The consequence for a consumer is immediate: anyone may advertise Pilates instruction, with any training or none.
Academics inside the field say the label has drifted further since. A 2024 position paper by two researchers, published in a non-selective open-access journal and read here as argument rather than evidence, reports a lack of clarity about what the term means and about which exercises and apparatus studies used, calls Pilates a branding catchphrase, and lists hybrids now sold under the name including High Intensity Pilates and Wall Pilates. The same authors note that apparatus manufacturers sponsor organizations that recognize practitioners, and that some of those bodies also sell training, which they describe as a possible conflict of interest. They also record what comprehensive teachers say about their own limits: for most, they are neither trained nor qualified to diagnose or treat clinical conditions.
Dance and movement therapy is a different kind of thing
Grouping this with the other three is the category error of the whole topic in miniature.
The Registered Dance/Movement Therapist credential is granted by the Dance/Movement Therapy Certification Board, with requirements set by the American Dance Therapy Association’s education committee, and the board has established the graduate level as the entry point to the profession. Candidates hold a master’s degree, through an approved program, another accredited program, or an alternate route. On top of that: 200 hours of supervised basic clinical exposure overseen by a licensed mental health professional, then an internship of at least 700 hours including a minimum of 350 hours of direct client contact, of which 150 must be leading sessions, plus 70 hours of clinical supervision by a board-certified therapist.
That is a serious credential by the standards of this group. It also has no examination. The handbook evaluates applications, transcripts, verification forms and references.
The sentence consumers most need is in the handbook itself. On the credentials it awards, it states plainly: “These are not equivalent to licensure.” It adds that licenses must be obtained through state licensing boards, and advises applicants to check their state’s requirements for licensure as a mental health practitioner. A dance/movement therapist practicing psychotherapy needs a separate state mental health license, and this credential is not one.
The evidence is thin and the best source is unimpeachable about it. In January 2015 Cochrane published a review of dance/movement therapy for psychological and physical outcomes in cancer patients, requiring that the therapy be delivered by a formally trained therapist or a trainee in a formal program. It found three studies with 207 participants between them. In women with breast cancer it found no effect on depression, a standardized difference of 0.02; none on stress, minus 0.18; none on anxiety, 0.21; none on fatigue, minus 0.36; and none on body image, minus 0.13. Two studies carried moderate risk of bias and one high, making the overall quality very low. Single studies suggested benefits for quality of life, vigour and somatisation. The reviewers concluded that the limited number of studies prevented conclusions either way.
The authors work in the Department of Creative Arts Therapies at Drexel University, and the senior author is a leading figure in the field. A null review written by a practice’s own academic leadership is about as credible as this kind of evidence gets.
The pattern nobody reports
Across the two practices in this group with real trial volume, the same asymmetry appears. For the Alexander Technique, disability is moderate certainty and pain is low certainty with an interval crossing zero. For Pilates, disability is moderate certainty and pain is low certainty.
In this family, the better evidence is for function, not for pain. Most writing on these practices says the opposite, leading with pain relief and treating improved function as a bonus. The large trial evidence runs the other way: the Alexander trial’s strongest result was a disability score and a count of pain-free days achieved through changed behavior, and the government review’s only moderate-certainty ratings, in both practices, are for function.
That is a useful thing for a reader to know before paying for a course of lessons. What the evidence best supports is doing more of what you want to do, not hurting less while doing the same amount.
No examination, no recognized accreditor, no license
Credentials in health care fall into four buckets: written into regulation, created by government as a voluntary certification, awarded by a private third party, or undefined. Here is where these four land.
None of the four has a certification examination. Alexander societies operate waiver routes allowing informally trained teachers to certify through peer review. The Feldenkrais Guild of North America’s certification policy, last revised on 27 January 2025, requires graduation from a program accredited by a training accreditation board recognized by the guild, but the policy does not name that board and states no program hours; it does require 100 hours of professional practice and 20 hours of learning a year after certification. Pilates has competing private bodies and no common standard. Dance/movement therapy has the graduate requirements above.
None of their accreditors holds federal recognition. The US Department of Education’s register of programmatic accrediting agencies names the bodies accrediting naturopathic, osteopathic, chiropractic, acupuncture and massage programs. Nothing on it relates to the Alexander Technique, to Feldenkrais, to Pilates or to dance therapy.
None is licensed in any US state. One is exempted from licensure by name in statute. One lost its name to genericness in federal court. And one, Feldenkrais, is a live registered trademark, enforced by the guild’s own marks policy rather than by any regulator acting for a consumer.
Two practices on the same shelf, opposite legal positions, opposite consequences when you read a studio’s advertising.
What is worth knowing
Harm is not the issue here. Two large pragmatic trials totalling nearly 1,100 patients reported no significant harm from Alexander lessons, and the trials of the other three are too small and too few to say anything in either direction, which is not the same as saying they are safe. The real risk is range of competence, because for three of the four nobody is checking.
So use the evidence as a filter rather than the label. Ask what the specific method is, what its own evidence covers, and who trained the person in front of you and for how long. If a practitioner implies that a movement class treats a diagnosed condition, that is a claim to check against the condition, not against the method.
Two situations need a clinician rather than a studio. New neurological symptoms, meaning weakness, numbness, loss of bladder or bowel control, or pain that wakes you at night, need assessment before any course of lessons. And psychological distress needs a licensed mental health professional, which a movement credential does not make anyone, by the credentialing board’s own statement.
Sources
- New Hampshire General Court. “RSA 328-B:10, Massage Therapists and Massage Establishments: Exemptions; Application of Chapter.” gc.nh.gov
- BMJ. “Little P, Lewith G, Webley F, et al. "Randomised controlled trial of Alexander technique lessons, exercise, and massage (ATEAM) for chronic and recurrent back pain." BMJ, 2008 (579 patients, 64 practices). PubMed. DOI: 10.1136/bmj.a884.” pubmed.ncbi.nlm.nih.gov
- BMJ. “Hollinghurst S, Sharp D, Ballard K, et al. "Randomised controlled trial of Alexander technique lessons, exercise, and massage (ATEAM) for chronic and recurrent back pain: economic evaluation." BMJ, 2008. PubMed. DOI: 10.1136/bmj.a2656.” pubmed.ncbi.nlm.nih.gov
- Annals of Internal Medicine. “MacPherson H, Tilbrook H, Richmond S, et al. "Alexander Technique lessons or acupuncture sessions for persons with chronic neck pain: a randomized trial." Ann Intern Med, 2015 (517 patients; 600 minutes per arm). PubMed. DOI: 10.7326/M15-0667.” pubmed.ncbi.nlm.nih.gov
- Clinical Rehabilitation. “Stallibrass C, Sissons P, Chalmers C. "Randomized controlled trial of the Alexander technique for idiopathic Parkinson's disease." Clin Rehabil, 2002 (93 participants). PubMed. DOI: 10.1191/0269215502cr544oa.” pubmed.ncbi.nlm.nih.gov
- Australian Government Department of Health and Aged Care / NHMRC / Cochrane Australia. “Natural Therapies Review 2024: Alexander Technique Evidence Evaluation, dated 4 November 2024 (searches to 6 April 2023).” health.gov.au
- Australian Government Department of Health and Aged Care / NHMRC / Cochrane Australia. “Natural Therapies Review 2024: Feldenkrais Evidence Evaluation, dated 4 November 2024.” health.gov.au
- Australian Government Department of Health and Aged Care / NHMRC / Health Technology Analysts. “Natural Therapies Review 2024: Pilates Evidence Evaluation, dated August 2022.” health.gov.au
- Australian Government Department of Health and Aged Care / NHMRC. “Natural Therapies Review 2024: Report and Recommendations (copyright 2025; NHMRC process report dated 23 January 2025).” health.gov.au
- Cochrane Database of Systematic Reviews. “Bradt J, Shim M, Goodill SW. "Dance/movement therapy for improving psychological and physical outcomes in cancer patients." Cochrane Database Syst Rev, 2015, CD007103.pub3 (3 studies, 207 participants). PubMed. DOI: 10.1002/14651858.CD007103.pub3.” pubmed.ncbi.nlm.nih.gov
- American Dance Therapy Association / Dance/Movement Therapy Certification Board. “R-DMT Handbook, revised January 2020.” adta.org
- Feldenkrais Guild of North America. “GCFP Certification policy E2.3.1.1, adopted October 1997, last revised 27 January 2025.” cdn.ymaws.com
- AlexanderTechnique.com. “Becoming an Alexander Technique Teacher (three-year course, 1,500 to 1,600 hours; peer-review waiver routes).” alexandertechnique.com
- Stites and Harbison PLLC (Trademarkology). “Pilates: A Deep Stretch of Trademark Law (S.D.N.Y. 2000, Cedarbaum J.; USPTO directed to cancel the marks).” stites.com
- Loeb and Loeb LLP. “Death by Popular Demand: genericide and the Pilates trademark (Pilates Inc v Current Concepts Inc, S.D.N.Y.).” loeb.com
- Journal of Alternative, Complementary and Integrative Medicine (via University of Strathclyde Strathprints). “Lewitt MS, McPherson LA. "Pilates in 2024: useful clinical tool or commercial catchphrase?" 2024. Position paper in a non-selective open-access journal; cited as argument, not evidence.” strathprints.strath.ac.uk
- US Department of Education. “Programmatic Accrediting Agencies.” ed.gov