Western physiotherapy
physiotherapy / physical therapyphysiotherapyEnglish
“treatment by physical means”
A profession forged in the polio epidemics and the First World War, built on the idea that tissue is a material with a load tolerance and that graded mechanical stimulus is a prescribable dose. It is included here on exactly the same terms as every other tradition, because it has a theory of the body with commitments and blind spots, and because an atlas that exempts the reader's own tradition from scrutiny is not an atlas.
Also called
- physical therapy (US)
- kinesitherapy
- medical gymnastics (the nineteenth-century antecedent)
Era of origin
1916CE
Regulation
Statutory
Origin region
Northern and Western Europe, then North America
Stockholm, London, and the American military rehabilitation programme
Evidence grade
Strong
Multiple independent lines converge, including controlled work, with a mechanism that holds.
The headline finding
The exercise arm has the best outcome evidence in this atlas, including trials where it matches or beats surgery. The passive arm includes modalities with weak or negative evidence that remain in daily use, which the profession's own literature says plainly.
Situating it
- Intellectual lineage
- Ling's Royal Central Institute of Gymnastics (Stockholm, 1813) supplied the template of prescribed graded movement, which is why this entry begins where the Swedish massage entry ends. The profession itself was created by two crises: the polio epidemics from 1916 and the First World War, which produced enormous numbers of paralysed children and injured soldiers and, with them, the reconstruction aides in the United States and the organizations that became the Chartered Society of Physiotherapy in Britain (chartered 1920). Its founding population was disabled civilians and wounded soldiers, not athletes.
- Era
- Antecedents from 1813; the profession forged 1916 to 1920
- Where it is practiced today
- A global statutory health profession with several million practitioners, embedded in hospitals, public health systems, private clinics, and professional sport at every level. In most high-income countries it is the default first-line provider for musculoskeletal complaints.
- Training
- Three to four years at degree or doctoral level: anatomy, physiology, pathology, biomechanics, exercise physiology, imaging interpretation, and extensive supervised clinical placement. Specialist certifications in sports, orthopaedics, and neurology on top.
- Regulatory status
- The most heavily regulated tradition in the atlas. Entry-level qualification in the United States is the Doctor of Physical Therapy, a three-year doctorate accredited by CAPTE, followed by the National Physical Therapy Examination and state licensure, with direct access (patients self-referring without a physician) available in all fifty states subject to varying restrictions. In the United Kingdom the title is protected and registration is with the Health and Care Professions Council. Degree-level entry and statutory registration are the norm across the OECD.
- Integration into modern sport
- The default. Physiotherapists staff essentially every professional team and national programme, run return-to-play decision-making, and own the rehabilitation pathway after surgery. In practice the quality range within the profession is wide, and the distinguishing variable is almost always how much of the session is spent loading tissue versus how much is spent on passive modalities.
The technique
Thirty to sixty minutes: a structured examination, a working hypothesis, a dosed exercise prescription with progression criteria, and an explanation. The valuable part usually happens between visits.
- 01
Subjective and objective examination
History and mechanism of injury, then observation, active and passive range measured in degrees with a goniometer or inclinometer, manual muscle testing graded 0 to 5 on the Oxford scale, orthopaedic special tests (Lachman, Spurling, Hawkins-Kennedy, Thessaly), neurological screening, palpation, and functional testing (single-leg hop, Y-balance, hop batteries).
- 02
Outcome measures
Validated patient-reported instruments recorded at baseline and repeated: KOOS for the knee, DASH for the upper limb, Oswestry or Roland-Morris for the low back, ACL-RSI for psychological readiness to return. The profession quantifies its own results, which almost nothing else in this atlas does.
- 03
Exercise prescription, the core
Specified exercise, load, sets, repetitions, tempo, frequency, and progression criteria. Isometrics for irritable tendon pain. Heavy slow resistance, typically three to four sets of six to fifteen repetitions with three-second eccentric and concentric phases, three times weekly for twelve weeks. Alfredson's eccentric protocol for Achilles tendinopathy: three sets of fifteen, twice daily, for twelve weeks. Then plyometric and change-of-direction progressions.
- 04
Manual therapy as adjunct
Joint mobilization graded I to IV, thrust manipulation, soft tissue work, neural glides, dry needling. Used to create a window of reduced pain and improved range in which loading becomes possible, which is the defensible rationale for all of it.
- 05
Modalities
Therapeutic ultrasound, TENS, laser, shockwave, traction, taping, electrical stimulation. Their evidence ranges from weak to absent, with a genuine exception for shockwave in some tendinopathies. Named here specifically because they occupy a large share of clinic time.
- 06
Education and load management
What the diagnosis means, what the timeline is, what to expect from flare-ups, how to modify training rather than stop it, and a graded return-to-sport pathway with explicit clearance criteria such as limb symmetry indices and hop-test thresholds. Often the most valuable twenty minutes of the whole course.
The dose distinction that matters for a reader comparing traditions: physiotherapy is the only entry in this atlas where most of the therapy happens when the practitioner is not in the room. The visit exists to prescribe, progress, and troubleshoot the home programme. A physiotherapy course in which nothing was prescribed to do daily has not really happened, which is also why adherence, rather than technique, is the profession's central practical problem.
Blood flow restriction training deserves a mention as a genuinely clever recent addition: a cuff at partial arterial occlusion allows hypertrophy-relevant adaptation at 20 to 30 percent of one-repetition maximum, which is what you need when a joint cannot yet tolerate heavy load.
The theory of the body
Tissue is a material with a load tolerance. Symptoms arise when applied load exceeds current capacity. Tissue adapts to load. Therefore the treatment is graded mechanical stimulus, dosed like a drug.
This is a real theory of the body and it should be read as one rather than as the absence of a theory. Its ontology is mechanical: the body is a set of materials under load, with measurable properties, adapting according to the stimulus applied. Its epistemology is measurement: what matters is what can be quantified, and the profession's examination is an apparatus for quantifying. Its therapeutics is deducible from its physiology in exactly the way Ayurveda's is deducible from guna theory: identify the tissue, establish its current capacity, apply a stimulus above it, progress on criteria.
It also carries commitments, and naming them is the point of including this entry. The model assumes that the relevant causes are local, mechanical, and measurable. It assumes symptoms track tissue state. It assumes function decomposes into components. Each of those assumptions is productive and each of them is sometimes false, and where they are false is precisely where physiotherapy has historically gone wrong.
The blind spots, named specifically. First, chronic pain: the tissue-and-load model predicts pain from tissue state, and for persistent pain that prediction is simply wrong. Imaging findings correlate poorly with symptoms, with disc degeneration present in the majority of pain-free adults over forty and asymptomatic rotator cuff and meniscal tears extremely common6. The profession spent decades chasing pelvic tilts, leg-length differences, core weakness, scapular dyskinesis, and posture as pain causes, on the strength of the mechanical model rather than on evidence, and much of that has not held up. The correction came substantially from outside, from pain neuroscience, and the profession is still absorbing it. Second, the implementation gap: passive modalities with weak or negative evidence remain in widespread daily use because they are billable, comfortable, and fill the first ten minutes of a visit. Third, examination reliability: many special tests have modest sensitivity and specificity, and palpation-based assessments have poor inter-rater reliability, the same criticism this atlas levels at chiropractic motion palpation and at Chinese pulse diagnosis. Fourth, specificity theatre: bespoke corrective-exercise programmes frequently perform no better than general strengthening, which suggests much of the prescribed specificity is not doing what it claims. Fifth, transfer: the profession was built for war wounds and polio, and its move into persistent nonspecific pain, where the driver is not tissue damage, was slower and rockier than its self-image admits.
Taken on its own terms
This model produced the most reliably effective non-surgical interventions in musculoskeletal medicine, and that is not a small thing to have done. It made exercise a prescribable dose, it generates predictions specific enough to be falsified individually, it quantifies its own outcomes, and it produced results strong enough to overturn surgical practice: exercise therapy matched arthroscopic surgery for degenerative meniscal tears4 and placebo surgery matched real decompression for subacromial pain5. It also, and this is the strongest thing that can be said for any tradition in this atlas, published the studies that embarrassed it. The critique of core-stability dogma, of therapeutic ultrasound, of the imaging-symptom mismatch, and of the profession's own overtreatment came from inside physiotherapy. That is what a healthy tradition looks like, and it is the standard the other eight entries are being held to.
Its vocabulary, in its own words
- Load tolerance
- The central object. Every tissue has a current ceiling for the mechanical stress it can absorb without symptoms, that ceiling is measurable in principle, and it moves in response to what you do.
- Capacity and demand
- The clinical frame. Pain is read as a mismatch: either demand rose (training spike, new surface, new volume) or capacity fell (detraining, illness, sleep loss, surgery). The intervention closes the gap from whichever side is cheaper.
- Progressive overload
- Adaptation requires a stimulus above habitual, applied repeatedly, with recovery between. Inherited from strength training and from Wolff's law of bone, and generalized to tendon, muscle, cartilage, and nerve.
- Dose
- The direct descendant of Ling's movement prescription. Exercise is specified in load, volume, frequency, tempo, and progression, because the profession's founding conviction is that movement is a prescribable quantity.
- Impairment decomposition
- Function is broken into measurable components (range, strength, power, endurance, motor control), each independently addressable and testable. Restore the components and function returns.
- Tissue healing timeline
- Inflammatory, proliferative, and remodelling phases with stated durations, used to gate what load is appropriate when. Progression is governed by the biology's clock and by symptom response.
The physiology
Mechanotransduction with stated timescales, which is why the intervention can be written as a schedule rather than as a course of treatment.
Tendon adaptation
Strain on tenocytes through integrin-cytoskeleton coupling upregulates IGF-1 and TGF-β signaling and type I collagen synthesis. Mechanical stiffness of the muscle-tendon unit changes measurably over eight to twelve weeks of heavy slow resistance. Collagen turnover in the tendon core is extremely slow, so most of the adaptation is in the interfibrillar matrix and in muscle properties, which is why tendon rehabilitation takes months and not weeks.
Muscle hypertrophy
Resistance load activates mTORC1; myofibrillar protein synthesis is elevated for 24 to 48 hours per bout; measurable hypertrophy appears over six to twelve weeks. Early strength gains, in the first two to four weeks, are predominantly neural rather than structural, which is why a rehabilitating athlete gets stronger before they get bigger.
Bone modelling
Osteocytes sense strain and initiate modelling above a threshold. Measurable bone mineral density change takes three to six months, and stress fracture occurs when load rate outruns remodelling capacity. This timescale is why return-to-running progressions are conservative and why they are frequently violated.
Exercise-induced hypoalgesia
Exercise raises pain thresholds through central mechanisms, and graded exposure reduces the threat value of a movement independently of any tissue change. This is the mechanism the mechanical model did not predict and the one that explains why loading helps chronic pain.
Where the passive modalities sit
Therapeutic ultrasound at clinical doses has not been shown to produce clinically important benefit in most musculoskeletal conditions3. Manual therapy produces the same transient neurophysiological effects documented in the chiropractic and tuina entries: real, short, and best used to open a window for loading rather than as the treatment.
The timescales are the profession's genuine intellectual asset and the thing an athlete should take from this entry regardless of which tradition they use. Neural strength gain in two to four weeks. Hypertrophy in six to twelve. Tendon stiffness in eight to twelve. Bone density in three to six months. Every recovery decision an athlete makes is a bet against one of those clocks, and no other tradition in this atlas states them.
It is worth being explicit that physiotherapy's manual arm and the manual traditions in this atlas are doing the same thing physiologically. A physiotherapist mobilizing a stiff thoracic spine, a tuina practitioner rolling the paraspinals, and a chiropractor thrusting the same segment are all delivering mechanoreceptor input that produces transient hypoalgesia and improved range. The difference is not the mechanism. The difference is what happens in the remaining forty minutes, and whether anything was prescribed for tomorrow.
What is not supported, stated precisely
Symptoms as a readout of tissue state is not supported for persistent pain: imaging findings and symptoms dissociate widely6. Postural correction, leg-length equalization, and pelvic-tilt normalization as treatments for pain are not supported. Core stability as specifically superior to general exercise is not supported. Therapeutic ultrasound and most low-level laser at clinical doses are not supported3. Stretching as injury prevention is not supported. Manual therapy as producing lasting structural change is not supported, on the same connective-tissue mechanics grounds as the Swedish massage entry. Stated precisely: several of physiotherapy's own proposed mechanisms do not survive, and the profession is unusual in this atlas mainly for having published the studies that showed it.
Side by side
The proposed mechanism, and the current physiology
The left column is the tradition’s own account, stated as it would state it. The right column is what can currently be said. The verdict grades the explanation only. A practice can survive a failed mechanism, and in this atlas it frequently does.
Question
Proposed mechanism
Current physiology
Standing
Tissue is a material with a load tolerance
Proposed
Symptoms arise when applied load exceeds current capacity; raise capacity and symptoms resolve.
Current physiology
Well supported for tendon, bone, and muscle, with stated mechanotransduction pathways and measured timescales. The profession's core claim is the one that holds.
ConvergesExercise is a prescribable dose
Proposed
Specify load, volume, frequency, and progression, and adaptation follows predictably.
Current physiology
Supported, and inherited directly from Ling. Dose-response relationships for hypertrophy, tendon stiffness, and bone density are real and quantified.
ConvergesSymptoms report tissue state
Proposed
Pain indicates what is happening in the tissue; find the mechanical fault and fix it.
Current physiology
Fails for persistent pain. Imaging and symptoms dissociate widely6, and central sensitization, threat appraisal, sleep, and load history contribute independently. This is physiotherapy's version of a mechanism that did not survive.
Not supportedSpecific corrective exercise
Proposed
Identify the impaired component and target it precisely; general exercise is a blunt instrument.
Current physiology
Bespoke corrective programmes frequently perform no better than general strengthening, and no specific exercise type reliably beats another for low back pain. The specificity is largely not doing the work.
Not supportedPassive modalities help tissue heal
Proposed
Ultrasound, laser, and electrical stimulation accelerate tissue healing and reduce pain.
Current physiology
Weak to negative evidence at clinical doses3, and still in daily use. The same standard this atlas applies to cupping applies here and produces the same answer.
Not supportedManual therapy creates a window
Proposed
Mobilization reduces pain and improves range so that loading becomes possible.
Current physiology
Supported as stated, which is why the framing matters. Transient hypoalgesia and range gain are real; the honesty is in calling it a window rather than a treatment.
Converges
- Converges The tradition's account and current physiology point the same way.
- Reframed The prediction survives. The explanation does not, and a different mechanism accounts for it.
- Not supported No mechanism, no route, no data. Stated about the mechanism, not about the practice.
The evidence
The strongest evidence in the atlas, and an unusually clean split within a single tradition between what is supported and what is merely customary.
Claim by claim
Exercise therapy for knee and hip osteoarthritis
SupportedCochrane-supported, with pain and function benefit comparable to NSAIDs and without their harms. One of the best-established non-drug interventions in medicine.
Progressive loading for Achilles and patellar tendinopathy
SupportedEccentric and heavy slow resistance protocols outperform passive care consistently. Timescale is twelve weeks and there is no shortcut, which is the whole clinical message.
Exercise-based injury prevention programmes in youth athletes
SupportedNordic hamstring protocols roughly halve hamstring injury rates in meta-analysis2, and neuromuscular programmes reduce ACL injury incidence. Prevention that actually prevents, which is rare.
Exercise therapy versus surgery for degenerative meniscal tear
SupportedKise's randomized trial found exercise therapy non-inferior to arthroscopic partial meniscectomy at two years4. A tradition that overturned a common operation with a randomized trial has earned its grade.
Structured exercise for chronic low back pain
SupportedSmall to moderate effects, and the best available. Notably, no specific exercise type reliably beats another, which is itself evidence against the specificity the model likes to claim.
Manual therapy as an adjunct
PartlyTransient analgesia and range improvement, real and short. Defensible as a window-opener for loading, indefensible as the treatment.
Therapeutic ultrasound, most laser, and traction
UnsupportedWeak to negative evidence, and still in daily clinical use3. The gap between the profession's own evidence base and its clinic behaviour is the single fairest criticism of physiotherapy, and physiotherapy researchers are the ones making it.
Postural correction and core stability as specific treatments for pain
UnsupportedDid not survive scrutiny. Both were adopted widely on mechanical plausibility and both are now substantially walked back inside the profession.
The summary sentence for this entry, and arguably for the whole atlas, is this: physiotherapy's exercise arm is the best-evidenced thing in the Global Recovery Atlas, and physiotherapy's passive arm is about as well-evidenced as everything else in the Global Recovery Atlas. Those are two different claims about one profession, and clinics blur them constantly, because a session that begins with ten minutes of ultrasound and ends with three exercises feels like one treatment to the patient.
Being explicit about the implication: a reader who came to this page expecting Western medicine to be the yardstick against which the others are measured should notice that the yardstick had to be split in half to be used. Physiotherapy earns its strong grade on the parts that load tissue. It does not earn it on the parts that resemble every other tradition here, and it should not be graded on the reputation of its best arm.
What it is genuinely good for
The verdict
The only tradition in this atlas that reliably changes what your tissue can tolerate rather than how it currently feels. Slow, effortful, unglamorous, and the one you cannot substitute.
Use it for
- Anything requiring more capacity than you currently have, which is most of sports medicine
- Post-surgical rehabilitation and criteria-based return to sport
- Tendinopathy, where nothing else works and the timeline is twelve weeks
- Injury prevention through neuromuscular and eccentric programmes, which genuinely reduce incidence
- Getting an accurate explanation of what is wrong and what the timeline is, which changes behaviour more than any modality
Do not use it for
- Passive modality sessions with no exercise prescription attached
- Chasing posture, pelvic alignment, or leg-length differences as causes of pain
- Expecting tissue change on a timescale shorter than the biology allows
Judge a physiotherapist by the same questions this atlas asks of every practitioner. Did they measure something at baseline that they will measure again? Did you leave with a specific prescription including load, sets, and frequency? Did they explain the timeline in weeks and name what would count as progress? Is more than half the session spent doing rather than receiving? A clinician who answers those well is delivering the best-evidenced intervention on this website. One who answers them badly is delivering a manual therapy with better regulatory paperwork than the others in this atlas and no better evidence.
Cautions
Loading progressed faster than the tissue's remodelling clock causes stress fracture and tendon flare. The timelines are biological and not negotiable by motivation.
Pain during rehabilitation is often acceptable and sometimes necessary, but that judgement requires a diagnosis. Do not self-apply it.
Red flags override the model: unexplained weight loss, night pain, fever, progressive neurological deficit, or inability to bear weight need medical assessment, not more exercise.
A course of care with no measured outcome and no endpoint is the same problem this atlas flags in the chiropractic entry, and it occurs here too.
Sources
References
A reading list rather than the output of a systematic search, and it is labelled that way deliberately. Where a claim on this page rests on a specific study, the superscript number links here.
- [1]
Fransen M, McConnell S, Harmer AR, et al. Exercise for osteoarthritis of the knee. Cochrane Database Syst Rev. 2015;(1):CD004376.
- [2]
van Dyk N, Behan FP, Whiteley R. Including the Nordic hamstring exercise in injury prevention programmes halves the rate of hamstring injuries: a systematic review and meta-analysis of 8459 athletes. Br J Sports Med. 2019;53(21):1362-1370.
- [3]
Robertson VJ, Baker KG. A review of therapeutic ultrasound: effectiveness studies. Phys Ther. 2001;81(7):1339-1350. (And subsequent reviews reaching the same conclusion.)
- [4]
Kise NJ, Risberg MA, Stensrud S, et al. Exercise therapy versus arthroscopic partial meniscectomy for degenerative meniscal tear in middle aged patients: randomised controlled trial with two year follow-up. BMJ. 2016;354:i3740.
- [5]
Beard DJ, Rees JL, Cook JA, et al. Arthroscopic subacromial decompression for subacromial shoulder pain (CSAW): a multicentre, pragmatic, parallel group, placebo-controlled, three-group, randomised surgical trial. Lancet. 2018;391(10118):329-338.
- [6]
Brinjikji W, Luetmer PH, Comstock B, et al. Systematic literature review of imaging features of spinal degeneration in asymptomatic populations. AJNR Am J Neuroradiol. 2015;36(4):811-816.
- [7]
Alfredson H, Pietilä T, Jonsson P, Lorentzon R. Heavy-load eccentric calf muscle training for the treatment of chronic Achilles tendinosis. Am J Sports Med. 1998;26(3):360-366.
- [8]
Moseley GL, Butler DS. Fifteen years of explaining pain: the past, present, and future. J Pain. 2015;16(9):807-813.
- [9]
Ottosson A. The manipulated history of manipulations of spines and joints? Rethinking orthopaedic medicine through the 19th century discourse of European mechanical medicine. Med Stud. 2011;3(2):83-116.
Continue
Elsewhere in the lab
Cross-links chosen for this tradition specifically, plus the rest of the atlas.
Education, not diagnosis. This is a student-authored science platform. Nothing here replaces a physician, a physical therapist, or an athletic trainer. Sudden severe pain, numbness, an inability to bear weight, or visible deformity means stop reading and get seen.