Idiopathic Inflammatory Myopathy

Stand up from a chair five times without using your hands.

If you genuinely cannot, if both sides feel the same, and if it is worse than it was a month ago, that is not laziness and it is not age. Book a doctor's appointment this week and take your numbers with you.

Do not wait for an appointment if you are choking on food or drink, newly short of breath, or your urine has turned dark. That is same-day medical care.

Takes 30 seconds. No equipment needed.

The Verdict

If you're getting weaker on both sides and training isn't fixing it, that's a doctor's question.

Think of your muscles as a workforce and your immune system as building security. Normally security just checks IDs at the door. In this condition it starts arresting the workers, so the muscle is not tired and not sore, there are simply fewer people on the job each week. That is why resting does not restore it, and why the fix is calling security off first and rebuilding the workforce second.

What this actually is: your immune system is attacking muscle tissue directly, so the weakness is real rather than pain holding you back.
What most people get wrong: a "negative" antibody panel does not clear you, because the commonest version of that lab test misses about a third of the highest-risk cases.
Start here: write down what you could do a month ago and what you can do today, and take that comparison to your doctor.

Best for

Adults noticing genuine, both-sided weakness in the shoulders or hips that is getting worse over weeks to months. Also for people already diagnosed and stable who want to know whether exercise is safe.

Skip if

You have trouble swallowing, new breathlessness, or dark urine. Stop reading and get same-day medical care.

Want the full evidence? Keep scrolling

What Works

Cinematic anatomical study of proximal limb girdle musculature

Tier 1 — Strong evidence

Recognition and referral.HIGH Telling genuine weakness apart from pain-limited weakness, recording dated objective measurements, and referring for blood tests and specialist assessment. In one subtype the median time from first symptom to diagnosis was 3 months, with a documented tail out to 156 months. That tail is what recognition shortens.

Confirming cancer screening has happened in a newly diagnosed adult.HIGH Dermatomyositis carries a pooled relative risk of 4.66 against the general population across 4,538 patients; polymyositis 1.75. Risk concentrates in the years around onset.

Progressive resistance and aerobic exercise once disease is established and stable.HIGH for safetyMODERATE for size Across 19 studies and 298 patients, strength improved (SMD 0.61, 95% CI 0.37–0.85) and aerobic capacity improved (SMD 0.82, 95% CI 0.29–1.34), with muscle enzymes unchanged and no flares reported.

Asking about swallowing and breathing at every review.HIGH Swallowing difficulty affects roughly a third of dermatomyositis patients. Lung involvement is the principal cause of early death. It costs two questions.

Exercise Prescription

For confirmed, stable disease only, and alongside the medical team rather than instead of it. If the diagnosis is not yet made, the plan is a single item: see the doctor, and take the dated measurements with you.

Sit-to-stand

3 × 8–12 · 2 days/week · effort is fine, stop if genuinely unsteady

Leg press or wall squat

3 × 8–12 · 2 days/week · hard work by the last 2 reps

Overhead press

3 × 8–12 · 2 days/week · effort, not strain

Seated row or band pull

3 × 8–12 · 2 days/week

Steady cardio

20–30 minutes · 2–3 days/week · able to hold a conversation

These are the loads used in the trials that tested this safely. They are not a proven ideal dose. No study in this literature has ever compared two exercise doses against each other, and an audit of 14 trials found only 6 reported all four of frequency, intensity, time and type. The right amount is the one you can do consistently and recover from, adjusted on what your strength testing actually shows.

Tier 2 and Tier 3 — moderate and emerging evidence

High-intensity resistance trainingMODERATE Two whole-body sessions per week for 16 weeks. Quality of life improved by 5.33 points (95% CI 0.61–10.05), muscle endurance by 11.49 (3.37–19.60) and strength score by 1.30 (0.09–2.51), with patients staying in remission. One trial, 32 patients.

Low-load training with blood flow restrictionMODERATE 30% of one-rep max, twice weekly for 12 weeks. Leg press up 19.6%, knee extension up 25.2%, thigh muscle size up 4.57%, enzymes unchanged, no adverse events. 13 patients, single-arm.

Aerobic and endurance trainingMODERATE Improved fitness, physical function and daily activity scores across two randomised trials. One systematic review concluded a programme should at minimum include endurance work.

Low-load home exercise in recent-onset active diseaseEMERGING 15 minutes of exercise plus a 15-minute walk, 5 days a week. Repeat muscle biopsy and scan after 12 weeks showed no increase in muscle inflammation. Safety is the demonstrated finding. Effectiveness at this stage is not established.

Preservation training in inclusion body myositisEMERGING The trial missed its primary endpoint: no improvement in self-reported or measured function. What it did show is that untrained patients lost 9.2% of leg strength while trained patients lost none. Preservation is worth having and it is a different promise from improvement.

What Doesn't Work

Treating it as deconditioning and progressing the load. This is the default error and it costs months. Progressive strength loss on an adequate programme is a diagnostic finding, not an adherence problem.
Using the classification criteria as a bedside rule-out. They discard 26.3% of amyopathic dermatomyositis and reach only 73% sensitivity for polymyositis. They exist to build research cohorts.
Reading a negative antibody panel as reassurance. Pooled sensitivity for anti-TIF1-gamma on the commonest platform is 63.8%, against a specificity of 94.7–99.3%. The two are not symmetric and they get collapsed into "the panel was normal".
Treating anti-HMGCR necrotising myopathy as a severe statin ache. Stopping the statin does not resolve it. It needs immune suppression.
Withholding exercise from stable disease for fear of a flare. Muscle enzymes did not rise across 19 studies and 298 patients, and the deconditioning cost of that caution is paid immediately.
Relying on MRI to make the diagnosis. The best combination of thigh MRI features reached only 55% positive predictive value. MRI maps extent and targets biopsy.
Promising inclusion body myositis patients that training will make them stronger. The trial that tested it missed its primary endpoint.

Red Flags

Dark cinematic study of respiratory and swallowing musculature

New or worsening trouble swallowing, choking on food or drink, or liquid coming back through the nose. Aspiration risk, and an independent predictor of death in inclusion body myositis. Same-day medical contact.
New breathlessness, dry cough, or dropping oxygen on exertion. Lung involvement is the principal cause of early death and the rapidly progressive form evolves over weeks. Same-day medical contact.
Genuine symmetrical weakness in the shoulders and hips getting worse over weeks to months, not explained by pain or disuse. Urgent GP or rheumatology referral.
A dermatomyositis rash that breaks down into a sore or ulcer. Cancer risk ratio 2.73, one of the strongest single clinical predictors found.
Any newly diagnosed adult who has not had cancer screening. Risk is concentrated in the three years around onset.
New focal weakness, dark urine, or a rapidly rising muscle enzyme level.
Quadriceps buckling with a weak grip in someone over 50. The inclusion body myositis pattern, chronically missed, with a median 10.5 years from first symptom to wheelchair.
Weakness that persists or worsens after a statin has been stopped. That is a different disease from statin ache.
Groin or hip pain in a patient on long-term steroids for this condition. Bone death prevalence in that group is 5.9%. Do not work it up as a hip flexor strain.

Refer to: GP or rheumatology for suspected new disease, urgently, with the weakness pattern and dated measurements documented. Same-day medical assessment for swallowing difficulty with aspiration risk, new or progressive breathlessness, or rapidly progressive weakness. Neurology where motor neurone disease is a genuine alternative.

Return to Training

Diagnosis established and disease documented as stable or in remission by the treating team
Cancer screening completed and documented (adult-onset disease)
No swallowing difficulty, or it has been formally assessed and managed
No new or progressive breathlessness or dry cough
Strength score stable or improving across two consecutive assessments at least 4 weeks apart
Timed functional measure stable or improving
Programme completed without post-session strength loss at 24 to 48 hours
Patient can state their own red flags and what to do about them

Conviction

MODERATE overall, and the sub-claims differ enough that a single number would mislead.

Recognition and referral pathway HIGH
Malignancy association and risk stratification HIGH
Exercise safety in established, stable disease HIGH
Size of the exercise effect MODERATE
High-intensity training in stable disease MODERATE
Exercise in active or recent-onset disease LOW to MODERATE
Preservation training in inclusion body myositis LOW
Any specific exercise dose, intensity or stopping threshold NO EVIDENCE
Any bedside test with published accuracy NO EVIDENCE

What would change my mind on the exercise dose

A multicentre trial of at least 120 adults with established, stable disease, randomised to three arms (high-intensity resistance at 70–85% of one-rep max twice weekly; low-load blood flow restriction at 30% twice weekly; and a graded aerobic-only comparator) for at least 24 weeks, with full reporting of frequency, intensity, time and type plus adherence to each. A between-arm difference would give this field its first real dose-response evidence. Nothing smaller will, because every trial to date is under 35 patients and single-armed on dose.

What would change my mind on recognition

A prospective cohort of at least 500 consecutive adults presenting with unexplained weakness in the shoulders and hips, in whom a standard bedside battery is recorded blind to the eventual diagnosis, with 12-month follow-up. That study would produce the accuracy figures this page currently has to report as unavailable, and it is the single most useful trial that could be run for physical therapy in this condition. Nothing like it exists in 292 papers.

The Full Picture — Anatomy, Diagnosis & Evidence

What's Actually Going On

$Dark cinematic rendering of skeletal muscle fibre architecture$

The immune system attacks skeletal muscle directly. Other organs are frequently involved, which is why these are best understood as whole-body diseases rather than muscle diseases with side effects.

The subtypes behave like different conditions and should be held apart. Dermatomyositis carries the skin signs (knuckles, eyelids, the V of the neck) and the strongest cancer link. Polymyositis is increasingly a diagnosis of exclusion, being dissolved into other categories as antibody testing improves. Immune-mediated necrotising myopathy is the most severe form, defined by antibodies rather than by inflammation on biopsy, and its anti-SRP variety brings more severe weakness, neck weakness, swallowing difficulty and breathing difficulty than the anti-HMGCR one. Sporadic inclusion body myositis is the commonest form over 50, does not respond to immune suppression, and breaks the usual pattern by hitting the quadriceps and the deep finger flexors, often unevenly.

Which muscles matter: the hip and shoulder girdles carry the burden in most subtypes, along with the neck flexors, which are commonly involved and rarely tested. The muscles you swallow with are striated too, which is why swallowing difficulty is common rather than exotic.

How to Identify It

Cinematic study of hip and shoulder girdle musculature under dramatic light

The honest headline is an absence. No bedside clinical test for this condition carries a published accuracy figure anywhere in the 292 papers reviewed for this page. Recognition rests on pattern, symmetry and trajectory, and the figures below all come from laboratory or classification instruments rather than from anything you can do with your hands.

Any bedside special test Sn: data unavailable | Sp: data unavailable — none exists. Borrowing a figure from a general musculoskeletal prediction rule was considered and declined, because a number developed in a different population is an invented number here.
2017 EULAR/ACR criteria, without biopsy Sn: 87% | Sp: 82% — derived on 976 patients against 624 look-alikes across 47 clinics. These are classification criteria for assembling research groups, not a bedside rule-out.
2017 EULAR/ACR criteria, with biopsy Sn: 93% | Sp: 88%
Anti-TIF1-gamma for cancer-associated dermatomyositis Sn: 78% | Sp: 89% — tells you about cancer risk, not about the muscle diagnosis. Negative predictive value 95% at a 17% background rate, but positive predictive value only 58%.
Anti-cN1A for inclusion body myositis Sn: moderate | Sp: high — the meta-analysis concluded these are not useful diagnostic markers. Present in 40% of patients against 3.6% of controls, so a positive result argues for it and a negative result argues for nothing.

What to do with your hands instead: watch the patient stand from a chair before you touch them, test strength formally across both girdles plus the neck flexors, time a functional measure, and then write it down and date it. A repeated, measured, dated strength record is the single most valuable thing a physical therapist holds in this condition, because it is the only thing that proves a trajectory.

The Debate

Classification criteria: endorsed instrument vs bedside misuse

2017: the EULAR/ACR criteria "generally perform better than existing criteria" and are endorsed internationally.

2024: a scoping review of 19 articles and 13 abstracts identifies significant limitations and poor coverage of necrotising myopathy, amyopathic dermatomyositis, antisynthetase syndrome and overlap myositis, and concludes a revision is warranted. Separately, 26.3% of 211 amyopathic dermatomyositis patients fail the recommended cut-off, and sensitivity for polymyositis in a 439-patient consecutive cohort was only 73%.

Use them to understand how a diagnosis was reached, never to exclude one. A patient who fails the cut-off has not been cleared.

Antibody panels: excellent at ruling in, poor at ruling out

Established view: myositis-specific antibodies are a cornerstone of diagnosis and subtyping.

2025 benchmark against the reference method: specificity is 94.7–99.3% across the board, but sensitivity swings from 95.7% for anti-MDA5 by one method down to 63.8% for anti-TIF1-gamma by the commonest one.

A positive panel is strong evidence. A negative panel is weak evidence, and for the antibody that quadruples cancer risk it misses roughly a third of cases. "The panel was normal" is not reassurance.

Muscle enzyme level and cancer risk: two meta-analyses, opposite directions

2014: a raised muscle enzyme level is among the factors predicting cancer in this population.

2021 (69 studies): a very high enzyme level is associated with reduced cancer risk.

Both hold at their own altitude. The 2014 analysis splits raised against not-raised; the 2021 analysis compares cancer-associated against non-cancer-associated patients already inside this population, and cancer-associated dermatomyositis skews toward the skin-predominant form with less muscle destruction. The clinician asking the question is already inside that population, so the 2021 reading governs, and the consequence reverses instinct: a modest enzyme level does not lower cancer suspicion.

Exercise: pooled across stages vs graded by stage

Pooled (19 studies, 298 patients): exercise is probably safe and moderately improves strength and fitness at all disease stages and ages.

Graded by stage: insufficient evidence for recent-onset disease and for inclusion body myositis, while supporting moderate-to-high intensity training in established disease.

Not a contradiction. One answers "does exercise help this condition", the other answers "does it help this patient at this stage". The stage-stratified reading governs prescription.

Honest Limitations

Every exercise trial studied stable, already-diagnosed disease

The research: pooled improvements in strength and fitness with no rise in muscle enzymes across 19 studies and 298 patients.

The gap: those trials recruited patients in remission, supervised, in academic centres, with samples between 9 and 32. An adult with unexplained progressive weakness and no diagnosis has no trial evidence attached to them at all, because such a person is a referral rather than a participant.

The adjustment: treat diagnostic status as the first branch of the whole plan. Before a diagnosis, the intervention is recognition, measurement and referral.

The prescriptions are incompletely reported

The research: a range of protocols all produced benefit without flares.

The gap: an audit of 14 studies found only 6 reported all four of frequency, intensity, time and type; progression and overload appeared in half; and reversibility and diminishing returns in none at all. No study compared two doses against each other.

The adjustment: present the doses as tested protocols, not optimal ones, and progress on the individual's measured response rather than on a schedule borrowed from a trial that did not fully describe itself.

The antibody panel performs worse commercially than its reputation implies

The research: antibodies define subtypes, drive cancer risk stratification, and carry excellent specificity.

The gap: pooled sensitivity for anti-TIF1-gamma on the commonest commercial platform is 63.8%. A patient can arrive with "the myositis screen was negative" in their notes and still carry the antibody that quadruples their cancer risk.

The adjustment: do not let a reported negative panel lower your suspicion in a patient whose pattern and trajectory still fit. Escalate on the clinical picture.

The Nuance

Cinematic anatomical contrast study of muscle groups

There is no surgical pathway for the muscle disease itself. This is managed medically, with steroids and steroid-sparing drugs, and immunoglobulin where the response is insufficient. Surgery enters only where an underlying cancer is found and removed, which is a separate decision.

That leaves the physical therapist's contribution concentrated at two ends and absent in the middle. Upstream, recognition, measurement and referral, where the evidence is strong and the clinician is uniquely placed, because nobody else in the pathway measures the same thing repeatedly. Downstream, once the disease is controlled, rebuilding what was lost, where the evidence is genuinely encouraging and widely under-used out of a fear of flares that the data does not support.

The subtype that breaks this framing is inclusion body myositis. It does not respond to immune suppression, it progresses regardless, and the honest goal is holding on rather than gaining. The one trial that tested training there missed its primary endpoint, and what it showed instead was that untrained patients lost 9.2% of their leg strength over 12 weeks while trained patients lost none. That is worth having. It is not the same promise, and telling a patient otherwise sets an expectation the disease will not meet.

Sources

Lundberg et al., 2017, Annals of the Rheumatic Diseases. PMID 29079590. 2017 EULAR/ACR classification criteria; 976 patients vs 624 look-alikes; Sn 87%/Sp 82% without biopsy.
Saygin et al., 2024, Clinical and Experimental Rheumatology. PMID 38436279. Scoping review; significant limitations identified; revision warranted.
Gono et al., 2025, Seminars in Arthritis and Rheumatism. PMID 41176843. 23 studies, low risk of bias throughout; commercial antibody assay sensitivity 63.8–95.7%, specificity 94.7–99.3%.
Oldroyd et al., 2021, Rheumatology. PMID 33599244. 69 studies; within-population cancer risk factors and the very-high-enzyme finding.
Oldroyd et al., 2023, Nature Reviews Rheumatology. PMID 37945774. International cancer screening guideline; 18 recommendations by modified Delphi.
Qiang et al., 2017, Journal of Cutaneous Medicine and Surgery. PMID 27534779. 5 population-based studies, 4,538 patients; RR 4.66 dermatomyositis, 1.75 polymyositis.
da Silva et al., 2022, Clinical Rheumatology. PMID 35821167. 19 studies, 298 patients; strength SMD 0.61, aerobic SMD 0.82, enzymes unchanged. Certainty downgraded by the authors.
Jensen et al., 2024, Rheumatology International. PMID 39192022. Randomised trial, 32 patients, 16 weeks high-intensity resistance training; remission maintained.
Alexanderson and Boström, 2020, Best Practice & Research Clinical Rheumatology. PMID 32819833. Graded review; insufficient evidence in recent-onset disease and inclusion body myositis.
Jørgensen et al., 2018, Scandinavian Journal of Rheumatology. PMID 29775118. Randomised trial, 22 patients; primary endpoint null; controls lost 9.2% leg strength.
Mattar et al., 2014, Arthritis Research & Therapy. PMID 25344395. 13 patients, blood flow restriction at 30% one-rep max; leg press +19.6%, enzymes unchanged.
Baschung Pfister et al., 2015, Rheumatology International. PMID 26271469. 14 studies audited; only 6 reported all four training components.
Cheng and Wong, 2023, Neurogastroenterology & Motility. PMID 37010885. 39 studies, 3,335 patients; pooled swallowing difficulty 32.3%.
Hannah et al., 2023, Journal of Rheumatology. PMID 36379584. 62 papers; anti-MDA5 odds ratio for death 6.20.
Lindgren et al., 2022, Annals of Neurology. PMID 35596584. 128 patients over 33 years; inclusion body myositis prevalence 32 per million, mean onset 64.4 years.
Shelly et al., 2021, Neurology. PMID 33879596. 40-year population study; two-thirds developed swallowing difficulty; median 10.5 years to wheelchair.
Pinal-Fernandez et al., 2017, Annals of the Rheumatic Diseases. PMID 27651398. 666 subjects; best thigh MRI feature combination only 55% positive predictive value.

Go Deeper

Most "I'm just getting weaker" is exactly that. Knowing which questions separate the ordinary from the serious is the whole job. The Verdict sends one evidence-scored protocol like this every week, free.

Join The Verdict — free