Exercise Prescription for Chronic Pain

The VerdictMODERATE CONVICTION

Doing some exercise beats doing none.

Pick one activity you have been avoiding because of chronic pain. Do a smaller version of it today. 5-10 minutes. Notice how you feel during the next 24 hours. If pain stays at 2/10 or under during and within 24 hours after, that is your starting dose. Build from there.

What this actually is: Exercise is first-line treatment for chronic pain. Not an add-on after passive treatments fail. The evidence for this is the strongest single piece of the chronic-pain evidence base.
The myth that won't die: "I have to find the perfect exercise type." Top-ranked modalities (Pilates, strength, aerobic, mind-body, motor control, mixed) all produce similar outcomes when delivered at the right dose. The only modality the evidence excludes is stretching alone.
The first thing to start doing: Hit the dose floor. For chronic low back pain, that is roughly 130-230 minutes of moderate-intensity activity per week, kept up for 8-12 weeks. Frequency (2-3 sessions/week) matters more than the length of any single session.

Chronic pain trains your nervous system to predict that movement will hurt, so you do less of it, your body de-conditions, the prediction comes true, and the loop tightens. Exercise is how you give your nervous system new evidence on a weekly schedule. The dose is what teaches the loop to loosen. Total weekly volume matters more than the brand on the studio door.

What Works

Exercise Prescription

Tier 1Supervised exercise at the dose floor for the patient's condition

This is the single highest-conviction recommendation. Hit the dose. Pick the modality the patient will adhere to within the supported set.

Chronic LBP: 520-920 MET-minutes per week (~130-230 min moderate or 65-115 min vigorous), across 8-12+ weeks. Pilates, strength, aerobic, mind-body, motor control, mixed all acceptable.
Chronic neck: 30-60 min per session, 2-3×/week, 7-12 weeks. Cervico-scapulothoracic strengthening + endurance preferred.
Fibromyalgia: Aerobic-dominant 50-60 min per session, 2-3×/week, ≥13 weeks. Strengthening as adjunct.
Chronic widespread: 2-3 sessions per week minimum. Frequency more predictive than per-session duration.

Evidence: Geneen 2017 Cochrane overview (~37,000 participants); Hayden 2021 NMA N=20,969; Fernandez-Rodriguez 2022 NMA N=9,710; Ferro Moura 2024 Bayesian NMA N=5,033 dose-response; Marris 2021 nociplastic stratification; Polaski 2019 frequency regression.

Tier 1Pain Neuroscience Education + exercise for high-fear or central-sensitisation patients

Dose: Single PNE session >60 minutes, then 4-8 sessions across 7-12 weeks integrated with the exercise programme. Group-based delivery preferred. Expect TSK-11 and PCS reductions by 8-12 weeks; pain magnitude lags slightly behind fear-avoidance markers.

Evidence: Siddall 2022 (kinesiophobia SMD -1.20 MODERATE certainty); Rufa 2024 (PNE+ex pain MD -1.14, disability SMD -0.80); Bonilla-Barba 2024 dose-response.

Tier 2 — Moderate Evidence (3 protocols)

Tier 2Cognitive Functional Therapy for disabling chronic LBP

Integrated framework: PNE + graded exposure + lifestyle + behaviour change delivered by a CFT-trained physical therapist over 12 weeks. Devonshire 2023 meta-analysis is very-low certainty due to imprecision; Kent 2023 RESTORE Lancet trial supports clinically meaningful effect sustained to 12 months. Evidence base is maturing.

Tier 2Pain-monitored progression rule across all loading

≤2/10 pain during session, ≤2/10 24-h post-session flare. Brief low-magnitude flares are normal and not a stop signal. Distal-symptom-migration adds an additional stop signal in nerve-pain presentations. Workable middle ground after the 2025 JOSPT update found painful vs non-painful exercise statistically equivalent.

Tier 2Adherence-support scaffolding

Digital reminders, group format, motivational interviewing built into delivery from week 1. Trial-grade 80-95% completion drops to 40-60% in community delivery without scaffolding. Pooled meta-analytic effect sizes assume the higher number.

Tier 2Aquatic exercise for older adults, severe OA, post-surgical, high-irritability presentations

Allows weight-supported loading when land-based exceeds tolerance. Particularly useful in the early weeks for patients with significant deconditioning.

Tier 3 — Emerging / Adjunct (2 protocols)

Tier 3VR-assisted active training

Cui 2024 reports large-magnitude effect outliers (SMD -2.94 to -8.15) that carry implausible magnitudes consistent with high heterogeneity. Direction-of-effect only. Use as engagement adjunct for tech-comfortable patients, not as standalone replacement for evidence-supported modalities.

Tier 3Home-based exercise with check-in support

Best as maintenance phase after supervised initiation. Not the right starting modality for high-fear or nociplastic-dominant patients. Lin 2021 home-based LBP and Hewitt 2019 unguided digital both show modest effects that improve when human-check-in scaffolding is added.

What DOESN'T Work

Stretching alone. Fernandez-Rodriguez 2022 NMA explicitly excludes from effective interventions for chronic-pain reduction. Use as warm-up or accessory only, never as standalone treatment.
Generic "core stability" prescription as superior modality. Saragiotto 2016 Cochrane: motor control exercise no better than other exercise forms. The "transversus abdominis activation" mechanism story is not supported.
Therapeutic ultrasound for chronic LBP. Cochrane 2020 found insufficient evidence to support; not recommended in current CPGs.
Calendar-based progression without pain-rule and function-marker check. Inherits the wrong signal. Patients regress when advanced past their tolerance window.
Single-modality dogma. NMA rank-position differences are small; patient preference and adherence carry more weight.
PNE delivered as a 5-10 minute "explanation" with no dose. The dose-response work is clear: single session >60 min, 4-8 sessions, integrated across 7-12 wk.
Loading patients with TSK-11 ≥41 without PNE scaffolding. Drop-out climbs and the condition reinforces.

Red Flags — Refer

Chronic pain prescription is safe in most people. These features mean a workup comes first, before any exercise loading.

Cauda equina symptoms. Saddle anaesthesia, bowel or bladder change, bilateral leg weakness — A&E immediately.
Progressive neurological deficit. Worsening weakness, sensory loss, reflex change — neurology / spinal surgery.
Suspected fracture. Trauma history, unexplained acute severe pain, history of cancer or osteoporosis with new severe pain — imaging + referral.
Suspected infection. Fever with new severe pain, immunocompromised patient, IV drug use, recent procedure — A&E or urgent GP.
Suspected malignancy. Unexplained weight loss, night pain unrelated to position, history of cancer, first-onset back pain age >50 with red-flag pattern — GP for workup.
Inflammatory arthropathy active flare. Morning stiffness >60 min, systemic features, raised CRP/ESR — rheumatology.
Symptomatic uncontrolled cardiovascular disease. Unstable angina, decompensated heart failure, severe valvular disease, uncontrolled arrhythmia — cardiology clearance before loading.
Suspected vascular emergency. Sudden severe back pain with abdominal pulsation (AAA), unilateral leg swelling with calf pain (DVT) — A&E.
Major mental-health crisis limiting capacity to engage. Severe depression with suicidality, acute PTSD, active substance-use disorder limiting engagement — mental-health-first co-management.

Refer to: GP for workup co-ordination; A&E for cauda equina, vascular emergencies, suspected infection; rheumatology for inflammatory arthropathy; oncology for suspected malignancy; cardiology for cardiac clearance; mental health for psychiatric crisis.

Pick one activity you have been avoiding because of chronic pain. Do a smaller version of it today. 5-10 minutes. Notice how you feel during the next 24 hours.

If pain stays at 2/10 or under during and within 24 hours after, that is your starting dose. Build from there. Brief low-magnitude flares are normal and not a stop signal.

The Verdict

Doing some exercise beats doing none. Hitting the dose floor and matching the approach to the patient beats chasing the "best modality."

What this actually is: Exercise is first-line treatment for chronic pain. Not an add-on after passive treatments fail. The evidence for this is the strongest single piece of the chronic-pain evidence base — Cochrane overview of 21 reviews and a network meta-analysis of 217 trials with nearly 21,000 participants both converge.
The myth that won't die: "I have to find the perfect exercise type." Top-ranked modalities (Pilates, strength, aerobic, mind-body, motor control, mixed) all produce similar outcomes when delivered at the right dose. The only modality the evidence excludes is stretching alone.
Start here: Hit the dose floor. For chronic low back pain, that is roughly 130-230 minutes of moderate-intensity activity per week, kept up for 8-12 weeks. Frequency (2-3 sessions/week) matters more than the length of any single session.

Best For

Adults with chronic pain ≥12 weeks, no red flags, who want a clear evidence-based plan rather than another round of passive treatment.

Skip If

You have any red-flag features above, you have an acute structural injury that requires offloading, or you have a major mental-health crisis that limits your capacity to engage with a programme.

Want the full evidence? Keep scrolling.

Return to Training Criteria

Use these as binary, criteria-based gates. Calendar-based progression without these checks is the most common failure mode.

Met weekly MET-minute floor (520+ MET-min/wk for chronic LBP) for 4 consecutive weeks.
Pain reduced by ≥2 points NPRS (chronic LBP MCID) OR functional improvement on patient-specific goal.
TSK-11 reduction (if intake TSK-11 ≥41) by 4+ points.
No symptom flare lasting >24 h in the last 2 consecutive weeks.
Adherence ≥70% to prescribed dose for 4 consecutive weeks.
Patient-defined functional goal achieved (specific to patient).
Specific load tolerance restored (lift, carry, sit, stand, walk task as relevant to patient).

Conviction

Overall: MODERATE-HIGH. MODERATE-HIGH

Exercise vs no-exercise / minimal for chronic pain. HIGH
Dose-response in chronic LBP (520-920 MET-min/wk). MODERATE-HIGH
PNE + exercise > exercise alone for fear-avoidance reduction. MODERATE-HIGH
PNE + exercise > exercise alone for pain magnitude. MODERATE
Pilates as top-ranked modality. MODERATE
Cognitive Functional Therapy for disabling chronic LBP. MODERATE
Painful vs non-painful exercise superiority. LOW
Specific exercise > placebo (mechanism-as-real claim). LOW-MODERATE
Stretching alone for chronic-pain reduction. DEBUNKED-LOW

What would change my mind on the dose-floor claim

A multi-centre pragmatic RCT of N≥800 chronic-pain patients (LBP, neck, fibromyalgia, OA stratified) randomised to four arms: (1) supervised exercise at 520-920 MET-min/wk × 12 wk + intensive adherence support, (2) same supervised exercise without intensive adherence support, (3) sham exercise (low-intensity, low-frequency, structurally similar) + intensive adherence support, (4) wait-list. Primary endpoints pain (NPRS), disability (RMDQ or NDI), adherence (% MET-min completed) at 12 wk and 12 mo, GRADE-assessed.

If arm 1 vs arm 3 effect is large (≥1.0 NPRS), the specific-exercise effect upgrades to HIGH. If small (<0.5 NPRS), the modality-ranking emphasis in current NMAs gets formally downgraded.

What would change my mind on the modality-ranking claim

A Bayesian network meta-analysis of N≥150 trials with formal GRADE-of-NMA-rankings methodology directly comparing all top-10 modalities head-to-head (currently sparse direct comparisons inflate rank uncertainty) would either confirm or refute the Pilates-leads-the-network claim.

Go Deeper

Pain that has stuck around for months is rarely fixed by another round of passive treatment. The Verdict reviews the protocols that actually move the dial, with conviction scores and the studies behind them — free, every week.

Subscribe to The Verdict

Sources

Geneen LJ, et al. (2017). Physical activity and exercise for chronic pain in adults: an overview of Cochrane Reviews. Cochrane Database Syst Rev. PMID 28436583. Cochrane overview, 21 reviews, ~37,000 participants.
Hayden JA, et al. (2021). Some types of exercise are more effective than others in people with chronic low back pain: a network meta-analysis. J Physiother. PMID 34538747. NMA, 217 RCTs, N=20,969.
Fernandez-Rodriguez R, et al. (2022). Best Exercise Options for Reducing Pain and Disability in Adults With Chronic Low Back Pain: Pilates, Strength, Core-Based, and Mind-Body. J Orthop Sports Phys Ther. PMID 35722759. NMA, 118 RCTs, N=9,710.
Ferro Moura Franco K, et al. (2024). The Best Exercise Modality and Dose for Reducing Pain in Adults With Low Back Pain: A Bayesian NMA. J Orthop Sports Phys Ther. PMID 38457134. 82 trials, N=5,033. Dose MCID 520 MET-min/wk; max 920 MET-min/wk.
Smith BE, et al. (2017). Should exercises be painful in the management of chronic musculoskeletal pain? Br J Sports Med. PMID 28596288.
Polaski AM, et al. (2019). Exercise-induced hypoalgesia: A meta-analysis of exercise dosing. PLoS One. PMID 30625201.
Marris D, et al. (2021). Prescription of exercises for the treatment of chronic pain along the continuum of nociplastic pain. Eur J Pain. PMID 32976664. SR/MA, 50 RCTs, N=3,562.
Siddall B, et al. (2022). Short-term impact of combining pain neuroscience education with exercise for chronic musculoskeletal pain. Pain. PMID 33863860. Kinesiophobia SMD -1.20 MODERATE certainty.
Rufa A, et al. (2024). PNE for chronic LBP: short-term outcomes. Physiother Theory Pract. PMID 37395152. SR/MA, 17 RCTs, N=1,078.
Bonilla-Barba L, et al. (2024). Optimal dose of PNE added to exercise for chronic spinal pain. PMID 38047772.
Devonshire JJ, et al. (2023). CFT for chronic LBP. J Orthop Sports Phys Ther. PMID 36812100.
Kent P, et al. (2023). RESTORE Trial. Lancet. [cite-unverified] Landmark pragmatic CFT RCT.
Shei C-S, et al. (2022). Attempting to Separate Placebo Effects from Exercise in Chronic Pain. Sports Med. PMID 34453277.
Belavy DL, et al. (2021). Pain sensitivity is reduced by exercise training. Neurosci Biobehav Rev. PMID 33253748.
Saragiotto BT, et al. (2016). Motor control exercise for chronic non-specific LBP. Cochrane Database Syst Rev. PMID 26742533.
NICE Guideline NG193 (2021). Chronic pain (primary and secondary) in over 16s. [cite-unverified]