Every comparison on this site cites published guidelines and peer-reviewed trials. This page collects all of them — what each found, and which condition it applies to — so you (or your clinician) can check the claims yourself.
The short version: for most back-pain presentations, the evidence favours structured, progressive exercise (which is what clinical Pilates is) over passive stretching as a standalone treatment. Stretching has a role — in the maintenance phase, after the supporting muscles have been rebuilt — but the research consistently shows it does not fix the motor-control deficits that cause back pain to recur.
The UK national guideline recommends structured group exercise programmes — explicitly including Pilates — as first-line care for low back pain and sciatica, before medication. Stretching as a standalone intervention is not the recommendation; the exercise must be structured and progressive.
Underpins the comparisons for lower back pain, sciatica, desk-worker back pain.
The international scoliosis society’s guidelines support condition-specific, asymmetric exercise over generic symmetric work. Symmetrical stretching treats both sides the same — scoliosis isn’t symmetric, and treating it that way can reinforce the rotational pattern.
Underpins the comparison for scoliosis.
Clinical Pilates produced up to 72% pain reduction in chronic low back pain — the headline RCT behind the protocols referenced across this site. The comparison arm was not structured Pilates, underlining that the structure and progression are what drive the outcome.
Underpins the comparisons for lower back pain, SI joint pain.
The Cochrane systematic review found Pilates probably more effective than minimal intervention for pain and disability in chronic low back pain. Minimal intervention in most included trials looked a lot like what people do unsupervised at home — including casual stretching.
Underpins the comparison for lower back pain.
Systematic review found neural mobilisation (specific nerve-gliding techniques) effective in reducing pain and improving function in nerve-related pain. Neural mobilisation is fundamentally different from passive stretching — sustained tension on an irritated nerve is the pattern that flares sciatica.
Underpins the comparisons for sciatica, herniated disc.
Neurodynamic mobilisation combined with motor-control exercise produced significantly greater reductions in pain and disability than motor-control exercise alone in disc-related radicular pain. Stretching alone was not the effective intervention in either arm.
Underpins the comparisons for herniated disc, sciatica.
The classic pair of studies showing the deep stabilising muscles (multifidus, transverse abdominis) do not recover spontaneously after a back-pain episode — they require specific motor-control retraining. Stretching does not address this deficit; structured Pilates is built around it. This is the single most important reason "I stretched and it came back" is so common.
Underpins the comparisons for lower back pain, SI joint pain, posture.
No single trial proves Pilates "beats" stretching for every person and every condition. What the body of evidence shows is more specific: back pain that recurs usually involves deep-stabiliser deficits that passive stretching cannot retrain (Hides et al), nerve-related pain responds to mobilisation rather than sustained tension (Ellis & Hing; Fernández-Carnero), and the interventions national guidelines actually recommend are structured and progressive (NICE NG59). That is why each condition comparison on this site reaches a different verdict depending on the mechanism — not a blanket "Pilates always wins."
If you have red-flag symptoms — numbness in the saddle area, bladder or bowel changes, unexplained weight loss, or severe night pain — see a clinician before starting any exercise programme.
Looking for just the numbers? The headline figures from these studies are collected on the statistics page.
Sophie Mercer's condition-specific Pilates protocols apply this research as structured, week-by-week programmes — decompression first, motor-control retraining second, integration last.