What it is
Physiotherapeutic scoliosis-specific exercise (PSSE) teaches you to correct your own curve in three dimensions, hold the corrected position, use it in daily life, and understand your condition. Schroth uses aids such as cushions and is usually practised for about an hour a day after an intensive course. SEAS (Scientific Exercise Approach to Scoliosis) uses self-correction without aids, for about 15 minutes a day, with a physiotherapy session once a month.
SOSORT says these programmes should be tailored to your curve and taught one to one by a therapist trained in the method. That is why this guide doesn't try to teach them with a 3D figure.
Sources 10, 3
The Cochrane reviewMixed evidence
A 2024 Cochrane review found 13 trials with 583 people, aged 12 to 15 on average. Compared with no treatment, exercise reduced curves by about 3.6 degrees (2 studies, 52 people; very uncertain), within the 5 degrees that X-ray measurements can vary. Added to a brace, it gave about 2.2 degrees more than a brace alone (low certainty).
Compared with general exercise, it probably reduced the twist of the trunk a little, and quality of life was slightly better. No study looked at harms. The 2012 version of the review found too little good evidence to recommend these exercises.
Sources 3, 22
Individual trialsEmerging evidence
Schroth added to usual care made curves 3.5 degrees smaller than usual care alone after 6 months (50 teenagers) and improved back-muscle endurance. Schroth supervised in a clinic made curves 2.5 degrees smaller, while the same exercises done at home without supervision did no better than no treatment (45 teenagers). In 110 teenagers with mild curves, self-correction and task-based exercise reduced curves by more than 5 degrees, and the effect lasted at least a year after treatment ended.
These were small trials, and the teenagers knew which treatment they were having.
Sources 23, 24, 25, 26
The largest trial of moderate curvesEmerging evidence
In CONTRAIS, 135 young people with curves of 25 to 40 degrees had scoliosis-specific exercise (3 individual sessions over 3 months, then 30 minutes a day at home) or a night brace, each on top of 60 minutes of daily physical activity, or the activity alone. Curves avoided progressing by more than 6 degrees in 58% with exercise and 53% with activity alone, a difference that could easily be chance. Two years later, 9 people in each group had had surgery. Differences in trunk twist, appearance and quality of life were small and unlikely to matter.
Sources 9, 27
What the expert bodies say
SOSORT's 2016 guidelines recommend scoliosis-specific exercise as the first step to stop a curve getting worse and to avoid a brace, and alongside a brace. The Scoliosis Research Society lists observation, bracing and surgery as the treatments; its patient pages say 'exercise does have a role in the treatment of scoliosis, along with more conventional treatments like bracing' and suggest discussing specific exercises with your doctor.
We found no formal statement from the Scoliosis Research Society on these exercises, so this summary comes from its patient pages and from trial authors' descriptions of North American care, which has used observation and bracing but not exercise; it may not reflect the society's latest view. The NHS says it isn't clear whether specific exercises or physiotherapy improve scoliosis, and they are not recommended by all specialists. The CONTRAIS trial and the 2024 Cochrane review both came after SOSORT's guidelines, and both support caution.
Sources 10, 11, 4, 23, 12
It can't replace a braceMixed evidence
In the one small trial comparing them, exercise alone allowed about 2.7 degrees more progression than a brace, with very uncertain evidence, and in CONTRAIS the night brace did better than exercise. Bracing has the strongest evidence for growing children with moderate curves.
Sources 3, 9, 8
The side plank and PilatesMixed evidence
Holding a side plank on one side reduced curves by 32% in an uncontrolled series of 25 people, but not at all in a randomised trial of 64 teenagers, in which only about a third kept it up. Its originator says it suits lower-back curves only and may exaggerate upper-back ones, so the side and the curve type matter. Pilates has only a few small, low-quality studies in scoliosis.
Sources 28, 29, 30, 31
Our reading
If your child's specialist team offers scoliosis-specific physiotherapy, it may help a little, especially with a brace; in the largest trial it caused only 3 minor problems. It doesn't replace a brace when one is recommended, and it can't straighten a curve. This is our editorial summary of the evidence above.
Sources 3, 9