Design: Randomised controlled trial, single tertiary centre
Population: Braced adolescents with idiopathic scoliosis, Cobb 25–45°, Risser 0–2
Primary outcome: Change in major Cobb angle and quality of life at 24 months
Effect: Between-group difference −3.23° (95% CI −3.51 to −2.96), below the 5° MCID

This randomised trial at a Chinese tertiary scoliosis centre enrolled adolescents with idiopathic scoliosis (Cobb angle 25–45°, skeletally immature) who were being braced. They were assigned to supervised Schroth-based scoliosis-specific exercises with daily home practice, or to self-directed general exercise.

At 24 months, supervised exercise reduced the main curve by about 3 degrees more than self-directed exercise. That fell short of the prespecified 5-degree threshold for a clinically important difference. The proportion of patients whose curves improved, stayed stable or worsened did not differ, and quality of life was similar.

Scoliosis-specific exercise programmes are widely promoted and can be costly in time and money for families. This trial suggests that, alongside a brace, they add a statistically measurable but clinically small effect.

Prioritise brace adherence, which remains the main conservative treatment for curves of 25–45 degrees.
Present supervised scoliosis-specific exercise as optional, with modest expected benefit, when families ask.
Encourage general physical activity in braced adolescents.
Refer progressive curves to a spinal specialist as standard.

Why it matters

It gives families a realistic picture before they commit to months of supervised sessions.

The statistics, in plain English

The 3.2-degree difference is precise (95% CI 3.0 to 3.5) and statistically significant, but below the 5 degrees that clinicians consider meaningful. That is a classic case of statistically significant but not clinically important.