Where this summary comes from
Wang L, Zhou C, Zhai C. Comparative effects of exercise modalities and dose parameters on chronic low back pain in adults: a systematic review and network meta-analysis. Frontiers in Physiology, 2026. PMID 42222290.
The paper itself — doi.org/10.3389/fphys.2026.1836061
Study design: Systematic review and network meta-analysis · 24 randomised controlled trials · 28 intervention arms
What the paper asked
For chronic non-specific low back pain, which kind of exercise — and how much, how often?
What was found
- All five kinds of exercise — stabilisation, resistance, traditional, combined and motor-control — significantly reduced pain compared with control.
- Which is to say that "which exercise?" has no decisive answer; what every arm had in common was that exercise happened.
- The best-ranked dose: 15–20 minutes a session, 6–7 days a week, for at least 16 weeks.
- A short, almost-daily session ranked above a long twice-weekly one — and the length of the programme mattered more than its type.
What does not follow from it
- ⚠️ SUCRA rankings carry weak conclusions. In a network meta-analysis the gap between first and third place can be clinically nothing. "Best" here means "highest in a ranking", not "proved better".
- 24 trials to compare five kinds of exercise at several doses is not much data; some comparisons rest on a handful of studies.
- Every participant had non-specific low back pain. The finding does not transfer to back pain with an identified structural cause.
What it means for someone here
This paper states the hardest part of the work plainly: the choice of exercise type matters less than it appears, while staying with the programme for four months matters a great deal. Programmes written here are short and daily for exactly that reason — a programme that is not done scores zero whatever type it is.
Related findings
This page summarises one study. It is not treatment advice, and no programme is written for a person from a single paper.