Pilates tops the rankings for chronic back pain

Chronic non-specific low back pain — pain without an identifiable structural cause — is extremely common, and exercise is a mainstay of treatment. But guidance rarely says which kind of exercise, and the options span Pilates, resistance training, stabilisation and motor-control work, aerobic exercise, stretching and McKenzie method. Standard meta-analysis can only compare each mode with a control; a network meta-analysis can rank all modes against each other by combining direct and indirect comparisons. This study applied that approach to establish which training modes are most likely to be the most effective.

The authors searched MEDLINE, CINAHL, SPORTDiscus, EMBASE and CENTRAL for randomised controlled and clinical trials of exercise training in adults with non-specific chronic low back pain. From 9,543 records, 89 studies covering 5,578 patients were eligible for qualitative synthesis. Network meta-analysis was possible for 70 studies on pain, 63 on physical function, 16 on mental health and 4 on trunk muscle strength. Treatments were ranked using the surface under the cumulative ranking curve (SUCRA), and evidence quality was graded using GRADE criteria.

Pilates had the highest probability of being best for pain (SUCRA 100%, standardised mean difference -1.86 versus true control). For physical function, resistance training (SUCRA 80%, -1.14) and stabilisation/motor-control exercise (SUCRA 80%, -1.13) ranked highest, while resistance (-1.26) and aerobic exercise (-1.18) ranked best for mental health. True control was most likely to be the worst option for all outcomes, and exercise also tended to outrank therapist hands-on and hands-off controls. Stretching and McKenzie exercise did not differ from true control for pain or function. The synthesised evidence was rated low quality overall.

This analysis supports the general principle that active exercise beats no treatment for chronic low back pain, and adds a provisional ranking of modes that earlier pairwise reviews could not provide. The suggestion that exercise training may outperform hands-on therapist treatment is a notable extension. However, the low GRADE rating, high heterogeneity between studies and shortage of low-risk-of-bias trials mean the specific rankings should be held loosely. Whether Pilates' top position for pain survives higher-quality head-to-head trials is the central open question.

For someone with long-standing, non-specific back pain, the pattern here is a reason to think that structured, active exercise — Pilates, resistance work, motor-control training or aerobic exercise — is where the evidence points, rather than passive approaches or stretching alone. The rankings suggest different modes may suit different goals, with pain, function and mental health each favouring slightly different options. The evidence quality is low, so this is a direction of travel rather than a verdict, and individual circumstances vary widely.