Article Text

Download PDFPDF

Infographic. What kinds of exercise are best for chronic low back pain?
  1. Patrick J Owen1,
  2. Clint T Miller2,
  3. Niamh L Mundell2,
  4. Simone J J M Verswijveren1,
  5. Scott D Tagliaferri2,
  6. Helena Brisby3,
  7. Steven J Bowe4,
  8. Daniel L Belavy1
  1. 1 Institute for Physical Activity and Nutrition, Deakin University, Geelong, Victoria, Australia
  2. 2 School of Exercise and Nutrition Sciences, Deakin University, Geelong, Victoria, Australia
  3. 3 Department of Orthopaedics, Institute of Clinical Sciences, University of Gothenburg, Gothenburg, Sweden
  4. 4 Faculty of Health, Biostatistics Unit, Deakin University, Geelong, Victoria, Australia
  1. Correspondence to Professor Daniel L Belavy, Institute for Physical Activity and Nutrition, Deakin University, Geelong, VIC 3217, Australia; belavy{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Back pain is the greatest cause of disability and lost productivity worldwide.1 Back pain generates significant financial costs for society in developed countries, such as the USA, Japan, Europe and Australia (eg, US$80 billion per year in direct and indirect costs in the USA).2 Chronic low back pain (ie, ≥12 weeks’ duration) presents the greatest challenge: it generates the greatest proportion of economic burden due to back pain3 and affects 20% of the global population.4 Of these, approximately 90% of cases of chronic low back pain are non-specific (NSCLBP), meaning that a definitive diagnosis that is agreed on between clinicians cannot be made.5 Current treatment recommendations for NSCLBP include exercise (high-quality evidence), manual therapy (low-quality evidence) and psychological therapies (moderate-quality evidence).5 6 There are a number of different modes of exercise training; however, prior pairwise meta-analysis7 concluded that there was no evidence that any particular mode of exercise was more effective than another for NSCLBP. Network meta-analysis is a method that can overcome limitations of pairwise meta-analysis by incorporating data from studies that do not necessarily have the same kind of comparator groups in a ‘network’ of studies.

In our recent network meta-analysis, published8 in the British Journal of Sports Medicine, we sought to identify the effectiveness of specific exercise training modalities in adults with NSCLBP. These modalities were also compared with ‘true control’ (ie, no intervention) and non-exercise treatments, such as treatment where the therapist provided ‘hands-on’ treatment (eg, massage, manual therapy) or ‘hands-off’ treatment (eg, education, GP management). Collectively, 89 studies and a total of 5578 patients were included. We assessed the quality of the evidence using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) criteria adapted for network meta-analysis.

For improving pain, there was low-quality evidence that Pilates, aerobic and stabilisation/motor control exercise training were optimal. For improving physical function (reducing disability), there was low-quality evidence that stabilisation/motor control and resistance exercise training were the most effective, and for improving mental health, there was low-quality evidence that resistance and aerobic were best. Only seven studies (8%) had low risk of bias. Few studies examined trunk muscle strength, trunk muscle endurance and analgesic pharmacotherapy use, making it difficult to compare the effect of different exercise modalities on these outcomes. Future studies examining these outcomes are warranted given the importance of considering these variables during the management of NSCLBP.9 Notably, stretching and McKenzie exercises did not differ to true control for pain or function, and in our analysis, stabilisation/motor control and Pilates had a clinically significant (>20-point pain reduction on a scale 0–100) greater effect on reducing pain than these exercise modes.

It is important to consider that exercise does not happen in isolation. Through exercise with an individualised approach, a person with NSCLBP can improve confidence in movement and address functional deficits concurrently. As we contend in our publication,8 understanding of the biopsychosocial aspects of (chronic) pain is important for all clinicians managing people with NSCLBP. Overall, we expect that there is no one particular kind of exercise that is ‘the best’ for NSCLBP. Rather, we interpret our findings to mean that a range of active exercise therapies are most effective for NSCLBP. Our interpretation is that active exercise, where a patient is progressed through a programme guided by an appropriately trained clinician to improve their function, is likely the best form of exercise training for NSCLBP.



  • Twitter @S1_Verswijveren, @BelavySpine

  • Contributors PJO and DLB drafted the infographic and the associated text. All other authors provided critical input into the infographic, the associated text and were contributing authors to the network meta-analysis which forms the basis of this infographic.

  • Funding The network meta-analysis was funded by Musculoskeletal Australia (formerly MOVE muscle, bone and joint health; CONTR2017/00399; not-for-profit sector).

  • Competing interests None declared.

  • Patient consent for publication Not required.

  • Provenance and peer review Not commissioned; externally peer reviewed.