Exercise and Back Pain – Hell’s own elephant

It’s getting cramped in here and I can’t work out why. I’m inside a room labelled ‘how to treat low back pain’ and something enormous is taking up all the space.

Exercise is at the heart of the physical therapies in the treatment of back pain. Most forms of therapeutic exercise boast arguably plausible theories and don’t require a belief in magic. Exercise is something that patients can be taught to do, is less likely to breed passive reliance on therapists and so should empower patients. Exercise Rocks! (right?).

Best Practice & Research in Clinical Rheumatology has just published a special edition on back pain research and one paper caught my eye. A very well respected group of researchers from the Netherlands have produced an overview of systematic reviews of exercise therapies for low back pain.[1] This research group have as much expertise at doing this as any in the world. Their findings are not so different from previous reviews but are no less striking for that. Here’s what they found:

Exercise therapy is not effective for acute low back pain. This is no surprise and is strongly reflected in most clinical guidelines. For chronic low back pain exercise is not more effective than no treatment or just leaving people on a waiting list. That is a humdinger of a finding and I want you to remember it when thinking about the next ones.

Exercise is more effective than usual care, not more or less effective than back school/education classes, behavioural therapy, electrotherapy, spinal manipulation or psychotherapy and no type of exercise therapy is clearly more effective than another.

Better than usual care but not in comparison with no treatment? That’s peculiar.  How much better than usual care? On a 0-100 point scale of pain the average difference probably lies between 2 and 16 points with similar results for disability. This estimate is very imprecise but still pretty low and the average hovers around the 10/100 mark. Is it likely to be clinically significant? Recent evidence[2] suggests that the average back pain sufferer would want to see around a 40% improvement to consider physical therapy worthwhile. Whilst change in an individual’s pain is a different thing to a group average, it doesn’t look good.

At least it is having some effect you might argue. Perhaps, but consider that you can’t properly blind trials of exercise – therapists and patients know what treatment they are getting. We have good evidence[3] that in clinical trials that measure subjective outcomes like pain incomplete blinding can artificially increase effect sizes by around 25%.  Many of the studies included in the review suffered from limitations in their methods that introduce a risk of bias. So it seems likely that these tiny effect sizes might actually be exaggerated. Which begs the question – what is left? You might conclude that exercise is hardly (if at all) better than doing little or nothing. That no form of exercise appears superior to another (and it really is a broad church of theories) leaves us with little indication that any specific approach has real merit for this condition.

It has been argued that clinical trials don’t capture the true power of physical therapy.  Trials often apply a one-size fits all treatment approach (although this is not always the case; clinicians can sometimes modify their treatments for different patients) and many believe that there are distinct subgroups of patients within chronic back pain that require different treatment approaches.

This reasonable argument suggests that only some of the people who got the “active” treatment in the trial were suited to that treatment. The real effect is washed away in the averaging by those who shouldn’t have been given that treatment in the first place. But if true it rests on an interesting premise. Where there is little or no effect on average, for every patient that the treatment made significantly better, there must have been one for whom it was not just ineffective, but it actually made them worse. I wonder whether that reflects most therapists’ clinical experience?  There is one more sizeable problem.  Another paper in the same edition[4] points out that we are a long way away from knowing what those subgroups might be, if they exist at all.

It seems hard to believe but the best estimates suggest that specific back exercises may not help back pain. Whenever you read evidence that calls a dearly held belief into question the reflex reaction is to find something wrong with it. This tendency has been called ‘rescue bias‘.[5] This is not so unreasonable as every trial has its flaws but when a bunch of trials that ask a similar question in a different way all find something similar – that’s evidence that shouldn’t be ignored.

In the abstract of the review the authors conclude that exercise is effective based solely on the comparisons with usual care. I find this odd as it doesn’t seem to reflect what we have just seen. That’s  a shame as most readers will probably never make it past the abstract to notice the enormous devil hiding in the detail. And that is why it is so cramped in here.  There is a huge horned elephant with a forked tail sitting next to me that nobody cares to look at. It’s time we noticed it, after all even inconvenient elephants deserve a bit of attention.

WATCH THIS SPACE – COMING UP: Peter O’Sullivan adds his two West Australian dollars (WAUD$1=AUD$1.3) to this. AND THEN….a bio-statistician rounds it off. STAY TUNED…..

Original Article Abstract

Exercise therapy for chronic nonspecific low-back pain

van Middelkoop M, Rubinstein SM, Verhagen AP, Ostelo RW, Koes BW, van Tulder MW.

Exercise therapy is the most widely used type of conservative treatment for low back pain. Systematic reviews have shown that exercise therapy is effective for chronic but not for acute low back pain. During the past 5 years, many additional trials have been published on chronic low back pain. This articles aims to give an overview on the effectiveness of exercise therapy in patients with low back pain. For this overview, existing Cochrane reviews for the individual interventions were screened for studies fulfilling the inclusion criteria, and the search strategy outlined by the Cochrane Back Review Group (CBRG) was followed. Studies were included if they fulfilled the following criteria: (1) randomised controlled trials,(2) adult (> or =18 years) population with chronic (> or =12 weeks) nonspecific low back pain and (3) evaluation of at least one of the main clinically relevant outcome measures (pain, functional status, perceived recovery or return to work). Two reviewers independently selected studies and extracted data on study characteristics, risk of bias and outcomes at short-term, intermediate and long-term follow-up. The GRADE approach (GRADE, Grading of Recommendations Assessment, Development and Evaluation) was used to determine the quality of evidence. In total, 37 randomised controlled trials met the inclusion criteria and were included in this overview. Compared to usual care, exercise therapy improved post-treatment pain intensity and disability, and long-term function. The authors conclude that evidence from randomised controlled trials demonstrated that exercise therapy is effective at reducing pain and function in the treatment of chronic low back pain. There is no evidence that one particular type of exercise therapy is clearly more effective than others. However, effects are small and it remains unclear which subgroups of patients benefit most from a specific type of treatment.

ResearchBlogging.org[1] van Middelkoop M, Rubinstein SM, Verhagen AP, Ostelo RW, Koes BW, & van Tulder MW (2010). Exercise therapy for chronic nonspecific low-back pain. Best practice & research. Clinical rheumatology, 24 (2), 193-204 PMID: 20227641

[2] Ferreira ML, Ferreira PH, Herbert RD, & Latimer J (2009). People with low back pain typically need to feel ‘much better’ to consider intervention worthwhile: an observational study. The Australian journal of physiotherapy, 55 (2), 123-7 PMID: 19463083

[3] Wood L, Egger M, Gluud LL, Schulz KF, Jüni P, Altman DG, Gluud C, Martin RM, Wood AJ, & Sterne JA (2008). Empirical evidence of bias in treatment effect estimates in controlled trials with different interventions and outcomes: meta-epidemiological study. BMJ (Clinical research ed.), 336 (7644), 601-5 PMID: 18316340

[4] Kamper SJ, Maher CG, Hancock MJ, Koes BW, Croft PR, & Hay E (2010). Treatment-based subgroups of low back pain: a guide to appraisal of research studies and a summary of current evidence. Best practice & research. Clinical rheumatology, 24 (2), 181-91 PMID: 20227640

[5] Kaptchuk TJ (2003). Effect of interpretive bias on research evidence. BMJ (Clinical research ed.), 326 (7404), 1453-5 PMID: 12829562

All blog posts should be attributed to their author, not to BodyInMind. That is, BodyInMind wants authors to say what they really think, not what they think BodyInMind thinks they should think. Think about that!