Exercise for Low Back Pain: What Works Best?

For a given, well-defined musculoskeletal disorder, what works best? Is a question like this even fair to ask? More to the point, is such a question answerable? As practicing chiropractors, we get asked this “what works best” question for any number of health-related topics. High on this list might be this one: “Doctor, besides coming here for treatment is there anything I can do to relieve my pain?” If you recognize this as one of the most pivotal opportunities you will ever have with this patient, you are wise indeed. There is much at stake, hinging upon how you answer. This may be your one chance to garner a “trusted-for-life” award from this patient. And these are, far and away, our favorite patients!

Dr. Carmichael

Correctly, most of us answer this type of question from our experience. But here’s the hidden opportunity: Does your experience line up with high-quality evidence? And the crucial linchpin: Can you unpack this evidence in an understandable way to inspire your patient to do exactly what they are asking you to help them do?

Unpack This: One of the Most Important Papers of 2021!

In March of this year a virtual conference (CARLoquium) took place, allowing the chiropractic community to “meet, chat, and share research.” ACA was a proud sponsor of this 2nd annual international CARL conference. One of the five keynote presenters was Professor Jill Hayden, PhD, DC, the lead author of a paper entitled: “Some types of exercise are more effective than others in people with chronic low back pain: a network meta-analysis.” (Hayden et al., J Physiotherapy, 2021) On behalf of her co-authors, Dr. Hayden concisely reported the results of a brilliant, methodologically rigorous study that actually does answer the question: What works best?

Hayden et al.’s study was a systematic review and network meta-analysis of randomized controlled trials (RCTs). Adults with chronic ( 12 weeks) non-specific low back pain (cLBP) were exposed to 11 categories of exercise treatment prescribed or planned by a health professional. Pain intensity was measured with a visual analogue scale (VAS) or numeric rating scale (NRS) standardized to a 100-point scale. Back-related functional limitations were also measured by the Oswestry Disability Index or the Roland Morris Back Disability Questionnaire, also standardized to a 100-point scale for analysis. The review included 217 RCTs with 24,486 participants (59% female, mean age 44 years, baseline pain score of 51/100, and baseline functional score of 38/100). Outcomes collected at 30 and 90 days were used for analysis, and clinically meaningful differences were interpreted as a 15-point reduction in pain score and a 10-point reduction in functional score. Forty percent of participants reported leg pain accompanying their cLBP at baseline. There were 426 exercise treatment groups reported in these trials with a mean of 8 weeks duration and median of 12 hours cumulative exercise treatment exposure. The most common type of exercise investigated was core strengthening (30%), followed by mixed exercises (26%), and general strengthening exercises (12%). Group exercises were performed in 40% of participants, while another 40% received one-to-one exercise sessions. Reporting on pairwise comparisons between the 11 categories of exercise this study answers the question: What works best?

The Graphic Results, and “The Reveal”

Network meta-analysis results were presented graphically as two forest plots in Figure 4 (page 259) and two league tables in Figure 5 on page 260 of the paper (click on link to see figures). Please open this free article and check them out! The forest plots are standard output for a meta-analysis. These plots provide an easy visual summary of results. In addition, the league tables are brilliant, using depth-of-color shading to show strength of effect sizes for each of the 11 categories of exercise intervention compared with each other. The tables clearly illustrate which exercises for cLBP show the strongest clinically important differences.

Dr. Hayden and colleagues showed that for patients with chronic non-specific low back pain, Pilates exercise provided the largest reductions in self-reported pain and largest improvement in self-reported function. McKenzie exercise finished a close second, posting statistically significant and clinically meaningful changes in both self-reported pain and functional domains nearly as good as Pilates therapy (with no statistically significant differences between them). Remember that McKenzie extension exercises are notably intended to centralize radicular pain and that 40% of participants in this study had lower extremity pain at baseline, so patient characteristics may not be entirely comparable between these two groups.

This study is well worth a slow, careful read. Implementing the findings in your practice will likely improve outcomes for your patients, and practice satisfaction for you. This is one of the most rigorous studies (and study designs) you will encounter. While experience tells you that no singular exercise type is right for all patients, you are adding ballast to your confidence and credibility as an evidence-based practitioner and back pain expert.

Joel Carmichael, DC, PhD practices at The Center for Spine, Sport & Physical Medicine in Colorado. He teaches at Universidad Central del Caribe in Puerto Rico and maintains volunteer clinical faculty status at the University of Colorado School of Medicine. His research seeks to improve the practice habits of sport and spine professionals. Dr. Carmichael loves teaching, latté art, C.S. Lewis, and an occasional morsel of chocolate. He and his wife enjoy morning strolls with their golden retrievers, Stella and Noli.