Wednesday, October 26, 2011

Do Yoga or Stretching Exercises Aid Back Pain?

Back Pain While there are many treatments for chronic low back pain, none are effective in all cases and the search is ongoing for helpful therapies. There is limited evidence that yoga instruction can be helpful for some patients, but it could be that only the stretching component of yoga is most effective. Another possibility is that a self-help book for back-pain relief could be just as beneficial for patients who follow its exercise recommendations. Which approach is best?

To help sort out the possibilities, Karen J. Sherman, PhD, MPH, and colleagues from Group Health Research Institute, Seattle, conducted a study to determine whether yoga is more effective than stretching exercises or a self-care book for patients with chronic low back pain [Sherman et al. 2011]. A total of 228 adults with moderately impairing pain were randomized to 12 weekly, 75-minute classes of either yoga (N=92) or conventional stretching exercises (N=91), or assigned a self-care book — The Back Pain Helpbook [info here] — for at-home use (N=45).

The book provided information on causes of back pain and advice on exercising, lifestyle modifications, and managing flare-ups. Stretching classes emphasized strengthening exercises in tandem with techniques for stretching major muscle groups, particularly in the trunk and legs. Yoga classes followed principles of viniyoga, involving breathing exercises, deep relaxation, and a range of simple stretching postures. Participants in both classes also were provided instructional materials (CD, DVD) for home use and asked to exercise 20 minutes on non-class days.

Overall, 64% of subjects were female, mean age 48 years, having experienced lower back pain problems for 11 years on average. The primary outcomes measured were back-related functional status (modified Roland Disability Questionnaire, RDQ; a 23-point scale) and bothersomeness of pain (on a 0-to-10 point numerical-rating scale) at 12 weeks. Outcomes were assessed in participants at baseline, 6, 12, and 26 weeks by interviewers unaware of treatment group.

The researchers reported the two primary outcomes in an early online edition of Archives of Internal Medicine [also see graphs below]

  • Back-related dysfunction declined over time in all groups; however, compared with the self-care group, yoga participants reported superior function at 12 and 26 weeks (mean difference on RDQ, -2.5 and -1.8, respectively) and stretching-group participants reported superior function at 6, 12 and 26 weeks (-1.7, -2.2, -1.5, respectively). There were no significant differences between the yoga and stretching groups.

  • There were less pronounced effects on bothersomeness of back pain. While there were small improvements from baseline in all groups, yoga produced a statistically significant reduction in bothersomeness scores compared with self-care only at 12 weeks (mean difference −1.1; 95% CI, −1.7 to −0.4, P<.001), and there were no other differences between the 3 groups at any other time point.

Mild or moderate adverse events were reported by 15% of yoga participants and 17% in the stretching group, mostly increased back pain. One yoga attendee experienced a herniated disk. Only 1 of 45 persons (2%) in the self-care group reported any problems; specifically, increased pain after doing recommended exercises in the book.

A majority (59%) of all participants were using pain medications at baseline, mostly NSAIDs. Fewer than 12% used acetaminophen, muscle relaxants, antidepressants, or opioids. During the study, twice as many participants in yoga or stretching reduced their use of medications as compared with those in the self-care group.

Globally, more participants in yoga or stretching groups were very satisfied with their overall care for back pain, and more patients said they would recommend yoga to a friend over a stretching class (85% vs 54%, respectively). Most, but not all, self-care participants (86%) reported reading some of the book, with nearly half of them reading more than a third of the book.

The researchers concluded that physical activity involving stretching, regardless of whether it is achieved using yoga or more conventional stretching exercises, has modest benefits in individuals with moderately impairing low back pain. However, in a news article, lead author Sherman cautioned, “Now we’re not talking about a person with severe back pain who is unable to move their body. But for the typical back-pain patient both approaches are certainly better than what people usually do, which is to take some meds and tough it out. And both seem to afford more clinically meaningful improvements than simply giving a patient a self-care book” [HealthDay News, October 24, 2011].

COMMENTARY: Self-management strategies, like exercise programs, for chronic low back pain are appealing because they seem relatively safe, inexpensive, accessible, and may have added beneficial effects on health. In small trials of fair quality, yoga demonstrated effectiveness for certain types of back pain, and it might be especially appealing because it includes a mental element that could enhance pain relief. The similar effects for both yoga and stretching approaches in this present study suggests that yoga’s benefits may be largely attributable to the physical benefits of stretching and strengthening the muscles and not to its mental components.

However, the study data reveal that the relative merits of yoga or stretching classes, compared with a self-help book, may not be extraordinary. The authors provide very helpful graphs for the two primary outcomes, displaying means and 95% Confidence Intervals (CIs) that clearly depict treatment differences (figure below, adapted from Sherman et al.). Several important points are illustrated, keeping in mind that the formal exercise program ended at 12 weeks…

  1. Yoga-vs-Exercise In terms of functionality (RDQ, top graph), benefits of an exercise program were evident at the 6-week assessment, yoga in particular, and results were most clear at 12-weeks with yoga and stretching classes equally better than self-care. These benefits compared with self-care persisted to a lesser extent at 26 weeks. Note the overlap vs separation of CIs for each group at each time point [see here for more info on interpreting CIs].

  2. Differences on bothersomeness scores are much less pronounced (bottom graph), with only yoga demonstrating significant benefits at 12-weeks compared with self-care, and neither yoga nor stretching having any advantage at 6 or 26 weeks.

  3. Reductions in all scores from baseline appear rather modest, especially on the bothersomeness scale. Although, in a separate analysis, the authors found that more than half of participants in yoga and stretching groups improved in their RDQ scores by at least 50%, compared with less than a quarter of those in the self-care group, and such improvements in functionality could be clinically meaningful to patients.

  4. Benefits of exercise appeared to persist several months beyond the end of classes, but it should be observed in the graphs that the self-care book also provided some reductions from baseline measures among participants in that group. An exercise program — whether yoga or stretching — must be maintained; once it is stopped the advantages appear to merge over time with those afforded by self-care (a regression toward the mean effect).

Note also that data in the graphs above were adjusted by the researchers to account for individual differences in baseline scores, sex, age, body mass index, days of pain during the prior 6 months, pain traveling down the leg, and employment-related exertion. However, unadjusted data also presented by the authors in their article followed the exact same patterns, so those factors did not appear to be confounding variables.

There were some limitations of this study that should be noted by readers…

  • Participants in the study were a select population with uncomplicated, nonspecific chronic lower back pain of only moderate severity. Average bothersomeness score at baseline was only 4.7 on the 0-to-10 scale. Also, the researchers excluded persons with back pain attributed to a specific cause (eg, fractured vertebra, etc.), and/or possibly associated with an underlying condition (eg, cancer, etc.) or prior back surgery, or complex in nature (eg, sciatica, stenosis). Persons for whom yoga would be contradicted (eg, severe disc disease) and persons with major depression also were excluded.

  • Comparative effectiveness research of this sort — assessing relative advantages of different treatments for pain conditions — is important and is now being encouraged by government agencies. However, the self-care intervention in this trial was in its own way an active therapy, rather than treatment as usual, no-treatment, or placebo condition. Subjects rigorously following recommended exercises in the book might have benefitted appreciably from the stretching component (as suggested in the graphed data above). It would have been better to also have a “wait list” group, not receiving any of the 3 treatments, to assess and compare the natural course of lower back pain in this select population during the 26 weeks of the study.

  • Patient compliance with and completion of any treatment regimen or study protocol can be a critical factor for success. For reasons not entirely clear, there were more study discontinuations among stretching-group participants. Also, nearly a third (30%) of participants in all groups visited other healthcare providers for their back pain (mostly chiropractors and massage therapists) during the first 3 months while classes were in session, and 40% did so during the post-class followup period. The type, quantity, and quality of extra care that each received might have confounded overall outcomes.

  • Furthermore, in their main paper the authors provide only an Intent-to-Treat (ITT) analysis, which is pragmatic but the results are diminished by including only partial data from subjects who dropped out of the study or did not fully comply. Sherman et al. report a complex sensitivity analysis (available online only, not in the published paper) examining the impact of missing data and report that no major differences in outcomes resulted; however, it would have been better if, along with the ITT perspective, they showed a Per Protocol analysis of only compliant subjects who completed the full program as it was designed. [This sort of dual presentation was discussed recently in an UPDATE comparing exercise with drug therapy for migraine here]. From a comparative research standpoint, a Per Protocol assessment could help reveal a more accurate portrayal of potential treatment effects.

News reports of this study touted the advantages and benefits of yoga or stretching classes for lower back pain. However, this research clearly suggests that patients need to be carefully selected and, at that, these therapies are not entirely without some risk. Furthermore, the research only tells us that patients with moderate pain and disability may gain modest benefits; perhaps, persons with much more severe nonspecific, uncomplicated, chronic low back pain would enjoy greater gains, but we do not know if this might be so from this study.

Also, as Sherman and colleagues advise, healthcare providers recommending stretching or yoga classes for their patients should carefully consider the type of exercises that will be taught and whether they will be modified for persons with physical limitations. Other considerations include patient preferences, class availability, and costs. And, data from this study also seem to suggest that even a suitable self-help book for home use might be of some benefit — if patients are sufficiently motivated to read and apply the recommendations.

REFERENCE: Sherman KJ, Cherkin DC, Wellman RD, et al. A Randomized Trial Comparing Yoga, Stretching, and a Self-care Book for Chronic Low Back Pain. Arch Int. Med. 2011(Oct 24), online ahead of print [abstract here].