By Jeffrey H. Baker, MD, FAAFP, DABMA, DABIHM

Assistant Professor, Department of Family and Community Medicine, Penn State College of Medicine, State College

Dr. Baker reports no financial relationships relevant to this field of study.

SUMMARY POINTS

  • Yoga as a medical therapy appears to offer improved functional outcomes compared to no exercise at six months.
  • Yoga appears to offer improved pain relief compared to no exercise at the first month of engagement.
  • Rigorous outcomes research protocols and trials are lacking in providing adequate data for conclusive analysis of yoga as a therapy for back pain dysfunction and pain.

SYNOPSIS: A review of randomized, controlled trials of treatments for chronic non-specific, low back pain revealed that yoga provides improvements in back-related function compared to non-exercise controls at intermediate time points, and in pain scores in the short term. Also, yoga seems to be comparable to exercise interventions, although the quality of evidence was low enough to preclude us from knowing for sure.

SOURCE: Wieland LS, et al. Yoga treatment for chronic non-specific low back pain. Cochrane Database Syst Rev 2017;1:CD010671.

Low back pain is a worldwide problem that has a reported prevalence of up to 84%, with a chronic prevalence of an estimated 23% of the population.1 The U.S. National Health Interview Survey estimated the three-month prevalence of low back pain as 26.4%.2 Treatment often is aimed at pain relief and dysfunction resolution, because there is often no known anatomical or pathologic cause for the problem.

The American College of Physicians (ACP) recently issued its clinical guideline for noninvasive treatments for acute, subacute, and chronic low back pain,3 listing yoga as a strongly recommended initial course of therapy for chronic low back pain. It noted that only low-quality evidence existed for Iyengar yoga to improve function and pain, based on the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) approach.

Yoga has become a popular form of lifestyle and exercise in the United States. A 2016 survey found that the top five reasons for students starting the activity were for flexibility, stress relief, general fitness, overall health, and physical fitness. It was estimated that 36.7 million people participate in yoga in the United States, a 79% increase from the previous study in 2012. Seventy-two percent of participants were women and 34% of the survey participants were likely to try yoga in the coming year.4

Wieland et al looked to assess the effectiveness of yoga in treating chronic, non-specific low back pain, defined as lasting three months or more and having an unknown metabolic or anatomic cause.5 The authors analyzed the effect of yoga practice compared to no treatment, waiting list or minimal (education) intervention, another active treatment, or yoga plus intervention, where it could be compared to the intervention alone. The review focused on the effect of yoga therapy on low back pain and function, in addition to adverse events that occurred. The review was limited to trials with participants 18 years or older and chronic non-specific low back pain was defined as three months or more duration. Of interest, the trials included yoga plus a component of meditation or controlled breathing. They excluded trials that examined yoga meditation or lifestyle without the physical component or trials with the physical component alone.

The authors searched MEDLINE, CENTRAL and Embase as of March 2016. They extracted data from 12 parallel, randomized, controlled trials published in English and carried out in the United States, the United Kingdom, and India. The trials included 1,080 participants, the treatment periods varied from one to 24 weeks, and the study periods varied from one week to 12 months.

The variation in treatment and study times created a dilemma for adequate assessment. The authors created and adjusted data groups to reflect these differences among the research trials. Outcome measures were grouped into four term measures: short (four weeks), short-intermediate (three months), intermediate (six months), and long (one year). The individual trials did not use the same pain scales to assess outcomes, so the authors transformed each study pain scale to a 0 to 100 scale and summarized the outcome data using the mean difference (MD). Multiple scales also were reported in the trials for back-related function; efforts were made to extract data from the Roland-Morris Disability Scale and the Oswestry Disability Index.

The authors used the GRADE approach in providing their strength of evidence assessments. In the analysis of the 12 trials, the authors believed the studies all had high risk of bias because the participants, therapists, and assessors were not blinded to treatment groups, and the outcomes were self-assessed. Therefore, all outcomes were downgraded to “moderate.” Some studies were very small and others had discrepancies between the methods and results section, downgrading them for risk of reporting bias. Four studies did not address class attendance, putting the trials at risk for compliance bias, while other studies did not address attrition rates clearly and had missing data. If other serious forms of bias were present, then the certainty of evidence was downgraded further.

In nine trials comparing yoga treatment to non-exercise controls in back-related function, there was low-certainty evidence that yoga was superior at 4-6 weeks (95% confidence interval [CI], -0.71 to - 0.19), 3-4 months (95% CI, -0.66 to -0.14), and 12 months (95% CI, -0.46 to -0.14) and moderate-certainty evidence that yoga was superior at six months (95% CI, -0.66 to -0.22). The reasons given for downgrading certainty of evidence included risk of bias, heterogeneity, and imprecision in measurement. Degree of improvement analysis was made difficult by the varied trial protocols, subjective response, study size, and duration, such that representative improvement was noted in terms of the Oswestry Disability Index or the Roland-Morris Disability Questionnaire of Pain Disability Index and reported in Table 1.

Table 1: Summary of Findings, Yoga Compared With No Exercise

Outcomes

Studies/
Participants

Comparative Effect*

Certainty
of Evidence

95% Confidence Interval

Function

Short term (1 month)

5/256

1.80 U lower

Low

-0.71 to -0.19

Short-intermediate (3 months)

7/667

2.18 U lower

Low

-0.66 to -0.14

Intermediate (6 months)

6/630

2.15 U lower

Moderate

-0.66 to -0.22

Long (12 months)

2/365

1.36 U lower

Low

-0.46 to -0.14

Pain

Short term

2/40

10.38 U lower

Very low

-20.85 to -0.81

Short-intermediate

5/458

4.55 U lower

Moderate

-7.04 to -2.06

Intermediate

4/414

7.81 U lower

Low

-13.37 to -2.25

Long

2/355

5.40 U lower

Very low

-14.5 to 3.7

Adverse Events

6/696

 

Moderate

 

*Back-specific function: Oswestry Disability Index/Roland-Morris Disability Questionnaire – lower score means better function.
Back-specific pain: Aberdeen Pain Scale or VAS (range 0-100) – lower scores mean less pain.

Adapted from: Wieland LS, et al. Yoga treatment for chronic non-specific low back pain. Cochrane Database Syst Rev 2017;1:CD010671.

Six trials were analyzed for pain improvement compared to non-exercise controls. Yoga did not meet the predefined criterion for “clinical importance” of pain change in any of the studies, although the change was noted as statistically significant. There was very-low-certainty evidence of pain change at 4-6 weeks (95% CI, -20.85 to -0.81), moderate-certainty evidence at 3-4 months (95% CI, -7.04 to -2.06), and low-certainty evidence at six months (95% CI, -13.37 to -2.25). There was very-low-certainty evidence for no statistically significant difference in pain at 12 months (95% CI, -14.5 to 3.7). Reasons for downgrading the certainty of evidence were similar to the functional analysis. (See Table 1.)

Four trials in the analysis compared yoga to exercise, finding very-low-certainty evidence of effect, and concluding that yoga provided little or no difference in back function at six weeks (95% CI, -0.41 to 0.37), three months (95% CI, -0.65 to 0.20), and six months (95% CI, -0.59 to 0.19). One study assessed the pain effect of yoga compared with exercise. Because there was very serious risk of bias and imprecision, the authors found very-low-certainty evidence for an effect of yoga on pain at one month (95% CI, -19.90 to -10.10) and seven months (95% CI, -25.48 to -15.31), although the reported difference in pain was statistically and clinically significant. (See Table 2.)

Table 2: Summary of Findings, Yoga Compared With Exercise

Outcomes

Studies/
Participants

Comparative Effect*

Certainty
of Evidence

95% Confidence Interval

Function

Short term (6 weeks)

2/248

0.11 U lower

Very low

-0.41 to -0.37

Short-intermediate (3 months)

2/249

0.99 U lower

Very low

-0.65 to -0.20

Intermediate (6 months)

2/249

0.90 U lower

Very low

-0.59 to -0.19

Pain

Short term (4 weeks)

1/54

15.00 U lower

Very low

-19.90 to -10.10

Intermediate (7 months)

1/54

20.450 U lower

Very low

-25.48 to -15.32

Adverse Events

3/314

 

Low

 

*Back-specific function: Oswestry Disability Index/Roland-Morris Disability Questionnaire – lower score means better function.
Back-specific pain: Aberdeen Pain Scale or VAS (range 0-100) – lower scores mean less pain.

Adapted from: Wieland LS, et al. Yoga treatment for chronic non-specific low back pain. Cochrane Database Syst Rev 2017;1:CD010671.

Adverse events were not reported consistently through the trials; there were increased adverse events in the yoga groups compared to the non-exercise controls, but they were similar to those that occurred with back-focused exercise. Adverse effects of yoga included fear of worsening the condition being treated, further back pain, disc herniation, and headache exacerbation.

COMMENTARY

The word “yoga” comes from the Sanskrit language and can be interpreted to mean “to join” or “to unite.” Yoga is an ancient lifestyle form, involving physical and mental practices with the goal of developing and harmonizing the individual’s mind, body, and spirit. The practice embodies diet, posture, breathing, cleansing, and meditative disciplines to achieve this. The traditional practices have evolved and branched over the centuries, becoming commercialized into many styles of practice today. Most of the therapeutic research in the Cochrane Review used the Hatha, Iyengar, and Viniyoga styles, emphasizing restorative practice rather than active aerobic activity.

Wieland et al found that yoga is more effective than non-exercise controls for back-related functioning at an intermediate time frame and for pain in the short term. They acknowledged that additional high-quality research is needed to improve knowledge of the effectiveness of yoga practice on non-specific low back problems, including specific populations, quality-of-life outcomes, depression, and long-term follow-up. There were too few good-quality studies available to complete the authors’ initial goals, which included study of subpopulations.

Overall, they suggested that further studies should be of low-risk bias and should include further reporting of adverse outcomes. A major concern with the certainty of evidence created in these trials was with performance and detection bias, as none of the studies were blinded and many of the results were from self-reported data gathering. It would be difficult to blind the participants from their activity, but the data-gathering aspect could be managed by blinded researchers, improving certainty of evidence of future work.

Saper et al published a study showing non-inferiority of yoga treatment to physical therapy for low back problems.6 They devised a randomized trial with study staff, who were blinded to treatment arms, analyzing paper surveys that were collected at designated study intervals.

Improving study size, terms of study, and reporting consistency are certainly doable for now, but fully blinded studies will be difficult to devise and carry out given the nature of yoga treatment. It is difficult to blind participants to this treatment, since it is so familiar to many. Therefore, it will be difficult to provide conclusive outcomes data for assessment.

Yoga practice seems safe by current studies, and appears to provide measurable improvements in function and pain relief for a common dysfunction. The ACP has strongly recommended nonpharmacologic treatment with yoga for chronic low back pain when compared to usual care and educational intervention.3 They believed that there were fewer harms associated with this type of treatment, while the benefits were evident. Until further evidence tells us otherwise, yoga appears as a viable option for improving the pain and dysfunction associated with chronic nonspecific back pain and can be added to the possible roads to improved health for our patients.

REFERENCES

  1. Balagué F, et al. Non-specific low back pain. Lancet 2012;379:482-491.
  2. Deyo R, et al. Back pain prevalence and visit rates: Estimates from U.S. national surveys, 2002. Spine 2006;31:2724-2727.
  3. Qaseem A, et al; for the Clinical Guidelines Committee of the American College of Physicians. Noninvasive treatments for acute, subacute, and chronic low back pain: A clinical practice guideline from the American College of Physicians. Ann Intern Med 2017;166:514-530.
  4. Yoga Alliance. 2016 Yoga in America Study Conducted by Yoga Journal and Yoga Alliance. Available at: https://www.yogaalliance.org/2016YogaInAmericaStudy. Accessed June 26, 2017.
  5. Van der Heijden GJ, et al. De effectiviteit van tractie bij lage rugklachten. De resultaten van een pilotstudy. Ned T Fysiotherapie 1991;101:37-43.
  6. Saper RB, et al. Yoga, physical therapy, or education for chronic low back pain: A randomized noninferiority trial. Ann Intern Med doi: 10.7326/M16-2579. [Epub ahead of print].