Yoga Versus Education for Veterans with Chronic Low Back Pain

Excerpt

Our study¹ found that 12-weekly hatha yoga classes were not more effective than an education intervention for improving pain and disability outcomes among 120 Veterans with chronic low back pain (cLBP). We appreciate Dr. Ozden’s comments regarding important methodologic considerations for studies of complex nonpharmacologic or behavioral interventions for cLBP. Blinding of participants to treatment allocation in a study of yoga vs. education is unfortunately not possible. However, we followed recommended guidelines for blinding those who collected, handled, or analyzed data to treatment group allocation.² We opted not to report p-values in Table 1 based on guidelines that discourage the use of p-values to assess baseline differences in parallel group randomized trials.^{1, 3, 4} We identified one baseline difference between groups of a clinically important prognostic factor (opioid use) which we adjusted for in our analyses. Also, Dr. Ozden highlights our multiple time points of data capture and recommends assessments for change at different time points. We had evaluated changes at additional endpoints with similar results, i.e., no between-group differences in primary outcomes at 6 and 24 weeks (shown in Appendix Tables 2 and 3 of previously published manuscript).¹ Furthermore, we used effect estimates and their 95% confidence intervals rather than p-values to interpret findings.³ Lastly, they raise the important point of whether patients with specific causes of back pain (e.g., stenosis, radiculopathy) were included. We identified Veterans with cLBP pragmatically using ICD diagnostic codes documented in health records, self-report of at least 12-week duration of pain, and current pain intensity at least 4 of 10 on the Defense and Veterans Pain Rating Scale. Participants answered screening questions on “red flags” to rule out serious specific causes of cLBP but were not required to have physical examinations or advanced imaging prior to enrollment. While this would have excluded most individuals with back pain due to fracture, cancer, infection, or severe spinal canal stenosis, heterogeneity of back pain subtypes is likely although this is a common challenge in cLBP intervention studies. Older age, male sex, and having more comorbid physical and mental health conditions have been found to be predictors of non-response to yoga.⁵ The negative findings of our trial in Veterans may be related to this.¹ Understanding the mechanisms underlying non-response and how to optimize yoga interventions for these individuals are fruitful areas of future study. …