Comparative Effectiveness of Exercise, Acupuncture, and Spinal Manipulation for Low Back Pain

This systematic review, published in Spine in 2011, investigated the comparative effectiveness of three widely used treatments for chronic low back pain: structured exercise, spinal manipulative therapy (chiropractic manipulation), and acupuncture. The primary objective was to determine which of these approaches offers greater benefits for patients with low back pain persisting for at least three months. The investigators conducted a comprehensive search of medical databases through December 2010, focusing specifically on randomized controlled trials that allowed direct comparisons among these treatments. To ensure analytic quality, the authors excluded studies with combined treatments, overly subjective selection criteria, or highly specific populations that would limit applicability of the results.

Of the many studies initially identified, only two met the rigorous inclusion criteria for the comparison between exercise and spinal manipulation. The first study, conducted by Ferreira and colleagues, enrolled 240 adults with chronic low back pain, comparing spinal manipulation, supervised general exercise, and specific motor control exercises over 8 weeks. The second study, conducted by Cecchi and colleagues, enrolled 137 patients comparing a back school with individualized exercise versus spinal manipulation delivered by experienced specialists. Both studies showed that structured exercise and spinal manipulation each produced significant improvements in patients' pain and function.

In the first study, all treatment approaches resulted in clinical improvement at 8 weeks, with no significant differences among groups at 12-month follow-up. In the second study, although spinal manipulation showed greater reductions in pain and functional improvement, about 50% of patients in that group required additional treatment during follow-up. When data from the two studies were combined statistically using random-effects models appropriate for between-study heterogeneity, no significant difference was observed between exercise and spinal manipulation for pain or function. This analysis suggests that both treatments offer equivalent benefits for patients with nonspecific chronic low back pain.

Surprisingly, the review identified no adequate study directly comparing acupuncture with structured exercise or spinal manipulation. Two potential studies were excluded: one because the comparison group included multiple modalities (medications, physical therapy, and exercise), and another because acupuncture was delivered in combination with exercise, preventing assessment of the isolated effect of acupuncture. This gap in the literature is particularly notable given that acupuncture is widely used for low back pain. The authors also sought evidence on which patient subgroups might respond best to specific treatments, but the few included studies did not provide sufficient data for that analysis.

This reflects a significant limitation of the current literature, which often treats patients with 'chronic low back pain' as a homogeneous group, when in reality this population is highly heterogeneous in terms of causes, characteristics, and treatment responses. Regarding cost-effectiveness analysis, the review found no adequate studies comparing the relative costs of these treatments for chronic low back pain, representing another important gap given the substantial costs associated with nonsurgical care for low back pain. The clinical implications of this study are important for clinicians and patients. The findings suggest that both structured exercise and spinal manipulation may be equivalent options for the initial treatment of chronic low back pain, with clinical benefits expected within 8 weeks.

If no perceptible improvement occurs within this timeframe, the treatment plan should be reassessed. Study limitations include the small number of eligible studies, significant heterogeneity in interventions among included studies, and the lack of data on acupuncture and subgroup analyses. The authors rated the overall quality of evidence as 'low,' indicating that future research is likely to change confidence in the effect estimates.