Study Question: Is spinal manipulation more effective than other treatments for acute or chronic low back pain?

Setting: Outpatient (any)

Study Design: Meta-analysis (randomized controlled trials)

Synopsis: All comparative studies of spinal manipulation for the treatment of acute or chronic low back pain were evaluated in this meta-analysis. After thoroughly searching the literature using the Cochrane Collaboration search strategy, Assendelft and colleagues identified 39 randomized controlled trials comparing spinal manipulation with sham manipulation, usual care (analgesics), physical therapy, exercise programs, or back school. They considered a 10-mm difference in pain scores on a 100-mm visual analog scale to be clinically important, which is a reasonable assumption.

Compared with sham manipulation, spinal manipulation was more effective for acute back pain (improvement of short-term pain: 10 mm; 95 percent confidence interval [CI], 3 to 17 mm) and long-term pain (19 mm; 95 percent CI, 3 to 35 mm). Function also improved. The authors lumped osteopathic and chiropractic manipulation together, although a recent study (not included in this review) found osteopathic manipulation to be effective (Spine 2003;28:1355–62). Spinal manipulation was no more effective than usual care, physical therapy, exercise, or back school.

In a separate systematic review by Cherkin and colleagues of the overall effectiveness of various approaches to back pain, the results were similar for spinal manipulation. The quality of the literature evaluating acupuncture is poor, and conclusions cannot be drawn. Massage also has good evidence of benefit compared with placebo therapy. Massage may even reduce the cost of care after the initial course of therapy.

Bottom Line: Spinal manipulation, whether chiropractic or osteopathic, is no more or less effective than usual care (with analgesics), physical therapy, exercises, or back school for acute and chronic low back pain. Spinal manipulation is more effective than sham (placebo) manipulation. There is not enough good research to support or refute the effectiveness of acupuncture. (Level of Evidence: 1a)