Acupuncture as an adjunct to exercise based physiotherapy for osteoarthritis of the knee: randomised controlled trial

This randomized controlled trial, published in the prestigious British Medical Journal, investigated an important practical question for knee osteoarthritis treatment: whether adding acupuncture to an exercise-based physical therapy program would provide additional benefits for patients. The study was conducted between 2003 and 2005 across 37 physical therapy centers in the United Kingdom, reflecting real-world clinical practice within the British healthcare system. The researchers enrolled 352 adults aged 50 or older who had clinical diagnosis of knee osteoarthritis and had been referred by their family physicians for physical therapy. Participants were randomized into three groups: a control group that received only advice and exercise (n=116), a group that received standard treatment plus true acupuncture (n=117), and a third group that received standard treatment plus non-penetrating acupuncture, using blunt-tipped needles that created the illusion of insertion (n=119).

The treatment program consisted of up to 6 sessions of 30 minutes over 6 weeks, including individualized strengthening, stretching, and balance exercises, plus a home exercise program. Acupuncture, when applied, used between 6 and 10 points per session, combining local points (such as ST-34, ST-35, ST-36, SP-9, SP-10, GB-34) and distal points (LI-4, TE-5, SP-6, LR-3), following Traditional Chinese Medicine protocols. The practitioners involved were experienced and trained in acupuncture according to national standards. The primary outcome was change in the WOMAC (Western Ontario and McMaster Universities) pain subscale score at 6 months.

The results were surprising and challenged common expectations about acupuncture efficacy. At 6 months, mean reductions in pain score were virtually identical between groups: 2.28 points for the control group, 2.32 for true acupuncture, and 2.53 for non-penetrating acupuncture. Differences between the groups that received acupuncture and the control group were clinically insignificant: 0.08 points for true acupuncture and 0.25 points for non-penetrating acupuncture, with confidence intervals that included zero. Interestingly, some small benefits were observed on secondary measures of pain intensity and discomfort, but these were more consistent and durable in the non-penetrating acupuncture group than in the true acupuncture group.

This finding suggests that the benefits observed were not due to specific effects of needle insertion or eliciting the qi sensation, but possibly to nonspecific factors such as patient expectations, additional therapist attention, or placebo effects. The study had several important methodological strengths: appropriately blinded randomization, excellent follow-up rates (94% at 6 months), high protocol adherence, and use of a credible placebo control. Participants who received non-penetrating acupuncture could not distinguish their treatment from true acupuncture, confirming successful blinding. In addition, the control group that received only exercise and advice had a notably high response rate (43% at 6 months by OMERACT-OARSI criteria), much higher than observed in previous studies, indicating that the physical therapy program was particularly effective.

The clinical implications are significant. The study suggests that when acupuncture is added to a well-structured and effective physical therapy program, it does not offer clinically relevant additional benefits. This is important because it raises questions about the cost-benefit ratio of integrating acupuncture into conventional care for knee osteoarthritis, especially considering that the exercise program alone already demonstrated substantial efficacy. Limitations include the use of only 6 acupuncture sessions, fewer than in some previous studies, although this number reflects typical practice in the British healthcare system.

In addition, the study included patients with clinical diagnosis, not necessarily confirmed radiographically, which better reflects real clinical practice but may include some conditions beyond osteoarthritis.