Acupuncture help, harm, or placebo?

This study is a critical reanalysis of the efficacy of acupuncture as an adjunct treatment in in vitro fertilization (IVF), fundamentally questioning whether the reported benefits are real or merely placebo effects. The authors, led by David Meldrum, reexamined a recent meta-analysis published in the same journal that included 23 studies with more than 5,000 participants.

The original work by Paulus et al. (2002) was the first to demonstrate a benefit of acupuncture in IVF, showing an odds ratio (OR) of 2.08 for clinical pregnancy. However, this study did not include an adequate placebo control, leaving open the possibility of a placebo effect. When the same researchers conducted a subsequent study with placebo control using Streitberger needles (which do not penetrate the skin), the benefit virtually disappeared.

In the current reanalysis, the authors identified significant methodological problems in the original meta-analysis. First, a non-randomized cohort study was included, violating basic criteria for high-quality meta-analyses. Second, three studies used inadequate controls with medications (alfentanil and lidocaine) that can directly affect oocytes and embryos. Third, some studies combined acupuncture with other therapies, such as Chinese herbal medicines, confounding the results.

When these problematic studies were removed and only pure acupuncture groups were analyzed, the odds ratio for clinical pregnancy fell to 1.14 (95% CI: 0.94-1.37), losing statistical significance. More concerning, when only studies with adequate placebo controls (Streitberger) were analyzed, the odds ratio was 0.89, suggesting an absence of benefit.

A particularly alarming finding was that three studies with Streitberger control showed a significant reduction in live birth rate (OR 0.74; 95% CI: 0.58-0.95), suggesting a potential harmful effect of acupuncture. Two individual studies showed statistically significant results with OR less than 1, indicating that acupuncture may, in certain circumstances, be harmful.

The authors propose that acupuncture may be harmful when applied aggressively, in patients unfamiliar with the treatment, or when it causes additional stress during critical moments of the IVF cycle. This is particularly relevant considering that stress can negatively affect reproductive outcomes.

The paper also addresses the question of placebo effects on 'objective' outcomes such as pregnancy. Contrary to popular belief, studies show that placebo effects can influence objective physical outcomes, not just subjective symptoms. The therapeutic environment, team expectations, and overall treatment experience can contribute to significant placebo effects.

The clinical implications are important. The authors suggest that patients should be informed about the current scientific uncertainty. If they choose acupuncture, they should be familiar with the treatment and the practitioner before the IVF cycle, and the procedure should be relaxing rather than stressful.

The study concludes that much larger and better-controlled clinical trials are needed to definitively determine whether acupuncture has real benefit beyond placebo. To detect a 10% increase in live birth rate, 3,829 participants would be needed in each group, requiring multicenter studies or multiple meta-analyses of smaller, uniformly designed studies.