Acupressure beat no treatment for insomnia but barely beat a fake version

Acupressure, reflexology and auricular acupressure apply physical pressure to acupoints or reflex areas instead of inserting needles, and their non-invasive nature has made them popular with people who want something other than sleeping tablets. Earlier randomised trials had suggested acupuncture itself may help insomnia, which raised the obvious question of whether the needle-free versions do too. Insomnia is a condition where expectation and attention alone can improve reported sleep, so the design of the comparison group matters more than usual. Establishing whether these therapies beat a convincing fake is the whole test.

Two independent researchers searched five English-language and 10 Chinese databases from inception to May 2010 for randomised controlled trials of acupressure, reflexology or auricular acupressure for insomnia. Forty trials were identified and analysed. The reviewers recorded how many used sham controls, how many were double-blind and how each scored on the Jadad quality scale, and assessed risk of bias across domains. Meta-analyses were then run on the trials of moderate quality, comparing the therapies against sham treatment, against routine care and against no treatment.

Only 10 of the 40 studies used a sham control, four used a double-blind design, nine scored three or more on the Jadad scale, and all 40 had at least one domain at high risk of bias. Pooling the moderate-quality trials, acupressure as a standalone treatment fared only marginally better than sham. Trials comparing auricular acupressure against sham produced equivocal results. Against routine care or no treatment, all three therapies, whether used alone or added to routine care, performed significantly better. The reviewers concluded that the methodological limitations and equivocal results prevent a clear conclusion about benefit.

This is a familiar pattern in trials of hands-on therapies: large apparent benefits against doing nothing, and much smaller ones once a plausible fake is introduced. It tempers the enthusiasm generated by earlier acupuncture research rather than contradicting it outright, because the sham-controlled evidence here is simply too thin to settle the matter. The gap between 40 trials conducted and 10 with a sham arm is the central problem. Better designed studies with credible sham procedures and blinded assessment would be needed before anyone could say these therapies work beyond expectation.

The honest reading is that nobody yet knows whether pressure-based therapies genuinely improve insomnia. Feeling better after a session is not evidence that the pressure points did it, since attention, routine and expectation all improve reported sleep. That does not make the experience worthless, but it does mean the claims made for these treatments run ahead of the evidence. If sleep is a persistent problem, this review offers no reason to prefer these approaches over the ones with sham-controlled support.