Is the location of the needling point in sham acupuncture for tension-type headache associated with outcomes? A systematic review and network meta-analysis.

Abstract

Objective: In trials of acupuncture for tension-type headache (TTH), sham acupuncture may not be physiologically inert, and its design may influence the estimated specific effect of acupuncture. Sham acupuncture was classified into the following 2 types: (1) sham acupuncture needling at the same acupuncture points as those in the acupuncture group, referred to as sham acupuncture therapy (verum) (SATV) and (2) sham acupuncture needling at points different from those in the acupuncture group, referred to as sham acupuncture therapy (sham) (SATS). Therefore, we aimed to evaluate whether the needling point location in sham acupuncture is associated with treatment outcomes for TTH.

Methods: Eight electronic databases were searched from inception to September 4, 2025. Randomized controlled trials (RCTs) comparing manual acupuncture with sham acupuncture or waiting-list control in adults with TTH diagnosed within the contemporaneous IHS/ICHD framework were eligible. Sham acupuncture was classified by needling location as SATV or SATS. The primary outcome was pain intensity; secondary outcomes were headache days and responder rate, defined as at least a 50% reduction in headache days. Random-effects frequentist network meta-analysis was performed after evaluating clinical similarity, transitivity, and statistical consistency. Certainty of evidence was assessed using the GRADE framework for network meta-analysis.

Results: Six RCTs involving 1,055 participants were included. Compared with waiting-list control, acupuncture, SATV, and SATS were each associated with improvements in pain intensity, headache days, and responder rate. No statistically significant difference was detected between SATV and SATS for any outcome. Acupuncture did not differ significantly from either sham approach for pain intensity or headache days; however, acupuncture was associated with a higher responder rate than both SATV and SATS. The certainty of network evidence ranged from moderate to low for continuous outcomes and from high to low for responder rate, mainly because of risk of bias and imprecision.

Conclusion: Within the available sparse network, the anatomical location of sham acupuncture alone did not explain differences in TTH outcomes. Sham acupuncture procedures used in previous TTH trials may carry physiological and contextual activity, regardless of whether they are delivered at the same acupoints or at different points. Future trials should move beyond simply debating "where to needle" and should quantify and report the full sham-acupuncture dose, including insertion depth, stimulation intensity, retention time, frequency, patient information, and blinding assessment.

Systematic review registration: https://www.crd.york.ac.uk/PROSPERO/view/CRD420251135928, identifier CRD420251135928.