Background In prospective experimental studies of neck pain patients, it is difficult to determine whether responses to sham acupuncture differ from responses to real acupuncture due to the heterogeneous methodologies in control/sham interventions. Here we aim to compare the specific and nonspecific effects of electroacupuncture with four types of sham acupuncture. Methods In this double‐blind, sham‐controlled study, we randomly assigned 175 patients with neck pain to receive 10 sessions of electroacupuncture, shallow puncture, nonacupoint deep puncture, nonacupoint shallow puncture, or nonpenetration acupuncture. We used the Northwick Park Neck Pain Questionnaire (NPQ) as our primary outcome, and Short‐form McGill Pain Questionnaire, visual analog scale (VAS), and Pain Threshold as secondary outcomes to measure the changes from baseline to a 3‐month follow up. Results All groups, except nonacupoint shallow puncture, had significant improvement in all outcome measurements. Electroacupuncture only showed superior improvements than the shallow puncture, nonacupoint shallow puncture, and nonpenetration groups when compared using the NPQ and VAS scale (*p < 0.001). Interestingly, the nonacupoint shallow puncture produced even less placebo response than nonpenetration acupuncture. Conclusion Our study demonstrates the high variability of placebo response among different types of sham controls depending on the depth of needle insertion and the puncture location. An important implication of our results is nonacupoint deep puncture produced similar analgesic effects as electroacupuncture. Our study may shed a new light on the predominant underlying mechanisms among different types of sham acupuncture controls, which can help with interpreting the effect of acupuncture in other studies. Trial registration Chinese clinical trial registry (ChiCTR‐IOR‐15006886). Significance This study compared the observed specific and nonspecific analgesia effect in four different types of sham acupuncture stimulation with neck pain patients, assessed by four outcomes. Although all of the sham controls produced significant reduction in neck pain, electroacupuncture had superior significant improvement. Importantly, placebo responses differed significantly between the sham controls and responses were inconsistent according to different outcome assessments. This study emphasizes the importance of taking into consideration which sham control and method of outcome measurement were used in a pain research study when evaluating its results.

BackgroundA large number of randomized trials on the use of acupuncture to treat chronic pain have been conducted. However, there is considerable controversy regarding the effectiveness of acupuncture. We designed a randomized trial involving patients with chronic neck pain (CNP) to investigate whether acupuncture is more effective than a placebo in treating CNP.Methods/designA five-arm, parallel, single-blinded, randomized, sham-controlled trial was designed. Patients with CNP of more than 3 months’ duration are being recruited from Guangdong Provincial Hospital of Chinese Medicine (China). Following examination, 175 patients will be randomized into one of five groups (35 patients in each group) as follows: a traditional acupuncture group (group A), a shallow-puncture group (group B), a non-acupoint acupuncture group (group C), a non-acupoint shallow-puncture group (group D) and a sham-puncture group (group E). The interventions will last for 20 min and will be carried out twice a week for 5 weeks. The primary outcome will be evaluated by changes in the Northwick Park Neck Pain Questionnaire (NPQ). Secondary outcomes will be measured by the pain threshold, the Short Form McGill Pain Questionnaire-2 (SF-MPQ-2), the 36-Item Short-Form Health Survey (SF-36) and diary entries. Analysis of the data will be performed at baseline, at the end of the intervention and at 3 months’ follow-up. The safety of acupuncture will be evaluated at each treatment period.DiscussionThe purpose of this trial is to determine whether traditional acupuncture is more effective for chronic pain relief than sham acupuncture in adults with CNP, and to determine which type of sham acupuncture is the optimal control for clinical trials.Trial registrationChinese Clinical Trial Registry: ChiCTR-IOR-15006886. Registered on 2 July 2015.Electronic supplementary materialThe online version of this article (doi:10.1186/s13063-017-2009-1) contains supplementary material, which is available to authorized users.