What Is Sham Acupuncture and How Is It Used in Research?

Sham acupuncture is a placebo control designed to mimic the experience of real acupuncture without delivering its theoretically active components. Researchers use it in clinical trials the same way a sugar pill stands in for a real drug: to separate the specific effects of needling from the psychological and contextual effects of simply receiving treatment. The challenge, and it is a significant one, is that no version of sham acupuncture has proven to be truly inert. Every approach carries trade-offs that shape what a trial can actually conclude, making the design of a convincing and biologically inactive sham one of the most debated problems in acupuncture research.

How Sham Acupuncture Works in Practice

In a drug trial, the placebo is straightforward: an identical-looking capsule with no active ingredient. Acupuncture does not lend itself to that kind of clean swap. It involves a practitioner, a needle, specific body locations, and physical sensations the patient expects to feel. A sham procedure has to reproduce enough of that experience to keep the patient unaware of their group assignment, while somehow removing the element believed to produce the therapeutic effect. Different research teams disagree on what that element even is, which is part of why so many sham methods exist.

The most common approaches fall into a few broad categories. Non-penetrating needle devices look and feel like real needles but retract into a handle on contact with the skin, never breaking the surface. Shallow or superficial needling inserts a real needle just a couple of millimeters into the skin at locations that are not recognized acupuncture points. And some protocols use real needles at genuine depth but deliberately choose “wrong” body locations far from the points traditionally indicated for the patient’s condition. Each of these controls for different variables and leaves different variables uncontrolled.

The Devices Researchers Have Built

The best-known sham needle is the Streitberger device, developed in the late 1990s. It uses a blunted needle that telescopes into its handle when pressed against the skin, giving the patient the visual and tactile impression of insertion without actual penetration. A validation study in a Chinese population found that about 82% of volunteers reported feeling that the Streitberger needle had penetrated their skin, compared to 75% who felt the real needle had penetrated, with no statistically significant difference between the two groups.¹ The fact that slightly more people believed the sham had penetrated than the real needle is a useful reminder that the perception of needling is more psychological than most people assume.

The Park sham needle works on a similar telescoping principle but uses a different housing. A separate design, developed by Takakura and colleagues, went further by encasing both the real and placebo needles inside opaque guide tubes so that neither the patient nor the practitioner could visually distinguish which needle was being used. The goal was to enable double-blinding, where even the person performing the acupuncture does not know whether they are delivering the real or sham treatment.² A crossover study using these double-blind needles for pain research reported that patients found the non-penetrating needles indistinguishable from the penetrating ones, and that practitioners were also successfully blinded.³

Another approach skips the special device entirely and uses a real needle inserted shallowly at a non-acupuncture point. One recent trial protocol for neck pain, for example, defined sham locations as spots about 4 centimeters away from the real acupuncture points, with needles inserted only 2–3 millimeters into the skin and no manipulation to produce characteristic acupuncture sensations.⁴ This shallow-needling-at-wrong-points method is cheap and requires no special equipment, which has made it popular. But it introduces its own complications.

Why No Sham Is Truly a Placebo

The fundamental problem with sham acupuncture is that even the least invasive versions do something to the body. A non-penetrating needle still touches the skin. A shallow needle at the wrong point still breaks the skin and stimulates local tissue. Researchers have identified a plausible mechanism for why this matters: gentle touch on the skin activates a class of slow-conducting nerve fibers called C tactile afferents, which can trigger feelings of calm and well-being and produce hormonal responses through what is sometimes called a “limbic touch” response.⁵ If your placebo is itself producing neurological and emotional effects, the bar for real acupuncture to beat it gets higher, and the comparison becomes muddled.

Brain imaging studies have shown that the muddling is real. A meta-analysis of fMRI studies found that real acupuncture and sham acupuncture both activate brain regions, but in different patterns. Real acupuncture produced greater activation in the primary somatosensory cortex, the cerebellum, and the middle cingulate gyrus compared to sham. Sham, meanwhile, showed greater activation in the supramarginal gyrus, superior temporal gyrus, and cuneus.⁶ A separate crossover study was more specific: real needling at a recognized acupuncture point activated brain areas involved in discriminating and processing pain signals, while sham needling activated areas tied to the emotional processing of pain.⁷ Both real and sham are doing something neurologically meaningful, just apparently different things.

This is not a minor technical detail. If sham acupuncture has its own therapeutic effect through touch, expectation, and the ritual of treatment, then comparing real acupuncture to sham does not tell you whether acupuncture “works.” It tells you whether real acupuncture works above and beyond an already active control. The distinction reshapes how trial results should be interpreted.

How the Choice of Sham Changes Trial Outcomes

One of the clearest findings in acupuncture research is that the type of sham used in a trial directly influences the size of the treatment effect. A secondary analysis of individual patient data from major chronic pain trials found that when the sham control involved a penetrating needle, the measured benefit of real acupuncture was smaller than when the sham was non-penetrating or did not involve a needle at all.⁸ The implication is straightforward: the more physiologically active the sham, the harder it is for real acupuncture to demonstrate superiority.

A large individual patient data meta-analysis covering thousands of patients with musculoskeletal pain, osteoarthritis, and chronic headache found that real acupuncture was statistically superior to both sham acupuncture and no-acupuncture controls. But the effect sizes told different stories depending on the comparison. Against no treatment at all, the effect sizes were moderate (ranging from about 0.42 to 0.57 across conditions). Against sham acupuncture, they shrank to roughly 0.15 to 0.37.⁹ A systematic review of pooled meta-analytic data reached a similar conclusion: acupuncture beat sham in the short term for back pain, knee pain, and headache, but the differences were relatively small.¹⁰

This pattern fuels a long-running debate. Skeptics argue the small advantage over sham means acupuncture’s benefits are mostly placebo. Proponents counter that the sham itself is an active treatment, so any measured advantage on top of it is clinically meaningful and understated. Both camps are making valid points about the same data, which is why the design of the sham matters so much: it determines which question the trial is even asking.

The Blinding Problem

For sham acupuncture to serve its purpose, patients need to believe they might be getting the real thing. A systematic review examining the credibility of common sham devices found that the Streitberger and Park devices may fail to blind participants when tested at a sensitive acupuncture point such as LI4, a commonly used point between the thumb and index finger. Participants could detect significant differences in penetration sensation at that location. The Takakura double-blind device showed more promise for blinding the practitioner as well, though results across different outcome measures were inconsistent.¹¹

A separate validation study of a non-penetrating sham device found that blinding success varied by body location. At one acupuncture point (ST36, on the lower leg), blinding worked well. At another (LI4, on the hand), over 40% of participants in the real acupuncture group correctly guessed which needle they received. People with more acupuncture experience were more likely to identify the real treatment at certain locations.¹² Interestingly, one device validation trial found that in acupuncture-naïve patients, no participant in either group could tell whether they had received the sham, and an expert panel confirmed that real needling produced the characteristic acupuncture sensation (de qi) far more reliably than sham.¹³

A large meta-analysis of blinding status across acupuncture trials quantified the pattern more broadly. In about 63% of trials, participants in the real acupuncture group tended to guess they had received real treatment while those in the sham group tended to guess they had received sham, suggesting imperfect blinding overall. Factors like the number of treatment sessions, the number of acupuncture points used, and the timing of blinding assessment all influenced how well the blinding held up.¹⁴ The implication is that blinding in acupuncture trials is leakier than in drug trials, and it varies substantially depending on methodological choices.

De Qi and the Sensation Gap

One reason blinding can fail is a phenomenon called de qi, the composite sensation of heaviness, numbness, aching, or tingling that acupuncture practitioners consider a sign of correct needle stimulation. In traditional practice, de qi is seen as essential for therapeutic effect. A literature review of clinical studies found that real acupuncture consistently elicited greater de qi than sham acupuncture across many trials.¹⁵ This creates a paradox for trial designers. If de qi is the very thing that distinguishes real from sham in the patient’s experience, designing a sham that suppresses de qi might make the control easier to distinguish from the real treatment. But designing one that produces de qi might make the sham itself therapeutically active.

Experienced acupuncture patients are more likely to notice the absence of de qi, which is one reason that patient population matters for blinding. Trials recruiting people who have never had acupuncture before generally report better blinding outcomes than those including experienced patients. This is a practical trade-off: recruiting only naïve patients improves the internal validity of the trial but limits how well the results apply to the broader population of people who actually seek acupuncture.

When Sham Acupuncture Is Used Beyond Needling

Sham controls are not limited to traditional needle-based acupuncture. Electroacupuncture, which passes a small electrical current through inserted needles, has its own sham variants. A review identified at least ten different methods used as sham electroacupuncture across 94 randomized trials involving over 6,000 participants. These methods varied across three dimensions: whether the needle was placed at a real or fake location, how deeply it was inserted, and whether the electrical stimulation was actually delivered or only mimicked.¹⁶ The diversity is itself revealing; there is no agreed-upon standard, which makes comparing results across trials difficult.

Acupressure, which uses pressure instead of needles, also presents sham challenges. In a trial of acupressure wristbands, participants were not told that one arm of the study involved a sham device. Yet most participants in a qualitative follow-up reported that they had assumed one of the treatments was a placebo anyway.¹⁷ This points to a broader truth about sham-controlled trials in physical therapies: patients are often savvier about trial design than researchers give them credit for, which can complicate blinding even when the sham device itself is convincing.

A recent trial of wrist-ankle acupuncture for preventing postoperative nausea and vomiting illustrates how sham controls function in newer acupuncture-adjacent techniques. Patients receiving real wrist-ankle acupuncture had significantly lower rates of nausea and vomiting within 24 hours compared to sham (about 42% versus 71%).¹⁸ That is one of the larger effect sizes seen in acupuncture-versus-sham comparisons, though it involved a specific clinical scenario rather than the chronic pain conditions where most sham debates play out.

The Role of Expectation and Practitioner Behavior

Sham acupuncture does not exist in a vacuum. The entire treatment ritual surrounds it: the practitioner’s demeanor, the conversation, the clinical setting, and the patient’s expectations. A randomized trial of acupuncture for knee osteoarthritis specifically tested how practitioner communication style influenced outcomes and found that the acupuncturist’s behavior had a significant effect on both pain reduction and patient satisfaction, independent of whether the patient received real or sham treatment.¹⁹ In other words, a warm, empathic practitioner delivering sham acupuncture could produce better outcomes than a brusque practitioner delivering the real thing. This makes sham controls in acupuncture trials fundamentally different from sugar pills, where the interaction with a pharmacist is minimal and standardized.

Reporting Standards and the Push for Consistency

The wide variety of sham methods, and the inconsistency in how they are described in published papers, led an international group of acupuncture researchers and clinicians to develop the STRICTA guidelines: Standards for Reporting Interventions in Clinical Trials of Acupuncture.²⁰ These were later revised and adopted as an official extension of the CONSORT statement, which governs how clinical trials across all of medicine are reported. The revised checklist includes six items and 17 sub-items covering everything from the rationale behind treatment to the practitioner’s training level.²¹

Crucially, STRICTA requires that any sham or acupuncture-like control be described in detail: whether it involves skin penetration, what locations are used, how deeply needles are inserted, and what sensation the patient was expected to experience.²² Before these guidelines, many published trials simply said “sham acupuncture” without specifying what that meant, making it nearly impossible for readers or reviewers to evaluate how the control might have influenced the results. Adoption has improved transparency, though compliance is still uneven.

Ethical Tangles in Sham-Controlled Acupuncture Trials

Consent in sham acupuncture trials has drawn pointed criticism. An ethical analysis found that participants are frequently not told they might receive a sham intervention. Instead, consent forms describe the study as comparing “different forms of acupuncture,” which technically avoids lying but deliberately withholds the relevant fact. The authors of this critique argued that this deceptive framing lacks a compelling methodological rationale and violates standard informed consent requirements that are routinely followed in placebo-controlled drug trials.²³

The counterargument, often made by trial designers, is that telling participants they might receive a sham would damage blinding and compromise the scientific value of the study. There is some irony in this reasoning, given the evidence that many patients assume a sham arm exists even when they are not told. The acupressure wristband study mentioned earlier found that most participants had guessed one treatment was a placebo despite not being informed of this, and they considered the non-disclosure acceptable.¹⁷ If patients are already assuming a placebo arm, telling them would bring the consent process into ethical compliance without necessarily harming blinding as much as researchers fear.

Animal Studies and the Placebo Question

One way to sidestep the psychological components of acupuncture entirely is to test it in animals, where expectation and belief are not factors. A preclinical systematic review and meta-analysis of acupuncture versus sham acupuncture in animal models of Alzheimer’s disease found that real acupuncture improved cognitive function beyond what sham produced, concluding the benefit was more than a placebo effect.²⁴ Animal studies cannot resolve the human clinical debate, but they offer a useful data point: if needling at specific locations outperforms needling at non-specific locations in rodents, the explanation is unlikely to be expectation or the therapeutic relationship. Whether the effect sizes seen in animals translate to humans, where psychology and ritual play outsized roles, is a separate question the animal data cannot answer.