Does Acupressure Work? Evidence by Use, and Where Claims Outrun Data

A wristband with a plastic stud sells in every airport pharmacy in the United States, and it is the only piece of acupressure most people ever buy. It presses PC6, three finger-widths above the inner wrist. Whether it does anything is the single best-studied question in the whole field, with dozens of randomised trials, and the answer is: probably a little, for nausea, in some settings. Everything else acupressure is marketed for rests on far thinner ground. This article lays the evidence out use by use, then says what a practitioner can honestly tell a client.

It does not describe the points themselves or how to press them; that is the job of our chart of the 12 points practitioners learn first. Nor does it cover training paths, which are in our explainer on acupressure certification without needles.

What does the evidence say, use by use?

It says different things for different uses, which is why a single "does it work" answer is wrong in both directions. The table below sorts the common claims by the strength of what has been published. "Strength" follows the language of systematic reviews: how many trials, how large, how well blinded, and whether the pooled result survives when only the better trials are counted.

Two things stand out. First, every use with evidence is a symptom a person can rate on a scale, not a disease that can be measured. Second, none of the reviews describe a large effect. The honest one-line summary is that acupressure sometimes produces small, short-term improvements in how people feel, and that is a reasonable thing to offer and an unreasonable thing to oversell.

Why is PC6 and nausea the strongest case?

Because it is the easiest thing to test. Nausea after anaesthesia is common, predictable and measured routinely, so trials can enrol hundreds of people, apply a wristband or a finger, and count episodes. That has produced more studies of one point than of the rest of the map combined. The pooled results favour real PC6 pressure over sham for nausea, less clearly for vomiting, and the reviews caution that many of the trials were small and not well blinded.

That is why our acupressure curriculum treats PC6 as the point students learn to explain first: it is the one where a practitioner can cite research without stretching. It is also a reminder of how narrow the evidence is. One point, one symptom, a modest effect. If a school or a practitioner extrapolates from PC6 to the whole tradition, that is a leap the data does not make.

What is the sham-point problem?

The problem is that pressing the wrong spot often works almost as well as pressing the right one. To test whether a point matters, researchers compare pressure on the traditional point with pressure on a nearby non-point, or with a wristband whose stud sits on the wrong side. In many such trials, both groups improve and the gap between them is small. In some, there is no gap at all.

There are three ways to read that. The sceptical reading: the points are irrelevant and the benefit is expectation plus a few minutes of calm attention. The traditional reading: sham points are not inert because the body has many active spots. The practical reading, which is ours: deliberate touch, slow breathing and the expectation of relief produce a measurable change in how people feel, and precise location adds a little on top for a few uses. A practitioner honest about that has nothing to hide.

Where do the claims outrun the data?

Almost everywhere the word "treat" appears. The gap between what acupressure has been shown to do and what it is advertised to do is wider than in most modalities we teach, because the tradition is vast and the internet is unfiltered.

• Weight loss. Ear-point and abdominal-point protocols are sold widely. No credible trial shows acupressure changes body weight beyond what the accompanying diet does.
• Fertility. Points are marketed to people trying to conceive. There is no evidence that finger pressure affects conception, and the emotional stakes make the claim especially unfair.
• Immunity and detox. Neither is a measurable outcome, which is precisely why the claims are unfalsifiable and why we do not allow them in our students' materials.
• Chronic disease. Any claim that acupressure treats hypertension, diabetes, depression or cancer symptoms is outside the evidence and, for a non-licensed practitioner, outside the law.

The further a claim moves from a symptom the client can rate on a scale, the less evidence stands behind it. That is a useful test for a client reading a practitioner's website, and for a practitioner writing one.

What can a practitioner honestly say to a client?

Three things, in this order: what the tradition holds, what the research found, and what the client should keep doing with their licensed provider. In practice that sounds like: "This point is traditionally used for nausea. Trials show a small benefit for some people, mostly after surgery. It is safe to try and it will not replace what your doctor has prescribed." Forty words, no promise, no claim, and the client can decide.

How do we teach clients to self-apply?

By sending them home with two to four points, written down, and checking at the next session whether they used them. The trials with the clearest results, PC6 for nausea and HT7 for sleep in older adults, mostly measured what people did to themselves several times a day, not what happened once a week on a table. If the small effects in the literature exist anywhere, they exist in daily self-care.

1. 1

   Pick two points the client can reach seated

   PC6 and HT7 on the wrist, LI4 on the hand or LV3 on the foot. Never more than four; people do not remember five.

2. 2

   Mark them on the client's own body

   Guide their thumb to the spot, have them find it three times, and describe the landmark in their words on an index card.

3. 3

   Set pressure and duration

   A dull ache, not pain, for one to two minutes per side, once or twice a day. Write the numbers down.

4. 4

   Name the limits

   Which points to skip if pregnant, when to stop, and that this sits alongside any medical care, not instead of it.

5. 5

   Ask at the next session

   Did they do it, what did they notice, on a scale of ten. That answer, not the practitioner's belief, decides whether to continue.

Does weak evidence mean acupressure is worthless?

No, it means the claims must be small. A modality with low-certainty evidence for short-term relief of nausea, sleep trouble and tension is still one many clients value, provided nobody pretends it is more. The fair comparison is not with a drug but with what a client actually has at 10 p.m. when they cannot sleep: nothing, a screen, or two minutes of pressure on the wrist while breathing slowly. The third option is safe, free, and has a few small trials behind it.

It is also why in-person training matters more here than the evidence debate suggests. The parts of an acupressure session that plausibly carry the effect, deliberate touch, calm attention, breathing and precise location, are exactly the parts that cannot be learned from a diagram. Our in-person acupressure program runs 10 classroom days (70 teaching hours) for $2,800, and every practice exercise in class is done with a real person reporting a real number.

What does this mean if you are deciding whether to train?

It means you should train for the right reasons. If you want a modality with strong clinical evidence, acupressure is not it, and neither is most of what a holistic school teaches; read our parallel review of what the Reiki research shows and you will see the same pattern. If you want a hands-on, needle-free practice that clients find calming, that has one honest evidence-backed use and several plausible ones, and that you can describe to a sceptical relative without blushing, it is a good fit.

The practitioners who last in this field read the trials, quote them accurately and stay inside their scope. The ones who leave usually promised something the evidence never supported. We would rather you knew that before you enrol than after.