Why Most Reiki Studies Cannot Be Compared With Each Other

Key Takeaways

When people ask what "the research" says about Reiki, they assume the studies can be added together. Most of them cannot, because they are not measuring comparable things.
The differences that block comparison are ordinary ones: how long a session lasted, who the participants were, what the control group received, and how the outcome was measured.
This is not an argument that Reiki research is worthless. It is an argument for reading each study as a description of one situation, rather than as a vote for or against.

Key Terms Explained

靈氣 (Reiki) / Reiki — a Japanese practice usually involving light touch or hands held just above the body.
Sham Reiki — a session that looks like Reiki but is given by someone without training or intent, used as a comparison group.
Outcome measure — the specific thing a study counts, such as a pain score or a blood pressure reading.
Meta-analysis — a study that combines results from several trials into one figure. It requires the trials to be comparable.

The Question I Cannot Answer Simply

Students ask me what the research shows. It is a fair question and I have never found a short answer that is also honest.

The short answers available are all misleading. "Studies show it reduces anxiety" ignores the trials that found nothing. "There is no good evidence" ignores trials that were reasonably designed and found something. Both summaries are the same mistake in opposite directions: they treat a pile of studies as one body of evidence.

I write about individual Reiki studies fairly often, and I have noticed that the honest thing to say about each is almost always specific to that study. The design of one says very little about what another would find.

This article is about why that is, because I think understanding it is more useful than any summary I could give.

Four Differences That Block Comparison

Session length varies enormously. Published trials use sessions ranging from around fifteen minutes to an hour or more, sometimes once and sometimes repeated over days. A single 20-minute session and five 60-minute sessions are different interventions. If two studies disagree, the disagreement may be about dose rather than about Reiki.

The participants are not the same kind of people. Reiki has been studied in cancer patients, dialysis patients, people recovering from surgery, ventilated patients in intensive care, nursing students, and community-dwelling older adults. These groups differ in baseline anxiety, in what a "reduction" would even mean, and in how much room there is to improve. A result in one says little about another.

The control group receives different things. Some trials compare Reiki against nothing at all. Some compare it against quiet rest. Some compare it against sham Reiki delivered by an untrained person. These three answer three different questions. Against nothing, you learn whether the whole package helps. Against sham, you learn whether the practice adds anything beyond the situation.

The outcome measures differ. Anxiety alone is measured by several different scales. Pain might be self-reported or assessed by observation, which matters enormously when participants cannot speak. Two trials can both report "reduced anxiety" while measuring things that do not line up.

Any one of these differences is enough to make two studies hard to compare. In practice, studies differ on all four at once.

Related: Your First Reiki Session, Step by Step: What Happens, and What I Never Promise explains this in detail.

What This Means When You Read a Headline

A headline saying research supports Reiki is usually describing one trial, and one trial describes one situation.

The useful questions are the ones the headline leaves out. How long was the session? Who were the participants? What did the other group get? What exactly was measured, and by whom?

If the answer to the third question is "nothing" or "usual care", the study cannot separate the practice from the attention. That is not a fatal flaw, but it limits the claim. If the answer is "sham Reiki", the study is testing something narrower and more specific.

This is also why meta-analyses of Reiki need reading with care. Combining trials into one figure assumes the trials are measuring comparable things. When session length, population, control condition and outcome measure all vary, that assumption is doing a lot of work.

I trained all the way to Reiki Shihan because this practice changed my life, and I teach it. None of that gives me a reason to overstate what has been demonstrated. If anything it gives me a reason not to, because a student who takes an overstated claim into a conversation with a doctor will be embarrassed, and rightly.

What I Tell Students Instead

I tell them the practice they are learning is a relaxation practice, not medical care, and that this is true regardless of what any individual study reports.

Then I tell them how to read a study if they want to: find the four things above before reading the conclusion. Session length, participants, control group, outcome measure. If a report does not tell you those, it is not telling you enough to judge.

And I tell them that a study finding nothing is not an attack on the practice. Trials find nothing for many reasons, including being too small to detect a modest effect. A null result constrains how large an effect could plausibly be. It does not erase the reason someone books a quiet hour.

What I ask them not to do is collect the favourable studies and ignore the rest. That is the habit that makes a practitioner unreliable, and it is easy to fall into without noticing.

FAQ

Q: So does the research support Reiki or not? A: It depends entirely on which study and which question. Some trials report reductions in anxiety or pain; others find no difference from the comparison group. Because the trials differ in session length, participants, control condition and outcome measure, there is no single answer that covers them all honestly.

Q: Why do meta-analyses exist if the studies are not comparable? A: Meta-analysis has statistical methods for handling variation between trials, and researchers apply them. The results still depend on how much variation there is and which trials are included. They are worth reading, with attention to what was combined.

Q: What makes a Reiki study well designed? A: A sham comparison group, enough participants to detect a modest effect, a clearly specified session protocol, and outcome measures appropriate to the participants. Blinding the practitioner as well as the participant would be better still, and is genuinely difficult in a practice delivered by a trained person.

Q: Should a practitioner cite studies to clients? A: Only with the limits attached. Citing a favourable result without mentioning the design invites a client to expect something the study does not support. If a study cannot be described accurately in a sentence or two, it may be better not to cite it at all.

Q: Does a null result mean the session was worthless? A: No, and it does not mean it was valuable either. A null result says the trial did not detect a difference on what it measured, in that group, at that dose. Someone who spent a quiet hour and felt better is not contradicted by that, and is also not evidence against it.

Key Insights to Remember

Reiki studies vary in session length, participants, control condition and outcome measure — usually all four at once — which is why they resist being added together into a single verdict.
The most informative question about any trial is what the comparison group received. Against nothing, you learn about the whole package; against sham, you learn about the practice specifically.
Collecting only the favourable studies is the habit that makes a practitioner unreliable. Reading each study for its design, including the unflattering ones, is the alternative.