This study investigated whether the perceived efficacy of healing crystals in reducing anxiety symptoms can be explained by classical conditioning mechanisms and belief-related cognitive biases, rather than genuine therapeutic effects. The aim was to disentangle placebo responses from true clinical outcomes in the context of pseudoscientific interventions.
A sample of 138 adults from the general population was classified as either believers or non-believers in the efficacy of healing crystals. Participants were randomly assigned to an experimental group (rose quartz crystal) or a control group (placebo crystal), following a standardized 14-day usage protocol. Anxiety symptoms were assessed pre- and
post-intervention using the Beck Anxiety Inventory and the Spanish version of the Kuwait University Anxiety Scale. A multilevel ANOVA and Bayesian analysis were conducted to evaluate main effects and interactions.
Significant reductions in anxiety were observed exclusively among believers, irrespective of whether they received the actual crystal or a placebo. No significant differences emerged between experimental and control groups, and the effects did not exceed those typically associated with placebo. Bayesian estimates further supported the null hypothesis for treatment effects. A strong correlation between pre-existing belief and perceived post-treatment efficacy suggested the presence of causal illusions shaped by classical conditioning.
The authors concluded that the findings indicate that healing crystals do not exert therapeutic effects beyond placebo. Observed symptom reductions were mediated by expectancy and conditioning mechanisms, particularly among participants prone to intuitive and magical thinking. Nevertheless, based on previous evidence, we do not rule out the possibility that this placebo effect could be amplified through interaction with other clinical variables
associated with the therapeutic alliance in the doctor-patient relationship.
I think this study is splendid! It may well be the first rigorous trial of crystal healing. In addition, it has a clever design that allows interesting conclusions not just about this particular form of so-called alternative medicine (SCAM) but also about placebo effects and the power of belief.
Some people claim that some SCAMs do not deserve to be tested scientifically. They claim that they are too implausible and that we therefore should not waste precious resources on them. This trial demonstrates that important insights can be gained by conducting proper science evn on such modalities.
Very well done, the authors, and my congratulations on this nice paper!
10 Responses to Healing crystals: do they work beyond placebo? … Short answer: No
Leave a Reply
This site uses Akismet to reduce spam. Learn how your comment data is processed.
Playing devil’s advocate- so these are beneficial to believers. If we are content to allow people their magical religious beliefs (indeed religion is a protected characteristic under UK law) then why not magic crystals?
There are several reasons; perhaps the most important is that any effective treatment would, when administered well [e.g. with empathy], also generate a placebo effect. In addition, it would generate the [usually much more important] specific effect of the therapy. Merely relying on a placebo will thus not achieve the optimal result; in fact, it prevents the patient from benefitting from the most important benefit medicine has to offer. In other words, it is unethical and [depending on the exact circumstances] actionable.
Couple of problems…
The study population is scarcely randomly selected, as they used contacft details from attendees at the “MAGIC International alternative therapies fair”;
One of the exclusion criteria is “having no history of a formally diagnosed mental disorder”. So no-one with a clinical anxiety state is included…
Looking at the actual BAI figures at the end seems to indicate support for a placebo effect for any crystals in Troo Beliebers.
This is nonsense.
Oh well, at least they used a respectable scale, which is rare in this sort of BS…
“The study population is scarcely randomly selected, as they used contact details from attendees at…”
I suggest that convenience sampling was a reasonable choice for this particular study.
“So no-one with a clinical anxiety state is included…”
That’s the ethically correct decision.
Fair does, but selecting your study population from a set of self-identified woo-believers when looking at woo beliefs seems to me to be stretching convenience sampling. The concept of crank magnetism has been long described, rendering this population, despite any protestations, far more likely to believe in crystals than either you or me.
Excluding folk with a clinically diagnosed condition is not an automatic disqualifier; it may be sometimes, but isn’t automatic (I’m about to participate in a study directly related to a diagnosed condition I have). In this instance, looking at supposed reduction in anxiety symptoms in a non-anxious population seems to have limitations.
But I take your points.
“Excluding folk with a clinically diagnosed condition is not an automatic disqualifier”
It damn well should be an automatic disqualification on ethical grounds when the treatment arm is known in advance to be a null treatment (a placebo); as was known to be the case for the study under discussion:
“experimental group (rose quartz crystal) or a control group (placebo crystal)”.
I applaud the authors for their avoidance of giving false hope to applicants who had “a clinical anxiety state” [your wording].
I missed a word or 2 out…
Should be “This demonstrates that this is nonsense”.
Sorry.
Prof. Ernst.
You forgot to include this research article on the importance and effectiveness of the placebo effect.
https://pmc.ncbi.nlm.nih.gov/articles/PMC3195867/
This makes your article on placebos and healing crystals not worth the paper it is written on.
thank you so much!
https://pubmed.ncbi.nlm.nih.gov/7663213/
It seems to be a rigorous clinical trial. But what made me laugh is your validation. Why do I say that? Because in your blog there are many examples of you using ad hominem when the authors are “believers in something” or have “affiliation with a department” if they published positive experiments. After doing a little research, I found that Alex Escolá is a religious believer and parapsychologist who rejects philosophers (he would say that your friend Nikil Mukerji writes nonsense) and has published things like this:
https://www.sciencedirect.com/science/article/abs/pii/S1550830723001696
https://onlinelibrary.wiley.com/doi/full/10.1002/brb3.3026
Ernst, it seems that according to your logic, Escolá would enter your bullying “hall of fame” alongside Harald Walach and Jonas Wayne (even though they have also published negative research). But seeing as Escolá agrees with you in an essay, it seems you won’t do that.