A study1 that claims to show that a homeopathic treatment can ease pain in rats has caused uproar after it was published in a peer-reviewed journal. Groups that promote homeopathy in Italy, where there is currently a debate about how to label homeopathic remedies, have held the study up as evidence that the practice works. But several researchers have cast doubt on its claims.
The authors acknowledge some errors flagged in an analysis of the paper by a separate researcher, but stand by its overall conclusions. Senior author, pharmacologist Chandragouda Patil of the R. C. Patel Institute of Pharmaceutical Education and Research in Dhule, India, also says that the results are preliminary and cannot yet be applied to people, and that he hopes that the team’s findings will encourage other researchers to conduct clinical studies.
Researchers have presented evidence in support of homeopathy before — famously, in a 1988 Nature paper2 by French immunologist Jacques Benveniste that was later discredited. This latest claim has attracted attention, in part, because it passed peer review at the journal Scientific Reports. (Nature’s news team is editorially independent of its publisher Springer Nature, which also publishes Scientific Reports).
A paper that claims something as exceptional as the corroboration of homeopathy but also contains errors “raises questions on whether the review process was adequate”, says Michelangelo Cordenonsi, a cancer researcher at the University of Padova in Italy.
“Either the paper is true, so it’s of extraordinary importance, or it’s false and should be closely scrutinized,” says Enrico Bucci, the researcher who carried out the analysis of the paper. Bucci is co-founder of the company Resis in Turin, Italy, which provides tools to uncover potential problems with scholarly articles, and a researcher in systems biology at Temple University in Philadelphia, Pennsylvania. “It’s worrying that a major journal like Scientific Reports didn’t pay close attention to a study that claims to show that homeopathy works," he adds.
A spokesperson for Scientific Reports, which published the paper on 10 September, says that the editors are looking into the criticisms, and will correct or retract the paper if necessary. On 1 October, the journal added an editors’ note to the homeopathy paper alerting readers to criticisms regarding the study.
Homeopathy is based on the claim that illnesses can be treated using substances that produce similar symptoms. Mostly, these have been heavily diluted in water or alcohol so that none or only a few molecules of the active ingredient are present. Some supporters of the practice say that the water or alcohol ‘remembers’ the substance, which triggers a healing response. But these claims aren’t backed up by scientific evidence, and the European Academies’ Science Advisory Council notes that homeopathic products are no more effective than placebos in treating health problems.
In the rat study, Patil and colleagues report that a homeopathic product — a heavily diluted extract from Toxicodendron pubescens, a plant commonly known as Atlantic poison oak — is as effective as the prescription drug gabapentin in reducing inflammation and pain responses in both cells grown in the lab and in animals.
Homeopathy groups worldwide have welcomed the study. And in Italy, where a recent proposal to label homeopathic products as ‘preparations’ rather than ‘drugs’ has provoked heated debate, homeopaths and their associations have said that the study’s publication demonstrates the effectiveness of homeopathy. On 24 September, the presidents of 11 Italian non-conventional medicine associations said in a press release that this is the “umpteenth proof” of the efficacy of the practice, which is opposed by “unscrupulous people” that don’t accept the evidence.
On social media and in the press, scientists in Italy have voiced concerns about the study, including that only eight animals were used, that the experimental design was not blinded (a process designed to stop researchers’ expectations from affecting the interpretation of the results) and that pain was measured indirectly by observing the withdrawal of the animals’ paws in response to heat or cold meaning that the results cannot yet be generalized to people.
In his analysis, Bucci used his company’s software to detect two identical images that supposedly describe different experiments in one of the paper’s figures. He also found inconsistencies in the way that doses of the homeopathic product given to the rats were reported. In the body of the text, he points out, the authors write that they had treated the animals with heavily diluted Toxicodendron pubescens (up to 1 × 10−30), but the data in one of the figures show the effects for dilutions up to 1 × 10–8. These discrepancies, as well as the image duplications, were also flagged by others on PubPeer, a platform to discuss scholarly articles. In another figure, Bucci spotted what seems to be the same data for two different experiments. He published his analysis online on 26 September and sent a detailed report to the editors of Scientific Reports on 3 October. (Bucci has also published a summary of his findings in English.)
When Nature's news team sent the concerns to the study’s corresponding authors, Patil said that his team had made some unintentional mistakes while preparing the manuscript, resulting in the duplicated images and the repeated data. The discrepancies between the text and the figures are the result of typos, according to Patil. The group will ask Scientific Reports to update the article with a correction. But “this does not change the scientific conclusions in any way”, Patil says. All the experiments were done “with utmost integrity”. The aim of the study was neither to criticize, nor support homeopathy, but to evaluate a homeopathic product using “pharmacological principles”, he adds.
In response to the other concerns, Patil says that small sample sizes are common in pharmacological studies and that, “although 100% blinding cannot be ensured”, the researchers who did the experiments were not aware of which rats received which treatment.
Bucci says that he has also found that some of the authors of the study, including Patil, had written another paper3 published in Scientific Reports in 2016 that he says also contains inappropriately duplicated images. Patil says that these occurred while converting the figures to high resolution when the researchers submitted the manuscript to the journal. The group will ask Scientific Reports to correct that article too, he says.
The spokesperson for Scientific Reports says that the editors are looking into the issues raised for both papers. “We take our responsibility to maintain the accuracy of the scientific record very seriously."
Nature 562, 173-174 (2018)