Does Homeopathy Work?

A new challenge has been offered to followers of homeopathic practice by two people well-versed in the subject: Professor of Complimentary Medicine Edzard Ernst and science writer Simon Singh. They have announced a £10,000 prize to anyone who can prove that homeopathy works, using standard clinical trials.

Professor Ernst appears to have strong credentials in this area. As well as being a medical doctor with a PhD, he has trained in acupuncture, autogenic training, herbalism and homoeopathy. Having studied alternative medicine for many years, he is of the opinion that approximately 5% of alternative medicines can be backed up with evidence. He believes homeopathy falls within the other 95%. It's important to note that he does support alternative remedies that come with evidential support, such as St John's Wort in the treatment of mild depression. In my opinion this gives him credibility (not to mention more than 700 peer-reviewed papers he's published).

Unfortunately homeopaths are reluctant to address the concerns raised by Ernst, Singh and others. In responding the the challenge, Steve Scrutton of the Alliance of Registered Homeopaths said "We have nothing to prove, and certainly not to people with closed minds." Dr Robert Mathie of the British Homeopathic Association, said "What is needed is more investment in homeopathy research, not facile enticements by scientists who should know better." Homeopath Frederiek Maddock added "Dr Ernst does not appear to have done his research properly and is very selective in what he decides to believe."

The problem for homeopaths is that the studies used by Ernst and Singh are open and peer-reviewed. Ernst and Singh are holding up evidence that homeopathy doesn't work and asking anyone to show where they've got it wrong. So far no one can point to anything that contradicts what they are saying.

Homeopaths can't just dismiss these concerns as being the biased opinions of closed-minded people. This is an issue of public health. It's not about prize money, it's about ethics. Most people would agree that knowingly selling a defective health product is morally wrong. Most people would agree that new drugs should be tested in clinical trials before being released to the public. Why does homeopathy get to bypass our normal requirements of accountability?

Here's my message to Steve Scrutton: I'm not the closed-minded person you refer to. I'm an open-minded person and I invite you to share any proof that homeopathy works. All I ask is the same level of evidential support I would expect from any health product. In particular, I want to see some evidence that homeopathy works better than the 10-30% positive result I would expect from a placebo.


  1. Seven double-blind, placebo-controlled studies involving a total of 752 participants have evaluated the potential benefits of Galphimia in relieving symptoms of hay fever.9 Potencies of the remedy used in these trials ranged from 2c to 3c. All of the studies evaluated improvements in allergic symptoms related to the eye, and some also considered symptoms related to the nose. Overall, the results were encouraging.

    In the best of these studies, researchers evaluated the effectiveness of Galphimia glauca in alleviating hay fever symptoms in 201 people.10 The treatment group received Galphimia glauca 2c in a liquid solution of 43% alcohol, and the placebo was in an identical solution. Improvement was noted after 2 weeks, and the relative benefit of treatment over placebo continued to increase throughout the 4-week study period.

  2. Jarad, you left out one rather important thing: The citation. Without knowing who performed these studies it is impossible to know if this study is real, reliable or completely made up.

    I did a quick google myself and the first meta-study I found was this (pollinosis means hay fever):

    Wiesenauer M, Gaus W. Double-blind trial comparing the effectiveness of the homeopathic preparation Galphimia potentiation D6, Galphimia dilution 10−6 and placebo on pollinosis. Arzneimittelforschung. 1985;35:1745–7.

    Finding: "no statistical significance was achieved"