The Hay Festival is not a place where you might expect to learn about the treatment of animals: it’s an annual literature festival held in Hay-on-Wye, Powys, Wales, for ten days at the end of May every year.

Caroline Ingraham has written an interesting book – “How animals heal themselves” –  which is presumably the reason she was given the opportunity to give an account of her subject at the Hay Festival last week. The BBC have created a podcast from her talk,  but I believe that the editors were wrong to give her this uncritical forum to propagate her views. Caroline has a controversial belief in the ability of animals to choose their own medicine. There’s nothing wrong with her having these beliefs, but there is a problem when her views are broadcast without any “public health warning”. There is a serious risk that animals could suffer unnecessarily if members of the public follow her advice to the letter.

In her talk, Caroline recounts entertaining anecdotes of animals (including elephants, horses, dogs and cats) that have recovered following her approach of allowing them to choose their own treatment from vegetation and other substances in their natural environment, or from herbal products offered to them by Caroline. In her words, the animals “guide her to help them make a full recovery”. Caroline stresses the importance of “letting animals lead the way” for behavioural and physical problems.

As a vet in practice, I know that 70% of the animals brought to see me will recover by themselves, with no intervention or medication. Animals have evolved with strong internal natural healing capacity (it’s called homoeostasis) Our aim as vets is to assist the healing process using scientific methods. There are many reasons why animals may fail to heal themselves, and science can often help. Serious bacterial infections are cured by giving antibiotics that kill the bacteria. Coughing caused by a failing heart is stopped by giving diuretics that remove the fluid gathering in the lungs. Cancer can be cured by surgical excision, followed sometimes by drugs to slow the regrowth of cancer cells. These are all treatments that are scientifically proven: in trials, it has been shown that if some animals are given the treatment, and some are not, a significantly greater number of animals improve.

Caroline offers no such evidence: her treatments are all anecdotal. My concern is that she may be witnessing the “regression towards the mean”  i.e. the ability of animals to heal themselves without human intervention. Since around 70% of animals may recover naturally, if you believe that any recovery that you witness was caused by your intervention, you will believe that your treatment “works” 70% of the time. While this may sound impressive, the truth is that your intervention is having zero effect. If Caroline wants to clearly demonstrate the efficacy of her methods, she needs to do what pharmaceutical companies are obliged to do: carry out trials that compare animals receiving her treatment with animals that receive no treatment (a so-called control group). If she does this, she will be able to say without question that any extra improvement in the treated group is due to her treatment. Without doing this, her claims have no scientific validation, and it’s hard for objective observers to take them seriously.

At the end of the talk, the presenter did ask Caroline if she was a scientist, and if her work was “evidence based research”. She replied “I am not a scientist but the subject of zoopharmacognacy is an academic science”.


Caroline says that in trying to develop her work, “resistance came in from a variety of different establishments that tried to make it really very difficult for me to continue this work.” There is a simple reason for this resistance: there is strong legislation in the UK to protect animals. Only vets are allowed to diagnose and treat animals. The law is there to stop (often well-meaning) unqualified people who may not be aware of their own ignorance  from accidentally harming animals because of their lack of knowledge.

There may be some truth in Caroline’s claims: animals may be able to choose certain forms of self treatment for some physical and behavioural issues. But this has not been proven, and it is wrong to state it as fact. It just does not seem right when an unqualified person suggests that pets should be allowed to choose their own “pain relieving herbal remedies” rather than the safe proven methods of pharmaceutical pain relief recommended by the vet attending the animal. And it does not seem right that the BBC should give such a person an uncritical platform to disseminate her viewpoint.

Update 25th November – Caroline has forwarded this response to the original blog:

I believe that your post undervalued the importance of case reports (referred to as anecdotes) that can provide important, detailed information about individual cases that can be very valuable, and it is much better to gather this information than do nothing at all, especially as cases typically come to us after other options have been tried. In fact some successful clinical trials have been started principally because of the promising findings of a collection of related interesting case reports; a topical example is in the human medicine world where the recent FDA approval of a new viral therapy against melanoma, which was inspired by case reports of some patients with cancer going into remission after a viral infection: (of course in this case viral therapy would have to be under the control of a qualified specialist with an appropriately modified virus!).

Other observational, non-experimental approaches have also been fruitful, such as with the work by Jane Goodall, among others.  As for Applied Zoopharmacognosy itself, after my lecture to veterinary students at Bristol University yesterday, many came up to me inspired by these case reports and want to explore setting up research projects, and possibly even small scale clinical trials, investigating self-medicative behaviours.

I take your point about regression to the mean, but I understand this likely applies more to certain, short term conditions such acute pain, acute infections, slight anxiety etc. that in many cases would be expected to resolve spontaneously. Many of the individuals that I have worked with have long term conditions, most commonly with long standing behavioural issues but also other symptoms as well. The responses can be dramatic and remarkably fast, and so regression to the mean is a less plausible explanation in these cases. It is it important to note that we do not make medical diagnoses, and I advocate self-medicative approaches as complimentary to veterinary treatment not as a replacement; veterinarians are highly trained and highly skilled individuals and their input is typically invaluable. It would be fantastic if veterinarians were to add Applied Zoopharmacognosy as part of their skill set.


Finally, I read with great concern in several veterinary blogs (though not this one) and tweets that I have been suggesting to pet owners to offer onions to dogs as a wormer. I would like to clarify that I have never offered, nor recommended giving an onion to a dog. It was merely an observation of an event that I found fascinating since it paralleled other documented occurrences of sick animals selecting typically poisonous plants including:

Michael Huffman when he witnessed chimpanzees selecting Vernonia amydalina (locally known as ‘Goat killer’), to rid themselves of nematodes (Huffman and Feifu, 1989; Jisaka, et al 1993), and in a study by Singer, et al. 2009, when caterpillars with parasitoid wasp larvae infestations changed their foraging on (normally) poisonous alkaloids such to rid themselves of parasites. The relationship between self-medicative behaviour and plant poisons is a complicated topic, which would take a lot of words to cover adequately. In brief, the most common plant poisons are due to certain chemicals called alkaloids and the reasons for why animals may poison themselves on these are not incompatible with self-medication (possible reasons include evolutionary separation of poisonous plants and companion animals, degradation of bitter warning compounds during drying such as with ragwort, reduced alternative foraging options etc.). Applied zoopharmacognosists typically work with extracts high in other compounds such as terpenoids, not alkaloids.

My observations have been presented at a scientific conference in South Korea last year following an invitation by my colleague Michael Huffman, Associate Professor, Kyoto University as well as at Bristol university. I would be more than happy to open dialogue on the subject with any interested parties.

Challenges to consider with self-medictaion and clinical trials: I would fully support clinical trials that look into the efficacy of self-medicative approaches. However, clinical trials are often prohibitively expensive and time intensive. The difficulty would also be that it is not immediately obvious how a trial that focuses on two different approaches to animal health could be set up without having to split it up into multiple, equally expensive trials that each focus on particular conditions. It is also important to remember that clinical trials are not completely fail-safe either.