In this third post I want to look in detail at the responses that Zoe Harcombe provided to the Advertising Standards Authority (ASA) to support the claims that Candida causes food cravings and that the Harcombe Diet can treat Candida.
Firstly, to avoid any misunderstandings further along lets start by defining our terms clearly. When Zoe refers to “Candida” she is referring specifically to Candida albicans 1 a type of fungus commonly present in our mouth and gut flora, which typically co-exists without harmful effects.
It is true that Candida albicans can cause adverse health effects, from opportunistic local infections of mouth or vagina (thrush) to more serious systemic infections which can be life threatening to those who are immunocompromised. These conditions are distinguished by having objective criteria and laboratory tests which can be undertaken to give a diagnosis of cause. Generally effective treatments to prevent or cure infections are scarce and have been subjected to only limited clinical trials 2.
What remains controversial however, is the idea that Candida overgrowth (alternatively referred to as chronic candidiasis, candidiasis hypersensitivity, yeast syndrome) occurs in otherwise healthy individuals who may not present verifiable diagnostic symptoms, and that Candida overgrowth is the cause of a wide variety of chronic, often unrelated health problems, including food cravings.
I challenged via the ASA complaint process 3 whether:-
…the claims in ads (a), (b) and (c) relating to Candida, hypoglycaemia and food intolerance misleadingly implied that they caused food cravings and could be treated by the Harcombe diet;
The response from Zoe Harcombe 3 was that:
…ads (a) and (b) did not mention food cravings. They said that ad (c) was a web page containing further background information on Candida and contained quotes from medical literature and peer reviewed references. They said it did not include claims or mention the Harcombe Diet, but factually reported work undertaken by Zoë Harcombe as a researcher.
The idea that no claims are being made is blatantly false and seems to be another attempt to avoid the burden of proof about her claims – that’s two now if we are counting – if you don’t believe me simply watch the video embedded on the same web page of the Harcombe Diet (from 4:03 onwards):
Phase I in the Harcombe Diet is a perfect diet to kill Candida…[…]…that is how Phase I came about, so you starve the Candida so it cannot proliferate any more and you actually get it back under control where it stops you craving all those bad foods…[…]…
It is also odd that Zoe asserts that she is not making any specific claims regarding food cravings, because this hypothesis is a central claim of her book 4 and during the previous investigation by the ASA 5 she gave the following response:-
ZH said the fact that there were three medical conditions that caused food cravings was ZH’s original discovery and the unique contribution of the Harcombe diet…[…]…ZH provided some brief extracts that referred to food cravings from a number of publications, which they listed…[…]…They said that the relationship between the conditions and cravings was now well known. They said that using an internet search engine returned a large number of results for links between the conditions and food cravings.
Ignoring the appeal to Google popularity – from the above response we might have a reasonable expectation that there would be a strong body of supporting literature showing a direct casual relationship between Candida and food cravings. So lets look at the references currently on the Harcombe Diet page 6 references which I note only appeared after the ASA began investigating these claims back in September 2013 7.
Taking them in the order they appear on The Harcombe Diet website:
- Hajjeh et al (2004) 8. While this paper supports the idea that Candida can have significant detrimental health effects, had Zoe not truncated its full title it would be clear that this paper is specifically about bloodstream infections of Candida in those admitted to hospital which were laboratory tested and verified by the Centre for Disease Control and Prevention (CDC). Candida albicans blood stream infections were identified in 12.6% of cases (45% of 28%) – is Zoe really asserting that the potential effects of Candida in the general population is in any way comparable to that of hospitalised patients, the majority of whom either had a central catheter or were recovering from surgery? This seems like scaremongering.
- Achkar & Fries (2010) 9. This review paper deals with infections of the genitourinary tract and states that both men and women can be colonised by Candida albicans without harmful effects, although it can lead to diseases which have clinical signs of inflammation. It also makes it clear that incidence numbers are the greatest in hospitalized patients. While this supports Zoe’s claim that Candida can have health consequences the distinction should be made that serious effects of Candida are not likely to the general population of dieters.
- Calderone & Fonzi, (2001) 10. I couldn’t obtain a copy of this, but it appears to be a paper about the technical details of just how Candida Albicans is able to successfully invade and sustain itself to cause localised or systemic infections in humans, not about symptoms. I may be wrong [Edit 05.01.2014, I now have the full paper, see the update underneath the post]
- Trofa, et al (2008) 11 This paper is about Candida parapsilosis, not Candida albicans. Regardless, there is no mention of food cravings anyway.
- Jobst & Kraft (2006) 12. I couldn’t obtain a copy of this paper but the part quoted is simply the introductory statement of the symptoms which are claimed to be associated with “Candida-Syndrome” not a conclusion. If I read this correctly 308 patients with unspecified symptoms filled in a questionnaire and provided a stool sample, of which about a third tested positive for Candida a “finding regarded as normal”. There were associations reported with candida-vaginitis and food allergies but “hints of a Candida-syndrome could not be found”. Even if the paper had found an association with food cravings, this is a cross sectional study so it wouldn’t be adequate to demonstrate causation anyway. [Edit 05.01.2014, I now have this paper, see the update underneath the post]
- Nusbaum (1986) 13. An opinion piece which suggests that Candida albicans perpetuates the problems of those with bulimia. The author clearly states that these are “theoretical explanations” for treatment of bulima and no direct studies showing causation are provided.
- Alcock et al. (2014) 14. This is an interesting paper which discusses potential mechanisms via which gut microbiota might manipulate host behaviour to bring about unhealthy eating behaviour such as increasing intake of foods which are high in sugar and fat. These are clearly identified by the author as alternative explanations and tempered with caution. There are no specific references to Candida albicans in this paper, or good reason to extrapolate the findings.
Finally there are a variety of books referenced, two of which I have copies of:-
- The Yeast Connection: A Medical Breakthrough by William G. Crook M.D 15. This is the book which really popularised the idea that Candida was a hidden health problem for the masses, containing a much copied self diagnosis checklist. Unfortunately by the authors own admission in the foreword “The book describes relationships which have been observed between the common yeast germ Candida albicans and human illness”. Yes – the whole book is basically a collection of uncontrolled, observational anecdotal case reports which do little to provide evidence of cause and effect.
- Beat Candida Though Diet, by Gill Jacobs16. Again, this book is simply a collection of anecdotes in which Candida is implicated as the cause of a range of symptoms. No robust evidence is presented to show food cravings are caused by Candida, and the author seems happy to make even more outrageous unverifiable assertions about the effects on health: “Other conditions which are thought, or suspected to have a candida connection are as follows:…[…]…heart disease…[…]…AIDS…”. I mean really?
I would suggest that robust evidence that Candida Albicans *causes* food cravings is scant to non-existent. I could find nothing in the references listed on the Harcombe Diet website, and in the light of this the ASA’s adjudication 3 seemed reasonable, and inevitable:
We acknowledged that the claims were based on a view expressed in her books, but considered that they were presented as objective fact. We had not been provided with any evidence that Candida, hypoglycaemia or food intolerance were associated with being overweight or food cravings…
So what about claims that the Harcombe Diet can treat Candida?
Well, although the response in the ASA adjudication shows no evidence was provided on this point, it was specifically addressed in the draft ruling from the previous investigation 5. Its worth looking at this to see the merits of the evidence and arguments presented. Firstly Zoe claimed that:
…doctors writing about the three medical conditions documented their medical evidence, including when the conditions were dealt with the symptoms cleared up…
The fact that it was “doctors” like Dr William D. Crook MD writing about the conditions in books like The Yeast Connection is an irrelevant argument from authority, in the absence of randomised controlled trials, doctors relying on observation are subject to exactly the same biases and errors as the next person.
The bigger problem here is that there are no objective criteria to establish that Candida overgrowth was even a valid initial diagnosis in these accounts. The diagnostic symptoms in Crook’s checklist reproduced on the Harcombe Diet website are ubiquitous and not in any way specific to Candida. I would expect most people to test positive for some of them – headaches fatigues, particular as no time-scales are specified on the majority of them. If you claim someone has condition “X” but cannot show objectively how to diagnose and measure this, how could you subsequently verify the impact of any intervention?
Even if you grant the premise that Candida overgrowth was the problem in these cases, how do we know that the cases highlighted are ‘typical’ experiences which have not been subject to observer and confirmation bias? Should we not be wary of the post hoc ergo propter hoc fallacy, and how do we attribute cause and effect when such wide ranging, radical changes to diet and lifestyle have been simultaneously implemented.
It is also worthy of note that none of the previously cited papers 8, 9, 10, 11, 12, 13 about Candida infection cite any dietary intervention treatment as having demonstrable efficacy.
Was any other evidence offered?
They said the research behind the Harcombe Diet analysed the dietary advise given by various doctors in relation the three medical conditions. They said the took the lowest common denominator of dietary advice for each condition to establish the core safe foods….
But the books cited don’t just offer dietary advice. The Yeast Connection for example makes it clear that the recommended treatment is dietary change accompanied by medication (the use of the anti-fungal Nystatin is mentioned frequently throughout the book) as well as cutting out alcohol, smoking and avoiding a number of supposed environmental triggers. You cannot rely upon someone else’s – rather poor – evidence for efficacy but cherry pick just the parts of the methodology related to diet. Incidentally the book also suggests hope and prayer* to treat Candida – why did these recommendations not make the cut for inclusion in the Harcombe Diet?
There was of course also a claim that:
…correspondence with people who had used the diet supported (sic) documented health improvements and they provided a selection of such correspondence.
*Sigh* We’ve been here before……
So in conclusion, we have seen that:-
- Candida albicans is a type of fungus which is generally present without harmful effects in our bodies
- Candida albicans commonly causes localised infections. Serious systemic infections are most likely in immunocompromised or hospitalised individuals.
- The theory that a wide range of unbiqutous, unrelated symptoms are due to candida overgrowth is unproven and accordingly there are no diagnostic criteria for this ‘condition’
- The claims that food cravings are caused by Candida albicans are not supported by robust clinical trials
- The claims that candida overgrowth can be effectively treated by dietary interventions such as The Harcombe Diet are not supported by robust evidence.
* Page 261 of the Yeast Connection says if your physician in unaware of the yeast connection you can take charge of your own health and suggests you:
Put your “emotional house” in order. Love, touch, faith, hope and prayer are all important in helping you get well and stay well”.
Big thanks to Rosie Norman (@GlutenFreeRosie) for providing me with full copies of the two papers I was missing. I can now give a more complete commentary.
Calderone & Fonzi, (2001) 10. As I suspected this is solely a technical paper. If you wish to understand what Candida is, how it can morph between unicellular yeast cells and filamentous growth it’s excellent. However citing this as evidence that Candida can “create havoc with our health and well-being” is borderline scaremongering because this paper is very clear that the clinical important of Candida is related to those with candidemia (i.e. an infection of the blood) patients with AIDs, or those suffering recurrent vaginitis. The former are serious health conditions, the latter has objective diagnostic criteria. How are these relevant to the general population of dieters looking to lose weight?
Jobst & Kraft (2006) 12. The paper does not support a link between Candida and food cravings, and if anything it simply demonstrates that the passage quoted on the Harcombe Diet page below is an example of egregious quote mining:
“Patients with unspecific symptoms were sometimes tested for Candida. In case of findings of this yeast-like fungus in their stools they often were labelled with the diagnosis of a ‘Candida-syndrome’. This comprises headache, weakness, flatulence, ravenous appetite for sweets, itching skin and several more unspecific symptoms.” [ii]
The emphasis on the above is not in the original paper and, as previously discussed this is a description of the symptoms ‘Candida-Syndrome’ is reported to be associated with i.e.. it is a description of the hypothesis they are investigating, not a conclusion of the paper. In fact it actually concludes that their study:
…indicates that Candida in stool is associated with smoking, Candida-vaginitis and allergies…[…]…these results are a statistical, not a causal finding.
There is nothing in this paper to support any link between Candida and food cravings. In fact the papers discussion section is quite dismissive of the whole idea of a “Candida-syndrome” as the following passages demonstrate:
Proponents of the so-called yeast-connection emphasise that consuming sweets are a major risk factor for the Candida-syndrome. This was not supported by our findings: distribution of positive and negative Candida tests were the same in patients who frequently were consuming sweets and patients who did not…[…]…Our results did not show headache/migraine, abdominal complaints, eczema/itching, tiredness/low general performances level, pollinosis, other skin diseases or the influence of intake of sweets to be associated to Candida occurrence in stools, each one considered as a part of the Candida-hypersensitivity-syndrome. Seen together with recent reviews [25, 26] our exploratory study leads far away from that idea.Candida in stool without evidence of a Candida-infection such as Candida-vaginitis is a normal finding. There is no need for further diagnostics towards a Candida-related disease or a therapeutic intervention than. Avoiding sweets because of Candida in stool does not appear to be useful.
1 Why do you Overeat? When all you want is to be slim. The Harcombe Diet, 2013, Columbus Publishing, ISBN-978-1-907797-24-8, Page 92.
4 Why do you Overeat? When all you want is to be slim. The Harcombe Diet, 2013, Columbus Publishing, ISBN-978-1-907797-24-8, Page 130.
5 Draft Recommendation, Advertising Standards Authority, Case Number 13-242778, June 2014
8 Incidence of Bloodstream Infections Due to Candida Species and In Vitro Susceptibilities of Isolates Collected from 1998 to 2000 in a Population-Based Active Surveillance Programme, J Clin Microbiol. Apr 2004; 42(4): 1519–1527
15 The Yeast Connection: A Medical Breakthrough, Third Edition, 1985, ISBN 0-933478-11-9
16 Beat Candida Through Diet, 1997, ISBN 978-0-09-181545-5.