[  Site Map  ]

Other Medical Papers


Not written by me, but included on this website by permission of the authors, because it has been denied publication elsewhere.

Eaton KK, Howard M, McLaren-Howard J, Pollard G, Anthony HM. British Society for Ecological Medicine


The causal role of yeasts in the Candida syndrome (more correctly fungal-type gut dysbiosis) is a previously unproven hypothesis (1). Patients are usually polysymptomatic, responding poorly to symptom suppression. They suffer from fatigue, catarrhal and/or allergic symptoms, but not always abdominal symptoms. In keeping with the hypothesis, they tend to show a raised blood ethanol after a fasting oral glucose challenge ?2?, and improve on antifungals and/or a diet low in fermentable foods. A double-blind trial ?3? tested nystatin against placebo in 116 patients who also chose between their normal diet and one low in yeasts, fermentable and mould-containing foods, giving four groups. All three treated groups showed reductions in symptom scores (p < 0.001 in each), the combined group showing most improvement. Nystatin gave a 23% reduction in symptom scores (p < 0.0001), but since antifungals have actions other than killing yeasts, this cannot prove yeasts are causative: nor have other studies done so ???.

Ergosterol (???, 25-? methylcholesterol), a little known phytosterol precursor of vitamin D?, is produced by yeasts: otherwise, trace amounts have been found only in maize, cottonseed, peanut and linseed oils ?4?. Measurement of environmental ergosterol is regarded as a robust method for estimating fungal biomass ???. Only 2-5% of ergosterol is said to be absorbed from the gut ?4?. On current knowledge, any ergosterol detected in the blood must originate from endogenous yeast if yeasts and these oils were excluded from the diet.


To estimate blood ergosterol, the blood was exposed to UV light (Nikon Labophot Fluorescent Microscope mercury-vapour lamp) to convert the ergosterol to vitamin D, measuring vitamin D before and after irradiation, using a standard fluorescent antibody technique and calculating the ergosterol concentration from the change. Water soluble ergocalciferol was used as the standard (Sigma-Aldrich E 8014). The recovery of standard additions of ergosterol was 99 ? 4% for concentrations from 20-200 ?g/l. The detection limit was 9 ?g/l.

Initial estimation of plasma ethanol levels, following a 3 hour fast and no alcohol consumption for 24 hours, was made on blood samples collected into fluoride-oxalate tubes 1 hour after ingesting 5g of glucose (4g dissolved in water and 2 x 500mg in gelatin capsules) ?2?. A range of alcohols and short-chain fatty acids were measured by headspace gas chromatography.

Participants and Results

Polysymptomatic patients were screened by measuring blood ethanol levels one hour after 5g oral glucose, using gas-liquid chromatography ?2?. Eighteen newly-referred patients with blood ethanol levels above 22?mol/l (median 77?mol/l, range 23-146) agreed to participate. They were asked to follow a yeast-free diet and to avoid maize, cottonseed, peanut and linseed oils to ensure that they took no known food source of ergosterol for 72 hours before venipuncture. Of the 18 patients, 16 had measurable ergosterol (median 25.5?g/1, range 10-70)

Ten healthy symptom-free volunteers, not on any medications or nutritional supplements, were studied as controls. None had measurable ergosterol in blood, even though no dietary restrictions had been imposed. The difference between patients and controls was significant by Fisher’s exact test (p = 0.001).

In this small preliminary study, the blood ergosterol test had a sensitivity of 89% and a specificity of 100%.


The detection of ergosterol, a compound produced by yeasts ?4?, in the blood of 16 out of 18 patients with symptoms regarded as ‘non-specific’ (or ‘not-medically-explained’) and positive gut fermentation tests, but not in 10 symptom-free controls, provides firm evidence of yeast involvement in fungal-type gut dysbiosis. In the light of the substantial symptomatic improvement in the trial with nystatin (which is not significantly absorbed), this must indicate that gastrointestinal yeasts play a causal role in this condition although it provides no information on which yeasts are involved or about the mechanism for symptom production.

Ergosterol is a component of fungal cell walls (4): different sterols (including cholesterol) play a similar role in plant and animal cell membranes. An experimental study has shown that ergosterol is primarily associated with the mycelial form of fungus and with hyphal length and not with sporulation (6). This, and the poor absorption from the gut, suggest that the symptoms of fungal gut dysbiosis may indicate mycelial penetration of the gut wall rather than merely fungal overgrowth within the gut. This would explain the observation that treatment usually needs to be prolonged and that some clinicians find they get a better response if they initiate treatment with an absorbed anti-fungal such as fluconazole (usually at 50mg a day for two weeks) with or without nystatin. Modern methods might now make it possible to detect whether mycelial penetration occurs.

We know of no other clinical studies of body levels of ergosterol. When confirmed, ergosterol measurement will offer a positive diagnosis for this overlooked but treatable condition. It should now be possible to use a blood test to identify patients with this condition among the large numbers of chronically-ill patients with multiple symptoms or symptoms regarded as ‘not-medically-explained’, but without conclusive proof of psychiatric aetiology. In addition to identifying the group which should now be labelled fungal gut dysbiosis, the test may contribute to the evaluation of fungal rhinitis, sinusitis and lung disease. It will be interesting to see whether fungal skin disease also gives positive results and if so how extensive the lesions need to be before this occurs.

In our experience fungal gut dysbiosis gives rise to much disabling symptomatology and repeated medical consultations until its basis is recognised ?1?.


1. Eaton K. Is there an allergic and fermentative gut condition, and does it relate to Candida? In: Food Allergy and Intolerance (eds. Brostoff J and Callacombe S.J.) Saunders, London. 2nd edn. 2002: 351-363.

2. Hunnisett A, Howard J, Davies S. Gut Fermentation (or the “autobrewery” syndrome): a clinical test with initial observations and discussion. J Nutr Med 1990; 1:33-38.

3. Santlemann H, Laerum E, Roennevig J, Fagertun HE. Effectiveness of nystatin in polysymptomatic patients. A randomised double blind trial with nystatin versus placebo in general practice. Family Practice 2001; 18: 258-265.

4. Kritchevsky D. Phytosterols. In: Dietary Fiber in Health and Disease (eds. Kritchevsky D, Bonfield C. Plenum, New York 1997: 235-243.

5. Miller JD, Young JC. The use of ergosterol to measure exposure to fungal propagules in indoor air. Am Ind Hyg Assoc J 1997; 58(1): 39-43.

6. Schnurer J. Comparison of methods for estimating the biomass of three food-borne fungi with different growth patterns. Appl Envir Microbiol 1993; 59: 552-5.


Dr Keith K Eaton was responsible for perceiving that the yeast metabolite ergosterol might be clinically relevant in this condition, for co-ordinating the project and drafting the paper in its original form: he unfortunately died in November 2002 and Dr Honor M Anthony took the lead in modifying the paper for publication. Mark Howard was responsible for statistical analysis, and read and criticised drafts of the text, including the final one. John McLaren-Howard was responsible for devising and performing the laboratory analyses. He also read and criticised drafts of the text, including the final one. Gail Pollard was responsible for researching and developing the yeast and ergosterol-free diet. She also read and criticised drafts of the text, including the final one. David Freed calculated the sensitivity and specificity.


This excellent short paper was submitted to the British Medical Journal in about 2000, but rejected on “ethical grounds” – the patients and volunteers had been invited to provide blood samples without approval from an institutional Ethics Committee. The then Editor, Richard Smith, reported the medical authors of the paper to the General Medical Council for this alleged malfeasance. The lead author Keith Eaton sadly died before his case could be heard and the case against Honor Anthony (since also sadly deceased) was dropped. The remaining (non-medical) authors give permission for this still unpublished paper to be displayed on this website and this is the only forum so far in which this crucial information is available. For the time being this website may be cited as the reference.

David LJ Freed MD

# Back to top of page

Return to Medical Information Index

Return to the Home Page