Irritable Bowel Syndrome
Hunter and his colleagues have studied patients with the irritable bowel syndrome in whom diarrhea, cramps and specific food intol- erances are major symptoms(24). They have found abnormal fecal flora to be a consistent finding, with a decrease in the ratio of anaer- obes to aerobes, apparently due to a deficiency of anaerobic flora (25,26). Previous exposure to antibiotics, metronidazole in particular, was associated with the development of this disorder(27).

Inflammatory Bowel Disease
Two decades ago, exaggerated immunologic responses to components of the normal fecal flora were proposed as possible mechanisms in the etiology of inflammatory bowel disease(28). Little progress has been made in confirming or disproving this theory, although bacterial overgrowth of the jejunum has been found in 30% of patients hospi- talized for Crohn's disease, in which it contributes to diarrhea and malabsorption(29).

The demonstration of increased intestinal permeability in patients with active Crohn's disease and in healthy first degree relatives sug- gests the existence of a pre-existing abnormality that allows an exag- gerated immune response to normal gut contents to occur(30).

It is interesting to note that elemental diets can induce remission in Crohn's disease as effectively as prednisone. The chief bacteriologic effect of elemental diets is to lower the concentration of Lactobacilli in stool drastically without altering levels of other bacteria(31). It is well-known that many patients with Crohn's disease can be brought into remission with metronidazole, tetracycline and other antibiotics. In ulcerative colitis, colonic damage from toxic metabolites of bile acids has been suggested(9). Alpha-tocopherylquinone, a vitamin E derivative that antagonizes vitamin K dependent bacterial enzymes reversed ulcerative colitis dramatically in one subject(32).

Drawing on much broader experience with inflammatory bowel dis- ease, Gottschall has proposed that gut dysbiosis plays the major etiologic role, with small and large bowel fermentation being a key component. She has used a specific carbohydrate diet restricted in disaccharide sugars and devoid of cereal grains to alter gut flora(33). Some will undoubtedly argue that Gottschall's success is due to food allergen elimination, but the time course of patients' responses is more consistent with the authors' contention that a gradual alter- ation of gut flora content is the mechanism.

McCann has pioneered a dramatic, experimental treatment for in- flammatory bowel disease which has induced a rapid remission in 16 out of 20 patients with ulcerative colitis. A two-day course of multi- ple-broad spectrum antibiotics to "decontaminate" the gut is followed by administration of defined strains of E. coli, and Lactobacillus ac- idophillus to produce a "reflorastation" of the colon(34).

Arthritis and Ankylosing Spondylitis
Immunologic responses to gut flora have been advanced by several authors as important factors in the pathogenesis of inflammatory joint diseases. It is well-known that reactive arthritis can be acti- vated by intestinal infections with Yersinia, Salmonella and other enterobacteria(35). In some cases bacterial antigens have been found in synovial cells(36,37) and may enter the circulation because of the increased intestinal permeability associated with the intestinal infec- tion(l5). Increased intestinal permeability and immune responses to bacterial debris may cause other types of inflammatory joint disease as well. but there is little evidence of the frequency with which this occurs(38-40). Several groups have proposed a specific mechanism by which Klebsiella pneumoniae may provoke ankylosing spondylitis (41-43). HLA-B27 is expressed on the lymphocytes and synovial cells of 97% of patients with ankylosing spondylitis. This antigen cross- reacts with antigens found on Klebsiella pneumoniae and possibly other enterobacteria. Patients with ankylosing spondylitis have higher levels of Klebsiella pneumoniae in their stools than controls and have higher levels of anti-Klebsiella IgA in plasma than do con- trols. Patients who are HLA-B27 positive but who do not have an- kylosing spondylitis do not have Klebsiella in their stools or Kleb- siella antibodies in their plasma.