In order to understand the devastating effects that may be caused by the widespread use of antibiotics, it is necessary to understand the role of microorganisms in the intestinal tract.
There are two main kinds of bacteria in the intestinal tract: aerobic and anaerobic. The aerobic bacteria need oxygen while the anaerobic bacteria don’t need oxygen to live and even may be killed if oxygen is present. Some bacteria grow faster with oxygen but can adapt to a low oxygen environment. Another major group of organisms in the intestine are the yeast and fungi. In the intestinal tracts of some individuals, there may be single-celled animals called protozoa as well. These organisms, in a normal intestinal tract, are usually found in a natural balance that is healthy. It is estimated that there are 500 or more different species of bacteria in the average human intestinal tract (1). Because there is limited oxygen in the intestinal tract, the anaerobic bacteria that don't require oxygen predominate. Of the 500 species, there are perhaps 30 or 40 species that constitute the majority of the bacteria present. It's estimated that there are about 10-100 trillion cells of bacteria in the intestinal tract at any one time (1). To give you an idea of the size of that number, there are about a 100 trillion human cells in the entire human body. Thus, in a normal individual who is not on antibiotics, 10-50 % of their total cell volume is composed of bacteria.
There are very few bacteria in the stomach because the stomach acid kills them. In comparison, the colon harbors a million times more bacteria than the stomach. In the normal individual, this acid is neutralized with bicarbonate from the pancreas as food passes into the small intestine, allowing greater microbial overgrowth. Bacteria constitute about 50% of the content of feces. These residents of the intestinal tract are always in a state of flux with new bacteria continuously being produced and old bacteria continuously being flushed out in the moving intestinal contents and later in feces.
A study that was reported in the Journal of Infection and Immunology (2) found that when oral penicillin was administered to experimental animals, the total population of anaerobic bacteria, including the beneficial bacteria, was reduced by a factor of 1,000. These bacteria, which are called Lactobacilli, are also present in yogurt. As the good bacteria are killed off, the potentially harmful bacteria increase rapidly. This study reported translocation of the harmful bacteria out of the intestinal tract and into the lymph nodes surrounding the intestinal tract. From these lymph nodes, these bacteria were then strategically placed to cause new infections throughout the body.
Another harmful effect of antibiotics is that killing off all the normal bacteria results in the proliferation of yeast. There are hundreds of articles in the scientific and medical literature indicating that yeast over-growth is associated with antibiotic use. Some of the most important are included in the references at the end of this chapter (3-13). There are two reasons for this. First, when the normal bacteria in the intestine are killed off, the yeast have no competition so they are able to get the lion's share of all the food that passes through the intestinal tract after a meal. Second, the yeast may actually be stimulated by many of the antibiotics (12, 13).
Scientific work on animals is relevant to yeast infection in humans. Infant mice were much more susceptible to Candida infection than older mice and, once exposed to Candida at an early age, developed persistent candidiasis (3). If these mice were given antibiotics at an early age, the Candida in the intestinal tract increased an average 130-fold. Exposure of infant mice to the hormone cortisone increased Candida in the intestine 8-fold. Similar results with antibiotics and cortisone are found in humans (5-11). Largely because of the overuse of antibiotics, the incidence of disseminated candidiasis has changed from a rare occurrence prior to 1960, to the fifth most common organism encountered in infections acquired at a hospital in Southern California (14). It is important to know that bacteria and yeast produce chemical byproducts in the body that are normally only present in very low concentrations. When yeast and bacteria, normally only present in small quantities in the intestinal tract, reach extremely high numbers, they produce these byproducts in much higher concentrations which are then absorbed from the intestinal tract into the blood. From there, they circulate throughout the body to all the tissues and are eventually filtered out of the body into the urine.
In addition to the production of these byproducts, the yeast cells may convert to their more invasive colony form. The yeast in this hypha form imbed themselves into the lining of the intestinal tract like ivy climbing a brick wall. This attachment is facilitated by the secretion of yeast digestive enzymes at the point of attachment. The intestinal lining is thus digested by a variety of yeast enzymes including phospholipase A2, catalase, acid and alkaline phosphatases, coagulase, keratinase, and secretory aspartate protease (15-17). The secretory aspartate protease is of special importance because it may destroy the lining of the intestinal tract and may also digest the IgA and IgM antibodies produced by the body to attack the yeast (15). The destruction of this gastrointestinal lining may be the reason for the abnormal secretin response discussed in the chapter on the digestive system.
As a result of multiple yeast attaching to the intestinal lining, some of the intestinal cells may die and the lining may appear like Swiss cheese on a microscopic level. Ordinarily, undigested food molecules would not be able to pass through this intestinal lining. However, because of the holes in the intestinal lining, undigested food molecules can pass through. This phenomenon is called the leaky gut syndrome. A major consequence of the leaky gut syndrome is a much greater susceptibility to food allergies. The undigested food is recognized as an invader by the immune system and as a consequence, antibodies of both the IgE and IgG types may start to be produced. After a while, both behavioral and allergic reactions may occur after eating these foods. Many times, patients with multiple allergies will be retested after anti-yeast therapy and find that their allergies have disappeared. When the yeast overgrowth has been eliminated, the intestinal lining heals, the intestine is no longer leaky, and the immune system may diminish its attacks against the offending foods.
As discussed in the first edition of my book, one of the chemical compounds in urine that I initially suspected was due to intestinal yeast overgrowth was called dihydroxyphenylpropionic, acid-like compound (DHPPA). Several years ago, I began a collaborative study with Dr. Walter Gattaz, a research psychiatrist at the Central Mental Health Institute of Germany in Mannheim to evaluate urine samples of patients with schizophrenia. These samples were very valuable since they were obtained from patients who were drug-free. Thus, any biochemical abnormalities would be due to their disease and not a drug effect. Five of the twelve samples contained a very high concentration of a compound identified by gas chromatograph-mass spectrometer (GC/MS) as a derivative of the amino acid tyrosine, which is very similar to, but not identical to 3,4-dihydroxyphenylpropionic acid. I have since then identified this compound as 3-(3-hydroxyphenyl)-3-hydroxypropionic acid or HPHPA (Figure 1).
Newborn infants tested at approximately one month of age had extremely low values of this compound in urine since newborns are not colonized with intestinal germs (Figure 2). In older children, the values are much higher. In children with autism, values may be extremely high. There is some degree of overlap in the normal and autism population but the median and the mean values are significantly higher in the children with autism. (The median is the middle value of a group of numbers while the mean is the average value of the group.) The mean value for all infants is 3.7 mmol/mol creatinine with a standard deviation of 3.6 mmol/mol creatinine and a range from 0.3 - 12.7 mmol/mol creatinine. In normal male control children, the mean value is 91.5-mmol/mol creatinine with a standard deviation of 90.4; the median value in this group is 51.1 mmol/mol creatinine. In autistic male children, the mean value is double that of the controls: 192.4 mmol/mol creatinine with a standard deviation of 90.4; the median value in this group is 143.5 mmol/mol creatinine, nearly triple the value of the control group. In normal female control children, the mean value is 85.5-mmol/mol creatinine with a standard deviation of 55.9; the median value in this group is 74.5-mmol/mol creatinine. In autistic female children, the mean value is double that of the controls: 182.4 mmol/mol creatinine with a standard deviation of 200.6; the median value in this group is 111 mmol/mol creatinine, a value 49% greater than the control females. In all groups the median values are smaller than the corresponding mean values indicating that the values are not normally distributed and that the populations are skewed by some samples with very high concentrations of HPHPA.
What was surprising to me was that there was not a significant decrease in HPHPA after antifungal drug therapy. The mean value for the HPHPA actually increased a little. This increase indicated to me that this compound could not be due to the yeast, but was probably due to a different microorganism. Several children and adults with Clostridium difficile infection of the intestinal tract had high values of HPHPA in their urine and also a similar compound, called monohydroxyphenylpropionic (18, 19), which I suspected was being produced by one or more species of the bacteria genus Clostridium.
There are many different species of Clostridium. Some of the common species are Clostridium tetani which causes tetanus; Clostridium botulinum which causes food poisoning (botulism); Clostridium perfringens and Clostridium difficile which cause diarrhea; Clostridium perfringens, Clostridium novyi, Clostridium bifermentans, Clostridium histolyticum, Clostridium septicum, and Clostridium fallax which may all cause gangrene (20). Many other species of Clostridium are normal inhabitants of the intestinal tract but may not even be scientifically described or named as a species. The major reason for a lack of knowledge about these organisms is that they are strict anaerobes that cannot tolerate oxygen. Since they must be processed in an oxygen free environment, many hospital laboratories do not have the capability to identify these organisms.
The exception is Clostridium difficile, which is identified by the toxin it produces in the stool rather than by the isolation of the organism itself. Clostridium difficile overgrowth of the intestinal tract causes a severe and potentially fatal disorder called pseudomembranous colitis (21). This overgrowth is frequently associated with the use of oral antibiotics, indicating that this organism is resistant to many of the common antibiotics such as penicillin, ampicillin, tetracyclines, cephalosporins, chloramphenicol, and others (22). This organism is usually treated with either metronidazole (Flagyl) or vancomycin followed by a replenishment of the intestine with Lactobacillus acidophilus (23). Since many bacteria can genetically transfer drug resistance to other similar species and even unrelated species, it is likely that multiple species of Clostridia may now be resistant to the most common drugs.
Another reason that I became interested in clostridia was due to the theory of Ellen Bolte (23) that the tetanus bacteria (Clostridium tetani) might be responsible for some cases of autism. Her child developed autism after a DPT (diphtheria, pertussis, and tetanus) immunization, which includes a tetanus toxoid. She was concerned that her child may actually have contracted tetanus from a contaminated vaccine. When the antibodies to tetanus were checked several years after this vaccination, the antibodies to tetanus were very high. Her child was extremely developmentally delayed and also had a high value of HPHPA in the urine. There are some interesting parallels between autism and tetanus. Individuals with tetanus, like many with autism, have extreme sensory sensitivity and often need to be placed in dimly lit rooms (24-25) and avoid loud noises. In addition, patients with this disorder, which is also known as lockjaw, might have difficulty chewing and swallowing. Likewise, children with autism frequently have difficulties eating foods with certain textures. Thus, Ellen’s idea was that perhaps her child had instead contracted “subacute” tetanus causing many of the sensory related symptoms of autism, but in a non-lethal form because of the immunization. Such cases of subacute tetanus have been reported even in individuals who had been immunized and had high levels of antibodies to the tetanus toxin (24, 25). I thought it was highly unlikely that her child contracted tetanus from the tetanus vaccine. However, I thought it possible that he might have a Clostridium tetani overgrowth of the intestinal tract or an overgrowth of another species of Clostridium that might also be producing toxins similar to that of tetanus. As a result, these toxins might have caused the high antibody levels in her child or it is possible that one or more of the toxins in the vaccine was much more toxic to her child than to the average child immunized.
Clostridium tetani overgrowth of the intestinal tract has been demonstrated in rats (26). The toxins produced by several different species of Clostridia (tetani, botulinum, barati, and butyricum) are very similar biochemically (27) and therefore antibodies produced against one Clostridium toxin would also probably react against the tetanus toxin. Also the gene for the tetanus neurotoxin is located on a plasmid (28), a piece of “naked” DNA that can be easily passed on to different species of Clostridia and perhaps even other species of bacteria which would confer on the new species the ability to make tetanus toxin.
Several of the patients with high urine concentrations of HPHPA had positive stool immunoassay tests for Clostridium difficile, leading me to suspect that Clostridia species were responsible for the production of this compound. Treatment of a number of patients with elevations of this compound with drugs that kill Clostridia such as vancomycin and Flagyl resulted in nearly complete elimination of this compound in urine samples. There is a marked decrease in the urinary concentration of HPHPA following the administration of standard age-appropriate doses of the antibiotic Flagyl (metronidazole). In all four patients, the concentrations of HPHPA decreased 99% or more after two to three weeks on this drug (Table 1). In the first patient in the above series, HPHPA rapidly increased following the cessation of metronidazole treatment. I suspect that this increase after stopping the drug was due to the fact that Clostridia are spore-forming organisms. Spores are extremely resistant forms of the bacteria that are difficult to kill and they “hatch out” when drug therapy ends and repopulates the intestinal tract. The first patient improved after Flagyl treatment but then regressed when the drug was discontinued. The same child was retreated with a six-week course of vancomycin. A developmental specialist estimated that the child had gained six months of development after the six weeks of drug therapy. Again, the child regressed after discontinuation of therapy.
In a clinical study (29), Dr. Richard H. Sandler from Rush Children's Hospital in Chicago, Illinois, along with a multi-center team recruited 11 children with regressive-onset autism who had a history of antimicrobial therapy. The children were given 500 mg of vancomycin per day for 8 weeks. Based on the Wilcoxon Signed Rank Z-scores and other measures during treatment with vancomycin, Dr. Sandler's group noted improvement in communication and behavior for the group as a whole. Their report indicated that "Although improvement was clear by several measures, unfortunately these gains did not endure." When the parents of the children were telephoned 2 weeks after the end of the trial, most reported "substantial behavioral deterioration." When the children were seen after 2 to 8 months, all but one had returned to baseline analog ratings.
Elevated levels of HPHPA is not only found in individuals with autism. Patients with values of HPHPA greater than 500 mmol/mol creatinine in the urine almost always have severe neurological, psychiatric, or gastrointestinal disorders. These types of disorders are often seen in autism, severe depression, chronic fatigue syndrome, tic disorders, psychotic behavior or schizophrenia, partial muscle paralysis, severe colitis, or sometimes a combination of these disorders. One young woman with an acute psychosis had the highest value I had ever seen, nearly 7500 mmol/mol creatinine, a value approximately 300 times the normal median value for adults! According to the patients’ physicians, treating psychotic individuals with vancomycin, who have elevated HPHPA in the urine, resulted in remission of symptoms without the use of neuroleptic drugs.
In his evaluation of vancomycin for the treatment of autism, Dr. Sandler did not use Lactobacillus acidophilus GG after the children were treated with vancomycin. Therefore, the Clostridia most likely returned after the spores hatched or perhaps more drug resistant forms of Clostridia were produced after this relatively intense drug therapy. The results of Dr. Sandler are very similar to those of the child in Table 1 whose urine HPHPA cleared up after metronidazole (Flagyl) but became abnormal again after metronidazole was stopped. In addition to an inability to kill spores of Clostridia by his treatment protocol, Dr. Sandler also failed to control the Candida overgrowth, which is extremely common following the use of all antibiotics. Candida overgrowth would, of course, eventually eliminate many of the gains made by the elimination of Clostridia.
An examination of how HPHPA fits into normal human metabolism helps to give an insight into how this metabolic route may be involved in autistic behaviors.
There are two possible sources of HPHPA: phenylpropionic acid and 3-hydroxytyrosine (Figure 1). Both of these compounds have significant neurochemical effects that may produce abnormal behaviors in both animals and humans. As shown in Figure 1, HPHPA is produced from the amino acid phenylalanine in the diet. Phenylalanine is a constituent of almost all proteins, which are broken down to amino acids by digestive enzymes in the intestinal tract. Phenylalanine is an essential amino acid that cannot be restricted from the diet without causing serious health problems. Phenylalanine is important because it is the raw material from which the neurotransmitters dopamine and norepinephrine are formed.
When certain bacteria of the Clostridium family (genus) are present in high numbers, phenylpropionic acid or 3-hydroxytyrosine may be formed from phenylalanine in the intestinal tract (Figure 1). Either of these compounds may then be further converted to 3-hydroxyphenylpropionic acid which, in turn, is converted to HPHPA by the enzymes in the human mitochondria that break down fatty acids. Occasionally, fatty acids such as adipic, suberic, methylsuccinic, and ethylmalonic are elevated when 3-hydroxyphenylpropionic acid is elevated. Presumably, these fatty acids may not be burned up efficiently by the body and become elevated when the mitochondrial enzymes are swamped by excessive 3-hydroxyphenylpropionic acid.
Phenylpropionic acid is important because it is an inhibitor of the enzymes in the brain that break down enkephalins (31,32) and administration of phenylpropionic acid to mice raises brain enkephalin concentrations (32) and causes analgesia (pain relief) when injected intraperitoneally (into the abdominal cavity) into mice. Enkephalins and endorphins are opioid peptides that are produced in the brain and adrenal gland. The enkephalins exert a whole range of biological effects including analgesia, regulation of the hypothalamus in the brain, modulation of emotions, stimulation of sexual and feeding behavior, and regulation of blood pressure, temperature, and intestinal function (33). These compounds also profoundly alter many functions of the immune system (34). Thus, the buildup of enkephalins after increased phenylpropionic acid in the blood might have extremely important effects on human physiology. I have found that phenylpropionic acid is frequently elevated when HPHPA is elevated in the urine. In the organic acid test done at The Great Plains Laboratory, measurements of Phenylpropionic acid in addition to HPHPA are included together with the metabolites of dopamine (HVA) and norephinephirine (VMA) which are discussed below.
3-Hydroxytyrosine (Figure 1), the other possible source of HPHPA, is important because it induces a characteristic behavioral syndrome in rats consisting of forepaw padding, head weaving, backward walking, splayed hind limbs, wet dog shakes, hyperactivity and hyper-reactivity in addition to depleting the brain of catecholamines (35). Thus, this compound might play a direct role in causing abnormal behaviors in autism, schizophrenia, and other disorders. We have noticed that the molar ratio of the urinary concentration of the dopamine metabolite homovanillic acid (HVA) to that of the epinephrine/ norepinephrine metabolite vanillylmandelic acid (VMA) in urine (tested by mass spectrometry) is commonly elevated when HPHPA is elevated. This elevation appears to indicate that a byproduct involved in the formation of HPHPA likely inhibits the conversion of dopamine to norepinephrine, leading to relative dopamine excess. Animal studies have indicated that dopamine neurons mediate behaviors such as hyperactivity and stereotypical behaviors common in autism. Of course, the drugs such as the phenothiazines and haloperidol, commonly used to treat autism and schizophrenia, are well known to block the action of excessive dopamine at the receptor level (36).
The following clinical implications result from this newly discovered metabolic pathway.
(2) The use of organic acid testing can provide a valuable tool in guiding therapy so that harmful microorganisms may be eliminated before treatment with amino acids like phenylalanine that might actually cause neuropsychiatric symptoms to worsen.
I want to emphasize that the die-off reaction with Flagyl or vancomycin therapies may be very severe. The die-off reaction appears to be a release of toxins by the Clostridia as they die that may last 3-7 days after drug therapy. A child getting this particular therapy should be under very close medical supervision because the side effects may be much more severe than those associated with the yeast die-off reaction and can include symptoms such as heart palpitations, fever, and extreme tiredness: (some children may not even move during the first several days of therapy). The severity of the die-off reaction indicates to me the potency of the toxins produced by these organisms. However, the die-off reaction may be minimized by the concomitant use of materials such as bentonite or powdered charcoal, which are available in health food stores to absorb the toxins. In addition to the use of Culturelle, Sacharomyces boulardii, a non pathogenic yeast, has also been proven clinically effective at controlling and preventing recurrent Clostridia. Probiotic control of Clostridia has been the most successful therapy to date.
Clostridium difficile appears to be one of the organisms that produce the HPHPA. According to Sidney Finegold MD, one of the world’s leading experts on anaerobic bacteria (Personal Communication), there are as many as 100 different species of Clostridium that may inhabit the intestinal tract. There is an immunological test for the toxin produced by Clostridium difficile that can be done on stool to confirm this organism. A negative test for Clostridium difficile does not rule out all species of Clostridium, only Clostridium difficile. There is no convenient method to confirm the identity of the other 99 species of Clostridium in the intestinal tract that may also produce this compound.
It has been found that injecting an animal with its own fecal matter, which consists of 50% bacteria by weight, only causes a mild immune response (1). This indicates that the normal flora (germs) of the intestine are given tolerance by the immune system or that the immune system does not mount an attack against these organisms. The immune system takes an “inventory” of all the cells present in the body during fetal development and shortly after birth. (I have relied on material by Teresa Binstock Ph.D. (38) at the University Of Colorado School Of Medicine as the primary source of this information.) In addition to the immune system taking inventory of its own cells, it seems increasingly likely that the immune system also takes an inventory of bacteria and yeast cells present in the intestinal tract. This inventory is performed by a group of cells called the CD5+ B-cells, which are among the very first immunological cells to appear in the developing embryo and appear to play a role in tolerance to intestinal microorganisms in postnatal life. These cells may also play a role in regulating the secretion of IgA, the antibody class that is secreted into the intestinal tract and may be involved in selecting which microorganisms are tolerated there. Furthermore, the eradication of normal flora by repetitive antibiotic use during infancy may cause the CD5+ B-cells to reject normal organisms as foreign invaders at a later age. Any cells that are on this early inventory may be given immune tolerance and as a result, will not be attacked later on by the immune system.
I have been impressed by numerous reports from parents of children with autism who indicate that their children used antibiotics at a very young age. I suspect that yeast and undesirable bacteria resulting from antibiotic therapy during early infancy have been “granted” immune tolerance; this immune tolerance may be one of the reasons why the yeast overgrowth in autism is so difficult to control and tends to recur even after months of antifungal therapy. Such an immune tolerance to yeast in the developing fetus may also occur if the mother has yeast infections during pregnancy. A woman, who had severe vaginal yeast infections during pregnancy, gave birth to a daughter with a severe yeast infection of the mouth called thrush. This daughter was later diagnosed with autism. Such cases may explain children who appear to behave abnormally even as young infants. Thus, a new direction for future research might be to find a way to reprogram CD5+ B-cells or to replace them with more suitable cells from a donor.