Chapter 16
Summing Up
By Dr. William Shaw


How effective are the nutritional and antifungal therapies discussed in the previous chapters? According to parents of children with autism who were surveyed by Bernard Rimland Ph.D. of the Autism Research Institute, some of these therapies were much more effective than all of the commonly used psychoactive drugs including neuroleptics and stimulants which are commonly prescribed for autism. (These data in Tables 1 and 2 were adapted from Dr. Rimland’s published study and are reprinted with permission on the following pages.) The better to worse ratio for antifungal drugs (nystatin and ketoconazole) in children with suspected yeast overgrowth of the intestinal tract is five times higher than the second most effective drug, the psychoactive agent Clonidine (Table 1). (The higher this ratio, the more effective the drug is rated.) I suspect that the effectiveness of the antifungal agents might be even higher if dietary restriction of simple sugars to reduce yeast overgrowth were employed.  Several antiseizure medications were ranked poorly (ratios less than 1.0) based on behavioral response (not on antiseizure effectiveness). All of the stimulant drugs including Cylert, Ritalin, and amphetamine were also rated very poorly. With nutritional supplements (Table 2), high ratios, indicating high effectiveness, were reported for vitamin B-6, DMG, zinc, niacin, vitamin C, calcium, and folic acid.

The idea that autism is made worse or caused by abnormal byproducts of microorganisms and  opiates from wheat and milk is not inconsistent with other research findings in autism such as abnormal neuroanatomical findings, abnormal EEG results, and abnormal brain scans. Similar abnormalities were found in the disease PKU even though the primary abnormality is a genetic defect in a single enzymatic reaction. There is no inherent reason that dramatic biochemical changes in multiple biochemical systems, caused by byproducts of microbes or abnormal peptides from wheat and milk, would not be expected to alter brain structure and function.

In PKU, correction of the metabolic defect by restriction of phenylalanine during infancy allows for normal development; retardation occurs if dietary intervention occurs too late. If abnormally elevated metabolites of yeast and bacteria cause autism, then it is reasonable to think that elevations of these compounds would have maximum negative impact during periods of critical brain growth and development. As in PKU, metabolic intervention in autism might only be possible in the early stages of the disorder before the brain has matured. The differences in severity of the disorder and individual differences in symptoms might be due to different combinations of metabolites, how elevated they are, the duration of the elevation, the age at which the metabolites become abnormally elevated, and the susceptibility of the individual developing nervous system to the different microbial metabolites. Indeed, these differences may even determine which disease is manifested. The concentrations of these microbial products are not trace amounts on the metabolic scale. One child with autism evaluated in my laboratory had a urine tartaric acid concentration (6000 mmol/mol creatinine) that was nearly 400 times the upper limit of normal (approaching a lethal dose) after the use of multiple oral antibiotics. Many of the concentrations of these microbial compounds found in the urine of children with autism frequently exceed even the concentrations of the predominant mammalian organic acids in urine.

Table 1
Parent ratings of behavioral results of drugs


Drug

Got
Worse

No
Effect

Got
Better

Better:
Worse

No. of
Cases

Aderall

41%

25%

34%

0.8:1

475

Amphetamine

47%

28%

25%

0.5:1

1217

Anafranil

32%

38%

30%

1.0:1

381

Antibiotics

31%

57%

12%

0.4:1

1799

AntifungalsC: Diflucan

5%

41%

55%

11:1

330

AntifungalsC: Nystatin

5%

46%

49%

10:1

986

Atarax

25%

53%

22%

0.9:1

477

Benadryl

24%

51%

25%

1.0:1

2711

Beta Blocker

18%

50%

33%

1.9:1

256

Buspar

26%

44%

30%

1.1:1

328

Chloral Hydrate

41%

38%

21%

0.5:1

418

Clonidine

21%

31%

47%

2.2:1

1280

Clozapine

40%

42%

18%

0.4:1

102

Cogentin

19%

54%

27%

1.5:1

162

Cylert

45%

35%

20%

0.4:1

600

Deanol

15%

56%

30%

2.0:1

200

DepakeneD: Behavior:

25%

43%

32%

1.2:1

957

DepakeneD: Seizures

12%

32%

57%

4.8:1

627

Desipramine

34%

31%

34%

1.0:1

67

DilantinD: Behavior

28%

49%

23%

0.8:1

1077

DilantinD: Seizures

14%

36%

50%

3.5:1

400

Felbatol

21%

53%

26%

1.2:1

43

Fenfluramine

20%

52%

28%

1.4:1

459

Halcion

37%

37%

26%

0.7:1

54

Haldol

38%

28%

34%

0.9:1

1154

IVIG

7%

51%

42%

6.3:1

45

KlonapinD: Behavior

28%

38%

34%

1.2:1

192

KlonapinD: Seizures

31%

60%

10%

0.3:1

42

Lithium

26%

43%

31%

1.2:1

410

Luvox

29%

35%

35%

1.2:1

161

Mellaril

29%

38%

33%

1.2:1

2062

MysolineD: Behavior

43%

43%

15%

0.3:1

136

MysolineD: Seizures

19%

59%

22%

1.2:1

64

Naltrexone

20%

46%

34%

1.7:1

221

Paxil

29%

30%

41%

1.4:1

283

Phenergan

29%

46%

24%

0.8:1

266

PhenobarbitalD: Behavior

47%

37%

16%

0.3:1

1076

PhenobarbitalD: Seizures

17%

43%

40%

2.4:1

480

Prolixin

31%

40%

30%

1.0:1

91

Prozac

31%

32%

36%

1.2:1

1123

Risperidal

18%

28%

54%

3.0:1

616

Ritalin

45%

26%

29%

0.7:1

3813

Secretin: Intravenous

7%

44%

48%

6.7:1

333

Secretin: Transdermal

10%

49%

41%

4.2:1

132

Stelazine

28%

45%

27%

1.0:1

420

TegretolD: Behavior

25%

45%

31%

1.2:1

1423

TegretolD: Seizures

12%

33%

55%

4.6:1

762

Thorazine

36%

40%

24%

0.7:1

919

Tofranil

30%

38%

33%

1.1:1

713

Valium

35%

41%

24%

0.7:1

824

ZarontinD: Behavior

34%

45%

21%

0.6:1

136

ZarontinD: Seizures

19%

53%

28%

1.4:1

93

Zoloft

33%

33%

34%

1.0:1

321


Table 2
Parent ratings of behavioral results of nutrients


Nutrients               

Got
Worse

No Effect

Got Better

Better:
Worse

No. of 
Cases

Vitamin A

2%

58%

41%

23:1

618

CalciumE:

2%

62%

36%

15:1

1378

Cod Liver Oil

3%

47%

50%

16:1

818

Cod Liver Oil with Bethanecol

16%

45%

39%

2.4:1

56

Colostrum

5%

58%

37%

8.1:1

345

Detox. (Chelation)C:

2%

22%

76%

35:1

324

Digestive Enzymes

3%

42%

56%

20:1

737

DMG

7%

51%

42%

5.6:1

5153

Fatty Acids

2%

42%

55%

23:1

626

5 HTP

10%

51%

39%

3.7:1

145

Folic Acid

3%

54%

42%

12:1

1437

Food Allergy Treatment

3%

37%

61%

21:1

560

Magnesium

6%

65%

29%

4.6:1

301

Melatonin

8%

30%

61%

7.3:1

573

P5P (Vit. B6)

13%

37%

51%

4.0:1

213

Pepcid

9%

63%

28%

3.2:1

93

SAMe

15%

66%

19%

1.3:1

62

St. Johns Wort

14%

64%

21%

1.5:1

84

TMG

14%

44%

42%

3.0:1

434

Transfer Factor

8%

53%

39%

4.8:1

98

Vitamin B3

4%

55%

41%

10:1

659

Vitamin B6 alone

8%

63%

30%

3.9:1

620

Vitamin B6 with Magnesium

4%

49%

47%

10:1

5780

Vitamin B12

4%

33%

63%

15:1

192

Vitamin C

2%

57%

41%

18:1

1706

Zinc

2%

51%

47%

20:1

1244


I think that the reason these abnormalities (some of which have been known for decades) have been ignored for so long, by almost all of the researchers in the field of metabolic diseases, is because the intense focus has been on finding new inborn errors of metabolism. By definition, abnormal microbial products are not due to a genetic defect in a human biochemical pathway. In addition, it is likely that most researchers in the field of metabolic disorders also made the unwarranted assumption that microbial metabolites are metabolically and physiologically inert. Instead, it appears to me that the human body and the microorganisms in the gastrointestinal tract are an integrated and interdependent biochemical system within the human body.

Early intervention is key to the treatment of PKU. Children with PKU who are put on the special diet low in phenylalanine have close to normal IQ’s while those who are untreated until later in life are impaired. The children of Pamela Scott and Karyn Seroussi, both who recovered from autism, were started on their therapies at the age of two years. Even this age may be too old for some children. I think that any child, under two years old having frequent infections treated with antibiotics, is at risk for autism, seizures, and/or ADD and should be tested and treated if abnormal microbial overgrowth is present.

An extensive body of research by Bauman, Corchesne, and others has documented numerous abnormal structures in the brains of children with autism. Children with autism caused by a defect in succinyl purine metabolism have the same kind of brain abnormalities reported by this group (see chapter 9). However, I wish to emphasize that finding abnormal anatomical structures in the brain proves nothing about the cause of these abnormalities. I suspect that some or all of these abnormalities may be due to the toxic effects of the microbial metabolites or the abnormal peptides from wheat and milk just as the drug thalidomide caused abnormal development of the limbs of children exposed to this drug in utero. Saying autism is a brain disease makes just as much sense as saying that the flippers in children exposed to thalidomide are due to an arm disorder; both statements have elements of truth but the oversimplification distorts the complexity of the truth. If autism is caused by microbial metabolites and peptides, then scientific studies, in which these compounds are given to animals, should be able to reproduce the symptoms of autism.

These abnormalities that I found in autism are not just specific for autism. I have found them elevated in Rett’s Syndrome, which is a separate disorder seen primarily in girls in which some autistic-like behaviors are exhibited and in the genetic disease Smith-Magenis syndrome. Elevated values have also been found in urine samples of children who have autistic symptoms with Prader-Willi syndrome, Fragile-X syndrome, Tourette’s syndrome, Williams disease, neurofibromatosis, and tuberous sclerosis. Autistic symptoms are common in all of these disorders. In addition, I have found the yeast byproducts to be elevated in the urine of children with Down’s syndrome who also exhibited autistic symptoms. In addition, I have found elevated yeast and/or bacterial metabolites common, both in adults and children with seizures, psychosis in severe depression, and in perhaps 80-90 % of children with attention deficit hyperactivity. These metabolites are also found in some people with hypoglycemia or low blood sugar.

All of my work leads me to discard the prevailing dogma that microbial metabolites are inert in human metabolism.
Based on all of my work and the work of many other researchers, I have developed a theory for autism:

Factors that impair the immune system can lead to recurrent infections such as ear infections, strep throat or bronchitis.  These factors can include genetic deficiencies of the immune system and inborn errors of metabolism. Other non-genetic factors include adverse reactions to immunizations such as gastrointestinal viral infection (from live vaccines) and metal and chemical toxicity (which can also be due to environmental factors). These infections are then treated with antibiotics.

A yeast overgrowth of the gastrointestinal (GI) tract occurs following the elimination of the normal flora of the gastrointestinal tract. The yeast produces abnormal compounds called gliotoxins and other immunotoxins such as mannan byproducts that are toxic to the immune system and make it weaker. Yeast overgrowth of the intestinal tract may persist in children who are exposed to any use of antibiotics as infants, especially if immune deficiencies are also present. Because of immunodeficiency, a child is more likely to be re-infected and be exposed to additional antibiotics until a vicious cycle has been established.

The yeast produces abnormal sugars which may interfere with carbohydrate metabolism or alter the structure and function of critical proteins through the formation of pentosidines. The yeast also produces analogs of the Krebs cycles that inhibit energy production and gluconeogenesis.  The yeast also produces enzymes such as phospholipase, which break down phospholipids, and proteases such as secretory aspartate protease which break down proteins. These enzymes may partially digest the lining of the intestinal tract itself.  This digestion of the intestinal tract takes place as the yeast cells attach to mucosa lining the intestinal tract. The digestion of the intestinal lining by the yeast and/or viral infection (perhaps from live virus vaccines) causes a leaky gut and may also limit the ability of intestinal cells to produce hormones such as secretin that is necessary for the production of sufficient pancreatic digestive enzymes. Undigested wheat products and other food molecules are more likely to be absorbed from the intestinal tract into the body and elicit an allergic response, a food allergy. Some of these food allergies may manifest as behavioral disorders.

Candida proliferation also elicits the production of antibodies that cross-react against many of the human tissues including the brain, pancreas and wheat proteins, perhaps leading to atrophy of the pancreas and disruption of key brain functions caused by myelin autoantibodies. Pancreatic atrophy may be associated with further impairment of digestive function with resulting malabsorption and malnutrition.  In addition to the yeast overgrowth, there may also be an overgrowth of certain bacteria of the Clostridia family.  The Clostridia share one common attribute with the yeast in that they are resistant to many of the common broad-spectrum antibiotics used to treat ear infections and Strep throat. Clostridia bacteria produce 3-(3-hydroxyphenyl)-3-hydroxypropionic acid and perhaps other neurotoxins that are absorbed into the body and which may also alter behavior.

The undigested peptides from wheat and milk, which react with opiate receptors in the temporal lobes of the brain that are responsible for auditory integration and language, disrupt the functions of this key area. Cade’s work established significant improvements in almost every aspect of autism in a group of 70 children with autism after only one month on the gluten and casein free diet (1). Areas of improvement included social isolation, eye contact, mutism, learning skills, hyperactivity, stereotypical activity, hygiene, panic attacks, and self-mutilation. The behavioral aspects of autism can be reproduced in rats by intraperitoneal injection of bovine casomorphin (2). Within seven minutes of injection, the following behaviors were observed in the rats: running violently around the enclosure, jumping behavior, wet dog shakes, dilation of pupils, raised hair, salivation, rapid respiration, teeth chattering, vocalization, circling behavior, reduced sound response, decreased social interaction, and abnormal postures. In addition, the peptides from wheat and milk may profoundly alter the metabolism of a large number of important hormonal peptides hydrolyzed by DPP IV.


 Implications for Gene-Searching

A large amount of money from the National Institutes of Health and other sources has been allocated to various academic centers throughout the country to discover the gene or genes that predispose to autism. If my theory is correct, there would not be a single gene but a whole host of genes, perhaps fifty or even a hundred or more that would lead to increased susceptibility to infection. The current approach is to extract DNA (the chemical basis for our genetic material) from the white blood cells of people with autism, break down the DNA to smaller pieces with enzymes called nucleases, and then separate the pieces by a process called gel electrophoresis. Computers are then used to determine if two siblings from the same family have an abnormal matching band of DNA that is not present in normal people. If my theory is correct, the matching band may very well be different in each family and could result in this information being misread by the researcher since they are looking for a small number of genes. The much higher incidence of autism in males compared to females naturally led geneticists to suspect that genetic factors influencing autism are linked to the X-chromosome. Since many of the genes for different immunodeficiencies are linked to the X-chromosome, it would seem worthwhile to focus DNA research on the well-documented phenomenon of immune deficiencies in autism.


 Where Do We Go From Here?

Recently, a psychological test called the CHAT (3), used for early diagnosis of autism, was developed in England. This test is extremely accurate in predicting autism in children at 18 months of age. No children with a normal CHAT score developed autism while all of the children with two or more major abnormalities in the CHAT test developed autism by the time they were thirty months old. I propose a relatively simple study of the effectiveness of the therapies in this book. Two hundred children with an abnormal CHAT test would be identified. With an incidence of autism at about one in a thousand, approximately 200,000 children would need to be screened. The test is relatively easy to administer, not very time consuming, and could easily be included in a routine well-baby examination with very little expense.

Half of the two hundred children identified as high risk by the CHAT test would receive the following as needed: a low sugar diet, antifungal therapy, therapy for Clostridia bacterial overgrowth, gluten and casein restriction, vitamin B6 and DMG supplementation, and food allergy desensitization. Children with immunodeficiency would be treated with gamma globulin and/or transfer factor.

The other 100 children would receive conventional medical treatment. Both groups would get any special services available such as Lovaas therapy, speech therapy, occupational therapy, etc., but the therapists would not be told which therapy group the child was in. At the end of one year, psychologists, who did not know which therapies the children had received, would evaluate all of the children. If early biochemical intervention in autism is most effective when started early (as in PKU), then the children with the “alternative” treatments would do much better than the children treated with conventional therapies. I propose that these younger children be tested because I suspect that autism is very much like PKU, in that the earlier a child is treated, the better the outcome. I have received many reports of benefits of these same therapies in adults and older children with autism and do not want to appear to “write-off” any group of people. But I believe our first priority should be an attempt to prevent of any new cases of autism.

There is obviously a tremendous amount of additional work that needs to be done to clarify the best dietary approach used with antifungal therapy; there is considerable disagreement about the best dietary approaches to autism, even among the contributors to this book. 

The overuse of antibiotics, especially for recurrent otitis media, needs to be completely re-examined and a large epidemiological study should be undertaken by the Centers for Disease Control to determine how much damage has been caused to our children by antibiotic use. A tax on antibiotics could be used to pay for such a study and other experimental studies on the role of abnormal microbial byproducts in human disease. A large group of infants, perhaps 10,000 or more, should be monitored by stool cultures for yeast and bacteria. Urine organic acid testing should also be done, on perhaps a monthly basis, for several years to evaluate the association between a wide number of disorders such as ADD, autism, and seizures and abnormal yeast and bacteria overgrowth caused by oral antibiotics.

However, I think it would be a tragic mistake to wait until all the data are collected before taking additional action.  It is better to act on preliminary findings, because the stakes are so high, especially when many safe alternatives are available.

Our children are our most precious resource.


References

  1. R. Cade et al. Autism and schizophrenia: intestinal disorders. Nutritional Neuroscience 3:5772,2000.
  2. Z. Sun and R. Cade.  A peptide found in schizophrenia and autism causes behavioral changes in rats. Autism 3: 85-95, 1999.
  3. S. Baron-Cohen, J. Allen, and C. Gilberg. Can autism be detected at 18 months? The needle, the haystack, and the CHAT. Brit. J. Psychiatry161: 839-843,1992.