Chelation of Mercury for the Treatment of Autism

Autism - Etiology:

By Amy S. Holmes, MD (written circa 2000)

Autism and disorders resembling autism can be caused by a number of disorders, including Fragile X Syndrome, tuberous sclerosis, and phenylketonuria, and by at least one notable chromosomal abnormality, an inverted duplication of a portion of chromosome 15. But for the vast majority of cases of autism today, there is no strictly genetic explanation. As with many chronic disorders, most cases of autism appear to be caused by some genetic predisposition coupled with some early environmental insult.

Several recently-released reports point to the occurrence of an autism "epidemic" with the latest incidence figures quoted to be on the order of 1 out of every 250 children. The Report on Autism to the California Legislature released in 1999 documents a large increase in full-blown DSM IV autism alone, with other disorders increasing at the same rate as population growth. F. E. Yazbak, M.D. found similar rates of increasing incidence in other states reported in his Autism 99:A National Emergency. The Center for Disease Control’s own investigation of Brick township, New Jersey found a very high incidence of autism as well. Some noted sources attribute the apparent increase in autism incidence to better diagnoses on the part of pediatricians and the various pediatric specialties. Most, however, are unable to fully accept this simplistic explanation because the diagnosis is strictly a behavioral one, and it is highly doubtful that the highly skilled diagnosticians of earlier years could have overlooked such obvious behavioral anomalies occurring in such a large proportion of children. Furthermore, since it is impossible to have a "genetic epidemic", one must examine possible early environmental insults for clues to explain the increase in autism cases.

Bernard, et al, have written an excellent article comparing autism with mercury poisoning. All aspects of both disorders are examined, including symptoms, signs and findings on laboratory tests. The parallels between the two disorders is disturbingly obvious, even to the most casual reader. This, coupled with many case reports of clinical improvement among autistic children upon removal of at least a small part of their whole-body load of mercury, seems to indicate that many cases of autism today are, in fact, cases of mercury poisoning. The early environmental insult, in these cases, is mercury exposure that overwhelmed the body’s attempts at detoxification.

How does mercury gain access to a fetus or an infant? First of all, mercury is ubiquitous. It is in our water supply. In this setting, it exists mainly in cationic (1+ or 2+) form. This form is largely unabsorbed. Fish and shellfish are a known source of organic mercury (methyl mercury). Organic mercury is absorbed reasonably well by the gastrointestinal tract. Exposure via these two routes is common, but it is far exceeded by exposure via dental amalgams and thimerosal-containing vaccines. Mercury vapor is known to be released from dental amalgams, and it is known to cross the placenta with ease. It is not too far-fetched to assume that some mercury vapor (Hg - 0) is released from the dental amalgams of the mother, she inhales the vapor, it enters her bloodstream, some crosses the placenta and enters the developing fetus. Once metallic mercury (vapor, Hg - 0) enters the cell, it can be easily converted to its cationic form, and in this form, readily binds to sulfhydryl groups on enzymes and other proteins. Once tightly bound via this mechanism, it is in the body for a long time. Thimerosal-containing vaccines are now given with abandon. Upon its arrival into our world, the newborn is greeted with a Hepatitis B vaccine. He then receives several more doses of this vaccine along with DPT and Hib vaccines. All three of these vaccines contain relatively large amounts of thimerosal, which is 49.6% ethyl-mercury by weight. It was not long ago that the only vaccine containing thimerosal was the DPT vaccine. But, the Hepatitis B vaccine was made "mandatory" in 1991 and the Hib vaccine a few years earlier. Is it a coincidence that the incidence rate of autism has soared in the 1990's? Is it better diagnosis or is it more mercury early in life? Add onto these noted exposures the thimerosal-containing RhoGam injection. A reasonable conclusion of greatly increased mercury exposure to developing fetuses, newborns and young infants being responsible for the obvious autism "epidemic" is almost inescapable.

Why isn’t every child equally affected? The answer remains unknown at the present time, although recent investigations point to the possibility of problems with at least one form of metallothionein. Studies further investigating the structure and amounts of various metallothionein proteins in autism will be done later this year.

Testing for Mercury Toxicity:

Poisoning with most heavy metals is detected easily with blood tests. For example, if a person has detectable lead in his body, he will have some detectable lead in his blood. In fact, the gold standard for the detection of poisoning for most heavy metals is a test of intracellular content using red blood cells. Hair and urine levels of heavy metals are a general reflection of blood levels. Also, getting rid of most heavy metals such as lead with chelating agents is not difficult. This is because most heavy metals in the body exist in a reasonable equilibrium between their preferred storage sites and the bloodstream.

This is not the case with mercury. After an exposure, detectable levels are present in the blood for only a short time, on the order of weeks to a few months. This is because mercury, unless eliminated, quickly becomes tightly bound to sulfhydryl-containing enzymes and other proteins in the liver, kidney, lining of the gastrointestinal tract, and brain. So, if any amount of time has elapsed after a significant mercury exposure, little if any mercury will be detected in the blood, urine or hair.

The only way of directly detecting the amount of mercury present in the liver, kidney, GI tract, and brain is via biopsy of these organs. This is NOT a recommended procedure. Besides, the real issue is not how much mercury is present, but how mercury-toxic the patient really is. Mercury has well-documented effects on different laboratory tests, so this is the preferred way of measuring mercury toxicity. The list below is only a partial list of helpful lab tests, and does not reflect at all the effect of mercury on the brain itself.

Useful Lab Tests for Assessing The Presence of Heavy Metal Toxicity (Partial List)                       

1. Urine Tests

    1. Indications of Mitochondrial Dysfunction
      1. Uncoupling of oxidative phosphorylation
        1. Elevated fatty acid metabolites
        2. Elevated lactate
        3. Elevated hydroxymethylglutarate
      2. Multiple partial blocks in Krebs cycle
    2. Elevated 3-methyl histidine
    3. Elevated sarcosine
    4. Elevated pyroglutamate
    5. Elevated vanilmandellate
    6. Elevated homovanillate
    7. Fractionated urine porphyrins
      1. Elevated coproporphyrin
      2. Elevated precoproporphyrin

2. Immune System Tests

    1. Elevated total IgE
    2. Low total IgG
    3. Low IgG subclasses
    4. Low CD8+ cells
    5. Low NK cells
    6. Elevated CD3+CD26+ cells

3. "Esoteric" Blood Tests

    1. Low superoxide dismutase
    2. Low reduced glutathione
    3. Low glutathione peroxidase
    4. Elevated lipid peroxides
    5. Elevated blood and/or platelet serotonin
    6. Elevated epinephrine and/or norepinephrine

4. Evidence of Urinary Sulfate-wasting - Low plasma sulfate with normal urine sulfate/creatinine ratio.

Note : No one, even the most toxic person, has all these lab abnormalities present.. Even the most mercury-toxic will have some normal results. 

Some signs of mercury toxicity affecting the brain and immune system can be found on physical exam. Below is a very partial list:

Abnormalities on Physical Exam That Might Be Found in Mercury Toxicity

  1. Dilated pupils
  2. Sweaty hands and feet
  3. Pathologic reflexes - Babinski most common
  4. Very brisk knee jerks
  5. Slight esotropia
  6. Rashes, eczema
  7. Elevated heart rate

And there are many others. Because of the known kinetics of mercury in the body, there are some abnormalities that will not be found unless the mercury exposure was recent. These are:

  1. Elevated hair mercury
  2. Elevated blood mercury
  3. Elevated intracellular (RBC) mercury
  4. Elevated urine mercury 

The natural history following exposure to mercury dictates certain findings. A few months after exposure ends, there will be no detectable mercury in the blood, urine or hair. For variable time periods after the end of mercury exposure, the non-CNS organs will gradually rid themselves of mercury, leaving just brain mercury behind. The half-life of mercury in the brain is estimated to be on the order of 20 years — far longer than its half-life in organs outside the central nervous system. For this reason, the only mercury left in a person whose exposure ended several years before testing is likely to be solely in the central nervous system.

This is the situation in many autistic children, especially older children, who have had no mercury exposure since vaccinations received at 18 to 24 months of age. They have no amalgam fillings, they eat no fish or seafood, and have received no further immunizations. In these patients, challenge tests with conventional agents commonly used for this purpose (DMSA, DMPS) show no mercury at all since these agents do not cross the blood-brain barrier, and therefore have no access to CNS mercury.

Treatment of Mercury Toxicity

In order to be a good chelator of mercury, a molecule must have two opposed (in 3-D structure) sulfhydryl groups or other groups that bind well to mercury. The effect of having these two opposed groups is to bind divalent cationic mercury (Hg 2+) in sort of a "pincer grasp", making it very difficult for mercury to leave the chelator to bind to another molecule. Some compounds that meet this requirement are:

  1. DMPS (2,3 dimercaptopropane sulfonate)
  2. DMSA (meso-2,3 dimercaptosuccinic acid)
  3. Lipoic acid

Those compounds which have only 1 sulfhydryl or other mercury-binding group are poor chelators simply because they do not bind mercury tightly enough to keep it from binding to other molecules. Among these compounds are MSM and cysteine. Their net effect may be simply moving mercury around to other sites in the body.

One substance that may have some good chelating properties is cilantro. The problem with this substance is that it is unknown at the present time exactly what the ingredient present in cilantro might be that may give it good chelating properties. Without knowing the identity of the actual chelating substance, it is impossible to know how much and how often it should be given.

Also it is impossible to determine if all cilantro has the same amount of the unknown substance. It is probably a good idea not to use cilantro for chelating mercury until more is known about the substances involved.

DMPS is a great mercury chelator. Unfortunately, it has never been tested in children, and for this reason alone, we do not advocate its use in children. It may prove to be safe for use in children, but until safety testing has been done, we prefer to use the one chelator that has been extensively tested and used for years in children — that is DMSA.

DMSA is an excellent chelator of most heavy metals including mercury. When used appropriately, it is safe and effective. DMSA has survived the testing necessary for FDA approval for use in children. This means it has been tested in children and was found to be both safe and effective. Despite the FDA’s poor record in testing and approving vaccines, the procedures for testing and approval of drugs are quite rigorous.

The only approved use for DMSA is for the treatment of lead poisoning in children. Fortunately, DMSA is not very selective about which heavy metal it chelates, and binds to mercury quite readily. Despite claims of DMSA’s ability to cross the blood-brain barrier (BBB), it is doubtful that it really does so. The study cited most often as proving DMSA’s ability to cross the BBB was done in rats. Rats are known to not have a good BBB. DMSA is water-soluble and not very lipid-soluble. This characteristic alone raises some doubts about its true ability to cross the BBB.

Lipoic acid fits the molecular criteria of a good chelator. It has two diametrically opposed sulfhydryl groups capable of tightly binding mercury in a "pincer grasp". It also has the advantage of being lipid-soluble which implies an innate ability to cross cell and mitochondrial membranes and the BBB more easily than DMSA.

An ideal course of chelation therapy for mercury poisoning should include the following:

1. First, stopping any ongoing exposure:

    1. No more fish or seafood (salmon is supposed to be OK).
    2. Replace any amalgam fillings in teeth with white composite material.
    3. Use only thimerosal-free vaccines.

2. Get rid of the loosely-bound body mercury.

3. Then, chelating the more tightly-bound mercury including that in the brain.

4. Appropriate nutritional support designed to counteract mercury’s known effects and to make the patient more comfortable while mercury is being moved around. The use of antioxidants is recommended.

5. Appropriate monitoring tests (especially important in non-verbal children) to check on blood counts, kidney and liver function, and mineral levels, and to gauge how much mercury is being excreted.

We have been using such an approach in the last 10 months with good results. The course of treatment we are currently using consists of:

1. DMSA — until provoked urine mercury is low

2. DMSA plus lipoic acid — until no more mercury is excreted in urine or stool

3. Appropriate nutritional support as determined by testing with particular attention to antioxidants

4. Monitoring tests — CBC, liver function tests, serum copper, plasma zinc, intracellular trace minerals — most every 2 to 3 months.

It appears that adding glycine to every dose of DMSA increases mercury excretion.

When undertaking a course of chelation for mercury, one important point is to do it in cycles consisting of "on" and "off" periods. Give the patient as much time off chelation as on chelation. If any abnormal results show up in the routine monitoring tests, it is best to stop chelation for a while, retest, and resume chelation when the results have normalized. The following are some "Reasonable Rules for Life" for during and after treatment for mercury toxicity:

1. No fish and no seafood (supposedly, salmon is okay).

2. No amalgam (metal) fillings in teeth. Use white composite material instead.

3. No more thimerosal-containing vaccines. By 2002, all US vaccines will be thimerosal-free. Until then, ask for thimerosal-free vaccines. For almost every possible vaccine given, there is at least one brand that does not contain thimerosal.

Early Results

We currently have over 500 autistic patients under treatment with DMSA ranging in age from 1 to 24 years old. In general, we do not expect to see any behavioral, language, or social improvements until at least some of the CNS mercury has been removed. As of 1/15/01, we had 85 patients who had finished DMSA alone and had completed at least 4 months of DMSA + lipoic acid. The results of treatment in these patients are presented below:

n = 85

Improvement (%)































Once lipoic acid is added, we usually track mercury excretion via tests of fecal mercury. We have noticed a large dependence of excretion on age of patient with the younger patients excreting much more mercury than the older patients. We think this difference in rapidity of excretion may explain the differences in response between the various age groups.

We have 6 patients, all 1 to 2 years of age who are finished with treatment by measurements of urinary and fecal mercury excretion. These 6 patients are "normal" by parent reports and repeat psychological testing. We have no children over the age of 2 who are finished with treatment. The rapidity of excretion seems to decrease markedly with each additional year of age. There are several children, mostly in the younger age groups, who have made remarkable progress to the point of being able to be mainstreamed in school, but who are still have some "oddities" of behavior — none of these children have completed treatment yet. These are very early results, but appear very promising. As more data is gathered, outcomes will be better able to be predicted, including length of treatment as well as ultimate prognosis.