Sign up to receive email updates, action alerts, health tips, promotions to support our work and more from EWG. You can opt-out at any time. [Privacy]


New Evidence in Mercury-Autism Link

December 13, 2004

Overloaded?: New Evidence in Mercury-Autism Link

The weight of the evidence supports a fresh look at the mercury-autism hypothesis

Both autism and mercury exposure are characterized by functional impairment to speech, language and behavior (Bernard 2001, Blaxill 2004b). Recent studies also suggest that the same key regions of the brain are affected in both cases (Limke 2004, Kates 2004). At the same time, episodes of severe mercury exposure reveal that there is no single manifestation of mercury poisoning. In fact, children exposed to high levels of mercury during gestation and infancy have suffered from strikingly different diseases.

Minamata disease resulted from in-utero exposure to mercury-contaminated fish. Children with Minamata disease had symptoms indistinguishable from mental retardation or cerebral palsy (Kondo 2000). Acrodynia resulted from mercury in infant teething powders in the early 1900s. Children with Acrodynia suffered peeling and reddened skin on their hands and feet, and heightened sensitivity to light (Warkany 1966).

Individual susceptibility played an important role in both disorders. Although thousands of children were treated with mercury-containing teething powders, only one in 500 to one in 1,000 children who were exposed developed Acrodynia (Warkany 1966). The role of individual sensitivity made it extremely difficult to link mercury exposure with what was, at the time, a new and bizarre disease. Similarly, when children have been exposed to high levels of mercury in foods, only a small group develop severe mercury poisoning while thousands are apparently unharmed (Jalili 1961, Kondo 2000).

Dr. James' findings clearly reveal a mechanism by which autistic children would be predisposed to mercury-related oxidative damage to their developing brain and nervous system. Several additional pieces of evidence strengthen the potential link between mercury exposure and autism in children with abnormal antioxidant capacity. These include:

  • The indisputable toxicity of mercury to the brain, particularly the developing brain (Limke 2004, Clarkson 2002, Mahaffey 1999).
  • Peer-reviewed reports showing that autistic children are extremely poor at ridding their bodies of mercury as measured by mercury hair levels (Holmes 2003).
  • The recent finding that autism-like symptoms are triggered by thimerosal in mice with a predisposition to autoimmunity (Hornig 2004).
  • The fact that the prevalence of autism in boys is four times that in girls, and that boys have elevated incidence of damage from mercury exposure in epidemiologic studies (Vahter 2002).

Mercury targets brain cells

Pre-natal and early life mercury exposures cause multiple impacts to basic brain development by disrupting the division and migration of neuronal cells (Mahaffey 1999). Mercury creates oxidative stress that directly kills brain cells. Human beings accumulate more mercury in the brain than in blood or other organs. Organic mercury actively transported through the blood—brain barrier accumulates in the highest concentrations in the cerebellum, especially the neuronal cells (Limke 2004). The cerebellum is the brain region associated with movement and cognition, and a key region targeted by toxic chemicals (Fonnum 2000), and a region of impairment in autistic patients (McAlonan 2004).

Dr. James investigated the effect of thimerosal on human brain astrocyte and neuron cells. She found that astrocytes have higher levels of glutathione compared to neurons and were more resistant to the cytotoxic effects of thimerosal (James 2005). Mercury was less toxic to human brain cells pretreated with glutathione or a glutathione precursor N-acetyl cysteine, which is used as a treatment for mercury intoxication and it is thought to speed mercury excretion from the body (Ballatori 1998). Similar studies have documented oxidative damages of mercury and glutathione protection to T cells, astrocytes, neurons and fibroblasts (Makani 2002, Shanker 2003).

Immune symptoms

In the summer of 2004, researchers at Columbia's Mailman School of Public Health reported symptoms similar to autism in thimerosal-exposed mice. Thimerosal was administered to three strains of laboratory mice at exposures that replicated infant exposure in the 1990s. Only the mouse strain with a predisposition to autoimmunity was affected by thimerosal exposure. These mice had significant growth delay, reduced locomotion, exaggerated response to novelty, and changes in the brain and nervous system that were suggestive of autism (Hornig 2004).

Metal metabolism and regression

Unlike most substances that are toxic to the brain, there is a significant lag time between exposure to either mercury or thimerosal, and the emergence of the symptoms of mercury poisoning. The length of the lag period depends on the severity of exposures. The delay between first exposure and on-set of symptoms of mercury poisoning has been attributed to the gradual depletion of the brain's compensatory responses, chiefly glutathione and other antioxidants (Weiss 2002).

Autism often is diagnosed after a period of seemingly healthy development. Regressive children lose previously acquired skills such as speech and mobility, or fail to progress in their development. The role that oxidative stress and environmental chemicals play in regressive autism is unknown, but studies finding reduced metal excretion in autistic children point to different dynamics in regressive autism.

A recent publication by physician Amy Holmes reported that autistic children had significantly lower levels of mercury in their hair relative to non-autistic children, suggesting greater accumulation of mercury in the body due to reduced excretion capabilities (Holmes 2003). Hair samples were analyzed from 94 autistic children and 45 non-autistic children between one and two years of age and the autistic children had significantly lower levels of mercury in their hair samples (0.47 vs. 3.63 ppm).

For non-autistic children the level of mercury in the hair sample was strongly correlated with the mother's exposure to mercury in dental fillings, fish consumption and mercury-containing RhoGAM vaccinations during pregnancy. In autistic children there was no correlation, hair levels were always low, even in cases where elevated maternal mercury exposure was reported.

Regressive cases had higher concentrations of mercury in hair indicating less impaired mercury metabolism. This indicates that children least able to excrete mercury might experience autistic symptoms immediately while regressive cases of those with some capacity to detoxify and get rid of mercury would at first appear normal, but would then build up mercury to a critical point before their excretion capacity was overwhelmed and autistic symptoms surfaced.

Gender difference in autism and mercury poisoning

Males are much more likely to be diagnosed with autism or learning and behavioral disorders, possibly due to a reduced capacity to combat oxidative stress. Interestingly, males are also more sensitive to early life mercury exposure than females. Human epidemiological studies find boys to be more susceptible to the cardiac effects of mercury. Dr. Philippe Grandjean reported effects on blood pressure and heart rate variability in Faroese boys with mercury concentrations between one and 10 µg/L (Sorensen 1999). EPA's 'safe' dose of mercury translates to approximately 5.8 µg/L. Studies of mass mercury poisoning in Minamata, Japan report a skewed birth ratio due to increased fetal death for males (Sakamoto 2001). The U.S. Centers for Disease Control found that boys had higher concentrations of mercury in hair than girls, but these differences were not statistically significant (McDowell 2004). These findings are mirrored in laboratory animals (Vahter citing Gimenez-Llort 2001).