Title: A Controlled Study
of Mercury Levels through Lab Analysis of Hair Samples of Children
with Autism as Compared to Their Typically Developing Siblings and
Age/Sex Matched Children with Other Developmental Disabilities.
Autism is a developmental disability characterized
by severe, pervasive deficits in social interaction, communication
and range of interests and activities. The neurobiologic basis of
autism is well accepted, although the specific etiology is unknown.
It has been theorized that autism may result from a combination
of predisposing genes and environmental factors. While autism has
a known association with some environmental factors such as rubella
and valproic acid exposure in utero, other proposed environmental
mechanisms such as mercury toxicity or other heavy metal exposure
have minimal research support. Despite this fact, interventions
including oral chelation therapy with potentially harmful side effects
are being used to treat autism. Treatment is initiated after hair,
blood or urine samples are analyzed by specialty labs revealing
elevated values consistent with heavy metal toxicity. The controls
and standards for these labs are unclear and similar treatment protocols
are prescribed for most of the children with autism.
We hypothesize that lab studies evaluating heavy
metal toxicity through hair analysis for children with autism as
compared to their typically developing siblings and other age and
sex-matched developmentally disabled controls will not be significantly
different. We will invite thirty children with autism who have typically
developing siblings to participate in the study. Subjects and controls
will be between the ages of 2 and 6 with subjects having met DSM-IV
criteria for autism. A second control group consisting of age and
sex matched children with other developmental disabilities will
also be used. All children will have been fully vaccinated and without
chronic health problems. Hair samples will be obtained from 30 subjects,
30 sibling controls, and 30 developmentally disabled controls according
to lab specifications and submitted in a blinded fashion to Doctor's
Data, Inc., a widely used specialty lab. Lab results will then be
compared for subjects and controls using analysis of variance to
determine whether there is a significant difference for values in
the groups. Introduction and Goals
Autism is a neurodevelopmental disorder diagnosed
clinically on the basis of pervasive deficits in social interaction,
communication, and range of interests and activities. While the
neurobiologic basis for autism is now undisputed, the specific etiology
is unknown. Research in the areas of immunology, genetics and neuroanatomy/physiology
are being conducted with a view otward establishing causal factors.
Many researchers have proposed that autism may result from several
predisposing genes which combine with environmental factors to produce
the autism phenotype. Environmental factors which have a known association
with autism include teratogenc effects of rubella infection and
valproic acid exposure in utero. Other environmental mechanisms
including dietary intolerance, vaccinations, and heavy metal toxicity
have been proposed as causal, but have minimal if any research support.
Despite this fact, interventions have been implemented based on
these presumed environmental causations. Some of these interventions
seem relatively harmless. The casein-gluten free diet based on a
theory of elevated blood peptide levels resulting in behavior changes
is one example; while time consuming and often costly, the diet
is unlikely to result in detrimental effects to the child. However,
other interventions are not so benign. Included in this category
is oral chelation therapy with DMSA used to treat "mercury poisoning".
Proponents state that the behavioral and developmental presentation
in autism is consistent with that of mercury poisoning and results
from exposure to mercury in vaccines and other sources. Parents
of children with autism have been urged to obtain mercury levels
on their children and pursue treatment if indicated. Hair samples
are sent to a variety of recommended specialty labs for this purpose.
It is unclear what controls are used in these laboratories and whether
there is data supporting differences in lab values for children
with autism and typically developing children. This information
would be helpful for both parents and professionals in making decisions
regarding treatment options.
The goal of this study is to determine whether
there are differences in mercury levels as measured by hair analysis
obtained from a widely used national specialty laboratory for children
with autism as compared to two control groups. One control group
will consist of typically developing siblings of subjects and the
other will consist of age and sex-matched children with other developmental
disabilities. This preliminary study will allow us the opportunity
to see if elevated mercury levels are present more frequently in
children with autism as compared to controls. If children with autism
demonstrate higher levels of mercury, this finding would have both
etiologic and treatment implications which need to be investigated
on a larger scale. If children with autism show no differences from
controls, this would be helpful in refuting generalizations regarding
mercury poisoning in children with autism and raising serious questions
about the basis for treatment strategies such as oral chelation
therapy.
Hypothesis
We hypothesize that there will be no significant
difference in lab studies evaluating mercury levels through hair
analysis for children with autism as compared to their sibling controls
and age and sex matched developmentally disabled controls. Background
and Significance
Given its neurobiologic basis, there has been
great interest in establishing specific cause(s) for autism. Many
researchers feel that autism may result from the interaction of
several predisposing genes in combination with environmental factors.
A few environmental exposures in utero have previously been identified
as having associations with autism based on case reports linking
the two entities; these include Fetal rubella syndrome, Fetal valproate
syndrome, and Fetal Alcohol syndrome. Many other environmental factors
have been hypothesized to play an etiologic role in autism. One
popular theory is that children with autism have difficulty breaking
down casein and gluten proteins, resulting in a buildup of peptide
byproducts which then act as natural opioids and result in the behavioral
manifestations of autism. Another hypothesis is that autism represents
a form of mercury poisoning as a result of environmental exposures,
particularly thimerisol-containing vaccines. Bernard et al note
the similarities in developmental, behavioral and physiologic presentations
in those with autism as compared to those with mercury poisoning.
Indeed, this correspondence of autistic and mercury poisoning traits
seems to be the only basis for this proposal as there are no case
reports or studies documenting higher levels of mercury in individuals
with autism as compared to controls. Nonetheless, national specialty
labs targeting families of developmentally disabled children offer
testing for heavy metal toxicity and several practitioners validated
by parent support groups offer oral chelation therapy as a treatment
option for autism. This treatment with DMSA has potentially serious
side effects including bone marrow suppression, but is being used
by a number of local families despite this.
Proponents of the mercury poisoning theory of
autism point to thimerosol as a primary source of mercury exposure.
Thimerosol is an organic compound which is 49.6% mercury that has
been used as a preservative in many vaccines since the 1930's, including
diphtheria, tetanus, haemophilus influenze type B, and hepatitis
B. They point to the 1999 AAP report recommending reducing or eliminating
thimerosol from vaccines and reports of exposures to cumulative
levels of mercury exceeding EPA recommendations during the first
6 months of life as support for their theory. However, a recent
review of thimerosol use in childhood vaccines revealed no evidence
of harm from thimerosol-containing vaccines other that local hypersensitivity
reactions. Case reports of neurotoxicity from thimerosol in humans
were found only at doses more than 100 times that found in vaccines.
However, it was noted that identification of the risk from thimerosol
was problematic because of gaps in knowledge of its toxicity. Of
interest is the fact that the vaccine which has been most controversial
in its link to autism is the measles-mumps-rubella vaccine which
has never contained thimerosol. Attempts to link widespread use
of vaccines to increased prevalence rates of autism has been controversial
with most researchers reporting lack of temporal association.
Given this background, it was our purpose to
conduct a small pilot study comparing mercury levels from hair samples
in children with autism to those of controls. Hair samples are chosen
over blood or urine samples as an effective, relatively noninvasive
measure of chronic exposure to mercury. The World Health Organization
and Environmental Protection Agency have indicated benefits of hair
testing for heavy metals including methylmercury. The laboratory
that we chose was one among several popular specialty labs and uses
reference ranges based on a 1996 Healthy American Population Study
as specified in a 1998 article in Biological Trace Elements Research.
Our study serves as an initial effort to validate
or refute anecdotal reports of mercury toxicity in children with
autism. By comparing hair samples of children with autism to controls,
we will obtain some initial objective data regarding this issue.
If there is a significant difference in values for subjects versus
controls, then this would warrant further study of mercury exposure
as a possible etiologic factor in autism. If, as suspected, there
is no difference in lab values for the groups, this would be important
information to share with parents and other professionals as they
contemplate diagnostic testing and treatment strategies. Methods
and Analysis
Thirty children with autism between the ages
of 2 and 6 will be recruited from the population of children evaluated
through the Weisskopf Center for the Evaluation of Children. These
children will have completed multidisciplinary team evaluations
and met DSM-IV criteria for autism. Children with associated medical
or genetic conditions such as Fragile X syndrome or inborn errors
of metabolism will be excluded. In order to be included, subjects
must have typically developing siblings between the ages of 2 and
6 . Normal development of siblings will be documented through administration
of the Vineland Scales of Adaptive Behavior by a psychologist at
the Center. Additional controls between the ages of 2 and 6 years
will be recruited from the population ofchildren diagnosed with
developmental disabilities other than autism. These children will
have had medical and psychological evaluations through the Weisskopf
Center for the Evaluation of Children and diagnoses of developmental
disabilities such as Attention Deficit Hyperactivity Disorder, learning
disabilities or mental retardation but not including Pervasive Developmental
Disorders. These controls will be matched to subjects by sex, age
(within two years of subject's age), and broad intelligence parameters
(cognitively impaired or not cognitively impaired based on IQ scores
less than or above 60).
Children and their families will be invited to
participate regardless of race or socioeconomic status. The subject
population is likely to be predominantly male based on the 4 to
1 predominance of males with autism. Families will sign consent
forms for participation in the study which have been approved by
the Human Studies Committee.
Hair samples will be obtained in accordance
with laboratory kit instructions. Hair is obtained by cutting samples
with stainless steel scissors from the back of the head with no
more than an inch of new growth. Small amounts of hair will be obtained
from 5 or 6 sites. One quarter gram of hair will be submitted for
each sample, placed in the envelope provided, and submitted with
a completed form stating the child's initials, age, and sex, but
without any specific developmental diagnosis. Samples will be obtained
from the thirty subjects, thirty sibling controls, and thirty age
and sex matched developmentally disabled controls. The hair samples
will be placed in mailers and shipped according to lab specifications
for heavy metal analysis. Arrangements have been made with David
Quig, Ph.D.,vice president of research at Doctor's Data, to submit
hair samples at one time in a blinded fashion.
After lab reports are received, an analysis
of variance will be conducted in order to determine whether there
are any statistically significant differences in mercury levels
between the three groups. Based on a review of the literature and
our communication with the laboratory, there is no data available
to determine expected group difference and group variability. Therefore,
the needed sample size cannot be determined through power analysis.
Due to this limitation, we will complete a post hoc power analysis
using the mean and standard deviation for the autism and control
groups to determine the number of subjects needed to find a significant
difference between groups. If we fail to find a difference, the
post hoc analysis will indicate the number of additional subjects
that would be needed for significance if we do a follow-up to this
pilot study.
Subjects will have already completed multidisciplinary
evaluations at the Weisskopf Center for the Evaluation of Children,
as will other developmentally disabled controls. The Vineland Scales
of Adaptive Behavior will be administered by Lonnie Sears, Ph.D.
to document normal development in siblings of subjects.
Hair samples will be obtained at the Center
and mailed off in the kits provided.
