The CHARGE Study: An Epidemiologic Investigation of Genetic and Environmental Factors Contributing to Autism

ArticleinEnvironmental Health Perspectives 114(7):1119-25 · August 2006with347 Reads
Impact Factor: 7.98 · DOI: 10.1289/ehp.8483 · Source: PubMed
Abstract

Causes and contributing factors for autism are poorly understood. Evidence suggests that prevalence is rising, but the extent to which diagnostic changes and improvements in ascertainment contribute to this increase is unclear. Both genetic and environmental factors are likely to contribute etiologically. Evidence from twin, family, and genetic studies supports a role for an inherited predisposition to the development of autism. Nonetheless, clinical, neuroanatomic, neurophysiologic, and epidemiologic studies suggest that gene penetrance and expression may be influenced, in some cases strongly, by the prenatal and early postnatal environmental milieu. Sporadic studies link autism to xenobiotic chemicals and/or viruses, but few methodologically rigorous investigations have been undertaken. In light of major gaps in understanding of autism, a large case-control investigation of underlying environmental and genetic causes for autism and triggers of regression has been launched. The CHARGE (Childhood Autism Risks from Genetics and Environment) study will address a wide spectrum of chemical and biologic exposures, susceptibility factors, and their interactions. Phenotypic variation among children with autism will be explored, as will similarities and differences with developmental delay. The CHARGE study infrastructure includes detailed developmental assessments, medical information, questionnaire data, and biologic specimens. The CHARGE study is linked to University of California-Davis Center for Children's Environmental Health laboratories in immunology, xenobiotic measurement, cell signaling, genomics, and proteomics. The goals, study design, and data collection protocols are described, as well as preliminary demographic data on study participants and on diagnoses of those recruited through the California Department of Developmental Services Regional Center System.

Full-text

Available from: Isaac Pessah
Autism is a serious neurodevelopmental
disorder characterized by impairments in social
interaction, abnormalities in verbal and nonver-
bal communication, and restricted, stereotyped
interests and behaviors (American Psychiatric
Association 1994). Although a large proportion
of individuals with autism manifest abnormal
development from birth, a subset of at least
20–30% experience a regression with onset
between 18 and 24 months of age after a
period of apparently normal development
(Lainhart et al. 2002). Autistic disorder is the
most severe form of autism spectrum disorders
(ASDs), which include Asperger’s syndrome
and pervasive developmental disorders (PDDs)
not otherwise specified. Approximately 70% of
individuals with autistic disorder have some
degree of mental retardation, and about half are
nonverbal or have very impaired speech.
Seizures are present by adolescence in about
30% of children with ASD, and between 5 and
10% of autism cases occur in association with
other serious medical conditions such as fragile
X, tuberous sclerosis, and Angelman’s syn-
drome (Fombonne 2003). Gastrointestinal
problems and sleep disturbances are also
thought to be common comorbidities;
however, population-based prevalence esti-
mates for these conditions are currently lack-
ing. Males are four times as likely as females to
have autism, but this ratio approaches one
among individuals with severe cognitive
impairment (Gillberg and Wing 1999). Most
individuals with autism cannot live indepen-
dently as adults (Rapin 1999). Over the past
20 years, the prevalence of autism has report-
edly risen, with much public debate surround-
ing the reasons for this increase. Early reports
estimated prevalence at 4–5 per 10,000 births
(Fombonne 1999). Data published in the last
few years suggest that autistic disorder occurs
in at least 1–2 per 1,000 births, and the preva-
lence of the broader autism spectrum may be
as high as 4–6 per 1,000 (Chakrabarti and
Fombonne 2005; Yeargin-Allsopp et al. 2003).
The causes and contributing factors for
autism are poorly understood. The number of
children with a diagnosis of autism as deter-
mined by the California Department of
Developmental Services (DDS) has been ris-
ing continuously for over a decade (California
DDS 2003). Although diagnostic changes
and improvements in detection probably
contribute to this increase (Chakrabarti and
Fombonne 2005; Croen et al. 2002), a true
rise in incidence may also be occurring
(Blaxill et al. 2003). Evidence for genetic
causes is strong, yet concordance in monozy-
gotic twins suggests that a minimum of 40%
of autism cases are likely to have an environ-
mental cause. No single gene has yet been
specifically linked to autism with replicability,
but the disorder is believed to be polygenic. A
few specific environmental factors are associ-
ated with autistic behaviors—prenatal expo-
sures to thalidomide (Rodier and Hyman
1998), valproic acid (Christianson et al.
1994), or rubella (Chess et al. 1978)—but
these are likely to play a negligible role, if any,
in incident cases in Western countries over
the last decade or so.
Mechanisms of pathogenesis have yet to
be delineated. Contrary to early beliefs that
autism resulted from bad parent–child inter-
actions (Bettelheim 1967), it is now widely
accepted that aberrant brain development
underlies autism pathogenesis (Bauman and
Kemper 2003; Courchesne et al. 1988; Piven
et al. 1990; Rodier et al. 1997). Autopsy stud-
ies demonstrate structural changes in the
brain, and imaging and electrophysiology
investigations reveal neurophysiologic differ-
ences in information processing between chil-
dren with autism and those with typical
development (Dawson et al. 2002; Maziade
et al. 2000; McPartland et al. 2004; Rapin
and Dunn 2003; Rosenhall et al. 2003).
Neuroimmunomodulatory factors may also
Environmental Health Perspectives
VOLUME 114 | NUMBER 7 | July 2006
1119
Research
|
Children’s Health
Address correspondence to I. Hertz-Picciotto,
Department of Public Health Sciences, TB #168,
University of California, Davis, CA 95616 USA.
Telephone: (530) 752-3025. Fax: (530) 752-3239.
E-mail: ihp@ucdavis.edu
We thank K. Jose for the data and subject tracking
systems, M. Rose for her excellent project manage-
ment, and L. Delwiche and P. Krakowiak for data
management and programming.
This work was supported by the National
Institutes of Health (1 P01 ES11269) and by the
U.S. Environmental Protection Agency through the
Science to Achieve Results (STAR) program
(R829388).
The authors declare they have no competing
financial interests.
Received 8 July 2005; accepted 6 April 2006.
The CHARGE Study: An Epidemiologic Investigation of Genetic and
Environmental Factors Contributing to Autism
Irva Hertz-Picciotto,
1,2
Lisa A. Croen,
3
Robin Hansen,
2,4
Carrie R. Jones,
1,2
Judy van de Water,
2,5
and Isaac N. Pessah
2,6
1
Division of Epidemiology, Department of Public Health Sciences, School of Medicine, and
2
Medical Investigations of
Neurodevelopmental Disorders (MIND) Institute, University of California–Davis, Davis, California, USA;
3
Division of Research,
Kaiser Foundation Research Institute, Kaiser Permanente, Oakland, California, USA;
4
Department of Pediatrics, and
5
Department of
Rheumatology/Allergy and Clinical Immunology, School of Medicine, and
6
Department of Molecular Biosciences, School
of Veterinary Medicine, University of California–Davis, Davis, California, USA
Causes and contributing factors for autism are poorly understood. Evidence suggests that
prevalence is rising, but the extent to which diagnostic changes and improvements in ascertain-
ment contribute to this increase is unclear. Both genetic and environmental factors are likely to
contribute etiologically. Evidence from twin, family, and genetic studies supports a role for an
inherited predisposition to the development of autism. Nonetheless, clinical, neuroanatomic,
neurophysiologic, and epidemiologic studies suggest that gene penetrance and expression may be
influenced, in some cases strongly, by the prenatal and early postnatal environmental milieu.
Sporadic studies link autism to xenobiotic chemicals and/or viruses, but few methodologically rig-
orous investigations have been undertaken. In light of major gaps in understanding of autism, a
large case–control investigation of underlying environmental and genetic causes for autism and
triggers of regression has been launched. The CHARGE (Childhood Autism Risks from Genetics
and Environment) study will address a wide spectrum of chemical and biologic exposures, suscep-
tibility factors, and their interactions. Phenotypic variation among children with autism will be
explored, as will similarities and differences with developmental delay. The CHARGE study infra-
structure includes detailed developmental assessments, medical information, questionnaire data,
and biologic specimens. The CHARGE study is linked to University of California–Davis Center
for Children’s Environmental Health laboratories in immunology, xenobiotic measurement, cell
signaling, genomics, and proteomics. The goals, study design, and data collection protocols are
described, as well as preliminary demographic data on study participants and on diagnoses of those
recruited through the California Department of Developmental Services Regional Center System.
Key words: autism, autistic spectrum disorder, developmental delay, environment, genetics, mental
retardation, pervasive developmental disorder. Environ Health Perspect 114:1119–1125 (2006).
doi:10.1289/ehp.8483 available via http://dx.doi.org/ [Online 6 April 2006]
Page 1
play a role (Silva et al. 2004; Vargas et al.
2005). Cytokine profiles, lymphocyte activa-
tion, and other immunologic parameters differ
between individuals with and without autism
(Ashwood and Van de Water 2004a, 2004b;
Croonenberghs et al. 2002). Distributions of
neuropeptides and neurotrophins at birth
appeared to be altered among children who
later developed autism (Nelson et al. 2001).
Results from twin and family studies sug-
gest a strong genetic contribution to the etiol-
ogy of autism. Beginning with the classic work
by Folstein and Rutter (1977), data from three
population-based twin studies have demon-
strated a higher concordance rate among
monozygotic compared with dizygotic twins
(Cook 1998). Strong familial aggregation of
autism has also been demonstrated. The sib-
ling recurrence risk (i.e., the probability of
developing autism given a person’s sibling is
autistic) has been estimated at 2–14% (Jorde
et al. 1990; Ritvo et al. 1989; Smalley et al.
1988), a 10- to 20-fold increase over the gen-
eral population prevalence. A family history of
social deficits, language abnormalities, and
psychiatric disorders has also been observed in
case–control and clinic-based studies (Folstein
and Rutter 1988; Piven and Palmer 1999).
Autism co-occurs with several known
genetic disorders, such as tuberous sclerosis
(Smalley 1998), Angelman syndrome
(Steffenburg et al. 1996), phenylketonuria,
Joubert syndrome (Ozonoff et al. 1999), and
Möbius syndrome (Johansson et al. 2001),
and chromosomal abnormalities such as frag-
ile X syndrome (Reiss and Freund 1990).
More than 90% of autism cases, however,
have none of the above syndromes.
Linkage, association, and cytogenetic stud-
ies have been conducted. Numerous candidate
genes for autism have been suggested based on
their functional role, location within candidate
chromosome regions, and positive associations
with the disease (Korvatska et al. 2002).
Replication of findings has been elusive
(Wassink et al. 2004), probably because of the
polygenic etiology, heterogeneity of the
phenotype, and, assuming a role for gene–
environment interaction, variation in exposure
distributions across populations. An epigenetic
mechanism related to Rett syndrome is also
plausible (Samaco et al. 2005). Genomewide
scans to identify regions marked by differing
gene expression are considered key at this
stage. One such scan hints at the possible
genetic basis for the well-established sex ratio
of four males to one female (Stone et al.
2004). A comparison of tuberous sclerosis
patients with and without autism demon-
strated 31 genes for which expression differed
(Tang et al. 2004); because both groups
shared the tuberous sclerosis diagnosis, the dif-
ferentially expressed genes may be related to
autism, although they are not necessarily
causal. It is plausible that a substantial propor-
tion of autism cases could be due to multiple
genes interacting with one or more environ-
mental factors (Cederlund and Gillberg 2004;
Glasson et al. 2004).
Neuroanatomic and epidemiologic investi-
gations support a prenatal or early postnatal
origin. Courchesne et al. (1988) observed cere-
bellar abnormalities consistent with abnormali-
ties in cell migration between the third and
fifth month of gestation. Magnetic resonance
imaging studies point to migrational errors that
result in disorganized columns of the cerebral
cortex (Casanova et al. 2002). Anthropometric
indicators, such as brain size and growth trajec-
tory (Herbert 2005), suggest overall cerebral
volume to be larger in mid-childhood, with
growth that accelerates early and then deceler-
ates, although this phenotype may apply to
only a subset of cases. Neuroimaging studies
indicate involvement of specific brain regions,
including the amygdala, hippocampus, and
corpus callosum (Brambilla et al. 2003;
Schumann et al. 2004).
Studies of environmental factors also
relate to the prenatal origin of autism. Chess
et al. (1978) reported that, within a cohort of
about 250 children with congenital rubella,
7% were later diagnosed with autism. A
case–control study using both maternal
reports and medical records of illnesses during
pregnancy showed relative risks of 4.1 for
influenza and 3.3 for rubella (Deykin and
MacMahon 1979). Daily maternal smoking
during early pregnancy was reported to be
linked to autism in a large case–control epi-
demiology study (odds ratio = 1.4; 95% con-
fidence interval, 1.1–1.8) (Hultman et al.
2002), although in our estimation, these
analyses may have inappropriately adjusted
for potentially intermediate variables. The
link between autism and early in utero expo-
sure to thalidomide places the timing of the
insult coincident with neural tube closure in
the fourth to fifth week of gestation (Rodier
and Hyman 1998). Case reports of autism in
children gestationally exposed to valproic acid
(Christianson et al. 1994; Rodier et al. 1997;
Williams et al. 2001) are concordant with
experimental animal studies (Ingram et al.
2000). A small number of cases of autism
after maternal infection with cytomegalovirus
(Markowitz 1983; Stubbs et al. 1984),
measles or mumps (Deykin and MacMahon
1979), or herpes (Ritvo et al. 1990) as well as
one case each of syphilis and toxoplasmosis
(Rutter and Bartak 1971) have been reported.
Taken together, the literature suggests a
prominent genetic component involving mul-
tiple gene loci, but also a likely contribution
Hertz-Picciotto et al.
1120
VOLUME 114 | NUMBER 7 | July 2006
Environmental Health Perspectives
Figure 1. Environmental exposures and sources of information in the CHARGE study. The left-hand box
indicates five classes of exposures that are candidates as environmental factors contributing to autism.
The right-hand box lists sources of data available on CHARGE study participants. Arrows show a few
examples of how specific exposures can be assessed. For example, pesticide exposures and/or their
metabolites can be assessed in several ways (black arrows): laboratory assays can be conducted on
blood (serum) and urine specimens; the interview collects information on applications in the home and
also obtains residential histories that can be linked to exposure databases on commercial pesticide
applications in California. Metals (blue arrows) can be measured in hair and in newborn blood spots
obtained from the State Genetics Diseases Branch biospecimen bank or assessed by interview questions
on fish consumption or use of household products. Exposures to infectious agents (dashed arrows) can
be determined from medical records, self-reports, and assays on serum samples to test for seropositivity
for antibodies to specific viruses.
Persistent
pollutants (flame
retardants, etc.)
Pesticides
Metals
Viruses, bacteria, and
other infections
Medical procedures
and pharmaceuticals
Biospecimens
Blood
Hair (past exposures)
Baby lock (first year of life)
Mother’s hair (if long enough, prenatal)
Urine
Buccal cells
Newborn blood spots
Interviews
Diet
Consumer products
Lifestyle
Residential information
Medical history
Linkage to exposure databases
Air, water, pesticides, hazardous waste
Medical records
Obstetric
Labor and delivery
Neonatal
Pediatric
Page 2
from both chemical and microbial agents. It is
likely that further understanding will require
consideration of critical windows during
gestation and possibly early infancy, as well as
interactions between genetic or epigenetic
predisposition and environmental factors.
CHARGE Study Aims
In light of the enormous gap in our under-
standing of the causes of both autism and
developmental delay (DD), a large epidemio-
logic study was initiated in 2002. The
Childhood Autism Risk from Genetics and the
Environment (CHARGE) study is addressing
a wide spectrum of environmental exposures,
endogenous susceptibility factors, and the
interplay between these two (CHARGE 2006).
To structure the search for etiologic factors,
we are beginning with known neuro-
developmental toxicants and hints from the
immunologic evidence. Additionally, physio-
logic differences that might provide clues about
susceptibility and mechanisms are being exam-
ined through characterization of metabolic,
immunologic, and gene expression profiles, as
well as genetic polymorphisms. Figure 1 shows
five broad classes of exposures of interest: pesti-
cides, metals, persistent pollutants with known
or suspected neurodevelopmental or immuno-
logic toxicity, medications and other treat-
ments, and infections. Exposures from both the
prenatal and early childhood periods are being
investigated, with data primarily from three
sources: a) extensive interviews with parents;
b) laboratory analysis of xenobiotics in blood,
urine, and hair specimens; and c) prenatal,
labor and delivery, neonatal, and pediatric
medical records.
CHARGE study specimens are analyzed for
immunologic, cell activation, xenobiotic,
lipomic, and genomic markers in laboratories of
the University of California–Davis (UC Davis)
Center for Children’s Environmental Health
(CCEH) (Table 1). Metals have been assayed
in blood samples from > 300 index children,
with a focus on mercury, lead, arsenic, cad-
mium, and manganese. Immunologic profiles
are being characterized, including cellular
responses to bacterial antigenic stimulation,
general immunoglobulins, and production of
chemokines and cytokines. Already, prelimi-
nary results have demonstrated significant dif-
ferences between children with autism and
children from the general population in leptin
concentrations (Ashwood P, Kwong C, Hansen
R, Hertz-Picciotto I, Croen L, Krakowiak P,
et al., unpublished observations).
A detailed lipomics screen is being applied
to the plasma from the first few hundred chil-
dren. Affymetrix GeneChip microarrays
(Affymetrix, Santa Clara, CA) have been gen-
erated from an initial sample of children and
analyzed to determine whether a genomic fin-
gerprint for autism can be identified; results
will be replicated on a further set. Brominated
flame retardants are being measured in 80–100
children, and metabolites of pyrethroid pesti-
cides will be evaluated in urine specimens.
The CHARGE study also benefits from
CCEH hypothesis-driven experimental research
on animal models for autism in mice and non-
human primates and in vitro investigations of
immune and neurogenic cells aimed at uncov-
ering molecular mechanisms. A common data-
base coordinates the archival, retrieval, and
analysis of samples, and the combination of
population-based epidemiology with state-of-
the-art molecular and cellular methods provides
a powerful basis for interdisciplinary collabora-
tive research. With future funding, the
CHARGE study will undertake targeted evalu-
ation of candidate genes, such as those responsi-
ble for regulation of xenobiotic metabolizing
enzymes, cell signaling in both neurons and
immune cells, and immune cell activation.
Currently, the study is also characterizing
phenotypic variation within the autism case
group and relating these phenotypes to the
exposures and physiologic profiles of interest.
For example, we have begun to compare
immune function in regressive autism (children
who have lost previously acquired social or lan-
guage skills) with those with early onset (chil-
dren who never acquired those skills). Other
phenotypic subtypes include, for example, high
versus low cognitive function and presence ver-
sus absence of gastrointestinal symptoms,
macrocephaly, and sleep disturbances.
Design and Subject Recruitment
The CHARGE study appears to be the first
large-scale, population-based epidemiologic
investigation focusing primarily on environ-
mental exposures, as well as their interactions
with genes, as underlying causes for autism. It
uses the case–control design, which provides
the most efficient sampling for studies of con-
ditions that are rare or of multifactorial etiol-
ogy. A further advantage is the focus on a
specific outcome, which translates into close
scrutiny of diagnoses and rigorous measure-
ment for the most highly suspect risk factors.
The CHARGE study population is sam-
pled from three strata: children with autism
(full-syndrome autism, not those with a “spec-
trum disorder”), children with DD but not
autism, and children selected from the general
population without regard for developmental
characteristics. All participating children (cur-
rently > 500, with an ultimate goal of between
1,000 and 2,000) meet the following criteria:
a) between the ages of 24 and 60 months,
b) living with at least one biologic parent,
c) having a parent who speaks English or
Spanish, d) born in California, and e) residing
in the catchment areas of a specified list of
regional centers (RCs) in California. No fur-
ther exclusions are made based on genetics or
family phenotype.
Children with autism and children with
mental retardation or DD are identified
through RCs that contract with the California
DDS to determine eligibility and coordinate
services for persons with developmental disabili-
ties. Eligibility in the DDS/RC system does not
depend on citizenship or financial status. Thus,
the system is widely used across socioeconomic
levels and racial/ethnic groups. Referrals are
from pediatricians, other clinical providers,
schools, friends, and family members.
The DDS/RC system is mandated to pro-
vide services for individuals with autism, as
well as for those with other PDDs who have
mental retardation (IQ < 70) or are substan-
tially handicapped. One investigation esti-
mated that 75–80% of the total population of
children with an autism diagnosis in the state
were enrolled in the DDS system (Croen et al.
2002). Among preschoolers, the figure may be
lower, with fewer mild cases. Additionally, this
proportion may decline with recent changes to
eligibility requirements that emphasize the
extent of disability. Children with Asperger’s
or PDDs not otherwise specified without
mental retardation are not generally eligible
for DDS/RC services and therefore are not
actively recruited into the CHARGE study.
Potential cases of autism for the CHARGE
study are defined as those who are eligible for
services based on a DDS/RC diagnosis of
autism. Families with a child who has received
a diagnosis but is not in the RC system are also
invited. The second study group, children with
Childhood autism: environment and genetics
Environmental Health Perspectives
VOLUME 114 | NUMBER 7 | July 2006
1121
Table 1. Biospecimen use for susceptibility and exposure markers.
a
Child’s blood Child’s urine Newborn blood spot Hair
Immune markers
Cytokines X X
Immunoglobulins (general) X X
Antigen-specific Ig responses X X
Cell activation X
Lipid profiles X
Brominated flame retardants X
Pesticide metabolites X
Metals X X X
Genomics X
Genetics X
a
Not an exhaustive list of assays.
Page 3
DD, is likewise drawn from those determined
eligible for services based on a diagnosis of
mental retardation or DD. Children 0–3 years
of age who are at risk for DD or disability can
receive RC services under the Early Start pro-
gram and are also eligible to be in the second
CHARGE study group. The DD children
must meet the above inclusion criteria but
are not age- or sex-matched to the children
with autism.
Staff of the RCs contact parents of chil-
dren with autism or DD, provide them an
information packet, and explain how they can
participate in the CHARGE study. For those
who are interested, permission is obtained for
the study staff to telephone the families and
schedule appointments. The children then
undergo further testing (see below) to confirm
their diagnoses.
The third group consists of children from
the general population identified from state
birth files. Throughout the study, we generate
random samples of children meeting the study
eligibility criteria according to their birth cer-
tificate information. This group is frequency-
matched to the age, sex, and broad residential
RC catchment area distribution of the autism
cases. Using names and social security num-
bers in birth certificate files, study personnel
attempt to locate current contact information
and then initiate a recruitment effort.
Data collection protocols. Participation
involves assessments of cognitive and social
development, a medical examination, biologic
specimen collection, and completion of an
exposure interview and several self-adminis-
tered questionnaires. Other components
include maternal and child medical records
review and abstractions. Table 2 summarizes
the protocols, other than specimen and
medical record collection.
CHARGE study children are assessed at
the UC Davis Medical Investigations of
Neurodevelopmental Disorders (MIND)
Institute; a small percentage were seen at the
UCLA Neuropsychiatric Institute. Stan-
dardized clinical assessments are administered
to confirm the child’s diagnostic group. Autism
cases are assessed using diagnostic tools widely
accepted for research: the Autism Diagnostic
Interview–Revised (ADI-R) (Le Couteur et al.
2003; Lord et al. 1994, 1997) and the Autism
Diagnostic Observation Schedules (ADOS)
(Lord et al. 2000, 2003). The ADI-R is a stan-
dardized, semistructured 2- to 3-hr interview
with caregivers of individuals with autism or
PDDs. It yields summary scores in the follow-
ing domains: qualitative impairments in recip-
rocal social interaction, communication, and
repetitive behaviors and stereotyped patterns.
Published values for interrater reliability are
good, with kappa values ranging between 0.62
and 0.89 (Lord et al. 2003).
The ADOS is a semistructured, standard-
ized assessment of children in which the
examiner observes the social interaction, com-
munication, play, and imaginative use of mate-
rials. The ADOS requires approximately
30 min and includes four possible modules;
the examiner chooses the one that best matches
the expressive language level of the individual
child to prevent a relatively low level of lan-
guage ability from impeding accurate measure-
ment. Diagnostic algorithms are available for
autism or for broader ASDs/PDDs (Lord et al.
2003). The ADOS provides measures in the
following domains: reciprocal social interac-
tions, communication, stereotyped behaviors
and restricted interests, and play. All kappa val-
ues for interrater reliability exceeded 0.60. All
CHARGE clinical assessment personnel are
trained and have attained research reliability on
the ADI-R and the ADOS.
Cognitive function is measured in all chil-
dren (those with autism or DD and the general
population controls) using the Mullen Scales
of Early Learning (MSEL) (Mullen 1995). The
MSEL is a standardized developmental test of
children 3–60 months of age. The MSEL con-
sists of five subscales: gross motor, fine motor,
visual reception, expressive language, and
receptive language. The MSEL allows for sepa-
rate standard verbal and nonverbal summary
scores to be constructed. The five MSEL scales
demonstrate satisfactory internal consistency
(0.75–0.83), internal reliability (0.91),
test–retest reliability (0.71–0.96), and inter-
rater reliability (0.91–0.99) (Mullen 1995).
Adaptive function is assessed by parental
interview using the Vineland Adaptive
Behavior Scales (VABS) (Sparrow et al. 1984).
The VABS is the most widely used instrument
for assessment of adaptive behavior across the
lifespan and covers the domains of socializa-
tion (interpersonal relationships, play and
leisure time, and coping skills), daily living
skills (personal, domestic, and community
skills), motor skills (gross and fine motor), and
communication (receptive, expressive, and
written communication), with developmen-
tally ordered skills for each area. The scale is
norm referenced, and recent supplemental
norms have been published for individuals
with autism (Carter et al. 1998). Psychometric
properties of the instrument include excellent
internal consistency (0.90–0.98), test–retest
reliability (r = 0.78–0.92), and interrater relia-
bility (r = 0.87 for young children).
Before the clinic visit, the parent is mailed
the consent form to review and several self-
administered forms to complete, including the
Aberrant Behavior Checklist, a standardized
checklist constructed to rate inappropriate and
maladaptive behaviors in developmentally
delayed individuals (Aman and Singh 1994);
Multiple Language Questionnaire to deter-
mine what languages are used at home; Child
Development Questionnaire (CDQ), consist-
ing of 31 questions regarding acquisition and
loss of language and skills, a subset of the
Early Development Questionnaire (Ozonoff
et al. 2005) to examine loss of developmental
skills; and structured questionnaires about
gastrointestinal symptoms and sleep habits of
the child (developed de novo). Parents are also
sent a list of autoimmune diseases with a
description of each, so that they can prepare to
respond to questions about family history of
these disorders during the clinic visit. All
instruments and forms are administered in
either English or Spanish, depending on the
language in which the parent or child feels
most comfortable. The CHARGE study
employs trained bilingual/bicultural staff for
every phase of the study.
At the clinic, the psychometric assessments
are administered, a family medical history with
an emphasis on mental health and autoim-
mune disorders is taken, and a family charac-
teristics questionnaire is used to document
developmental and other aspects of the broader
phenotype in immediate family members.
Hertz-Picciotto et al.
1122
VOLUME 114 | NUMBER 7 | July 2006
Environmental Health Perspectives
Table 2. Data collection protocol for CHARGE study: three developmental groups of children.
Instruments administered Administered to AU, DD, and GP children (except where noted)
In clinic
ADOS (Lord et al. 2000) AU only
ADI-R (Le Couteur et al. 2003) AU only
MSEL (Mullen 1995)
VABS (Sparrow et al. 1984)
SCQ (Rutter et al. 2003) DD or GP only
Child’s medical history
Family autoimmune history
Family medical history
Physical, neurological, and dysmorphology exams
CDQ
Family early developmental characteristics
Self-administered questionnaires completed at home
Aberrant Behavior Checklist (Aman and Singh 1994)
Multiple language questionnaire
Gastrointestinal disorders survey
Sleep history survey
Telephone-administered exposure questionnaire
Abbreviations: AU, autism; GP, general population.
Page 4
Physical and neurologic exams are completed;
dysmorphology and growth or neurologic
abnormalities are recorded. Finally, blood speci-
mens are collected at the end of the clinic visit.
The parent is asked to bring in urine specimens
for the child and immediate family members.
For families of children recruited from the
nonautistic groups, the protocol is essentially
identical, except that the ADI-R and ADOS
are not routinely administered. The Social
Communication Questionnaire (SCQ) was
developed from the ADI-R to screen children
for evidence of features of ASDs. If the score
on the SCQ is above 15, the ADI-R and
ADOS are administered on a second visit.
Final autism case status is defined as
meeting criteria on the communication,
social, and repetitive behavior domains of the
ADI-R and scoring at or above the total cut-
off for autistic disorder on the ADOS module
1 or 2. Analyses will be conducted for cases
meeting criteria for autistic disorder, as well as
for a broader definition of impairment
encompassing ASDs. A similar approach will
be used for mental retardation/DD: Children
obtaining an MSEL composite score of < 69
and a VABS composite score of < 70 will be
classified as meeting strict criteria for DD.
Separate from the clinic visit, we conduct a
telephone interview with the primary caregiver
regarding periconceptional, prenatal, and early
childhood exposures and experiences. The
interview of approximately 1 hr 40 min covers
the following areas: demographics; mother’s
medical history; mother’s reproductive and
pregnancy history; index pregnancy, including
use of reproductive technology for conception;
maternal illnesses and medications during index
pregnancy; metals, diet, and household product
use; child’s illnesses and medications; maternal
lifestyle information; residential history; and
occupational history of the mother and father.
An index time period is defined as 3 months
before pregnancy to the end of pregnancy or, if
the child was breast-fed, until weaning.
Information on medications, metals, household
products, and the occupational and residential
histories focuses on this index period.
Blood and urine specimens are collected
from the index child, parents, and siblings.
For any family member from whom blood is
not obtained, an attempt is made to collect
buccal swabs for DNA extraction. Hair speci-
mens are collected from the index child and
from the mother if her hair is long enough to
potentially contain information about expo-
sures during the pregnancy or lactation period.
If the parent saved locks from the child’s first
haircut, we request a few strands. Additionally,
neonatal blood spots from the index child will
be obtained from the newborn screening
specimen archive maintained by the Genetic
Disease Branch of the California Department
of Health Services (Richmond, CA).
Medical records are procured and
abstracted for information about procedures,
medications and other treatments, and condi-
tions at birth of the index child. Obstetric/
gynecology/prenatal clinic and mental health
provider records are obtained for the mother.
Similarly, labor and delivery, neonatal, pedi-
atric, and specialty clinic medical records are
procured. Dental records are sought for the
confirmation of mercury amalgams.
The study complies with all applicable
requirements regarding human subjects and is
approved by the institutional review boards
for the State of California and the University
of California. Informed consent is obtained
before collection of any data.
Preliminary data on participants and
their diagnoses. Full recruitment into the
CHARGE study began in late 2003. More
than 520 children and their families have
enrolled in the CHARGE study at the time of
this writing. This includes > 360 recruited
because of an RC diagnosis of autism, > 50
with an RC diagnosis of DD (recruitment
began later for this group), and > 120 from
the general population. By the end of the first
5 years of funding, we expect to have a total
of approximately 650–700 children enrolled.
Among contacted families of children with
autism, 20% were ineligible, 22% refused,
and 58% agreed to participate. Among gen-
eral population families with whom we made
contact, 22% were ineligible, 41% refused,
and 36% agreed to join the study.
Among children with a diagnosis of autism
recruited from RCs, after assessment by
CHARGE study personnel, 64% met criteria
on both the ADOS and ADI-R. Among those
3 or 4 years of age who are California DDS eli-
gible based on their diagnosis of autism, 64%
meet criteria on both instruments, another 9%
meet criteria on the ADOS alone, and a fur-
ther 14% on the ADI-R alone, for a total of
87%. Additionally, among the remainder, 6%
meet criteria for ASD based on both examina-
tions (scores at least 7 on ADOS module 1 or
at least 8 on ADOS module 2, and meets cut-
off in ADI-R for section D and either section
A or B, and falls within 2 points on the other
section of A or B); another 5% meet criteria for
ASD based on either ADI-R alone or ADOS
alone. Fewer than 2% would not be classified
as being on the spectrum.
Among those recruited through RCs with
a diagnosis of DD, the percentage that
showed delay in both adaptive and cognitive
domains was 64%, with another 6% that met
the cutoff on at least one of the tests. Among
those who entered the study with a diagnosis
of DD, 3% met criteria for autism and
another 8% met criteria for ASD.
Another phenotypic distinction we are
investigating is early-onset versus regressive
autism, as defined by the language and social
regression questions on the ADI-R and the
CDQ. Using a broad definition of regression
that includes loss of previously attained lan-
guage and/or social skills (Ozonoff et al.
2005), close to 50% of the CHARGE chil-
dren with a confirmed diagnosis of autism
had regression.
Finally, Table 3 provides basic demo-
graphic information about the CHARGE
study sample, based on data from the birth
certificate. This table also provides compari-
son information about the pool of births from
which we recruit the general population con-
trols. Compared with this pool, mothers who
participate are older, more highly educated,
and more likely to have private health insur-
ance. Participant mothers of general popula-
tion controls are also more likely to have been
born in the United States. The children were
more likely to be twins. In further work, the
autistic and DD participants will be com-
pared with their respective pools.
Community partnership. A community
advisory council (CAC) was formed early in
the development of this project to maximize
participation in the research by parents,
clinicians, service providers, advocacy organi-
zations, and RC and DDS staff. Parental
Childhood autism: environment and genetics
Environmental Health Perspectives
VOLUME 114 | NUMBER 7 | July 2006
1123
Table 3. Demographics in CHARGE study (%).
CHARGE study participants
AU (
n
= 341) DD (
n
= 54) GP
a
(
n
= 101)
GP pool (
n
= 1,240)
Nonsingletons 6.2 0 3.0 1.6
Mother’s age 35 years at delivery 25.5 18.5 28.7 16.0
Mother’s education < 12 years 6.8 14.8 12.1 29.8
Mother’s education 16 years 41.8 27.8 41.4 23.1
Mother born in United States 72.4 68.5 70.3 54.5
Mother born in Mexico 10.3 25.9 14.9 24.1
Mother born outside
United States and Mexico 17.3 5.6 14.9 21.4
Payment method for delivery
Public 17.6 37.0 19.8 42.9
Private 82.4 63.0 80.2 57.1
Male child
b
88.0 66.7 83.2 79.4
Abbreviations: AU, autism; GP, general population.
a
From birth certificates; pool consists of a stratified random sample selected to have 80% boys, to match the overall age
distribution of the autism cases, and from the same geographic catchment area as the other two groups.
b
The general
population pool was selected with odds of 4:1 male-to-female ratio.
Page 5
suggestions regarding the collection of speci-
mens and information from younger siblings
of affected children were incorporated into
the study design. The CAC meets regularly to
hear updates on study progress and provide
input. CAC members have given critical
advice on data collection instruments, ways to
make the clinical protocol as child-friendly
and special-needs–friendly as possible, and
strategies to enhance recruitment.
Discussion
The CHARGE study is building an infrastruc-
ture that will support multiple investigations
of autism and related neurodevelopmental dis-
orders. The psychometric evaluations and
clinical examinations combined with extensive
exposure information and biologic specimens
represent rich resources for research on etiol-
ogy and phenotypic expression of these disor-
ders and make possible the comprehensive
approach needed to advance understanding of
autism and DD. In our clinical assessments of
> 300 children identified with autism in the
California DDS system, we have confirmed
the diagnosis in 87%, suggesting that the large
increases in DDS system clients with autism
over the last decade or two is unlikely to be
due to overdiagnosis in younger cohorts.
Although several large birth cohort studies
recently initiated or in progress will be able to
examine factors that predict autism, the num-
ber of cases of autism in the CHARGE study
may be comparable with what is expected in
birth cohorts of 100,000 (i.e., we have enrolled
> 360 children with autism and are continuing
recruitment). In contrast with large cohort
studies with dispersed populations, we are able
to confirm diagnoses using standardized instru-
ments administered by a small, well-trained
clinical assessment team. Additionally, in
cohort studies attempting to address a wide
range of health and developmental outcomes,
the exposures and factors measured will not
necessarily have been chosen for relevance
to autism.
The specimen bank is currently being used
by several laboratories that are part of the UC
Davis CCEH. In this first stage, xenobiotic
and biochemical profiles of children with
autism are being compared with those of unaf-
fected children, and comparisons are being
made between different autism phenotypes. As
distinguishing features emerge, the second
stage will be to determine whether any differ-
ences in biomarkers were present at birth,
using the neonatal blood spots where possible.
Data and specimens will be made available to
qualified researchers with targeted, worthwhile
hypotheses not being addressed by CCEH and
CHARGE investigators.
Limitations of this study must be
recognized. Much of the information will be
gathered retrospectively. The only biologic
specimens prospectively collected (i.e., before
diagnosis) are the newborn blood spots and, for
some children, baby hair locks. Similarly, ques-
tionnaires on use of pesticides and other house-
hold products will be retrospective and hence
subject to reporting/recall bias. Thus, the large
birth cohort studies under way or in prepara-
tion will complement the CHARGE study by
providing fully prospective data, although they
are subject to the limitations described above.
Nevertheless, in the CHARGE study, medical
records will yield prospectively recorded data on
treatments, illnesses, and prescription medi-
cations. Other unbiased, relevant sources of
information on xenobiotics include blood
measurements that represent cumulative expo-
sures for persistent compounds and California’s
Pesticide Use Reporting system, which docu-
ments commercial pesticide applications that
can be linked to participant residences during
critical time windows.
Conclusion
Although sporadic studies have identified
specific environmental factors that have been
associated with autism, no previous effort has
attempted to address the broad spectrum of
environmental factors that may, in combina-
tion with genetic susceptibility, affect develop-
ment and severity of this condition in the
population. The CHARGE study is charting
new territory in the investigation of etiologic
factors for autism and DD. The goal of the
CHARGE study is to understand causes of
autism and DD, both genetic and environmen-
tal, in order to reduce their incidence in the
future. The design combines a large popula-
tion-based sample of children with different
patterns of development; standardized diagnos-
tic assessments of autism, cognitive develop-
ment, and adaptive behavior by trained
assessors; medical and neurologic examinations;
detailed reviews of medical records; and an
extensive set of questionnaires describing phe-
notypic characteristics and environmental expo-
sures from preconception through early
childhood. Currently, it is unique in its empha-
sis on environmental factors and its tight link-
age with state-of-the-art laboratories of the UC
Davis CCEH that enable us to address biologic
markers of xenobiotic exposures, immunologic
responses, and gene expression. Other features
include active community involvement, an eth-
nically diverse pool of participants, and inclu-
sion of developmentally delayed children in
addition to general population controls. Finally,
the collaboration by CHARGE study investi-
gators with other population-based autism
epidemiologic efforts currently under way, such
as the national Centers for Autism and
Developmental Disabilities Research and
Epidemiology (CADDRE) study, will create
valuable opportunities for replication and
perhaps data pooling.
REFERENCES
Aman MG, Singh NN. 1994. Aberrant Behavior Checklist—
Community. Supplementary Manual. East Aurora,
NY:Slosson Educational Publications.
American Psychiatric Association. 1994. Diagnostic and
Statistical Manual of Mental Disorders. 4th ed.
Washington, DC:American Psychiatric Association.
Ashwood P, Van de Water J. 2004a. Is autism an autoimmune
disease? Autoimmun Rev 3(7–8):557–562.
Ashwood P, Van de Water J. 2004b. A review of autism and the
immune response. Clin Dev Immunol 11(2):165–174.
Bauman ML, Kemper TL. 2003. The neuropathology of the
autism spectrum disorders: what have we learned?
Novartis Found Symp 251:112–122.
Bettelheim B. 1967. The Empty Fortress: Infantile Autism and
the Birth of the Self. New York:Free Press.
Blaxill MF, Baskin DS, Spitzer WO. 2003. Commentary: Blaxill,
Baskin, and Spitzer on Croen et al. (2002), the changing
prevalence of autism in California. J Autism Dev Disord
33(2):223–226.
Brambilla P, Hardan A, di Nemi SU, Perez J, Soares JC, Barale F.
2003. Brain anatomy and development in autism: review of
structural MRI studies. Brain Res Bull 61(6):557–569.
California Department of Developmental Services. 2003. Autistic
Spectrum Disorders. Changes in the California Caseload,
an Update: 1999 through 2002. Sacramento:Department of
Developmental Services, California Health and Human
Services Agency.
Carter AS, Volkmar FR, Sparrow SS, Wang JJ, Lord C,
Dawson G, et al. 1998. The Vineland Adaptive Behavior
Scales: supplementary norms for individuals with autism.
J Autism Dev Disord 28(4):287–302.
Casanova MF, Buxhoeveden DP, Switala AE, Roy E. 2002.
Minicolumnar pathology in autism. Neurology 58(3):428–432.
Cederlund M, Gillberg C. 2004. One hundred males with
Asperger syndrome: a clinical study of background and
associated factors. Dev Med Child Neurol 46(10):652–660.
Chakrabarti S, Fombonne E. 2005. Pervasive developmental dis-
orders in preschool children: confirmation of high preva-
lence. Am J Psychiatry 162(6):1133–1141.
CHARGE. 2006. Childhood Autism Risk from Genetics and the
Environment. Available: http://beincharge.ucdavis.edu/
[accessed 26 May 2006].
Chess S, Fernandez P, Korn S. 1978. Behavioral consequences
of congenital rubella. J Pediatr 93(4):699–703.
Christianson AL, Chesler N, Kromberg JG. 1994. Fetal valproate
syndrome: clinical and neuro-developmental features in
two sibling pairs. Dev Med Child Neurol 36(4):361–369.
Cook EH Jr. 1998. Genetics of autism. Ment Retard Dev Disabil
Res Rev 4:113–120.
Courchesne E, Yeung-Courchesne R, Press GA, Hesselink JR,
Jernigan TL. 1988. Hypoplasia of cerebellar vermal lobules
VI and VII in autism. N Engl J Med 318(21):1349–1354.
Croen LA, Grether JK, Hoogstrate J, Selvin S. 2002. The chang-
ing prevalence of autism in California. J Autism Dev Disord
32(3):207–215.
Croonenberghs J, Wauters A, Devreese K, Verkerk R, Scharpe S,
Bosmans E, et al. 2002. Increased serum albumin, gamma
globulin, immunoglobulin IgG, and IgG2 and IgG4 in autism.
Psychol Med 32(8):1457–1463.
Dawson G, Carver L, Meltzoff AN, Panagiotides H, McPartland J,
Webb SJ. 2002. Neural correlates of face and object recog-
nition in young children with autism spectrum disorder,
developmental delay, and typical development. Child Dev
73(3):700–717.
Deykin EY, MacMahon B. 1979. Viral exposure and autism. Am
J Epidemiol 109(6):628–638.
Folstein S, Rutter M. 1977. Genetic influences and infantile
autism. Nature 265(5596):726–728.
Folstein SE, Rutter ML. 1988. Autism: familial aggregation and
genetic implications. J Autism Dev Disord 18(1):3–30.
Fombonne E. 1999. The epidemiology of autism: a review.
Psychol Med 29(4):769–786.
Fombonne E. 2003. Epidemiological surveys of autism and other
pervasive developmental disorders: an update. J Autism
Dev Disord 33(4):365–382.
Gillberg C, Wing L. 1999. Autism: not an extremely rare disor-
der. Acta Psychiatr Scand 99(6):399–406.
Glasson EJ, Bower C, Petterson B, de Klerk N, Chaney G,
Hallmayer JF. 2004. Perinatal factors and the development
of autism: a population study. Arch Gen Psychiatry
61(6):618–627.
Hertz-Picciotto et al.
1124
VOLUME 114 | NUMBER 7 | July 2006
Environmental Health Perspectives
Page 6
Childhood autism: environment and genetics
Environmental Health Perspectives
VOLUME 114 | NUMBER 7 | July 2006
1125
Herbert M. 2005. Large brains in autism: the challenge of per-
vasive abnormality. Neuroscientist 11(5):417–440.
Hultman CM, Sparen P, Cnattingius S. 2002. Perinatal risk fac-
tors for infantile autism. Epidemiology 13(4):417–423.
Ingram JL, Peckham SM, Tisdale B, Rodier PM. 2000. Prenatal
exposure of rats to valproic acid reproduces the cerebellar
anomalies associated with autism. Neurotoxicol Teratol
22(3):319–324.
Johansson M, Wentz E, Fernell E, Stromland K, Miller MT,
Gillberg C. 2001. Autistic spectrum disorders in Mobius
sequence: a comprehensive study of 25 individuals. Dev
Med Child Neurol 43(5):338–345.
Jorde LB, Mason-Brothers A, Waldmann R, Ritvo ER, Freeman
BJ, Pingree C, et al. 1990. The UCLA-University of Utah
epidemiologic survey of autism: genealogical analysis of
familial aggregation. Am J Med Genet 36(1):85–88.
Korvatska E, Van de Water J, Anders TF, Gershwin ME. 2002.
Genetic and immunologic considerations in autism.
Neurobiol Dis 9(2):107–125.
Lainhart JE, Ozonoff S, Coon H, Krasny L, Dinh E, Nice J, et al.
2002. Autism, regression, and the broader autism pheno-
type. Am J Med Genet 113(3):231–237.
Le Couteur A, Lord C, Rutter M. 2003. Autism Diagnostic
Interview–Revised (ADI-R). Los Angeles:Western
Psychological Services.
Lord C, Pickles A, McLennan J, Rutter M, Bregman J,
Folstein S, et al. 1997. Diagnosing autism: analyses of data
from the Autism Diagnostic Interview. J Autism Dev Disord
27(5):501–517.
Lord C, Rutter M, DiLavore PC, Risi S. 2000. The Autism Diagnostic
Observation Schedule (ADOS). Los Angeles:Western
Psychological Services.
Lord C, Rutter M, DiLavore PC, Risi S. 2003. Autism Diagnostic
Observation Schedule Manual. Los Angeles:Western
Psychological Services.
Lord C, Rutter M, Le Couteur A. 1994. Autism Diagnostic
Interview-Revised: a revised version of a diagnostic inter-
view for caregivers of individuals with possible pervasive
developmental disorders. J Autism Dev Disord 24(5):659–685.
Markowitz PI. 1983. Autism in a child with congenital cyto-
megalovirus infection. J Autism Dev Disord 13(3):249–253.
Maziade M, Merette C, Cayer M, Roy MA, Szatmari P, Cote R,
et al. 2000. Prolongation of brainstem auditory-evoked
responses in autistic probands and their unaffected rela-
tives. Arch Gen Psychiatry 57(11):1077–1083.
McPartland J, Dawson G, Webb SJ, Panagiotides H, Carver LJ.
2004. Event-related brain potentials reveal anomalies in
temporal processing of faces in autism spectrum disorder.
J Child Psychol Psychiatry 45(7):1235–1245.
Mullen EM. 1995. Mullen Scales of Early Learning. Circle Pines,
MN:American Guidance Services Inc.
Nelson KB, Grether JK, Croen LA, Dambrosia JM, Dickens BF,
Jelliffe LL, et al. 2001. Neuropeptides and neurotrophins in
neonatal blood of children with autism or mental retarda-
tion. Ann Neurol 49(5):597–606.
Ozonoff S, Williams BJ, Gale S, Miller JN. 1999. Autism and
autistic behavior in Joubert syndrome. J Child Neurol
14(10):636–641.
Ozonoff S, Williams BJ, Landa R. 2005. Parental report of the
early development of children with regressive autism: the
delays plus regression phenotype. Autism 9(5):461–486.
Piven J, Berthier ML, Starkstein SE, Nehme E, Pearlson G,
Folstein S. 1990. Magnetic resonance imaging evidence
for a defect of cerebral cortical development in autism.
Am J Psychiatry 147(6):734–739.
Piven J, Palmer P. 1999. Psychiatric disorder and the broad
autism phenotype: evidence from a family study of multiple-
incidence autism families. Am J Psychiatry 156(4):557–563.
Rapin I. 1999. Autism in search of a home in the brain.
Neurology 52(5):902–904.
Rapin I, Dunn M. 2003. Update on the language disorders of indi-
viduals on the autistic spectrum. Brain Dev 25(3):166–172.
Reiss AL, Freund L. 1990. Fragile X syndrome, DSM-III-R, and
autism. J Am Acad Child Adolesc Psychiatry 29(6):885–891.
Ritvo ER, Jorde LB, Mason-Brothers A, Freeman BJ, Pingree C,
Jones MB, et al. 1989. The UCLA-University of Utah epi-
demiologic survey of autism: recurrence risk estimates
and genetic counseling. Am J Psychiatry 146(8):1032–1036.
Ritvo ER, Mason-Brothers A, Freeman BJ, Pingree C, Jenson
WR, McMahon WM, et al. 1990. The UCLA-University of
Utah epidemiologic survey of autism: the etiologic role of
rare diseases. Am J Psychiatry 147(12):1614–1621.
Rodier P, Hyman S. 1998. Early environmental factors in autism.
Ment Retard Dev Disabil Res Rev 4(2):121–128.
Rodier PM, Ingram JL, Tisdale B, Croog VJ. 1997. Linking etiolo-
gies in humans and animal models: studies of autism.
Reprod Toxicol 11(2–3):417–422.
Rosenhall U, Nordin V, Brantberg K, Gillberg C. 2003. Autism and
auditory brain stem responses. Ear Hear 24(3):206–214.
Rutter M, Bailey A, Berument SK, Lord C, Pickles A. 2003.
Social Communication Questionnaire (SCQ). Los
Angeles:Western Psychological Services.
Rutter M, Bartak L. 1971. Causes of infantile autism: some con-
siderations from recent research. J Autism Child Schizophr
1(1):20–32.
Samaco RC, Hogart A, LaSalle JM. 2005. Epigenetic overlap in
autism-spectrum neurodevelopmental disorders: MECP2
deficiency causes reduced expression of UBE3A and
GABRB3. Hum Mol Genet 14(4):483–492.
Schumann CM, Hamstra J, Goodlin-Jones BL, Lotspeich LJ, Kwon
H, Buonocore MH, et al. 2004. The amygdala is enlarged in
children but not adolescents with autism; the hippocampus is
enlarged at all ages. J Neurosci 24(28):6392–6401.
Silva SC, Correia C, Fesel C, Barreto M, Coutinho AM, Marques C,
et al. 2004. Autoantibody repertoires to brain tissue in autism
nuclear families. J Neuroimmunol 152(1–2):176–182.
Smalley SL. 1998. Autism and tuberous sclerosis. J Autism Dev
Disord 28(5):407–414.
Smalley SL, Asarnow RF, Spence MA. 1988. Autism and genetics.
A decade of research. Arch Gen Psychiatry 45(10):953–961.
Sparrow SS, Balla DA, Cicchetti DV. 1984. Vineland Adaptive
Behavior Scales Interview Edition Expanded Form Manual.
Circle Pines, MN:American Guidance Services Inc.
Steffenburg S, Gillberg CL, Steffenburg U, Kyllerman M. 1996.
Autism in Angelman syndrome: a population-based study.
Pediatr Neurol 14(2):131–136.
Stone JL, Merriman B, Cantor RM, Yonan AL, Gilliam TC,
Geschwind DH, et al. 2004. Evidence for sex-specific risk
alleles in autism spectrum disorder. Am J Hum Genet
75(6):1117–1123.
Stubbs EG, Ash E, Williams CP. 1984. Autism and congenital
cytomegalovirus. J Autism Dev Disord 14(2):183–189.
Tang Y, Schapiro MB, Franz DN, Patterson BJ, Hickey FJ,
Schorry EK, et al. 2004. Blood expression profiles for
tuberous sclerosis complex 2, neurofibromatosis type 1,
and Down’s syndrome. Ann Neurol 56(6):808–814.
Vargas DL, Nascimbene C, Krishnan C, Zimmerman AW, Pardo
CA. 2005. Neuroglial activation and neuroinflammation in
the brain of patients with autism. Ann Neurol 57(1):67–81.
Wassink TH, Brzustowicz LM, Bartlett CW, Szatmari P. 2004.
The search for autism disease genes. Ment Retard Dev
Disabil Res Rev 10(4):272–283.
Williams G, King J, Cunningham M, Stephan M, Kerr B, Hersh JH.
2001. Fetal valproate syndrome and autism: additional evi-
dence of an association. Dev Med Child Neurol 43(3):202–206.
Yeargin-Allsopp M, Rice C, Karapurkar T, Doernberg N, Boyle
C, Murphy C. 2003. Prevalence of autism in a US metropoli-
tan area. JAMA 289(1):49–55.
Page 7