- About ASF
- What is Autism?
- How Common is Autism?
- Early Signs of Autism
- Autism Diagnosis
- Following a Diagnosis
- Treatment Options
- Beware of Non-Evidence-Based Treatments
- Autism and Vaccines
- Autism Science
- Quick Facts About Autism
- What We Fund
- Resources for Grantees
- Funding Calendar
- ASF Funded Research
- ASF Supported Findings
- Apply for a Fellowship
- Apply for a Research Accelerator Grant
- Apply for an Undergraduate Summer Research Grant
- Apply for IMFAR Travel Grant
- Get Involved
- Participate in Research
- Student Clubs
- Live Chat with Scientists
- Jobs & Internships
- Apply for a Grant
- Day of Learning & Evening of Celebration
- Contact Us
Research by Topic: Diagnostic Disparities
A study by researchers with the UC Davis MIND Institute studied children with chromosome 22q11.2 deletion syndrome, who as a group have a prevalence of autism between 20 and 50 percent according to parent reports. This study found that these children may be getting misdiagnosed because the symptoms of the chromosome 22q11.2 deletion syndrome, including social impairments, are very similar to symptoms of autism.
With the release of the DSM-5 comes an updated definition of autism and a major change to the way it is diagnosed.
“Why do boys get diagnosed with autism four times as often as girls? New research, including some of the latest data from the International Society for Autism Research annual conference last week, addresses this question, one of the biggest mysteries in the field.”
In effort to stimulate more research to better understand ASD trends, ASF President Alison Singer and other stakeholders discuss the increase in ASD prevalence and share their knowledge and opinions.
The age at which a child with autism is diagnosed is related to the particular suite of behavioral symptoms he or she exhibits, according to this study led by an ASF Grantee. Certain diagnostic features, including poor nonverbal communication and repetitive behaviors, were associated with earlier identification of an autism spectrum disorder. Displaying more behavioral features was also associated with earlier diagnosis.For more information about this study, read the guest blog from the lead author here: http://autismsciencefoundation.wordpress.com/2013/04/09/identifying-asd-in-community-settings/
Researchers at Virginia Tech examine M-CHAT performance in a very low socio-economic status setting and find it lacks internal consistency across ethnic and educational groups. Caregivers who reported a low maternal educational level or with minority status were more likely to mark items suggestive of autism compared to those with higher maternal education or non-minority status
Researchers found that estimates of the number of children with ASD might be lower using the current DSM-5 criteria than using the previous criteria. This study looked at information collected by CDCs Autism and Developmental Disabilities Monitoring (ADDM) Network. This is the first population-based study in the United States to look at what effect the updated ASD criteria in the DSM-5 might have on estimates of the number of children with ASD. One of the advantages of the ADDM Network method is that it does not rely solely on the presence of an ASD diagnosis, but also includes review of records for children who have behaviors consistent with ASDs, even if they do not have a diagnosis. Because of the way the ADDM Network collects data, in the future CDC will be able to use both the previous DSM-IV-TR and the current DSM-5 criteria to estimate the number of children with ASD. CDC will also continue to evaluate the effect of using the DSM-5 on trends in how doctors and other health professionals diagnose ASD and how service providers evaluate and document symptoms as they transition to using the new criteria.
DSM-5 Field Trials in the United States and Canada, Part II: Test-Retest Reliability of Selected Categorical DiagnosesPublished October 30, 2012 in American Journal of Psychiatry
“OBJECTIVE The DSM-5 Field Trials were designed to obtain precise (standard error <0.1) estimates of the intraclass kappa as a measure of the degree to which two clinicians could independently agree on the presence or absence of selected DSM-5 diagnoses when the same patient was interviewed on separate occasions, in clinical settings, and evaluated with usual clinical interview methods.”
Application of DSM-5 Criteria for Autism Spectrum Disorder to Three Samples of Children with DSM-IV Diagnoses of Pervasive Developmental DisordersPublished October 1, 2012 in The American Journal of Psychiatry
“OBJECTIVE: Substantial revisions to the DSM-IV criteria for autism spectrum disorders (ASDs) have been proposed in efforts to increase diagnostic sensitivity and specificity. This study evaluated the proposed DSM-5 criteria for the single diagnostic category of autism spectrum disorder in children with DSM-IV diagnoses of pervasive developmental disorders (PDDs) and non-PDD diagnoses.CONCLUSIONS: These results suggest that most children with DSM-IV PDD diagnoses would remain eligible for an ASD diagnosis under the proposed DSM-5 criteria. Compared with the DSM-IV criteria for Asperger’s disorder and PDD-NOS, the DSM-5 ASD criteria have greater specificity, particularly when abnormalities are evident from both parents and clinical observation.”
“We have followed up a 2002 population study of autism prevalence in 15-24-year olds in the Faroe Islands. The rate of ASD grew significantly from 0.56% in 2002 to 0.94% in 2009. Although these results are within the range of typical findings from other studies, there were some interesting details. There were-in addition to 43 originally diagnosed cases in 2002-24 newly discovered cases in 2009 and nearly half of them were females. It is possible that unfamiliarity with the clinical presentation of autism in females have played a significant role in this context. There was diagnostic stability for the overall category of ASD over time in the group diagnosed in childhood (7-16) years, but considerable variability as regards diagnostic sub-groupings.”
How Different Are Girls and Boys Above and Below the Diagnostic Threshold for Autism Spectrum Disorders?Published August 1, 2012 in Journal of the American Academy of Child and Adolescent Psychiatry
A study finds that despite showing similar autistic traits, girls are less likely than boys to meet diagnostic criteria for ASD if no other intellectual or behavioral issues are present. The authors suggest the results might reflect biased diagnosis or better adaptation in girls.
A brain imaging study out of Boston Children’s Hospital suggests that autism and Asperger’s syndrome are biologically distinct conditions. The study analyzes the patterns of brain connectivity in children with ASD and found that children with autism might have distinct neural signaling patterns. This study follows the release of the APA’s new DSM-5 that classifies Asperger’s under an umbrella diagnosis of ASD.News Article: http://www.medicaldaily.com/aspergers-syndrome-and-autism-are-biologically-distinct-according-new-brain-study-video-249255
“Early identification of autism has become a national priority but, despite efforts, there are children who are being identified at a later age. In this study, children of Hispanic and African American origin, foreign-born children, and children born to foreign mothers were more likely to be diagnosed later.”
Swedo spoke at the American Psychiatric Association’s (APA) annual meeting, in her role as chairperson of the work group developing new diagnostic criteria for neurodevelopmental disorders in DSM-5, the forthcoming fifth edition of the APA’s Diagnostic and Statistical Manual of Mental Disorders.
Evidence shows an increased number of autism diagnoses. There is the possibility that the increase in cases is entirely the result of better detection. Scientists must work to uncover the truth.
A new study in the journal Pediatrics suggests that moms who are obese or have diabetes are more likely to have a child with autism or another developmental problem.
Three teams of scientists working independently to understand the biology of autism have for the first time homed in on several gene mutations that they agree sharply increase the chances that a child will develop the disorder, and have found further evidence that the risk increases with the age of the parents, particularly the father.
For many families, the quest for the causes of autism has grown more urgent with the news that the estimated prevalence of the condition grew by 23% from 2006 to 2008, a Centers for Disease Control and Prevention report said last week.
From Autism Science Foundation(March 29, 2012New York) The Centers for Disease Control in Atlanta today reported that 1 in 88 children (1 in 48 boys) is diagnosed with an autism spectrum disorder (ASD). Previous reports had pegged the number at 1 in 110.
Prevalence of Autism Spectrum Disorders Autism and Developmental Disabilities Monitoring Network, 14 Sites, United States, 2008Published March 29, 2012 in MMWR (CDC)
Full text of today’s CDC report indicating 1 in 88 children is now diagnosed with an autism spectrum disorder.
The Centers for Disease Control in Atlanta today reported that 1 in 88 children (1 in 54 boys) is diagnosed with an autism spectrum disorder (ASD). Previous reports had pegged the number at 1 in 110.
When compared to non-Hispanic white children, Hispanic youth are more likely to go undiagnosed for developmental issues and autism, says a study from UC Davis MIND Institute.
“OBJECTIVE: To determine whether the relationships between behavioral phenotypes and clinical diagnoses of different autism spectrum disorders vary across 12 university-based sites.CONCLUSION: Clinical distinctions among categorical diagnostic subtypes of autism spectrum disorders were not reliable even across sites with well-documented fidelity using standardized diagnostic instruments. Results support the move from existing subgroupings of autism spectrum disorders to dimensional descriptions of core features of social affect and fixated, repetitive behaviors, together with characteristics such as language level and cognitive function.”
Early diagnosis is considered key for autism, but minority children tend to be diagnosed later than white children. Some new work is beginning to try to uncover why and to raise awareness of the warning signs so more parents know they can seek help even for a toddler.
The first prospective study of ethnic differences in the symptoms of autism in toddlers shows that children from a minority background have more delayed language, communication and gross motor skills than Caucasian children with the disorder. Researchers at the Kennedy Krieger Institute concluded that subtle developmental delays may be going unaddressed in minority toddlers until more severe symptoms develop.
Older maternal and paternal age are jointly associated with having a child with autism, according to a recently published study led by researchers at The University of Texas Health Science Center at Houston (UTHealth).
Concerns that proposed changes to autism criteria in the upcoming Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) will exclude many individuals from diagnosis and treatment are unfounded, says the American Psychiatric Association (APA).
Although the genetic basis of autism is now well established, a growing body of research also suggests that environmental factors may play a role. Using a unique study design, a new study suggests that low birth weight is an important environmental factor contributing to the risk of autism spectrum disorder (ASD).
According to the Centers for Disease Control, nearly one percent of U.S. children have some form of autism, 20 times higher than the rate in the 1980s. Alan Zarembo of The Los Angeles Times and clinical psychologist Catherine Lord discuss what’s behind the growing number of diagnoses.
As more children are diagnosed with autism, researchers are trying to find unrecognized cases of the disorder in adults. The search for the missing millions is just beginning.
Children exposed to the epilepsy drug valproate have a nearly three times higher risk of having an autism spectrum disorder, new research finds.
A new study published in the journal Archives of General Psychiatry suggests that the current gold standard of “best-estimate clinical diagnoses” for the diagnosis of autism spectrum disorders may not be the best method of diagnosis.
University of Missouri researchers have found distinct differences between the facial characteristics of children with autism compared to those of typically developing children.Published October 22, 2011 in Medical News Today
The face and brain develop in coordination, with each influencing the other, beginning in the embryo and continuing through adolescence. Now, University of Missouri researchers have found distinct differences between the facial characteristics of children with autism compared to those of typically developing children…
Having A Child With Autism Linked To Genetic Variant And Autoantibodies: Finding May Lead To Screening TestPublished October 20, 2011 in Medical News Today
A study by researchers at UC Davis has found that pregnant women with a particular gene variation are more likely to produce autoantibodies to the brains of their developing fetuses and that the children of these mothers are at greater risk of later being diagnosed with autism.
Autism is far more common in low-birth-weight babies than the general population, researchers are reporting, a significant finding that nevertheless raises more questions than it answers and illustrates how little is known about a group of disorders that affect nearly 1 percent of American children.
New findings from a 16-year study confirm that the Diagnostic and Statistic Manual of Mental Disorders (DSM), the gold-standard for the classification of mental health conditions, can be used to accurately identify autism spectrum disorders (ASD) in children with Down syndrome, according to research from Kennedy Krieger Institute.
This study estimated the ASD prevalence in a psychiatric hospital and evaluated the Social Responsiveness Scale (SRS) combined with other information for differential diagnosis. Chart review, SRS and clinical interviews were collected for 141 patients at one hospital. Diagnosis was determined at case conference. Receiver operating characteristic (ROC) curves were used to evaluate the SRS as a screening instrument. Chi-squared Automatic Interaction Detector (CHAID) analysis estimated the role of other variables, in combination with the SRS, in separating cases and non-cases. Ten percent of the sample had ASD. More than other patients, their onset was prior to 12 years of age, they had gait problems and intellectual disability, and were less likely to have a history of criminal involvement or substance abuse. Sensitivity (0.86) and specificity (0.60) of the SRS were maximized at a score of 84. Adding age of onset
Researchers at the University of Southern California examined the association between autism and proximity of residence to freeways and major roadways during pregnancy and near the time of delivery, as a surrogate for air pollution exposure. Using the mother’s address recorded on the birth certificate and trimester-specific addresses derived from a residential history, measures of […]
A long-awaited study of autism prevalence in Korea came out today in the The American Journal of Psychiatry. Results showed a much higher prevalence estimate than previously found, along with a large fraction of the autistic students previously unidentified and being educated in regular education programs.
An examination of the birth records of the more than 7 million children born in the state of California during the 1990s and early 2000s has found a clear link between the month in which a child is conceived and the risk of that child later receiving a diagnosis of autism. Among the children included in the study, those conceived during winter had a significantly greater risk of autism, the study found.
Dr Brugha, who is also a consultant psychiatrist working in the NHS with the Leicestershire Partnership NHS Trust, said none of the cases with autism found in the community survey throughout England knew that they were autistic or had received an official diagnosis of autism or asperger syndrome.
While there is an increasing equality in terms of the likelihood that children from communities and families across the socioeconomic spectrum will be diagnosed with autism, a new study finds that such factors still influence the chance of an autism diagnosis, though to a much lesser extent than they did at the height of rising prevalence.
A study conducted in 1998 found that autism occurred in 0.05 per cent of Norwegian children. The figures from the “Barn i Bergen” project could therefore be interpreted to mean that the incidence of autism has risen dramatically. However, Ms Posserud thinks it is important to downplay the difference in results. Her conclusion is that the rise in ASD can be explained mainly by the use of more thorough mapping methods and, consequently, that we are not seeing the emergence of an autism epidemic
George Washington University researchers have found that male and female sex hormones regulate expression of an important gene in neuronal cell culture through a mechanism that could explain not only higher levels of testosterone observed in some individuals with autism, but also why males have a higher incidence of autism than females. The gene, RORA, […]
Medical journalist recaps his struggle to determine if vaccinations caused his son to be diagnosed with Asperger’s Syndrome concluding that it would not have made the slightest bit of difference if we had refused to vaccinate when our son was small, claiming any conspiracy theories are not based on any compelling data.
This study, performed by researchers at the University of Pennsylvania, analyzed the level of gene expression in children with autism, compared with a control group. The researchers hypothesized that the variability in the pattern of the overall of gene expression levels would be associated with variability in hippocampal-dependent behaviors, which include short-term memory and spatial […]
A second child is three times more likely to be diagnosed with autism if they are born within twelve months of their siblings, compared to those born three or more years apart, researchers from the Lazarsfeld Center for the Social Sciences at Columbia University, New York revealed in the journal Pediatrics. The investigators gathered information on 660,000 second children born in California between 1992 to 2002.
This study examined the rates of autism according to maternal immigrant status and ethnic origins based on the vitamin D insufficiency hypothesis, which proposes that maternal vitamin D insufficiency during pregnancy could be associated with autism. The study provided further support to the association between maternal immigrant status and an increased risk of autism. In […]
Describing the Brain in Autism in Five Dimensions-Magnetic Resonance Imaging Assisted Diagnosis of Autism Spectrum Disorder Using a Multiparameter Classification ApproachPublished December 1, 2010 in Journal of Neuroscience, Ecker et al
The study tested a group of 20 high functioning adults with autism, together with 20 control adults, to determine whether MRI scans can detect autism. Using left hemisphere cortical thickness, the algorithm could achieve 90% accuracy, however the right hemisphere was worse at differentiating between the two groups. The study shows that it is feasible […]
Researchers have uncovered a prominent genetic risk factor for autism spectrum disorders and schizophrenia is a small genomic deletion. Remarkably, they found the same deletion on chromosome 17 in 24 separate patients. This CNV was absent in 52,448 controls, making the finding statistically significant. Someone with this deletion is 13.58 times more likely to develop ASD or schizophrenia than is someone lacking this CNV. This gene mutation is also known to cause kidney disease (renal cysts and diabetes syndrome, RCAD).
Studies have suggested associations between immune response and idiopathic disorders (such as autism). This study explores associations between parental autoimmune disorders and children's diagnosis of autism by linking. The study found associations between parental autoimmune disorders and autism spectrum disorders, suggesting parental autoimmune disorders may represent a pathway that warrants more detailed investigation.
Full-term neonates with jaundice are at greatly increased risk of later being diagnosed with a disorder of psychological development, a Danish study found. Neonatal jaundice typically is caused by increased bilirubin production and inadequate liver excretory function. Recent research has suggested that even moderate bilirubin exposure in very young children can be harmful, possibly leading to impairments in their development. They found that jaundice was more common among boys, infants born preterm, infants with congenital malformations, and low-birthweight infants.
A new study, led by researchers at Washington University School of Medicine in St. Louis, found mild traits, not strong enough to provoke a diagnosis of autism, seem to be present in the siblings of affected children at significantly higher rates than seen in the general population. Siblings of children with autism have more frequent language delays and other subtle characteristics of the disorder than previously understood. Girls also may be mildly affected more often than recognized in the past.
If a boy’s X-chromosome is missing the PTCHD1 gene or other nearby DNA sequences, they will be at high risk of developing ASD or intellectual disability. Girls are different in that, even if they are missing one PTCHD1 gene, by nature they always carry a second X-chromosome, shielding them from ASD.
Study sought to describe autism spectrum disorder (ASD) population characteristics and changes in identified prevalence across 3 time periods. Children with a potential ASD were identified through records abstraction at multiple sources with clinician review based on Diagnostic and Statistical Manual (DSM-IV-TR) criteria. Multisite, population-based data from the Autism and Developmental Disabilities Monitoring (ADDM) Network […]
The first “test tube baby” was born in 1978. With advances in reproductive science, an estimated one percent of all American babies are now born each year through in vitro fertilization (IVF). But IVF and other assisted fertility treatments may be solving one problem by creating another, suggests new evidence from Tel Aviv University.
A study, conducted by a team at the Harvard School of Public Health, found that autism was nearly twice as common among the children of women who were treated with the ovulation-inducing drug Clomid and other similar drugs than women who did not suffer from infertility, and the link persisted even after researchers accounted for the women’s age. Moreover, the association between fertility drugs and autism appeared to strengthen with exposure: the longer women reported being treated for infertility, the higher the chances their child had an autism spectrum disorder (ASD).
Infections during infancy or childhood do not seem to raise the risk of autism, new research finds. The study found that children who were admitted to the hospital for an infectious disease, either bacterial or viral, were more likely to receive a diagnosis of autism spectrum disorder. However, children admitted to the hospital for non-infectious diseases were also more likely to be diagnosed with autism than kids who were never hospitalized, the study found.
Significant advances in the neonatal intensive care have resulted in increased survival rates of children who are born at less than 26 weeks of gestation, so termed “extremely preterm children.” Notably, however, improved survival rates have been accompanied by a higher risk for later cognitive, neuromotor, and sensory impairments in these children.
Children with autism whose social and communications skills regress around age 3 tend to have more severe autism than children who show signs of the neurodevelopmental disorder at younger ages, new research finds.
A new study by the Kennedy Krieger Institute suggests that the long-term outcome of autism disorders is linked to when and how symptoms first appear. Surprisingly, researchers discovered children with early developmental warning signs may actually be at lower risk for poor outcomes than children with less delayed early development who experience a loss or plateau in skills.
There has been a major increase in the incidence of autism over the last twenty years. While people have differing opinions as to why this is (environment, vaccines, mother’s age, better diagnostic practice, more awareness etc.) there are still many children who have autistic traits that are never diagnosed clinically. Therefore, they do not receive the support they need through educational or health services.
The American Psychiatric Association, with its release this week of proposed revisions to its authoritative Diagnostic and Statistical Manual of Mental Disorders, is recommending that Aspergers be dropped. If this revision is adopted, the condition will be folded into the category of autism spectrum disorder, which will no longer contain any categories for distinct subtypes of autism like Aspergers and pervasive developmental disorder not otherwise specified (a category for children with some traits of autism but not enough to warrant a diagnosis).The change is welcome, because careful study of people with Aspergers has demonstrated that the diagnosis is misleading and invalid, and there are clear benefits to understanding autism as one condition that runs along a spectrum.
A study of the development of autism in infants, comparing the behavior of the siblings of children diagnosed with autism to that of babies developing normally, has found that the nascent symptoms of the condition — a lack of shared eye contact, smiling and communicative babbling — are not present at 6 months, but emerge gradually and only become apparent during the latter part of the first year of life.
Advanced maternal age is linked to a significantly elevated risk of having a child with autism, regardless of the father’s age, according to an exhaustive study of all births in California during the 1990s by UC Davis Health System researchers. Advanced paternal age is associated with elevated autism risk only when the father is older and the mother is under 30, the study found.
California-based studies suggest that local environmental or social factors are driving the high autism-diagnosis rates. And they conclude that childhood vaccinationswhich some people fear is a factor behind rising autismare not to blame. Otherwise, diagnoses of the disorder would be more evenly dispersed, they say.
Researchers have known for years that when one identical twin has autism, the other is also likely to be diagnosed with it – evidence that autism likely has a genetic component. Recent studies support that theory. Researchers at Kennedy Krieger Institute studied 277 pairs of twins and found that when one identical twin had the disorder, the other developed it 88 percent of the time; for fraternal twins, that figure was 31 percent.
The Childhood Autism Risk from Genetics and the Environment (CHARGE) Study enrolled children 2-5 years of age. After diagnostic evaluation, they analyzed three groups: AU/ASD, non-AU/ASD with developmental delay (DD), and population-based TD controls. Mothers were interviewed about household, medical, and dietary exposures. Blood Hg was measured by inductively coupled plasma mass spectrometry. Multiple linear […]
Nobody has been able to show consistent differences between what clinicians diagnose as Aspergers syndrome and what they diagnose as mild autistic disorder. Thus, there is a proposal to remove Aspergers syndrome from the next edition of psychiatrys diagnostic manual. But the change, if approved by the manuals editors and consultants, is likely to be controversial. The Aspergers diagnosis is used by health insurers, researchers, state agencies and schools not to mention people with the disorder, many of whom proudly call themselves Aspies.
A new study provides more proof that childhood vaccines with mercury as a preservative — no longer on the market — did not cause autism. It found that the number of autism cases continued to rise through that period even though the preservative thimerosal — nearly half of which is made of ethylmercury — was removed from most vaccines in 2001.
About 1 in 100 8-year-old children in the U.S. have been diagnosed with autism spectrum disorder, according to U.S. Centers for Disease Control and Prevention researchers who will be releasing details of their study later this year. The rate — significantly higher than the government’s 2007 estimate of 1 in 150.
Reviewing a larger population than in any other study of its kind, the California Department of Public Health (CDPH) has found that as parents age their risk of giving birth to a child with autism increases modestly. Published in the American Journal of Epidemiology, the new CDPH study shows that for each 10-year increase in a mothers age, the risk of autism increased by about 38 percent. For each 10-year increase in a fathers age, the risk of autism increased by about 22 percent.
On Sept. 22, England’s National Health Service (NHS) released the first study of autism in the general adult population. The findings confirm the intuitive assumption: that ASD is just as common in adults as it is in children. Researchers at the University of Leicester, working with the NHS Information Center found that roughly 1 in 100 adults are on the spectrum the same rate found for children in England, Japan, Canada and, for that matter, New Jersey.
Although Autism Spectrum Disorders (ASD) are generally assumed to be lifelong, we review evidence that between 3% and 25% of children reportedly lose their ASD diagnosis and enter the normal range of cognitive, adaptive and social skills. Predictors of recovery include relatively high intelligence, receptive language, verbal and motor imitation, and motor development, but not […]
Variation in Season of Birth in Singleton and Multiple Births Concordant for Autism Spectrum DisordersPublished December 31, 1969 in Paediatric Pernatal Epidemiology, Lee, Newschaffer
Patterns of seasonal variation in births in some neuropsychiatric conditions have been found in previous research; however, no study to date has examined these disorders for seasonal variation in singletons and multiple births separately. This study aimed to determine whether the birth date distribution for individuals with autism spectrum disorders (ASD), including singletons and multiple […]
Fifty-eight percent of children had a documented autism spectrum disorder. In adjusted analyses, children who were Black (odds ratio [OR] = 0.79; 95% confidence interval [CI] = 0.64, 0.96), Hispanic (OR = 0.76; CI = 0.56, 0.99), or of other race/ethnicity (OR = 0.65; CI = 0.43, 0.97) were less likely than were White children […]
Continuing Increases in Autism Reported to Californias Developmental Services System: Mercury in RetrogradePublished December 31, 1969 in Archives of General Psychiatry, Schechter, Grether
California Department of Developmental Services (DDS) data do not show any recent decrease in autism in California despite the exclusion of more than trace levels of thimerosal from nearly all childhood vaccines. The DDS data do not support the hypothesis that exposure to thimerosal during childhood is a primary cause of autism.