February 9, 2010
Top Ten Autism Research Findings Of 2009
Leading Autism Advocacy Organization Documents Progress to Discover Causes and Treatments for Autism Spectrum Disorders
Autism Speaks, the world's largest autism science and advocacy organization, has released its annual list of the 10 most significant research achievements to have impacted autism during the previous year. Every year, Autism Speaks documents the progress made toward its mission to discover the causes and treatment for autism spectrum disorders, and compiles a list of the 10 most significant research achievements to have impacted autism during the previous year. The 2009 list contains important results from clinical and epidemiological research together with advances in gene discovery and effective treatments which will combine to shape the future of autism research for 2010 and beyond." Compelling autism research comes from across the United States and around the world with funding from a variety of public sources, such as the National Institutes of Health, and private sources such as Autism Speaks," explained Autism Speaks Chief Science Officer Geraldine Dawson, Ph.D. "These findings illustrate that we are indeed making progress toward understanding the causes and effective treatments for autism spectrum disorder. At the same time, we recognize that progress is not nearly fast enough and we need more answers."
"This past year we witnessed several important advances in autism research, from the discovery of autism susceptibility genes to alarming new autism prevalence estimates to novel findings about effective treatments for autism spectrum disorders," said Vice President of Research, Sophia Colamarino, Ph.D. "While a few of the studies selected may not have achieved a large amount of publicity, each advanced how we think about autism in some important way."
With input from Autism Speaks' Scientific Advisory Committee, Autism Speaks science staff culled through thousands of publications to arrive at these choices. The 2009 compilation is designed to be considered in its entirety and is not arranged to suggest a ranking.
1. Autism Epidemiology "“ converging findings show that autism prevalence is now one percent
In 2009, two major studies using different research methodologies yielded strikingly similar and eye-opening results showing that ASD affects approximately one percent of children in the United States. Based on data collected just four years earlier, it was found that ASD affected 1 out of every 150 children in the U.S. The new results represent a 57% increase in ASD prevalence in a relatively short period of time. Both studies also found that ASD continues to be four times more common in boys than girls.
The first study, published in Pediatrics, was authored by U.S. Department of Health and Human Services (HHS) and the CDC and collected data through the National Survey of Children's Health (NSCH) on parent-reported diagnosis of ASD. Among a nationally representative sample of 78,000 children aged 3 to 17 years, the investigators found that 1 in 91, or an estimated 673,000 children in the U.S. had an ASD. While concerns lingered over the parent-reported nature of the data, this large-scale study set the stage for another major publication on ASD prevalence with similar results. A second study from CDC researchers released new prevalence data collected by the Autism and Developmental Disabilities Monitoring Network (ADDM), a series of surveillance sites throughout the U.S. that maintain medical and service records on children with autism. By abstracting data and subjecting those records to stringent clinical evaluation, the authors found that approximately 1 in 110 children, 1 in 70 boys, met the criteria for ASD. This 1 in 110 statistic, based on data collected in 2006, represents a 57 percent increase from the 1 in 150 statistic which was based on data collected by the ADDM network in 2002 using identical research methods to the current study.
Through 2009, a number of publications on autism epidemiology sought to investigate the reasons for the dramatic increase witnessed in autism prevalence. Researchers from Columbia University reported that approximately 25 percent of the rise in autism caseload in California between 1992 and 2005 could be directly attributed to changes in diagnostic criteria that resulted in a shift from mental retardation diagnosis to autism diagnosis. Thus, converging evidence from this study and others around the world suggests that while changes in diagnostic practice may account for a portion of the increase, they cannot alone explain the rise in autism prevalence, and that other factors, including environmental factors, likely play a role.
According to Dr. Colamarino, "The prevalence studies of 2009 helped shed additional light on the immense nature of the autism public health crisis. With one percent of the U.S. population affected by ASD, and emerging data suggesting that one percent of the global population may be affected by ASD, not only is there tremendous need for funding to support research into the causes and treatments of ASD, these findings call attention to the necessity for more accessible diagnostic and intervention services for the growing population of those affected."
2. Early Intervention for Toddlers with ASD "“ first randomized clinical trial of early intervention for toddlers shown to be effective for improving cognition, language and adaptive behavior
Although previous studies have found that early intervention can be helpful for preschool-aged children, interventions for children who are toddlers are just now being tested. As 2009 came to a close, results were published in Pediatrics from the first controlled study of an intensive early intervention appropriate for children with ASD who are less than 2Ã½ years of age, showing that a novel early intervention program was effective for improving IQ, language ability, and adaptive behavior in children as young as 18 months.
The intervention, called the Early Start Denver Model, combines applied behavioral analysis (ABA) teaching methods with developmental "Ërelationship-based' approaches, blending the rigor of ABA with play-based routines that focus on building a relationship with the child. Children in the study were separated into two groups, one that received 20 hours a week of the intervention "“ two two-hour sessions five days a week "“ from University of Washington specialists. They also received five hours a week of parent-delivered therapy. Children in the second group were referred to community-based programs for therapy. Researchers closely monitored the progress of both groups.
At the beginning of the study there was no difference in functioning between the two groups. At the conclusion of the study, the IQs of the children in the intervention group had improved by an average of close to 18 points, compared to only seven points in the comparison group. The intervention group also had a nearly 19-point improvement in receptive language (listening and understanding) compared to approximately 10 points in the control group. Whereas only one child in the community-based intervention group had an improved diagnosis, seven of the children in the intervention group had enough improvement in overall skills to warrant a change in diagnosis from autism to the milder condition known as pervasive developmental disorder not otherwise specified (PDD-NOS).
While the youngest children in the study were 18 months old, this particular intervention is designed to be appropriate for children with ASD as young as 12 months of age. Given that the American Academy of Pediatrics recommends that all 18- and 24- month-old children be screened for ASD, it is crucial that we are able to offer parents effective therapies for children within this age range. "This study strongly affirms the positive outcomes of early intervention and the need for the earliest possible start," concluded Dr. Colamarino.
3. The First Successful Genome-Wide Association Studies for Autism "“ new technologies lead to discovery of new autism genes involved in brain development
Advances in technology and analytical methods over the past several years have enabled a better understanding of genetic risk factors for ASD. New methods, called Genome-Wide Association Studies (GWAS), have now made it possible to perform comparisons of populations of individuals affected by a condition versus a non-affected group and identify single changes in DNA nucleotides as specific genetic risk factors. In Spring and Fall of 2009, autism researchers reported the first successful application of GWAS technology to ASD.
GWAS is a powerful analysis technique that allows researchers to sift through hundreds of millions of bits of genetic data to identify changes to the genetic code that are associated with a disease. Because the approach is not based on any specific biological hypothesis, scientists can cast the broadest experimental net possible, and use sophisticated statistical methods to establish the disease association. In April 2009, a large team of scientists led by investigators at Children's Hospital of Philadelphia, reported results from the first successful GWAS study in autism. Because tens of thousands of DNA samples are required for GWAS to produce meaningful results, the researchers collaborated with members of the Autism Genome Project, and pooled samples from the Autism Speaks-funded Autism Genetic Resource Exchange (AGRE) combined with many other collections. The result was identification of a DNA variant associated with the genes cadherin 10 and 9, which are responsible for creating molecules that facilitate the formation of neuronal connectivity. This finding is consistent with accumulating evidence suggesting abnormal interactions between neurons may be at the core of the deficits seen in autism.
The idea that faulty connections between neurons plays a major role in ASD was further supported with the publication of the second autism GWAS study in October. Also working with AGRE and members of the Autism Speaks-funded Autism Genome Project, a collaboration led by investigators from Boston's Autism Consortium and Johns Hopkins University used a very different statistical approach to discover an association between ASD and the gene semaphorin-5A. Similar to the cadherins identified in the first study, semaphorin 5A is thought to play an important role in neural development.
These two groundbreaking studies confirm the potential for GWAS to make successful contributions to our understanding of autism genetics. "Remarkably, out of the approximately 20,000 different human genes the experiments could have identified, the genetic variations that were uncovered are genes involved in brain development, serving to expand and reinforce our current thinking about the biological mechanisms underlying autism," explained Dr. Colamarino.
4. Studies of Genetic Copy Number Variations Reveal New Biological Pathways "“ the ubiquitin pathway is linked to autism for the first time
One of the newest different types of genetic variation to be appreciated are submicroscopic DNA deletions or duplications called copy number variants (CNV). In a companion paper to the GWAS study published in Nature in April 2009, researchers led by Children's Hospital of Philadelphia scientists took advantage of their expansive set of genetic data to conduct additional analyses that would explore CNVs in the autism genome. Along with revealing genes involved in biological pathways previously connected to ASD, their results also surprisingly implicated a new cellular pathway, known as the 'ubiquitin pathway,' in the pathology of autism.
Genes are the instructions that create proteins. Proteins are in turn the functional end of our biology. By studying autism genetics, researchers are not only uncovering factors that can be used to assess the risk of developing autism, they are simultaneously unlocking the mystery of what biological pathways underlie ASD. To date, the majority of autism-associated CNVs have been found to lie in genes whose proteins help neurons adhere to each other and establish proper connections. Results of this newest and largest CNV study continued to confirm this association, finding several CNV in neuronal adhesion and synapse-related genes.
The second major class of genes in which the investigators uncovered CNV is the ubiquitin pathway. The ubiquitin system is primarily involved with altering protein function and disposing of unused molecules within cells. At first glance, the report of CNVs associated with the ubiquitin pathway was somewhat surprising, even if one of the genes in the pathway, UBE3A, has previously been linked to ASD because of its involvement in Angelman Syndrome. However, the authors note that one of the major roles of the ubiquitin pathway is to regulate the turnover of synaptic components, especially those related to plasticity and learning, including the cell-adhesion molecules identified by both their GWAS and CNV analyses. Thus, the authors speculate that the two different types of gene networks their CNV analyses identified may actually be functionally related.
"Taken together, the genetic findings of 2009 are helping to refine our understanding of autism," explained Dr. Colamarino. "Not only is the discovery of CNV in genes that govern the ubiquitin pathway a new finding in autism, the fact that this can be functionally related to other autism genetic discoveries, specifically through a role in synapse activity, suggests that many of these seemingly disparate and individually rare mutations are converging at the level of biology." Identification of the various different types of genetic mutations may define different subclasses of autism and reveal individual opportunities for future therapeutic development, as well as suggest the possibility of developing a more general approach to treatment by focusing on the common biological pathways.
5. Combined Therapies Hold Promise for More Effective Treatments "“ medication and parent training are more effective for reducing serious behavioral problems in children with autism than medication alone
Large-scale autism studies have sought FDA approval for drugs that target core or associated symptoms for autism, however few have proven successful. In 2009, a paper in the Journal of the American Academy of Child and Adolescent Psychiatry demonstrated the first successful randomized controlled trial for a combined treatment approach for ASD, showing that combined pharmacological and behavioral treatment was more effective than pharmacological treatment alone for reducing challenging behaviors.
A new multi-site study by the Research Units on Pediatric Psychopharmacology Autism Network, the same group that conducted the pivotal studies leading to the approval of risperidone for reducing aggression and irritability in children and adolescents with autism, investigated whether combining risperidone treatments with a simultaneous behavioral intervention would be more effective than medication alone. Their 24-week study of 124 children ages 4-13 compared a treatment regime of risperidone alone with a combined treatment regimen of risperidone and a parent training program that followed the principles of applied behavioral analysis. While both the combined and medication-only treatments reduced the severity of non-compliant behaviors, the combined therapy resulted in a significantly greater reduction while using lower doses of risperidone. The combined therapy was also better at reducing other challenging behaviors, such as irritability and hyperactivity.
"This study provides hope for a wider range of available treatments and greater flexibility for clinicians who should be encouraged to use combined approaches in cases where medications or behavioral interventions are not effective on their own," explained Dr. Colamarino. Confirming the effectiveness of coordinated treatments that take full advantage of the benefits of both pharmaceutical and behavioral approaches also demonstrates the continued need to support research establishing the most effective treatments in all realms. "The vast majority of clinical trials conducted to date have only addressed how an individual treatment compares to a placebo. Very few studies have been conducted that make head-to-head comparisons of two or more treatments as was done here, so the success of this trial will also serve to highlight the utility of "comparative effectiveness trials" for determining the best treatments for ASD," she concluded.
6. Genetic Findings Lead to a New Mouse Model of Autism "“ mice carrying a mutation in a gene linked to autism display some subtle and specific features of the disorder
In the most recent wave of autism genetic studies, the gene neurexin-1alpha has been linked to ASD perhaps more often than any other, drawing attention to our need to better understand how its biological function relates to ASD. In 2009 researchers completed the first detailed behavioral characterization of mice lacking the neurexin-1alpha gene, discovering analogies to at least one core domain of ASD.
Over the past few years, researchers studying ASD from a variety of angles have honed in on what appears to be a central role for proper synaptic functioning in the biology of ASD. Synapses are the specialized sites of nerve cell communication within the brain. Importantly, it is believed that the ability of synapses to regularly change their communication properties is what underlies learning and memory and other forms of neuroplasticity. Scientists have discovered that several genes responsible for producing molecules that are active at synapses are altered in individuals with ASD. The first synapse genes linked to ASD were the "neuroligins," and in the last two years, no less than eight genetic studies have "“ incredibly "“ also implicated abnormalities in a gene, neurexin-1alpha, that produces binding partners of the neuroligin proteins. Neurexin proteins work in tandem with the neuroligin proteins to govern proper operation of synapses. This provides independent support for the role of synaptic function in ASD, and it also makes researchers fairly confident that understanding the complicated biological role of the neuroligin-neurexin signaling pathway will provide important insight into brain function in ASD.
Researchers from Stanford University and UT Southwestern Medical Center, Dallas studied mice lacking neurexin-1alpha, looking for signs of any ASD-related symptoms. First, focusing on the synapse, they confirmed subtle disruptions in neurotransmission between brain cells that may be expected to impact brain function. Then in examining behaviors of the mice, they found that although the mice appeared fairly typical, they actually had a nearly two-fold increase in stereotyped grooming behaviors (considered a mouse version of human repetitive behaviors), linking the gene to at least one core behavioral feature of ASD. Finally, in a somewhat unexpected result, the researchers tested the animals in several different situations, but found no deficits in social behaviors or anxiety. Fascinatingly, this result contrasts with their earlier work in mice with a specific neuroligin mutation, which showed abnormalities in social behavior but no changes in repetitive behaviors. This could mean that the neurexin-1alpha gene plays a unique role in stereotyped behaviors. Alternatively, the currently generated neurexin-1alpha mouse mutant may not fully mimic the human ASD mutation. All of these are possibilities that future research will now be in position to address.
Dr. Colamarino reflected on how many animal models exist of the genetic disorders that share overlap with ASD (such as Fragile X, Tuberous Sclerosis etc.) and how these models have been responsible for some of the most exciting results to emerge in recent years regarding the treatment of neurodevelopmental disorders. "However the creation of model systems of so-called 'idiopathic' autism, autism that is not related to another known medical disorder, is still in the beginning stages, so the characterization of each novel model adds a significant new tool for the research community." Because the animals display some unusual characteristics that are conceptually similar to individuals with ASD, lessons from such mice stand to improve our ability to understand and treat ASD in humans. They also demonstrate that ASD can be dissociated into individual 'phenotypes' which may be independently modeled and studied. "In this way," explained Dr. Colamarino, "characterization of mouse models with mutations in genes for neuroligins and neurexins, both of which regulate synapses in the brain, highlights how genetic findings can be successfully and immediately leveraged into new autism research opportunities."
7. Mitochondrial Dysfunction, Autistic Regression"¦ and Fever "“ individuals with mitochondrial dysfunction and autism found to have high rate of autistic regression
Mitochondria are responsible for producing most of the energy the body uses for every day metabolic functions. More attention has recently been focused on a potential link between ASD and dysfunctional mitochondria. A study published in 2009 in the Journal of Child Neurology further examined this link, finding that a subgroup of patients with mitochondria disorders may be at increased risk for autistic regression, especially around periods of fever.
Mitochondria are intimately tuned to the environment in which they reside and are built to respond quickly to fluctuations in the state of that environment. To characterize a relationship between mitochondria disorders and ASD, researchers from Atlanta identified a group of 28 children who had been diagnosed with both ASD and mitochondrial disease. The most common clinical observation in children with both ASD and mitochondria disorder was "hypotonia," or muscles with low tone, followed closely by "fatigue with activity." They also found that approximately 60 percent (17 of 28) of these children experienced a regressive form of ASD, a rate of regression that is over two times greater than what is observed in ASD in general. Notably, 12 of those 17 regressions occurred in conjunction with having suffered a fever within a two week period of the regression. However, this regression did not appear to be necessarily linked to vaccinations, as two-thirds of the children that regressed with fever had not received vaccination, and of those who did receive a vaccination, none regressed without also having a fever.
Although a small study, this report illuminates potentially useful new commonalities between children with both an ASD and mitochondria disorder, suggesting that children with mitochondrial disease may be at increased risk for autistic regression and that increased risk may be associated with some fever-response pathway. Although this paper did not establish the temporal relationship between fever and autistic regression, fever-induced regression is a well-known feature of metabolic disorders overall, and the study brings another angle to the already intriguing relationship of fever and autism. In 2008, researchers reported that some children with ASD actually improve around periods of fever, suggesting that subgroups of ASD exist in which the individuals react differently to fever.
"In light of this new data, it is clear we need more research into the body's complex cascade of metabolic and immune actions that accompany fever, how those relate to the biology of autism, and the appropriateness of fever management," commented Dr. Colamarino. "By showing that a subgroup of individuals with mitochondria disorders may be at risk for autistic regression, the publication highlights the continued need for enhanced awareness of the clinical signs of mitochondrial dysfunction as well."
8. Later Language Acquisition in Nonverbal Individuals with Autism "“ contrary to early beliefs, individuals with autism can acquire spoken language after age five
A common belief of many parents and clinicians is that, if a child with ASD has not developed communicative speech by 5 years of age, the prognosis for future development of speech is extremely poor. In 2009 scientists challenged this belief by conducting a comprehensive review of the research literature to search for reports of individuals who were reported to have acquired speech at age 5 or older. One-hundred sixty-seven such cases were identified, changing the way in which we view language development in individuals with ASD.
Early theories of brain development held that the period before age five represents a unique time in development during which language acquisition is possible, a critical period for language. Yet, recent longitudinal neuroimaging research has shown that the brain has a prolonged development, with major changes occurring during adolescence, and we now know that the capacity for neural generation extends even into adulthood. While the field of neuroscience has revised its notions of neuroplasticity and development accordingly, the field of ASD has held onto the notion of an early critical period for language acquisition. This paper published in the Journal of Cognitive and Behavioral Neurology, however, provides a very different perspective.
The authors identified in the published literature 167 individuals with ASD who used speech for the first time after age five. Many of these children had been offered language intervention based on either traditional or naturalistic applied behavior analysis during the elementary school years, with the intensity of intervention ranging from 30 minutes/week to 30 hour/week. Others had been taught sign language or provided with Picture Exchange Communication System (PECS) training, computer-based training, or speech-language therapy. Children who developed phrase speech were found to have been in treatment longer than those who only achieved single word speech. In virtually all cases, significant time and effort put into treatment was necessary for speech to develop. According to their records, many of these children learned to "use phrases," "answer simple questions," "make spontaneous requests," use "complete sentences," and "speak in spontaneous, complex sentences."
Although the age at which speech developed was variable (ranging from 6-12 years), once the child began speaking, subsequent improvement was often quite rapid. This suggests that achieving initial sound production and words can provide an important springboard for the development of subsequent speech. According to Dr. Colamarino, "This important paper offers hope for the many children who have not yet developed speech by age five, dispelling the belief that older individuals with ASD cannot respond well to speech interventions and providing a much more positive prognosis for individuals with ASD."
9. Language Regression in Autism - losing language skills is found to be specific to children with ASD
Similar to ASD, language learning difficulties are associated with children with Specific Language Impairment (SLI), developmental language disorder that is not associated with deafness, autism, or general developmental delay. Overlapping behaviors and challenges have been shown in children with ASD and those with SLI, making people wonder what features of language impairment, if any, are specific to autism. An important new study was published in 2009 investigating language development in children with ASD and SLI, which showed that the loss of language skills is highly specific to children with ASD.
Published in the Journal of Child Psychology and Psychiatry, the study included 368 children with ASD and SLI from the United Kingdom. Parents were interviewed using the Autism Diagnostic Interview "“ Revised, which provided a detailed developmental history of the child and included information on loss of language skills (i.e. the milestone of acquiring either single words or short phrases was reached but then language development plateaued or was lost). The authors found that whereas 15 percent of children with ASD showed loss of language, only 1% of children with SLI showed similar patterns of loss. Curiously, they found that those children with ASD who later lost language skills acquired either single words or short phrases significantly earlier than those children with ASD who showed no signs of language loss.
The study reveals a surprisingly strong specificity for language regression and ASD, perhaps indicating that even though language impairment is common to both ASD and SLI, they may involve different biological mechanisms. "Most importantly, these findings have immediate clinical implications," explained Dr. Colamarino, "suggesting that the presence or absence of language regression can be useful information in the differential diagnosis of ASD versus SLI for children presenting to their doctors with language difficulties."
10. Association of Family History of Autoimmune Disease and Autism Spectrum Disorders "“ use of large Danish database adds another link between autism and the immune system
A 2009 study published in Pediatrics used the nationwide psychiatric health registry in Denmark to re-examine the potential link between autoimmune diseases, particularly in the mother, and ASD. Using a sample size over ten times larger than previously studied, their findings confirmed an association between autoimmune disease and ASD.
Multiple research approaches have been used throughout the years to uncover a potential relationship between immune function and autism, and while producing very intriguing data, the size of these studies has been limited and the findings have not always been replicated. This study utilized the nationwide health care system in Denmark to focus on the relationship between familial autoimmune diseases and ASD, including virtually all of the children born in the country between 1993 and 2004 (over 680,000 children). After locating those with diagnoses of an ASD, the authors used other databases to track down information on the presence or absence of twenty-six different autoimmune disorders in the parents or unaffected siblings. Their results confirmed previously reported links between familial type 1 diabetes and ASD, as well as between rheumatoid arthritis and ASD, discovering for the first time that rheumatoid arthritis in the mother, but not the father, is associated with increased risk for ASD. The authors also uncovered the first association between ASD and untreated celiac disease in mothers.
By using one of the largest and most comprehensive national health databases, this study provides additional clues regarding the association between immune system dysregulation and ASD. The associations found with specific autoimmune diseases and whether the disease is present in mothers versus fathers provide important clues about the biological mechanisms that may lead to autism. For instance, because the risk of ASD was increased only when the mothers, but not the fathers, had rheumatoid arthritis, the authors hypothesize that the link to autism may be due to exposure to maternal antibodies secreted during pregnancy, or other alterations within the prenatal environment. A relatively new focus in the search for the causes of ASD is the complex interaction between the immune function of a mother during pregnancy and the biological impact this may have on the early brain development of her child. In contrast, the association between type 1 diabetes and ASD was found if either parent was diabetic indicating that this link may be explained by a genetic factor that is related to both diabetes and autism. "This study suggests that looking at the associations with familial autoimmunity may be used to narrow down the search for autism risk factors, both genetic and environmental," said Dr. Colamarino.
"Not only does this research provide insights into causes and treatments, it provides a scientific context to prioritize funding for further research" remarked Dr. Dawson. "Autism research is a slow and incremental process. We see that epidemiological, genetic and environmental discoveries proceed simultaneously, with findings in one, advancing the other. To proceed in all areas of autism research concurrently requires a tremendous level of funding support." Autism Speaks has committed more than $141 million to date to fund research into the causes, diagnosis and treatment for autism through 2014. It is currently funding research into potential genetic and environmental risk factors involved with autism, identification of the biological pathways that underlie autism, as well as improved methods of early diagnosis and new treatment models."
Further Dr. Dawson explained that money spent on research is well spent. According to a 2007 Harvard School of Public Health study, it costs approximately $35 billion each year to care for people with autism "“ a number that has clearly increased over the past two years with the rising prevalence among the youngest people with ASD and a growing demand for housing, work skills and opportunities, healthcare, and other services that simply do not exist for adults with ASD. In FY 2008, total federal spending on autism research was just $177 million, expected to increase to $282 million in FY 2009 due to a one-time infusion of $89 million in stimulus spending. During his campaign, President Obama committed to $1 billion of annual federal spending on autism by 2012. In October 2009, the President identified autism as one of his administration's top three public health priorities. Increased funding for autism research that defines causes and leads to effective treatments will clearly offset the growing cost to the public associated with caring and services for the still increasing number of individuals with autism.
Autism is a complex neurobiological disorder that inhibits a person's ability to communicate and develop social relationships, and is often accompanied by behavioral challenges. Autism spectrum disorders are diagnosed in one in 110 children in the United States, affecting four times as many boys as girls. The prevalence of autism increased 57 percent from 2002 to 2006. The Centers for Disease Control and Prevention have called autism a national public health crisis whose cause and cure remain unknown.
On the Net: