VOLUME 114 | NUMBER 7 | July 2006 • Environmental Health Perspectives
Tracing the Origins of Autism A Spectrum of New Studies
The etiology of a medical condition might seem an unlikely subject to arouse intense feelings. Yet few medical disorders have stirred up as much passion and divisiveness among scientists and the general public as autism has in recent years. The heat of the controversy has even attracted attention from periodicals such as The Wall Street Journal, the Columbia Journalism Review, and Wired magazine—seemingly improbable forums for a medical debate. Why all the furor?
At the eye of the storm is the startling climb in the numbers of children who have been diag- nosed with one of the autism spectrum disorders (ASDs). The most severe ASD
is autistic disorder (which often is called simply “autism”); other forms include Asperger syndrome and the much rarer childhood disintegrative disorder. In the United States, the diagnosis of ASDs increased roughly 10-fold over the course of a decade, from 4–5 children per 10,000 in the 1980s to 30–60 children per 10,000 in the 1990s, according to a report in the August 2003 Journal of Autism and Developmental Disorders.
The 5 May 2006 issue of Morbidity and Mortality Weekly Report describes the results of two parent surveys from 2003 and 2004, which suggested that 55–57 children per 10,000 had autism (however, an editorial note points out that, due to the nature of the surveys, parents of children with other ASDs may have reported their children as having autistic disorder).
Some scientists believe that much of the upsurge is the result of increased awareness of ASDs or changes in diagnostic criteria, which would suggest that the true preva- lence of the disorders has been stable over time. Others disagree. “It is premature to state that there is no increase in prevalence,”says W. Ian Lipkin, a professor of neurolo-
gy, anatomy, and neurobiology at Columbia University. “None of the studies todatehasbeen designed to definitively address the issue.”
The prevalence of ASDs plays into the fundamental question of what causes these disorders. If the number of cases is truly on the rise, then it would seem likely that some change in the environment is driving up the total. That’s partly what has divided scientists into opposing camps—they can- not agree on the relative importance of genetic and environmental factors in the disorders’ etiology.
Alas, answering the prevalence question might not end that debate. “Even if the prevalence of autism were stable,” says Lipkin, “you would not be able to rule out the possibility of an environmental trig- ger.” That’s because very little is known about the mechanisms that cause autism, be they environmental or genetic.
“The study of autism was, until recent- ly, largely dominated by the field of psychology, where characterizing the behaviors and developing reliable instruments for diagnosis have been major areas of research over the past few decades,” says Irva Hertz- Picciotto, an epidemiologist at the University of California, Davis.
Indeed, the core symptoms of ASDs— social disinterest, repetitive and overly
focused behavior, and problems in com- munication, usually appearing before 3 years of age—have been well described. Much less research has focused on the causes of these symptoms.
Several investigations dating back to the 1970s indicate that identical twins have a much higher concordance rate of ASDs than fraternal twins, according to a report in the Spring 1998 issue of Mental Retardation and Developmental Disabilities Research Reviews. Those studies provide some of the best evidence that these disor- ders have a strong genetic component. But the identity of the genes involved, much less how they produce ASDs, has not been established. Moreover, the concordance rate for identical twins is not 100%, which suggests that at least some cases must be associated with environmental or epigenetic factors.
A few cases of ASDs have been clearly linked to environmental insults. These include prenatal exposure to chemical agents such as thalidomide and valproic acid, as well as to infectious agents such as the rubella and influenza viruses. Here again, the concordance rate is not 100%, which suggests that a genetic predisposi- tion is necessary for chemical and microbial factors to act as triggers.
Tantalizing clues like these are prompt- ing scientists to reconsider the research
agenda for ASDs. Martha Herbert, a pediatric neurologist at Harvard Medical School, and her colleagues have been applying the methods of genomics to identify environmentally responsive genes that might be important in these disorders.
“When you realize that the widespread changes we’re seeing in autistic brains may occur in parallel with or even downstream from widespread changes in the body— such as in the immune system—and that these changes may be environmentally trig- gered, you start looking for ways to think more broadly about genetic vulnerability. It can’t be just about ‘brain genes,’” Herbert says.
Some new epidemiological studies also are looking for gene–environment interactions. According to Diana Schendel, an epidemiologist and project officer for autism research at the CDC, which funds one of the projects, these initiatives will be able to examine many possible causal path- ways to ASDs, including both genetic and environmental causes that may lead to the development of the disorders in different subgroups of children.
Some of these projects are already under way, whereas others will begin soon. All of the scientists involved, however, believe their research will finally provide some of the answers that everyone has been looking for.
The Childhood Autism Risks from Genetics and the Environment (CHARGE) project is unique among the large ASD epidemiological studies. It focuses solely on autistic disorder, and it emphasizes a search for environmental factors—including a broad array of chemicals in food, consumer products, and ambient air, as well as infectious and medical exposures—that might be linked to the disorder. The study is funded by the NIH.
CHARGE is a case–control study in which a group of autistic children aged 2 to 5 years is compared to a group of age- matched controls in a population-based study. “Because of the California Department of Developmental Services’ system of Regional Centers [nonprofit corporations that coordinate health care services and support for citizens with developmental disabilities], we have a handle on enumerating a high proportion of the children newly diagnosed with autism in our defined area over a specific time period,” says Hertz-Picciotto, the principal investi- gator of the CHARGE study. “Similarly, we can enumerate the children in the same area and time period who are not cases. We then sample from both.”
The project was initiated in 2002 with the goal of recruiting 1,000 to 2,000 children. Half of the children will be autistic. The other half will make up two control groups: one group of children with devel- opmental delays (but not an ASD) and a second group of children selected from the general population without regard to developmental characteristics.
The advantage of the case–control design is that scientists can acquire large numbers of children with the disorder. By comparison, in a cohort design researchers would need a very large sample size, given the prevalence of autism, to acquire the same number of cases.
Hertz-Picciotto expects to have enrolled nearly 700 children by August 2006, the end of the first funding period. “I’ve applied for another five-year grant,” she says, “and I hope to be funded to enroll nine hundred in that round, which would bring us to sixteen hundred children.”
The CHARGE team is looking at possible exposures during the prenatal period and early childhood. Some of the data will be gathered through comprehensive interviews with parents, but Hertz-Picciotto admits that this is not the best way to look for exposures. “You ask people questions, and their answers may be colored by the fact that they know they have a child with a condition,” she says. “They may spend a lot of time thinking about what they might have done or what might have gone wrong, and they may have preconceived ideas about what caused [the disorder]. They might not be as objective.” Such problems with postdiagnosis interview information are recognized as a weakness of retrospective studies.
The scientists are getting around this issue by examining each child’s medical records and those of the mother during pregnancy and delivery—nonsubjective data gathered in the course of routine obstetric care. They are also collecting blood, urine, and hair specimens that will be analyzed in the laboratory.
The study has already provided some intriguing leads. “We’re finding that the immune system seems to function at a lower level in autism,” says Hertz-Picciotto. “That’s an important clue. It could mean that whatever causes autism also disrupts the immune system, or it could be that the immune system disrupts neural develop- ment so that something goes awry in laying down brain circuitry prenatally or in the early postnatal period.” [For more information on the CHARGE study, see p. 1119, this issue.]
The Autism Birth Cohort (ABC) Study, now under way in Norway, is a large prospective design that is expected to gather information on 100,000 babies. The work is being led by scientists at the Mailman School of Public Health at Columbia University, who are collaborat- ing with colleagues at the Norwegian Institute of Public Health, with funding from the U.S. National Institute of Neurological Disorders and Stroke.
“When you want to know why some people are more at risk than others in a population, then that’s best answered using a cohort design,” says Ezra Susser, an epidemiologist at Columbia University and a co-investigator on the ABC project. “When we think about environmental causes of [ASDs], we’re probably interested in phenomena that occur prior to birth or perhaps shortly after birth. So you want to collect prospective data from people as early as possible in pregnancy.” Because ASDs are not common, the study will need large numbers of children to have enough statistical power, according to Susser.
So far the ABC team has recruited 75,000 pregnant Norwegian mothers, but Susser is hoping for more. “We’ve got enough to look for an environmental risk factor, but you need larger numbers for studying gene–environment interactions, which could turn out to be important,” he says. It’s possible the team could acquire greater numbers by collaborating withother studies. One candidate for collaboration is the Avon Longitudinal Study of Parents and Children in the United Kingdom, which is looking at the complex ways in which environmental features may relate to optimal development and health in children. But there’s been no agreement yet, Susser says.
Even so, the ABC scientists are opti- mistic about their study. “Little is known about the natural history of [ASDs],” says Lipkin, who is the principal investigator of the project. “By starting prenatally, we’re collecting detailed, critical information about environmental exposures in an unbi- ased fashion.”
The scientists are also collecting plas- ma, serum, RNA, and DNA. “We have extraordinary biological materials,” says Lipkin. “We can pursue biomarkers as well as exposure to toxicants and infection. We also have maternal DNA, paternal DNA, and the child’s DNA [so-called trio data]; thus we can look for the appearance of novel mutations,” he adds.
The ABC researchers will follow the children through time, with parents answering questionnaires about the health and social interactions of their children as they reach 6, 18, and 36 months of age.
“It may be that the developmental trajectory tells us much more than a single time point can ever tell us about the pathogenesis of [ASDs],” says Mady Hornig, a physician- scientist at Columbia University who participates in the project.
Despite their enthusiasm for the pro- ject’s potential, the ABC scientists feel they could accomplish much more if they only had the funding. “The pity of it is we have no money to do the biological work,” says Lipkin. “We can collect the samples and do the questionnaires, but we’ve been unable to get funding to look for any of the envi- ronmental factors. We’re collecting blood, but we won’t know whether there’s a bio- marker until we do a biomarker analysis. We have funds to collect RNA, but in order to do the transcript profiling we need approximately four hundred dollars per sample,” he says.
Lipkin adds that there’s only so much that one can do with questionnaire data. “We do ask about infection and diet, but that’s not the same as having a lab value that can validate what was reported, and then look at a direct correlation with the outcome,” he says.
Lipkin believes that part of the prob- lem is that searching for environmental factors goes against the current research paradigm in ASDs. “The focus is on genet- ic factors,” he says. “Infectious diseases, toxicology, and immunology receive short shrift. The ABC is clearly the right oppor- tunity to pursue these other leads because we have the ideal samples to survey prena- tally and postnatally,” he says.
The scientists are just now receiving the responses to the 36-month question- naire. “It’ll probably be another two years before we have our first report,” Hornig says. Funds are now in place to study the children at 36 months; however, the team hopes to follow them for a lifetime, accord- ing to Hornig.
In response to the Children’s Health Act of 2000, the CDC established and funds six Centers for Autism and Developmental Disabilities Research and Epidemiology (CADDRE) to investigate potential risk factors for ASDs. The multisite approach offers a study group that is geographically and demographically more representative of the general U.S. population than a smaller regional study could provide, according to Craig Newschaffer, an epidemiologist and principal investigator at the Johns Hopkins Bloomberg School of Public Health CADDRE site.
According to Newschaffer, the CADDRE sites will use a case cohort design in which the exposure patterns of the ASD cases are compared to a random sample of children living in the same geographic area. A third study group, consisting of neurodevelopmentally impaired children who do not have an ASD, will round out the sample populations. The investigators hope to enroll a total of 650 to 900 children, aged 3 to 5 years, in each study group across all the sites, making CADDRE the largest study of its kind in the United States, says Newschaffer. A uniform protocol across the sites will allow the scientists to pool their data.
CADDRE will collect and archive blood, cheek cell, and hair samples from the children in order to investigate a broad range of potential risk factors. “We’re not focused on the environment as much as CHARGE is,” says Newschaffer, “but we are collecting data on questionnaires and reviewing medical records on exposure, in addition to the biosampling for exposures.”
The scientists should have sufficient numbers to look at gene–environment inter- actions. “We are collecting DNA from the parents and the kids from each of the groups. We’ll have trio data in each of the three groups, a potentially powerful design,” says Newschaffer.
CADDRE scientists will also character- ize the behavior of the children, as well as describe any comorbid medical conditions and atypical physical features. The goal is to sort out different etiologic subgroups with- in the autism spectrum. As Newschaffer explains,
“There are a lot of possible reasons why we’ve had a hard time coming up with genetic and nongenetic risk factors. One of them is that autism is likely a heterogeneous condition, with different etiologies producing kids with what appear to be similar phenotypic profiles. If you don’t separate out the different etiologic groups, it’s going to be very hard to find an association with a gene or an exposure. If we limit our analyses to kids that have a certain profile, we’re going to be able to make some informed guesses about what profiles might allow risk factors to emerge,” he says. The CADDRE sites will begin recruiting children into the study in the fall of 2006.
More Studies, More Acronyms
There are several other smaller epidemiological studies in the works. In California, scientists are tapping into specimen banks that have stored blood samples taken from mothers during pregnancy and from their children at birth. The Early Markers for Autism (EMA) study employs a case– control design, with about 100 children with an ASD (primarily autism), 100 who are developmentally delayed, and 200 from the general population. “We can correlate what’s happening in the mom and the baby, which is really exciting,” says Lisa Croen, a perinatal epidemiologist at the Kaiser Permanente Division of Research in California and the project’s principal investigator.
EMA is a multidisciplinary collabora- tion with epidemiologists, geneticists, immunologists, neurovirologists, and endo- crinologists, according to Croen. “Because autism is so complex, it’s important for all these researchers to communicate with each other. I think EMA is a model for how to do research in autism,” she says.
EMA is unique, according to Croen, because the study will be looking for biological markers of ASDs very early in development, during gestation, and at birth. “This allows us to focus on mechanisms that may be leading to autism rather than mechanisms that are consequences of having autism,” she says.
The EMA scientists are investigating genetic and nongenetic factors, with a focus on the immune dysregulation hypothesis of ASDs.
“We’re measuring different kinds of immune markers, including immuno- globulin levels and antibodies to specific infectious agents, cytokines, and autoanti- bodies,” says Croen.
“We’re looking for things that distinguish kids who are subse- quently diagnosed with autism from those who aren’t. This will help us understand the pathobiology of autism—the mechanisms that are leading to the dysregulation in development.”
The three-year EMA is currently in its last year. “We still have lots of analyses to do,” says Croen, “but we’re beginning to write some papers. We’re finding differences between the children in levels of certain proteins measured in the circulating blood collected from mothers during pregnancy. I think the study has much to contribute to our understanding of the biology of what might be going wrong.”
Croen is also an investigator on the California Autism Twin Study (CATS), which expects to recruit 300 identical and fraternal twin pairs born between 1987 and 1999 in which at least one of the twins has an ASD. Comparing the twin pairs will allow the scientists to estimate the heritability of ASDs—the relative genetic and envi- ronmental contributions to the disorder. “Knowing the behavioral and developmen- tal differences between the twins might help us understand the effects of gene expres- sion, the in utero environment, and envi- ronmental triggers,” Croen says.
Hertz-Picciotto is also excited about a five-year study that she and her colleagues hope to begin soon. Unlike CHARGE, the new effort, called MARBLES (Markers for Autism Risk in Babies—Learning Early Signs), will be a prospective study in which data will be gathered before the children are diagnosed. Pregnant women who already have at least one child with autism will be enrolled right at the beginning of pregnancy. The mothers will keep diaries about their symptoms and health-related events, and the researchers will collect cord blood samples and placentas.
Based on previous research, Hertz- Picciotto expects that about 1 in 10 siblings of the autistic children will also have the disorder, and perhaps 1 in 4 or 5 will be “on spectrum” with a related but less severe con- dition such as Asperger syndrome, or with some symptoms of the broad behavioral phenotype, such as language delays and atypical social skills. “This work is comple mentary to the case–control approach, and should provide us with a lot of information that will build on what we find in CHARGE. It should be a phenomenal resource,” she says.
You Say You Want a Revolution
In April 2004, the U.S. DHHS issued a publication, Congressional Appropriations Committee Report on the State of Autism Research, describing recommendations
made by a panel of expert scientists convened by the Interagency Autism Coordinating Committee (IACC). The IACC panel suggested an ambitious agenda, which included the goal of identifying envi- ronmental risk factors and their associated developmental windows within a four- to six-year period, as well as identifying genetic and nongenetic causes of ASDs and their interactions within seven to ten years.
Hertz-Picciotto, a member of the IACC panel, thinks these goals should be taken with a grain of salt. “I’m optimistic that we will have identified some environmental risk factors, and may have excluded a few others, between 2008 and 2010—but by no means will we have the final word. The genetics and the gene–environment interactions may be even tougher. Unfortunately, I don’t see enough groups working on the environmental contribution to autism, so it may be slower than projected,” she says.
Mark Blaxill, vice president of SafeMinds, a parent-led advocacy group, also believes that environmental risk factors don’t receive enough consideration. “The CDC has not addressed the crisis in autism responsibly,” he says. “They should be raising the alarm, and they have failed to do so. They should be asking why so many children are sick. Instead, they’ve tried to suggest a degree of doubt about the increases, and that diverts attention and funding from environmental causes.”
Schendel responds, “It is clear that more children than ever before are being classified as having an ASD. It is important that we treat common developmental disorders, and especially the ASDs, as conditions of urgent public health concern. The CDC’s efforts in addressing this public health con- cern include funding for ASD monitoring programs to understand ASD trends, fund- ing for research into the genetic and envi- ronmental causes of ASDs, and education and outreach programs to promote early identification and timely intervention for all children with developmental problems.”
Despite the promise of the new epi- demiological studies, some researchers are still dismayed, as one scientist put it, that “geneticists are running the show, and ignoring the environmental aspects.” What would it take for things to change?
Blaxill invokes the ideas of philosopher Thomas Kuhn, who suggested that scientific revolutions occur when an old paradigm is replaced by a new one. “I believe we’re in the middle of a paradigm shift,” Blaxill says. “The dramatic explosion of autism rates does not fit the genetic model. It’s an anomaly that will kill the old paradigm.”
Autistic people have fought the inclusion of ABA in therapy for us since before Autism Speaks, and other non-Autistic-led autism organizations, started lobbying legislation to get it covered by insurances and Medicaid.
ABA is a myth originally sold to parents that it would keep their Autistic child out of an institution. Today, parents are told that with early intervention therapy their child will either be less Autistic or no longer Autistic by elementary school, and can be mainstreamed in typical education classes. ABA is very expensive to pay out of pocket. Essentially, Autism Speaks has justified the big price tag up front will offset the overall burden on resources for an Autistic’s lifetime. The recommendation for this therapy is 40 hours a week for children and toddlers.
The original study that showed the success rate of ABA to be at 50% has never been replicated. In fact, the study of ABA by United States Department of Defense was denounced as a failure. Not just once, but multiple times. Simply stated: ABA doesn’t work. In study after repeated study: ABA (conversion therapy) doesn’t work.
What more recent studies do show: Autistics who experienced ABA therapy are at high risk to develop PTSD and other lifelong trauma-related conditions. Historically, the autism organizations promoting ABA as a cure or solution have silenced Autistic advocates’ opposition. ABA is also known as gay conversion therapy.
The ‘cure’ for Autistics not born yet is the prevention of birth.
The ‘cure’ is a choice to terminate a pregnancy based on ‘autism risk.’ The cure is abortion. This is the same ‘cure’ society has for Down Syndrome.
This is eugenics 2021. Instead of killing Autistics and disabled children in gas chambers or ‘mercy killings’ like in Aktion T4, it’ll happen at the doctor’s office, quietly, one Autistic baby at a time. Different approaches yes, but still eugenics and the extinction of an entire minority group of people.
Fact: You can’t cure Autistics from being Autistic.
Fact: You can’t recover an Autistic from being Autistic.
Fact: You can groom an Autistic to mask and hide their traits. Somewhat. … however, this comes at the expense of the Autistic child, promotes Autistic Burnout (this should not be confused with typical burnout, Autistic Burnout can kill Autistics), and places the Autistic child at high risk for PTSD and other lifelong trauma-related conditions.
[Note: Autism is NOT a disease, but a neurodevelopmental difference and disability.]
Fact: Vaccines Do Not Cause Autism.