[an error occurred while processing this directive]
Skip to Main Content

Government & Policy

May 14, 2007
Volume 85, Number 20
pp. 32-35

The Conundrum Of Autism

Institute of Medicine symposium examines genetics, environmental triggers, treatments for this baffling illness

Bette Hileman

AUTISM AND RELATED health problems, collectively known as autism spectrum disorders (ASDs), are a group of widespread developmental disorders that impair an individual's ability to communicate or fully engage in social activities. According to the Centers for Disease Control & Prevention, ASDs now affect one out of every 150 children in the U.S. Ratings of the thinking and learning abilities in autistic children range from severely challenged to gifted, CDC notes, and autism is four times more common in males than in females.

Center for morphometric analysis/Massachusetts General Hospital
Picturing the Brain In this color-coded magnetic resonance image of the brain of an autistic patient, white regions are the enlarged white matter and red regions are the relatively smaller gray matter.

Most researchers in the field believe the prevalence of ASDs has increased over the past two decades, but a minority say the apparent increase could result primarily from an expanded definition of the family of disorders and more of an effort to identify cases.

Some facts are clear: Between 1994 and 2004, the number of six- to 17-year-old children classified in the U.S. public schools as having an autism spectrum disorder rose from 22,664 to 193,637. More children are diagnosed with ASDs each year than with diabetes, AIDS, and cancer combined. According to a Harvard University School of Public Health study, autism now costs the U.S. economy more than $35 billion per year in direct and indirect expenses.

To improve understanding about the prevalence and causes of and treatments for autism and to design new research strategies, the Institute of Medicine held a symposium on April 18-19 in Washington, D.C. Researchers there examined possible environmental causes for ASDs, particularly the interactions between genetic susceptibility, environmental chemicals, and exposures to microbes. The speakers at the meeting made it clear that questions about autism far outnumber answers.

It is not known, for example, whether autism is primarily a disease of the industrialized world or whether the incidences in developing and developed countries are similar. Genetic susceptibility seems to be a necessary component of autism, but the number of estimated genes involved varies from three to 10 on the low end to as many as 100 on the high end. Also unknown is whether environmental causes, such as exposure to chemicals, microbes, or vaccines, are important factors.

Some doctors have developed treatments that, they claim, cure or alleviate many, but not all, cases of autism. However, little research has been done to assess the efficacy of these treatments. Some researchers have pointed to the measles-mumps-rubella (MMR) vaccine as a cause of autism, but several studies indicate that the two are merely correlated. Other ongoing studies designed to assess such a relationship have not yet come to a conclusion.

Similarly, several research projects found no relationship between autism and another alleged risk factor, thimerosal (a mercury preservative in vaccines), or any other source of mercury exposure. Still, some physicians claim to have treated many cases of autism successfully by chelating mercury from children's bodies.

A complicating factor is that ASDs occur in many different forms with varying degrees of severity. One element ASDs share is that symptoms involving delays in the development of many basic skills—the ability to socialize or form relationships with others and the ability to communicate and use imagination, especially in fantasy play—appear before the age of three.

Children in one subgroup of ASDs appear abnormal from the day they are born. They have low scores on the Apgar test (the test given immediately after birth), and most are profoundly retarded.

Children in another subgroup of ASDs are labeled regressive. These children appear normal at birth and develop normally for at least 12 months. Then, in some cases almost overnight, the children begin to regress, losing language and other skills they had already acquired. Often, this happens shortly after inoculation with the MMR vaccine, which is how it became suspect. Because of these differences, researchers believe ASDs have a variety of causes.

At the IOM meeting, Craig J. Newschaffer, an epidemiologist at the Drexel University School of Public Health, explained that studies conducted since 2000 show ASD prevalences of 23 to 40 cases per 10,000 in North America, 18 to 25 cases per 10,000 in Japan, and 10 to 35 cases per 10,000 in Western Europe. Before 1980, the prevalence rates were much lower, he said, but that fact can be partly explained by the narrower definition of ASDs used before 1980. Recent analyses show that some, but not all, of the increase in the U.S. statistics results from the reclassification of large numbers of mental retardation cases as ASDs, he observed.

INTERNATIONAL STUDIES can help in establishing the causes of ASDs, Newschaffer said. For example, studies of immigrant groups in which ASD prevalence goes up after migration to another country can suggest an environmental contribution. Such studies are ongoing or in planning stages in both developed and developing countries, he said.

Mark F. Blaxill, vice president of SafeMinds, a nonprofit think tank, pointed out that his review of the literature on autism "found evidence of large increases in prevalence rates in both the U.S. and the U.K. that cannot be explained by changes in diagnostic criteria or improvements in case ascertainment." Prevalence of ASDs in both countries rose from five to 10 per 10,000 in the 1980s to 50 to 80 per 10,000 in the 1990s, Blaxill said. "The rising incidence of autism should be a matter of urgent public concern."

Mug Title
Mug Title Bette Hileman/C&EN (both)

Susan E. Swedo, chief of the Pediatrics & Developmental Neuropsychiatry Branch at the National Institute of Mental Health (NIMH), explained how clinical manifestations of autism can shed light on environmental causes. "About 10% of autism cases are caused by a known genetic defect; 1% are caused by a teratogen, such as valproic acid used to treat epilepsy; and the rest have no known cause," she said. "All of us are looking for an environmental trigger in a genetically susceptible population."

Some potential triggers include infectious agents; exposures to household chemicals; foods, dietary supplements, vitamins, and minerals; drugs, medications, and herbal remedies; and other medical interventions, such as ultrasound, Swedo said.

Environmental causes are suspected for a number of reasons. Researchers know that some cases of autism are induced by teratogenic agents like valproic acid, Swedo explained, so they realize that chemicals can trigger autism. Also, because the prevalence of autism seems to be increasing, it is unlikely that genes alone are responsible for the increase. And the onset of many cases occurs at 12 to 18 months of age, when advances in verbal and physical skills stop and regress. About 15 to 50% of autism cases fall into the regressive subtype, she said.

Several factors make it very important to study children with the regressive subtype of ASD, Swedo said. One is that the children appear normal at birth, so theoretically their brain circuits may be in place for recovery of function. Also, with these children, it may be easier to detect possible causes because whatever triggers the illness may occur after birth. "Just as researchers found an environmental trigger for obsessive-compulsive disorder, the Streptococcus bacterium, they may be able to find one for regressive autism," Swedo said.

Uncovering the genetic landscape associated with ASDs is an ongoing effort that might help reveal environmental triggers of autism, said Pat R. Levitt, director of the Vanderbilt University Kennedy Center.

Autism "has the highest heritability of any [mental] disorder," he said. If one identical twin develops the disorder, the other has a 90% chance of having some aspects of it. And the sibling of a child with autism has at least five times the average risk of acquiring the disorder. Several studies have found areas of the chromosome that link with autism, he said.

Researchers know that genes contribute to ASDs, but that information alone is of limited value when it comes to preventing, treating, or managing the disorders. For example, Levitt said, scientists "do not know how much of the risk is due to the direct impact of mutations on brain development or how much is due to the impact of mutations on peripheral functions that influence brain development," he said. Also, "they do not understand how much of the risk is due to genetically established sensitivities to environmental insults," such as chemical exposures.

Because nature and nurture are intimately connected, questions about ASDs do not involve a conflict between genes and the environment but do involve connections between the genetic makeup and environmental influences, Levitt said. "The brain develops by reacting to the environment. What is needed to advance research on autism is more gene sequencing to identify additional gene variants that contribute to the disorder and functional characterization of the variants," he said.

Another important research area, Levitt continued, is wise investment in model systems, such as "humanized" mice (transgenic mice with some human genes) to test the effects of various chemicals and microbes on brain development when genetic risk is introduced.

Multiple genes are involved in conferring autism susceptibility, noted Isaac N. Pessah, director of the Children's Center for Environmental Health & Disease Prevention at the University of California, Davis. But the development of autism "is likely to be more profoundly affected by the environment than by other disorders and diseases," he said.

To help understand the gene-environment interactions that affect the risk of autism and to guide research, Pessah has developed a general formula: genetic susceptibility times environmental exposures times timing equals the prevalence and severity of autistic developmental disorders.

"What we don't know about environmental triggers that could play a role in autism far exceeds what we do know," Pessah said.

Of the more than 53,000 commercially important chemicals, Pessah said, about 80% lack adequate toxicity data. Food additives are not well-understood, he said, and less is known about the additive and synergistic effects of environmental chemicals.

One major hypothesis that could explain the core symptoms of autism is that these emerge from imbalances in the activities of excitatory and inhibitory circuits within the developing brain, Pessah said. Several studies show that GABAA receptors, which sit on neuron membranes and respond to the neurotransmitter γ-aminobutyric acid, are deficient or defective in those with ASDs, he said.

This hypothesis may give a clue for one possible cause of autism, Pessah observed. Several widely used pesticides are neurotoxic "because of their ability to promote imbalances in the electrical activity patterns within the central nervous system," he said. An example is fipronil (4-alkyl-1-phenylpyrazole), which inhibits activation of the GABAA receptor in insects and vertebrates. Sold under trade names such as Regent and Frontline, more than 800 tons of fipronil is used annually in U.S. homes against ants, fleas, cockroaches, and termites.

In her presentation at the IOM symposium, Martha R. Herbert said autism probably involves complicated interactions between the nervous, immune, and metabolic systems. Herbert, a neurologist at Harvard Medical School, said ASDs can affect multiple physiological systems even shortly after birth. During the first year, some autistic children never have a normal appetite, some vomit a great deal, and some have repeated colds and ear infections.

"We need biomarkers to diagnose autism, to identify treatable features, and to predict and track responses to treatment," Herbert said, "but we have none." Moreover, it is challenging to find biomarkers because the illness is so complicated and manifests itself in many different ways with variable severity, she said.

The severity of autism can oscillate even in the same individual, Herbert added. "There are often striking 'good-hair day and bad-hair day' changes," she explained. Autism can show transient improvement with fever, antibiotics, and allergy medications and can vary with food, allergen, and toxic exposures. In some cases, "stable improvement can follow treatment. And there are published reports of loss of diagnosis," Herbert said, adding that studies now are in progress to document such cases of full recovery.

Although autism seems to involve multiple bodily systems, we do not know whether the perturbations in the digestive, immune, and sleep systems cause the problems in the brain or simply coexist with the brain problems, Herbert said.

For Herbert, another interesting point is that there seem to be abnormal bacterial species in the stools of autistic children. More research is needed in this area because abnormal species could deplete vital nutrients and alter the metabolism of chemicals and other environmental substances to which the child is exposed, she explained. "Beyond characterizing autism, the point is to treat it."

Mug Title
Mug Title Bette Hileman/C&EN (both)

Adding to that line of thought, Claudia S. Miller, professor of environmental and occupational medicine at the University of Texas Health Science Center, San Antonio, noted that some autistic children have improved dramatically after removing certain foods from their diets or chemicals from their environments. However, many parents who have deliberately tried to make such changes "found little or no improvement," she said.

To rigorously investigate whether autistic children could improve after certain foods or chemicals are removed from their environment, Miller proposed the use of a specially designed and constructed residential or hospital facility called an environmental medical unit (EMU). Those in the EMU would have virtually no exposure to air pollution, cleaning agents, allergens, or any other toxicants. A child and a parent could stay in the unit for several weeks while the child is exposed to one substance at a time, including particular foods and everyday chemicals at low levels, to see if any of them produce positive or adverse reactions.

"The EMU would allow us to see the role of exposures in ASDs, just as the microscope allowed scientists to see germs responsible for infectious diseases," Miller explained. But even though several bills authorizing an EMU have been passed in Congress, it has not been built in the U.S. Four such units are operating in Japan.

Arthur L. Beaudet, chairman of the department of molecular and human genetics at Baylor College of Medicine, proposed a number of factors that may be partially responsible for the apparent rise in autism. First, he said, about 40 to 50% of autism has a genetic cause, either directly inherited from the parents or caused by chromosomal rearrangement or copy-number and single-gene point mutations.

"There can be duplications and deletions in virtually every single chromosome associated with autism," Beaudet said. "We see many genetic causes in very severe autism. With paternal age above 40, the offspring are 5.75 times more likely to have an ASD than offspring of men younger than 30 years."

THE ETIOLOGY of the remaining 50 to 60% of ASD cases remains unknown, Beaudet said, pointing to still undiscovered interactions between genes and the environment as likely players. He mentioned a number of possible environmental insults that could interact with genes: childhood vaccines; thimerosal; prenatal use of ultrasound; steroids given after birth; epidural anesthesia given during labor; magnesium sulfate and terbutaline, which are given to stop preterm labor; chemical exposures; and folate supplements taken by the pregnant mother to prevent neural tube defects.

Rodent studies show that prenatal exposure to ultrasound or folate supplements can affect offspring, Beaudet said. "Ultrasound waves impact neuronal migration in mice during brain development," he said. In mice, prenatal exposure to folic acid supplements changes the color of the fur by altering gene expression, he said. "The folic acid intake of pregnant women has increased dramatically over the past few decades," Beaudet noted. "This could be one factor affecting autism."

At the end of the IOM meeting, Pessah presented the participants' consensus on what a research agenda for autism and the environment might look like. Multidisciplinary teams are essential, including epidemiologists and trained assessors to administer behavioral and cognitive assessments, he said. In addition, exposure experts and analytical chemists are needed to prioritize which chemicals and physical agents to measure, optimize environmental and biological sampling, and choose analytical methods.

Furthermore, expertise in molecular genomics, as well as toxicology and cell biology, is necessary for identifying biomarkers that are selective or unique to autism. Toxicologists who can develop appropriate animal models of autism should also be part of this multidisciplinary effort.

"Especially urgent are randomized controlled trials of novel therapeutics and innovative interventions," Pessah said when outlining what a near-term research agenda should include. Trials of chelation therapy, and of supplementation with antioxidants, vitamins, and docosahexaenoic acid (an omega-3 essential fatty acid mainly found in fish oil) should be undertaken, he noted.

Laura Bono, a board member of the National Autism Association, implored the attendees "to declare autism a national health emergency and treat it with urgency." Roughly one in 65 families are affected by autism, she said. Regressive autism is now so common that 36,000 otherwise normal toddlers will regress into autism during this year alone, she observed.

"There are treatments we can use now," Bono noted. "We can treat autistic children with enzymes and improve their digestion. As a result, their brains may improve because they get more nutrition. We can treat with vitamin B-12 and get a response in some children."

AUTISM IS a baffling disease, and extensive research is needed before it is fully understood. Indeed, the government and researchers are giving ASDs more attention now than they did even five years ago.

According to Thomas R. Insel, director of NIMH, National Institutes of Health funding for autism increased nearly fivefold since 1997. NIH has established a National Database for Autism Research to provide an "open-access platform to facilitate sharing of raw research materials, foster collaborations and public-private partnerships, and enhance rapid dissemination of research findings into clinical practice," Insel told a Senate appropriations subcommittee on April 17.

NIH, along with the advocacy group Autism Speaks, is supporting a research consortium to improve early detection and diagnosis. In addition, the National Institute of Environmental Health Sciences, in partnership with the Environmental Protection Agency, is supporting research on environmental factors that may cause ASDs.

Despite the recent increases in support, ASD research funding remains lower than spending for far less prevalent diseases. ASD research receives only $108 million in annual funding at NIH, compared with the $310 million earmarked each year for leukemia, which affects only one in 25,000 people, and the $394 million per year that the government spends on research in pediatric AIDS, which affects one in 8,000 children.

Chemical & Engineering News
ISSN 0009-2347
Copyright © 2011 American Chemical Society

Related Stories

Adjust text size:

A- A+

Articles By Topic