network-like background pattern


Autism can suddenly devastate a child’s ability to talk and interact with others. What causes it and how can it be stopped?

By Melissa Hendricks

Alex Bush wants a second bowl of Goldfish snacks. He stands in front of his mother in the living room of his family’s suburban Baltimore home, bouncing on the balls of his feet and pressing together the fingertips of his left and right hand to form the sign language gesture for “more.”

“No, Alex, you’ve already had some,” says Sandy Bush. Alex repeats the gesture and looks at his mom with intense, dark brown eyes. He is a handsome 11-year-old, with dark hair shaved close to his head and a tanned, muscular body.

“Okay,” says Sandy Bush, relenting. “But first give me a kiss. A kiss, Alex. Give me a kiss.” Alex slowly leans forward and lightly kisses his mother who then goes to the kitchen. When she returns with a bowlful of crackers, Alex places his hand on his chest—the gesture for “thank you.”

Alex likes a lot of the stuff that other kids his age do. He loves holding a ball, bouncing on his backyard trampoline, watching videos, and eating cookies and junk food. But Alex cannot tell you any of these things. He is autistic, and for the past six and a half years, he has not spoken. He communicates with the three sign language gestures he knows, or by taking his mother’s hand and placing it on the thing he wants. 

A few miles away from Alex’s home lives another boy just a year older than Alex who also has autism. John Gillin can speak and loves to read, especially books about colonial America, novels by Roald Dahl and anything about geography. For relaxation, he’ll read the atlas and almanac. When a visitor comes to his home, he knows to say, “Hey, how’s it going? I’m John. Who are you?” But beyond such scripted lines, John has trouble making casual conversation. He responds slowly to people’s questions and misses the non-verbal aspects of social interaction.

The big differences between Alex and John illustrate one of the many challenges confronting researchers who are trying to figure out what causes autism. What sort of model could explain a disorder that deprives one child of language but affects another in ways that might be apparent only to a trained professional?

Scientists know that autism is a brain disorder that impairs the ability to communicate and socialize, and can cause a person to engage in repetitive or ritualized behaviors or to be unusually sensitive to light, sound or touch. Beyond that, much about autism remains a mystery.

“We know very little about the causal risk factors of autism,” says Craig Newschaffer, an associate professor of Epidemiology and Mental Health. “It’s critical we find some answers.”  

Newschaffer directs the Bloomberg School’s Center for Autism and Developmental Disabilities Epidemiology (CADDE). It is one of six regional centers for autism research established by the Centers for Disease Control and Prevention (CDC) in response to a dramatic rise in autism over the past decade. Some states have reported autism climbing 25 percent per year. Those rising rates have generated a lot of publicity, most of it focused on a debate over whether the measles-mumps-and-rubella vaccine routinely given to American toddlers has fueled an “autism epidemic.” In May, a report by the National Academy of Sciences’ Institute of Medicine concluded that the evidence did not support the vaccine-autism hypothesis. Still, some parent groups see the vaccine as the culprit.

Newschaffer and  colleagues at CADDE and the other CDC-funded autism research centers are conducting a broader investigation into autism’s causes and risks. They are about to launch one of the largest autism epidemiology studies ever conducted. It will involve almost 2,000 children, ages 3 to 5, and their parents, and include children with autism, those without it, and a group with other neurodevelopmental disabilities whom researchers will study to identify differences that could increase autism’s risk. Investigators will also give the mothers detailed questionnaires about their medical histories and pregnancies in hopes of uncovering events that might have increased their risk of having an autistic child. Results of the far-reaching study are not expected until the end of the decade.

Meanwhile, researchers have no shortage of leads in the search for autism’s cause. Gene-hunting studies suggest dozens of different genes with possible ties to autism; according to one model, 10 to 20 genes play a role in any one case of autism. But so far, no one has clearly demonstrated an “autism gene.” Researchers also suspect a number of different environmental factors. “Nothing has been definitively linked,” says Newschaffer. “Zero.” Scientists also know that autism strikes four times as many boys as girls, but that fact hardly narrows the search for a cause. (more)

If researchers knew more—even if they could pin autism’s roots to a region of the brain—then they could focus on finding treatments that relieve or prevent it. For Newschaffer, that hope is personal. He has a 12-year-old son who has autism, and his experience as the father of an autistic child helped convince him to focus his epidemiological research on the disorder.

“We always like to point to home runs,” Newschaffer says of his profession. He notes one “homer” that epidemiologists are particularly proud of: making the link between a lack of folic acid in a woman’s diet and the resulting neural tube defect in her offspring. Women are now advised to take folic acid before and during pregnancy to reduce their risk of having a child with a neural tube defect by as much as 70 percent. 

Autism research is unlikely to yield such a home run. Because so many different genes—possibly in combination with various environmental factors—may cause autism, researchers would be happy to get a base hit. When Andrew Zimmerman, a pediatric neurologist and CADDE co-investigator, announced that he was going to study the underpinnings of autism, an older faculty member who had had a successful career investigating a disorder caused by a single gene tried to dissuade him. “Why do you want to get into that swamp?” he asked. But Zimmerman was determined. He had seen the full spectrum of the illness. Alex Bush and John Gillin are both his patients.

It’s Christmas Day, 1995, in a house in Cockeysville, Maryland. A toddler with dark eyes and dark hair wearing red sleeper pajamas excitedly plunges his hand into an enormous stocking, pulls out a package and unwraps it. “A big new ball!” he exclaims. “I want to play catch!” He tucks the small ball safely under one arm while he hunts for more presents, issuing a steady stream of commentary about each gift.

The boy is Alex Bush, as he appears in a videotape his parents recorded in their former home when their son was 2. Then, Alex spoke in complete sentences. He knew the words to “Jesus Loves Me” and could sing a funny song about bubbles. Already sharing the family’s passion for baseball, he could recite the names of some of the Baltimore Orioles.

But that was about as far as Alex ever got in learning to speak. Fast-forward the video to a clip taken a few months later. Alex and his sister Chelsea are tucked in bed listening to their mother read a book about baseball. Sandy Bush pauses frequently to ask her children questions about the story they have read dozens of times before. Alex answers his mother’s questions, but he is quieter than he was at Christmas, and he does not talk unless his mother asks him a question. Sometimes Sandy has to repeat the question.

The signs were subtle, but Alex was starting to withdraw and lose his ability to speak. Over the next few months, he spoke less and understood less. His parents took Alex to specialists, including Zimmerman at the Kennedy Krieger Institute, who diagnosed autism.

But such dramatic regression does not always occur in children with autism. For John Gillin, autism was less obvious and took longer to diagnose. 

At 18 months, John seemed precocious in some ways. On a trip to Florida, his mom recalls, John kept saying, “Eight! Eight!” It baffled his parents until they realized their son was reading the numeral on their beach cabana. Before he was 2, John could count to 20 in English and Spanish, and knew his alphabet. He counted so much, his family called him “The Count,” after the Sesame Street character. By age 3, he had begun to read.

But something was not right. Although John knew a lot of words, he couldn’t put them together to form a sentence. He could say “cookie,” but not “I want a cookie.” The Gillins took their son to several specialists, eventually arriving at a diagnosis of autism.

“I think the heterogeneity of autism is one thing that has kept many researchers out of the field,” says Zimmerman.

Researchers and clinicians have been studying and debating the cause of this enigmatic disorder ever since Johns Hopkins child psychiatrist Leo Kanner first coined the term. In a paper published in 1943, Kanner described the cases of 11 children who were extremely socially withdrawn, lacked language skills and engaged in ritualized behaviors. Kanner proposed that these behaviors sprang from brain abnormalities present before birth. But his idea ended there; Kanner was never able to find a biological source for autism.

Then psychologists started developing their own theories, most famously the psychodynamic explanation popularized by Bruno Bettelheim. In his book The Empty Fortress, Bettelheim concluded that cold and emotionally distant parents (called “refrigerator mothers”) gave rise to autistic offspring by depriving their children of the emotional and social connections required for normal development.

Later, abandoning those theories for lack of evidence, scientists again contemplated the biological underpinnings of autism. Research began to suggest that autism had its roots in the development of the brain, and family and twin studies demonstrated that autism has a genetic component.

What is now becoming clear is that understanding this curious and puzzling disorder may require new “outside-the-box” scientific models.

Zimmerman began thinking about autism in a different way in the mid-1980s, after he began an autism clinic at the University of Tennessee. While seeing dozens of patients with autism, Zimmerman also got to know his patients’ relatives. “Certain things would stand out,” he says. Many of the mothers themselves were ill, particularly with autoimmune diseases.

He conducted an epidemiological study that confirmed his observations. Mothers of his autistic patients were eight to nine times more likely than people in a control group to have rheumatoid arthritis. Since joining the Hopkins faculty in 1994, Zimmerman, along with epidemiologist Li-Ching Lee, an assistant scientist of Epidemiology, has reported that many of the mothers from Tennessee carry an antigen that frequently occurs in people with autoimmune disorders. These findings have led Zimmerman to believe that the autism disorder may arise from a sort of autoimmune reaction.

“The fetus is something like a tissue transplant,” says Zimmerman. Normally (for reasons that remain a mystery to scientists) the mother’s immune system tolerates the fetus. Zimmerman is exploring the hypothesis that autism may result from failures in this mechanism. If the mother’s immune system reacts against the fetus, for instance, it might target the fetal brain. Such a reaction could fundamentally change connections among the brain’s cells or cause inflammation, possibly leading to autism.

In autopsy studies conducted with Hopkins neuropathologist Carlos Pardo, Zimmerman has found signs of the inflammation this model predicts. The team analyzed brain autopsies of nine autistic people, ranging in age from 6 to 55. Without exception, every one showed inflammation. Other researchers have found additional differences in the brains of people with autism—subtle changes in regions that develop during the second trimester. “To me, this is the most exciting thing I’ve seen in the field,” says Zimmerman.

Other researchers are studying variations on this theme, if other forms of incompatibility between mother and fetus are linked to autism. For example, Newschaffer and genetic epidemiologist Peter Zandi, PhD, an assistant professor in the Bloomberg School’s Department of Mental Health, are studying whether maternal-fetal differences in Rh factor or blood type are linked to an increased risk of autism.

Geneticists are also exploring new models of autism. Finding a gene for autism has challenged these scientists because there is probably no single cause for autism, says genetic epidemiologist Dani Fallin, PhD, a co-investigator with CADDE and an assistant professor of Epidemiology. In addition, the genes involved in autism may not follow conventional rules. Normally, the chemical code inscribed in a stretch of DNA determines a gene’s function. But Fallin and other geneticists now recognize that factors other than a gene’s code can, in some cases, influence that gene’s expression. In one type of interaction called methylation, for instance, chemical units attach to a gene and alter or squelch its expression. If methylation occurs in autism, it may have been foiling geneticists’ attempts to pin down an autism gene. Fallin and other researchers are also investigating the possibility that autism involves a constellation of genes and environmental factors. One candidate gene they are studying is called the beta-2 adrenergic receptor (B2AR) gene.

The researchers hypothesize that various environmental factors—chemicals, drugs, stress—can overstimulate B2AR. They also suspect that some people have a slightly different genetic version of B2AR that makes them more sensitive to these factors.

One drug they are focusing on is called terbutaline. It was brought to their attention by a woman in western Massachusetts named Beth Crowell.

Crowell is the mother of 16-year-old triplets who have all been diagnosed with autism. “Every parent of an autistic child goes through a phase where she asks herself, ‘What did I do?’ ” says Crowell. For Crowell, that question drove her to search the medical literature for answers. Her search led her to terbutaline, an asthma drug also prescribed to prevent premature labor. Crowell, who had taken terbutaline during her pregnancy, was struck by her findings. 

Crowell found studies showing that rats exposed to terbutaline developed neurological problems that, to her, echoed autism. She mentioned her observations to Susan Connors, a physician and mother of an autistic child, and eventually the pair teamed up with Zimmerman to study the connection between terbutaline and autism.

The team conducted a small study of fraternal twins drawn from a database of families of autistic children and from autism clinics. They compared twins born to women who had taken terbutaline for at least two weeks during their pregnancies to the twin children of women who had not taken the drug. They then calculated the rate of concordance for each group of twins, with concordance meaning that both siblings in a pair of twins had autism. The study revealed that women who had taken terbutaline were significantly more likely to have two twins who developed autism.

The researchers believe that terbutaline accounts for the increased risk. However, not every child exposed to terbutaline in the womb developed autism, and the researchers have conducted further studies suggesting that a variant of the B2AR gene increases the risk.

Newschaffer calls these findings “an excellent example of how a gene-environment interaction could work” to cause autism. However, he adds, “I don’t want to leap. All of this stuff is very speculative.” He is acutely aware of how easy it is to give people false hope of finding the key cause of autism.

Autism clearly has no single cause. The many factors that contribute to the  disorder may feed into the same pathway—or they may not. There could be multiple pathways. “All of the different disciplines have been looking for a grand unifying theory that explains autism,” says Newschaffer. No one has yet found such a theory. Just as an autistic child might struggle to form words into sentences, autism researchers continue to grope for a meaning that will assemble the scattered fragments of this enigmatic disorder.

Sandy Bush and Pam Gillin don’t know what caused their sons to have autism, and they don’t spend a lot of time thinking about that question. They mainly focus on the here and now.

John Gillin is attending school at the Kennedy Krieger Institute. He is working on learning the verbal and non-verbal skills of social interactions. John has a terrific memory, says his mom. “He can tell you all about Lewis and Clark’s trip. But it’s not conversation.”

Alex Bush attends a public elementary school that has a special autism outreach program. At home, therapists are teaching him occupational skills. Another goal, says Sandy Bush, is to look for ways to help Alex do more of the fun things that other kids do, like riding a bike. When Alex is ready, a new bike is waiting for him. Sandy Bush thinks about how she and her husband will teach Alex to ride. They will start by showing Alex how to operate the hand brake, although Sandy does not know how they will get their son to perform that challenging task. Then, perhaps Alex’s dad will hold onto the handlebars and push Alex along. He will help Alex place his feet on the pedals. Maybe as Alex coasts along, he will get the hang of pedaling.

Big efforts begin one step at a time.