Skip to Navigation
University of Pittsburgh
Print This Page Print this pages

February 4, 2010

On Health: Autism

Increasingly, studies have shown, applespeople are seeking health-related information. A recent Pew Internet study found that 80 percent of Americans with Internet access turn to the web for answers to their medical and health questions.

But three-quarters of consumers fail to check how reliable and how current that information is, the study revealed.

In an effort to detangle some of the overload of health information that is out there, this new occasional University Times series, On Health, is turning to Pitt experts for current — and reliable — information on some of today’s major health-related topics.


More people than ever are being diagnosed with the developmental disorders classified as autism.

The condition, which impacts social interaction and communication, affects an estimated 1.5 million Americans, and has been labeled an “urgent public health concern” by the U.S. Centers for Disease Control and Prevention (CDC).

The CDC recently increased its estimate of the prevalence of autism. A December 2009 CDC report indicates that 1 in 110 American children have autism, up from an estimated 1 in 150 in 2007.

A study in the October 2009 journal Pediatrics found the rate for parent-reported autism to be one in 91 children, including one in 58 boys, based on data from the 2007 National Survey of Children’s Health.

The causes behind autism and effective treatments for it are not yet well understood, but Pitt researchers are contributing to the quest for answers through the University’s Center for Excellence in Autism Research, one of five such National Institutes of Health centers.

Directed by psychiatry and neurology professor Nancy Minshew, Pitt’s center is dedicated to seeking the causes of and treatments for the condition.

Because autism is a spectrum disorder, its effect on an individual can range from mild to severe. While symptoms can vary widely, common attributes include engaging in repetitive behaviors and difficulties with social interaction and communication.

According to the CDC, there are three types of autism spectrum disorders (ASDs):

Autistic Disorder, or classic autism, which includes social and communication difficulties, language delays, repetitive behaviors, unusual interests and, in many cases, intellectual disabilities.

Asperger Syndrome, which is marked by milder symptoms of autistic disorder, but typically without language problems or intellectual disabilities.

Pervasive Developmental Disorder-Not Otherwise Specified (PPD-NOS) or atypical autism, which includes people who meet some of the criteria for other ASDs.

Why the number of autism diagnoses is increasing is the subject of much debate. “It’s not that the disorder is burgeoning. It may be increasing. We don’t know,” said center researcher and psychology professor Mark Strauss.

Mark Strauss

Mark Strauss

One factor, said center researcher and psychology faculty member Sue Campbell, is that many children who previously would not have been diagnosed with autism now are getting an autism diagnosis. “The net got wider,” she said.

Strauss acknowledged that awareness is changing because autism is becoming a better-studied and better-diagnosed condition. Earlier detection and broader criteria for an autism diagnosis both impact the numbers.

He noted that the history of autism is brief; autism has been known as a disorder only since the 1940s. “Even back then, most kids who had the disorder were labeled as schizophrenic,” he said.

As late as the 1960s and ’70s, it was thought that autism came from bad parenting by distant “refrigerator moms,” adding further stigma to the condition.

“In terms of being a disorder that we recognize as a clearly neurologic disorder that is not due to bad parenting and that needs to be studied, that’s only happened within the last 20 years,” Strauss said. “That’s very new, so of course you’re going to have this burgeoning of recognition that there are many more kids with it and a better notion of how to define it.”

Strauss noted that individuals with autism vary widely in their intellectual capacity, ranging from children with IQs in the 20-30 range and no language skills to a small number at the other extreme with genius IQs.

Many high-functioning individuals previously may not have been identified as having autism, Strauss noted. “When I think back to my elementary school days, the kids we used to ostracize socially — I’m sure many of those kids probably had autism,” he said.

And, added Jana Iverson, a center researcher and psychology faculty member, lower-functioning children with autism might have received a diagnosis of mental retardation.

Early diagnosis

Campbell noted that autism is classified as a disorder of infancy, “but we can’t diagnose it in infancy, even though parents can tell you there’s something wrong.”

Sue Campbell

Sue Campbell

Currently, it’s difficult to diagnose autism reliably before 30-36 months of age, Iverson said. That means there can be a substantial gap between the time parents notice something amiss in their child’s development and an actual diagnosis is made.

“Many parents go to a pediatrician with concerns about their child’s development, sometimes as early as 12 months, but at the same time, many children don’t receive an autism diagnosis until they’re 3 and sometimes even older,” she said. “This is a very stressful time for parents because they don’t get services, they don’t get anything for their child, but they know something’s wrong. The work we’re doing is designed to address that issue and to reduce this gap by coming up with a set of early risk markers.”

How early could professionals diagnose autism? “This is the open question now,” Iverson said, noting that for some children there are indications by the age of 18 months that a diagnosis is on the horizon.

“It’s not perfect by any means and that’s a big concern,” she said, noting that it can be difficult to distinguish autism from other developmental delays.

Campbell added, “It’s partly difficult because early in development kids’ development is uneven. So, we may notice something that looks a little different about a child, but it might be something that’s going to resolve, or it might be something that’s really telling us there is something wrong. So it’s a balancing act to decide when you are over-diagnosing, when you are under-diagnosing.”

However, Strauss noted, their goal isn’t simply earlier diagnosis, but also how to chart differences in early development.

“The evidence clearly suggests autism has early origins, pre-age 2 or 3,” he said. “There are markers that show up very early,” such as differences in brain volumes or head sizes, which are identifiable well before 2 years of age.

“Partly what we’re trying to do is understand what is happening early and what’s different in children who end up with diagnoses of autism. So, even though it may not lead to diagnoses as early as 6-12 months, we’ll have a handle on how the course of the development is different.”

Early markers

Babies and toddlers should be screened for developmental milestones during their regular well-child checkups, but parents should not wait to request an evaluation if they have a concern about developmental delays.

For an interactive chart of developmental milestones for children through age 5, visit

Jana Iverson

Jana Iverson

Iverson said the early markers most widely replicated in current research literature are in the realm of early social communication.

Most infants are saying a few words by the age of 18 months, but nearly all are using gestures to convey their interests and wants and needs to the people around them, she said.

“Infants who end up with an autism diagnosis generally aren’t doing that even at 18 months. And if they are, it’s extremely infrequent.”

Campbell added, “The important part of that is it’s sharing — the social part of the pointing.”

Research has shown if by that age a baby isn’t pointing at objects to show interest in them, it’s a potential red flag. “That in particular is the marker that most people would feel most comfortable saying is pretty good. It’s not perfect, but it’s a good one,” Iverson said.

Fine-tuning the list of observable behaviors is one aspect, but underlying them may be much more subtle differences, Strauss noted. “You’re not going to be able to study those underlying subtle differences unless you take a research approach by studying the babies early and following them up,” he said.

One diagnosis or many?

While there may be a common set of diagnostic features, one problem in diagnosis is that children may have those features for different reasons — reasons that aren’t yet well understood, Strauss noted.

Iverson said, “One of the issues in DSM (Diagnostic and Statistical Manual of Mental Disorders) criteria is it takes a Chinese menu approach — one from here, one from there. You have a group of children who all have social communication deficits, but some of them have it in the area of early pretend play, but their language is developing fairly normally. Others are not developing any language at all. And they’re all kind of subsumed into this one large category. Most people who have experience with autism would tell you they get a definite sense that there are subgroups of kids who function in very different ways.

“We are absolutely not there at all in understanding what those subgroups are about in terms of the etiology. Just that they exist,” Iverson said.


Autism research today aims for a better understanding of the causes underlying the condition as well as ways to identify and treat autism sooner. Several teams of Pitt researchers are working to further what is known about risk factors in hopes of contributing to earlier diagnosis and the development of earlier treatments for people with autism.

“There’s no cure, but intervention does help,” said Strauss, who studies infants at risk for autism as well as older children and adults who have been diagnosed with autism.

“There are indications that early interventions do help these children. It’s not that you’re curing them, but you’re giving them skills to cope with the problems they have. You’re helping them learn alternative ways to do things, compensatory ways to do things, so in fact the outcome in some children looks fairly decent if you can get the intervention early enough,” he said.

Pitt is part of an international consortium of universities where researchers are studying babies at risk for autism in hopes of finding ways to identify them sooner.

The research teams are working with infants who have an older sibling who has been diagnosed with autism. Studying babies at higher risk means that a smaller group of infants is needed in order to yield a group who eventually will be diagnosed with autism, Iverson said, noting that the recurrence risk for families in which there already is a child with autism is about 15-20 percent.

“We would need thousands from the general population versus needing a group of about 100 infants to end up with 20 infants who will have an autism diagnosis,” she said.

Strauss noted that Pitt researchers are studying various aspects of the babies’ development. “We’re somewhat unique in that we are taking a broad approach,” he said. His part of the research is focused on cognitive ability, attention, memory and how knowledge develops. Iverson is focusing on communication skills such as language and gesturing, and Sue Campbell is focusing on early social development with a particular attention to mother-infant interactions. Strauss and Campbell are studying at-risk infants and a control group in the lab; Iverson’s research observes the babies at home.

“So we’re really covering all the major things that are developing in infants among the three of us,” Strauss said. “You can take one common thing that’s developing, like language and interaction, where all three of us are looking at components or subcomponents of it. We aim to bring it together to get a picture of why it looks wrong in particular kids who end up with the diagnosis.”

Iverson acknowledged theirs is a unique approach. “We’ve converged on this one problem all together.  Each of us is coming at it from our own perspectives, but in combination. By the time we’re finished, we’ll have so many measures on this common group of children that we can really start to look at not just specific profiles in our individual domains of interest, but look across domains and try to think about how early behavior in one domain might be related to patterns that we see in another domain later on.”

She said the practical goal is to develop an early screening tool for autism. “The basic question that we’re all addressing from different perspectives is can we come up with a set of signs that are apparent in infants at 12 months, or even younger if possible, that are good predictors of a future autism diagnosis?”

Part of the reason for studying infants, Strauss added, is that by the time a child is older and has exhibited symptoms, it’s too late to learn how they developed the symptoms or to trace their origins. “So if you want to try to sort out whether there are different ways to get there, you really have to study the developing abilities,” he said.

Another reason for working with babies, Iverson said, is that intervention research isn’t moving in lockstep with early identification research. While some professionals argue for intervening as early as 12-18 months, current practices are appropriate for use with 3-year-olds, not babies who have different skill sets and different interests.

“We make it to a point where we feel confident identifying infants who might be at risk, but then the problem is the intervention we have to offer doesn’t match. That’s one outstanding problem,” Iverson said.

Recruitment of babies is continuing, but Strauss said some preliminary data should be available within the next year.

He is optimistic that a set of early markers will emerge.

“Assuming that in fact we can pick these up, obviously they’ll need to be studied in a lot more detail in the future, but it’s certainly conceivable that we could start talking about interventions based on the problems we’re finding.

“The other critical thing we’re geared for doing is that even if you pick up differences, as we expect, very early — the big issue is what’s the long-term implication of the differences you’re picking up?”

That determination will require longer study of the infants  — at least through early to middle childhood — because children’s development can take many different trajectories. “This long-term picture is so critical so we can look to see: The kids we identified as kind of being suspect, do they actually stay that way? Or is it the case that we see a totally different child emerging several years down the road? We just don’t know that yet,” Strauss said.

Long-term study is needed to show whether the screener researchers develop is related reliably to the diagnosis. “If you don’t have this long-term picture you don’t know if your screener is going to be worth it,” he said.

Addressing the causes

Getting to the root causes of autism is complicated. “It’s a multiple pathway, with multiple indicators to this hodgepodge of symptoms,” Campbell said. “There’s lots of variability in the causes, in the clinical picture, but it’s subsumed under this group of autism.”

Although many potential genetic markers have been identified, Campbell said, “There are going to be multiple genetic pathways with multiple genes in each pathway as well as potentially other events, like maybe some kind of pregnancy complication, or illness during pregnancy.”

The controversial debate linking autism to vaccines continues, although the CDC states that “evidence from several studies examining trends in vaccine use and changes in autism frequency does not support such an association.”

While vaccines are not completely safe — some people do have reactions to them — Strauss said there is no indication they cause autism.

He noted that the initial study that ignited the controversy linking autism and vaccines’ mercury-based thimerosal preservative has been seriously questioned and recent studies show no association. Now some parents wonder whether administering multiple vaccines at one time may be problematic.

A 2008 federal court decision found the administration of nine vaccines in one day significantly aggravated underlying mitochondrial disease in a young girl, causing autism-like symptoms.

Autism and mitochondrial disease organizations are joining forces to look into anecdotal reports that mitochondrial disorders may be more prevalent in people with autism spectrum disorders.

Environmental exposures or stress also may contribute to autism, Iverson said.

A 2008 Harvard Medical School study of babies born in Louisiana found that autistic disorders were more prevalent in those who were exposed to hurricanes in utero, with the prevalence increasing not only with the severity of the storm, but also depending on their gestational age when the storm occurred.

Iverson said the incidence was highest during the 5th and 6th months and the 8th and 9th months of pregnancy. “That suggests something about stress that potentially could be important, but also the timing of the stress,” she said.

“There’s never going to be a universal cause,” Iverson said, noting that an overemphasis of attention on one potential cause can short-change research into other factors.

Long-term outlooks

Although intervention is not a cure, earlier intervention can have long-term implications, the researchers say.

While autism is considered a disorder of childhood, it’s actually a lifespan disorder, Iverson said, adding that few services — such as help with finding and keeping a stable job, or succeeding in college, for instance — exist for adults with autism.

Individuals with autism, Strauss said, “tend to be very good at any course that requires any sort of logic or deductive reasoning. They do very well at mathematical courses, computer science and engineering. On the other hand, they have difficulties with abstract or inductive reasoning. So English and philosophy courses can be very difficult,” he noted.

“So you have very high-IQ individuals sometimes, but getting through college can be very hard,” he said.

Likewise, people with autism tend to be very detail-oriented, which would be a strength in certain jobs. “There are a lot of things they could potentially bring to the table that would be very useful to many employers,” Strauss said. “The problem is helping them stay there,” he said, citing the case of one study participant who holds a civil engineering degree and is a math savant, yet recently was fired from his job stocking grocery store shelves.

“It’s a terrible problem and there’s been very little publicity on older individuals with autism and their needs and outcomes,” Strauss said.

—Kimberly K. Barlow

Some signs & symptoms

According to the Centers for Disease Control and Prevention, those with an autism spectrum disorder might:

Not respond to their name by 12 months of age.

Not point at objects to show interest (for example, point at an airplane flying over) by 14 months.

Not play “pretend” games (pretend to “feed” a doll) by 18 months.

Avoid eye contact and want to be alone.

Have trouble understanding other people’s feelings or talking about their own feelings.

Have delayed speech and language skills.

Repeat words or phrases over and over (echolalia).

Give unrelated answers to questions.

Get upset by minor changes.

Have obsessive interests.

Flap their hands, rock their body or spin in circles.

Have unusual reactions to the way things sound, smell, taste, look or feel.

Leave a Reply