Can Early Intervention Prevent Autism?

Even before he was born, Kristin Hinson knew her son Noah’s odds of developing autism were high. When a child is diagnosed with the disorder, a younger sibling has a heightened risk of receiving the same diagnosis. That risk is further magnified when one has two older siblings on the spectrum. Noah has three older siblings, and two of them have autism.

When I reach Hinson over the phone, she’s cutting the kids’ hair (she has 5). She’s currently working on Justin, her second oldest, now 11, whom she periodically shushes as she tells her story. Her oldest child Amelia, says Hinson, a stay-at-home mom in Lincoln, California, was a typical child so it never occurred to her to worry about Justin. In retrospect, she says, he showed some warning signs as a young toddler, such as a weird obsession with things that opened and closed, like buckles, windows, and doors, and delayed talking, but he wasn’t diagnosed until age three. By the time Simon came along a few years later, Hinson had started to recognize those early red flags and she soon enrolled him in a study focusing on the younger siblings of autistic children at the University of California, Davis MIND Institute. Younger siblings like Simon are almost 15 times more likely to wind up on the spectrum than children in the general population. Simon was diagnosed with autism at just 15 months.

When Noah came along in 2010, Hinson took him to the MIND Institute when he was just 6 months old. Researchers there soon observed that Noah didn’t turn in response to his name and wasn’t babbling by nine months like other babies his age. Sally Rogers, a psychologist at the MIND Institute, suggested enrolling in a 12-week early intervention program aimed at helping babies at high risk of developing autism. Hinson readily agreed.

In recent years, developing and testing early intervention programs for babies and young toddlers at risk of developing autism has become a hot area of research. A firm diagnosis in some cases is now possible by 18 months of age or even slightly earlier, and as research into the disorder progresses, scientists are finding that some early indicators, such as delayed babbling and a loss of ability to visually track a caregiver around a room, might appear in infancy.

At 12 months, the signs of autism are often so subtle that they can be easily missed.

Yet established intervention programs are geared at older children, leaving parents like Hinson in limbo. That gap is particularly frustrating as the brain is thought to become less malleable with age. Very early intervention programs, then, hold tremendous promise. They have the potential to mitigate autism’s severity or, in theory, prevent the disorder from manifesting at all. Gordon Ramsay, director of the Spoken Communications Laboratory at the Marcus Autism Center in Atlanta, Georgia, likens babies showing early signs of autism to trains that are headed for a crash. “Can you,” he asks, “get the engineer to flip the switch” in time?

However, current methods can only identify at-risk babies—they don’t offer a definitive diagnosis. Moreover, early childhood development is naturally and notoriously uneven. Many babies outgrow those early warning signs while others who appear normal in infancy regress later on. Treating babies who may or may not develop a lifelong and debilitating disorder is fraught with ethical challenges. Asking families to participate in trial early intervention programs could cause undue anxiety at best, or they could even throw development off kilter in unexpected ways, says Jonathan Green, a child psychiatrist at the University of Manchester in the U.K. As one researcher put it to me: “We can’t be cavalier about this. False positives really matter.”

Translating Treatments

Up until a few decades ago, autism was thought to result from frigid parenting. That’s not the case anymore as researchers have shown that autism’s roots are biological rather than environmental. Nonetheless, parental involvement remains pivotal to autism treatment. When Michael Siller, a developmental psychologist at Hunter College in New York City, was in graduate school, for instance, he found that adult autistics’ language ability was heavily influenced by how much their parents communicated with them at age four. While parents aren’t to blame for their child’s autism, the reasoning now goes, perhaps, with the help of specialists, they can do something to limit its severity.

That logic seems especially valid for very young children, for whom the parental relationship is central. But it’s not as easy as simply shifting early intervention programs for preschoolers over to infants and younger toddlers, wrote Lonnie Zwaigenbaum, a developmental pediatrician at the University of Alberta in Edmonton, Canada, in a 2009 commentary. The two cohorts play and communicate in vastly different ways. Programs geared at autistic preschoolers, which often entail 20 to 40 hours a week of classroom time, would not translate well to babies and younger toddlers. Very early intervention programs, Zwaigenbaum and colleagues stressed, should occur in “natural learning environments” such as the home.

Researchers designing programs geared at babies and young toddlers have largely taken that advice to heart. Most trial programs—and there are a lot of them—involve parents and children meeting with a trained clinician in their own homes or in a lab setting meant to mimic a home environment. Sessions are short and parents are expected to implement the concepts they’ve learned through daily activities, such as during diaper changes or while reading a book. “Maybe a two or three year old can handle 20 hours a week or even 40 hours a week. How much can a 12 month old really take?” asks Amanda Steiner, a clinical therapist for kids with autism in Placerville, California, who worked on one such pilot project while on faculty at Yale University in New Haven, Connecticut. “Working with the parent or caregiver seems the way to go.”

Early Hope

For Sally Rogers, developing an intervention program for babies seemed the natural extension of her MIND Institute work with the infant siblings of children with autism. From her observations of babies like Noah, she knew that it was possible to detect some very early warning signs of the disorder. “It seemed an appropriate time to test whether those symptoms were at all plastic,” Rogers says.

Rogers recruited seven families whose babies, ranging in age from seven to 15 months, seemed to be at risk of developing autism. The families then visited for an hour a week for 12 weeks with a trained clinician in a lab setting made to resemble a home. There, clinicians helped parents address atypical behaviors. For instance, sometimes babies would coo quietly to themselves rather than interact with others in the room, Rogers says. So the clinician would have the parents mimic the sound to convey to the baby that “language” is a way of engaging with other people.

To the uninitiated, such training may sound mundane or even banal, but when children act in unexpected ways, parents often respond in kind rather than forcing a behavior that feels unnatural. Consequently, negative or repetitive behaviors solidify over time. For instance, autistic babies often play better by themselves, so parents are often inclined to just leave them alone, Hinson says. “Parents get in the habit of doing whatever makes their babies happy.”

Rogers’ findings were striking. At 36 months of age—when a reliable autism diagnosis becomes possible—six of the seven infants tested normal, while one showed mild signs of the disorder. The media went wild. “Treating infants for autism may eliminate symptoms,” read the headline from NBC News. “Early intervention in kids with autism eliminates symptoms, developmental delay,” wrote the Business Standard.

Yet Rogers’s study was small and her findings preliminary. “A lot of the publicity after Sally’s study was really hugely overblown,” says Jonathan Green of the University of Manchester. “That’s one of the dangers in our field is that it’s such an interesting area that people hype up the results.”

Green has just published his own larger, randomized trial of 54 families with infants who had autistic older siblings. Half of the parents were trained to adapt to their infant’s communication style while the other half were not. The babies, ranging in age from seven to ten months, were randomly selected from a pool of infant siblings of children with autism, which meant that 20 to 30% would be expected to develop the disorder.

The parents in the intervention group went through 12 training sessions. After the babies reached 14 months, their behavior was assessed again. Of the 16 different measures Green and his colleagues recorded, only one was significant—the amount of time it took a baby to shift his or her attention to a parent or caregiver. (A slow response has been linked to autism.) In the study, infants whose parents had received the intervention training responded 50 milliseconds faster than those whose parents hadn’t, a substantial change. In the 15 other behaviors and markers, though, the researchers could not detect a significant difference.

A limitation of the study is its limited sample size. Though it is large by autism research standards, the number of participants is still less than half what Green would like it to be. Another caveat is the early age at which the intervention took place. Green and his colleagues won’t know for sure whether their approach worked until the children are three, the age at which the disorder can be reliably diagnosed.

The Challenges Ahead

Establishing a firm autism diagnosis in babyhood is never going to happen, says Michael Siller of Hunter College, because early child development is too unpredictable. “You’re not going to get any autism diagnoses below 18 months. There’s just no way.” As such, all one can do is screen for babies at risk with the knowledge that many of those babies will go on to develop normally and won’t need any intervention. Even between 18 and 24 months, when a firm diagnosis is possible, Siller says, parents are overwhelmed and often unready to even think about early intervention.

It also poses an ethical quandary. In Green’s study, there were hints that intervention may have impeded language development, though the data were not statistically significant. Yet it does highlight the possibility that early intervention could affect a child who would have gone on to develop normally.

In some ways, focusing on the younger siblings of children already diagnosed with autism circumvents some of these issues. In addition to being higher risk, such families “know much more and may be more likely to accept the news,” Siller says. “That’s a very different group from an average family who has their first child with autism.”

But it isn’t clear that the inherited form of autism seen in infant siblings is representative of autism in the general population, which is likely several disorders under a single name. Moreover, infant siblings represent only a small fraction of kids who go on to develop autism. Identifying only those children ignores the vast majority of babies without an older sibling on the spectrum who will go on to develop the disorder.

Yet with the promise so high, researchers have been looking for ways to get parents on board without causing undue stress. Since the late 1990s, Grace Baranek, an occupational therapist and developmental psychologist at the University of North Carolina in Chapel Hill, has been analyzing recordings of babies who later went on to develop autism. She was able to observe atypical visual and sensory behaviors, such as fixating on peripheral stimuli like lights or spinning ceiling fans to the exclusion of all else.

Now, Baranek has developed an early screening questionnaire for one-year-olds with questions like: “When your child sees other children, does your child seem interested in them?” Or “Does your child spend a lot of time spinning, wobbling, or rubbing objects?” She has found that the questionnaire correctly identifies one-year-olds with autism about 30% of the time and more general developmental delays, including autism, 85% of the time. For Baranek, that suggests that the focus should not be on whether or not a child will go on to develop autism but on how to “intervene with very young children without a firm diagnosis.”

Baranek herself is trying to answer that question through early intervention studies at UNC. Because the questionnaire is optional, Baranek suspects that parents for whom it would cause too much anxiety simply opt out. Other families can ignore her when she reaches out to say their child may be at risk for developmental delays. For families that do participate, the response has been largely positive, Baranek says.

Cautious Optimism

For her part, Hinson is convinced researchers like Rogers are onto something. As a baby, she says, Noah wouldn’t look people in the eye or gaze at them as they wandered about the room. And if Hinson sang him a song and paused to build up suspense, he’d quickly lose interest. But then at 15 months, she says, something changed. Noah began interacting with people and tracking them with his eyes. “Everybody noticed,” Hinson says.

Now Noah is a typical 4-year-old. He’s never been diagnosed with autism. Hinson says she’s skeptical by nature, but she’s cautiously optimistic about such programs. “I don’t know that it’s a cure…but my son was headed down that road,” Hinson says. Then “all of a sudden, this light turned on.”