The Pitfalls of Early Screening for Autism
Pulling no punches, Dr. Jose Cordero, the founding director of the National Center on Birth Defects and Developmental Disabilities recently went on record about the dire need to screen for autism at younger ages: "For healthcare providers, we have a message that's pretty direct about autism. And the message is: The 4-year-old with autism was once a 3-year-old with autism, which was once a 2-year-old with autism."
On the face of it, promoting early autism screenings is a nothing short of a public health imperative. The science is stacking up to show that catching and treating autism earlier leads to better outcomes. Dr. Laura Schreibman, who directs the Autism Intervention Research Program at the University of California, San Diego last year told American Psychological Association’s Monitor Magazine columnist Eve Glicksman: “Psychologists need to advise parents that the ‘wait-and-see’ approach is not appropriate when autism is expected. Delaying a diagnosis can mean giving up significant gains of intervention that have been demonstrated before age 6.”
What neither parents nor many doctors realize is that the younger the age at which we assess for problems, the greater the potential a slow-to-mature kid will be given a false diagnosis. In fact, as we screen ever younger children for autism, we need to be reminded that the period of greatest diagnostic uncertainty is probably toddlerhood. A 2007 study out of the University of North Carolina at Chapel Hill found that over 30 percent of children diagnosed as autistic at age two no longer fit the diagnosis at age four. Based on both my direct clinical experience working with children, and my reading of the data, Dr. Cordero’s statement greatly oversimplifies the issue.
Let’s be realistic. The likelihood that a slow-to-mature toddler will be misperceived as severely autistic is slim. On the extreme end, autism is, more often than not, a conspicuous, lifelong, disabling neurological condition. Roy Richard Grinker, in his acclaimed book Unstrange Minds, masterfully documents the challenges he faced raising Isabel, his daughter with pronounced autism. At age two, she only made passing eye contact, rarely initiated interactions, and had trouble responding to her name in a consistent fashion. Her play often took the form of rote activities such as drawing the same picture repeatedly, or rewinding a DVD to watch identical film clips over and over. Unless awakened each morning with the same greeting, “Get up! Get up!” Isabel became quite agitated. She also tended to be very literal and concrete in her language comprehension: expressions like ‘I’m so tired I could die” left her apprehensive about actual death. By age five, Isabel remained almost completely nonverbal.
When the signs of autistic spectrum disorder are indisputable, as in Isabel’s case, early detection and intervention are crucial to bolster verbal communication and social skills. The brain is simply more malleable when children are young. Isabel’s story in Unstrange Minds is a heroic testament to the strides a child can make when afforded the right interventions at the right time.
Diagnostic conundrums enter the picture when we frame autism as a spectrum disorder, (as it is now officially designated in the newly minted Diagnostic and Statistical Manual 5th Edition) and try to draw a bold line between a struggling toddler and one on the mild end of the spectrum. What is a doctor to make of a chatty, intellectually advanced three-year-old patient presenting with a hodge-podge of issues, such as poor eye contact, clumsiness, difficulties transitioning, overactivity or underactivity, tantrumming, picky eating, quirky interests, and social awkwardness? Does this presentation indicate mild autism? Or, does it speak to a combination of off-beat developmental events and difficult personality traits, resulting in an inflammatory parent-child relationship and a struggling kid?
The parallels between a slow-to-mature toddler and a mildly autistic one are so striking that there is a real possibility a false diagnosis. In an Australian study covered by Linda Carroll, a contributor to NBC News’ TODAY.com, it was shown that almost one in five two year olds are late talkers. They fall below the requisite 50 word expressive vocabulary threshold and appear incapable of stringing together two and three word phrases. This sort of irregular pattern of language acquisition is one of the hallmarks of early autism. Yet it is notoriously difficult to unscramble signs of autistic spectrum disorder from signs of delayed language development. The situation is further complicated by the fact that toddlers with delayed language development share other features in common with autistic spectrum kids. Scientific findings at the famed Yale Child Study Center point to how toddlers with delayed language development are almost identical to their counterparts with autistic spectrum disorder in their poor use of eye contact to gauge social interactions, the range of sounds and words they produce, and the emotional give-and-take they are capable of. Many tots who simply don’t meet standard benchmarks for how quickly language should be acquired and social interactions mastered are likely to be slotted into the spectrum.
Expanding autistic phenomena to include picky eating and tantruming only creates more befuddlement when applied to small children. Several years ago a study published in the Journal of the American Dietetic Association tracking over 3,000 families found that 50% of toddlers are considered to be picky eaters by their caregivers. The percentage of young children in the U.S. who are picky eaters and have poor appetites is so high that experts writing in the journal Pediatrics in 2007 remarked, “it could reasonably be said that eating-behavior problems are a normal feature of toddler life.” Tantrums also are surprisingly frequent and intense during the toddler years. Dr. Gina Mireault, a behavioral scientist at Johnson State College in Vermont studied kids from three separate local pre-schools. She discerned that toddlers tantrumed, on average, once every few days. Almost a third of the parents surveyed experienced their offspring’s tantrums as distressing or disturbing.
A dearth of pretend play is also thought to be a red flag for autism in early childhood. But in one of the few naturalistic studies that exist, researchers Rachana Karnick and Jonathan Trudge, associated with the Cultural Ecology of Young Children Project, ran the data on preschoolers who were watched in their everyday surroundings at play and discovered that the typical 3 year-old only pretend plays about 4% of the time. Lastly, it is ironic that the very procedures used by most evaluators to assess for autism can stress a vulnerable toddler to the point that he or she resorts to autistic-like coping behaviors in the examination room. Nobody has made this point more compellingly than the late Dr. Stanley Greenspan, the internationally recognized child psychiatrist who devised the popular “floortime” approach to treating autistic spectrum disordered kids. In a web-based radio show several years ago he cited an alarming statistic: of the 200 autism assessment programs his team surveyed across the country, most of which were housed in prestigious medical centers, only 10% emphasized the need to observe a child along with a parent or guardian for more than 10 minutes as they spontaneously interacted together. The norm was to swiftly separate a little one from his or her caregivers and go about the business of conducting a formal assessment. Dr. Greenspan himself was apt to sit back and keenly watch parents and children for 45 minutes or longer, waiting for choice points to delicately reach out to a child and probe for greater sustained eye contact, elaborate verbalizations, or shared emotional reactions. He was of the mindset that these conditions of safety and sensitivity were basic ways to get an accurate read on a child’s true verbal and social skills. Void of such conditions many toddlers simply clam up, fail to respond to their name, crawl under nearby furniture, stare blankly, or flail around—in short, look like they might just be on the spectrum.
Up-to-date figures out of the Centers for Disease Control and Prevention put rates of autism among kids in the U.S. at one in eighty-eight. That’s a 78% increase since 2002 alone. More children than ever before are being diagnosed by age 3. Twice the number of new cases are of the mild versus the severe variety, as pinpointed by a 2007 National Survey of Children’s Health data set. It wouldn’t be naÃ¯ve to deduce that we are overdiagnosising due to earlier screening and a softening of autism criteria. Make no mistake about it, screening for autism at earlier ages should be a top public health priority. But experts and practitioners need to do a better job of staying on speaking terms with the broad range of normal during early childhood. And like a sturdy toddler, they should be fussy—that is, about detecting clear-cut cases.