Medical Examiner

Could a 3-Year-Old Just “Disappear”?

Why parents shouldn’t worry so much about regressive autism.

regressive autism.

Fear over regressive autism is unwarranted.

Photo by Ingram Publishing/Thinkstock

A few days ago, an old friend sent me a panicked email. She had just finished reading Ron Suskind’s beautiful essay in the New York Times Magazine about raising a son with autism: “Reaching My Autistic Son Through Disney.” Suskind describes how, at almost 3 years of age, his son Owen “disappeared.” The child was once “engaged, chatty, full of typical speech,” but then he stopped talking, lost eye contact, even struggled to use a sippy cup.

Owen was eventually diagnosed with a regressive form of autism, which Suskind says affects about a third of children with the disorder. “Unlike the kids born with it,” he continues, “this group seems typical until somewhere between 18 and 36 months—then they vanish.”

That was the line that alarmed my friend, whose son is nearing his third birthday. “What is this ‘regressive autism?’ ” she asked me, the resident autism expert in her peer group. (I conducted research on autism and regression in autism before becoming a freelance writer.) “I thought we were out of the woods!”

I’m sure many parents of young children who read the piece had the same reaction, and it’s completely understandable. It’s also unwarranted. The claim that many kids with autism develop typically for almost three years and then experience a near-complete loss of language, social skills, and motor ability—a claim I’ve read many times before—simply isn’t true. It’s time to set the record straight.

The kind of developmental pattern that Suskind’s son experienced is very rare. Most children with autism show signs of the disorder in the first two years of life. Yes, studies have suggested that about one-third of children with autism experience some kind of regression, but most of these children do not have typical development to begin with. Instead, they have early delays and lose some of the skills they had attained. This finding shows up time and time again; it’s what I found when I published a paper on regression in 2006, and it’s what researchers are finding today. Losses also usually occur considerably earlier than Owen’s. A recent meta-analysis of 28 studies found the average age of regression to be a bit older than 21 months; studies consistently report a range of 15 to 30 months.

What’s more, recent research strongly suggests regression is not an all-or-nothing phenomenon. In one study of 167 children on the autism spectrum, the majority acquired and lost some skills within the first two years of life. Some children lost only one or two of the 15 skills the study examined (most often first words, responding to their names, and smiling) and some lost more than that. Only 6 percent of the sample lost all of the skills they had acquired, as it sounds like Owen did, and those who did often had few skills to begin with.

Suskind’s story, and stories from other parents who’ve watched their children regress so dramatically, are real and haunting. As a parent, I can imagine few things more devastating than watching my exuberant daughter, nearing her second birthday, simply vanish.

But these cases are rare. Stories such as Owen’s make it seem as though unusual experiences are commonplace, and they unintentionally feed parents’ already acute autism-phobia, a creeping anxiety about autism that leads parents to worry when their children don’t meet developmental milestones precisely on schedule. I blame this anxiety partly on alarmist ad campaigns by autism research and advocacy agencies, which often describe the prevalence of the disorder in sensationalist terms.

Suskind’s piece should evoke compassion and hope: Owen, now a young adult, is doing remarkably well, in no small part because of the creativity and determination of his parents and therapists. But it should not evoke fear. As parents, we already agonize plenty over our kids. Let’s not worry about things we don’t need to worry about.