Predicting Autism

Of all the misfortunes a child can suffer, few provoke as much dread as autism. The condition--a neurological disorder that impedes language and derails social and emotional development--has become ever more common in recent decades, thanks partly to better diagnosis. Experts now suspect that one person in 160 lives with some degree of autism. That's three to four times the rate in the 1970s. But while the outward manifestations are well known, science is just beginning to illuminate the underlying biology. What goes wrong in the autistic brain? What defect or injury leaves it largely incapable of empathy? A growing body of evidence, capped last week by new findings from the University of California, San Diego, raises a tantalizing possibility. The new study, published in The Journal of the American Medical Association, links the condition to abnormally rapid brain growth during infancy--and it raises new hopes for diagnosis and treatment.

The key to last week's finding was not a million-dollar imaging device but a tape measure. Past studies have shown that autistic toddlers have abnormally large brains for their age. But because autism is rarely detected in kids younger than 2 or 3 years old, researchers have never known quite how that situation arises. Two years ago the San Diego team realized that children's old medical records might hold important clues. Led by neuroscientist Eric Courchesne, the researchers tracked down early-childhood head measurements for 48 autistic preschoolers, and compared them with national norms. As it turned out, the kids' heads had been smaller than average at birth but had grown explosively during infancy, shooting from the 25th percentile to the 84th in roughly a year's time. And faster growth predicted greater impairment. Mildly autistic subjects reached only the 59th percentile (chart), but the severely afflicted kids reached the 95th percentile.

The implications are hard to miss. Autism, the new findings suggest, is not a sudden calamity that strikes children at the age of 2 or 3 but a developmental problem that can be traced back to infancy. That alone should help allay the suspicion that autism is caused by vaccines or pollutants that kids encounter later in childhood. But the new findings say less about the causes of autism than about its dynamics. The current study focuses on the first year of life, but the trouble isn't confined to that period. Other recent studies suggest that the early growth spurt is followed by several years of slower expansion, giving the autistic child an adult-size brain by the age of 4 or 5. During adolescence and adulthood, autistic brains are generally no larger than normal ones. Unfortunately, they exhibit a range of other anomalies, including dense clusters of underdeveloped cells in the hippocampus and amygdala--structures that are critical for integrating emotional and sensory information.

Does rapid growth actually cause all this damage? It's still an open question. "The abnormal growth patterns give you a clue that something is amiss," says Dr. Margaret Bauman, a neurologist at Harvard Medical School and the ladders Autism Research Foundation, "but we can only guess at the underlying process." Courchesne believes it can be summed up in three words: "growth without guidance." Normal brain development is not a monologue but a dialogue, in which the brain generates neural circuits and the child's experiences determine which ones survive. The first year of life is a critical period for this "experience-guided growth"--and it's not hard to see how a sudden shift into high gear might derail it. The brain's circuitry would expand haphazardly as cell growth outpaced experience, creating a chronic sensory overload. Courchesne hopes researchers will now confirm the dangers of unregulated brain growth by inducing it experimentally in animals. "Once we know what causes this growth defect," he says, "it may be possible to use biological treatments to counter it."

The more immediate goal is simply to recognize autism at earlier stages, and to give affected kids the support they need to grow and learn and cope. Will the new findings advance that cause? Dr. Janet Lainhart, an autism expert at the University of Utah, is skeptical. "The findings... are most useful to researchers attempting to define the underlying developmental neuropathology of autism," she writes in |a commentary on the San Diego study, "rather than to physicians trying to identify young children with autism." That's because rapid head growth can signal other childhood maladies, including tumors and hydrocephalus, and often means nothing at all. Lainhart calculates that if doctors used head circumference as a screening test for autism, they would pick up 60 healthy children for every autistic one. Courchesne concedes the point, but he still believes it's prudent for pediatricians to monitor head growth. "The world's oldest measurement tool still has the power to amaze," he says. It may not provide a definitive diagnosis, but it's inexpensive, noninvasive and objective--and most of the concerns it raises can quickly be resolved. Where autism is concerned, that's still as good a goal as any.