The study, published in the Journal of the American Academy of Child and Adolescent Psychiatry, is among the first to glean the pattern of autism's emergence in very young children by following babies from the age of 6 months. At that age, babies who would go on to be diagnosed as autistic and babies who would develop typically showed no significant differences in social behaviors, including smiling, making eye contact and vocalizing responsively.
In its detailed comparison of 50 babies - half of whom would go on to be diagnosed with autism - the researchers found a steady loss of sociability and responsiveness in the babies who would progress to an autism diagnosis. Those babies' loss of social skills looked more like regression and less like a slowing of progress that allowed normally developing babies to pull far ahead of them. And that regression was most marked between 6 and 18 months, though it continued more gradually to the 3-year mark, where the study left off.
But while the reduced rates of face-gazing, vocalizations and social engagement were evident to researchers who systematically evaluated the babies every six months, 83 percent of the parents did not observe the changes chronicled by researchers - not, at least, in the first year they were happening.
In an accompanying editorial, Tony Charman, an autism researcher at London's Institute of Education, expressed surprise that so few parents of children who would go on to be diagnosed with autism observed the changes in their babies. That is especially unexpected because the babies considered to be at high risk of developing autism came into the trial because they had an older sibling who already had been diagnosed with autism-spectrum disorder. Since parents enrolling a baby in such a trial generally recognize his or her higher risk of autism, "we might expect worried parents to be hypervigilant for early signs that something is not right with their younger child," Charman wrote.