A commonly misunderstood aspect of autism and autism spectrum disorders (particularly by antivaccinationists and believers in the quackery known as “autism biomed”) is that autism is not a condition of developmental stasis. It is a condition of developmental delay. Autistic children can and do exhibit improvement in their symptoms simply through growth and development. However, parents who subject their children to “autism biomed” quackery of the sort championed by Jenny McCarthy and others seem to view autism as a condition of developmental stasis. That’s why they so easily and predictably attribute any improvement in their children to whatever quackery du jour they are using on them. It’s also why, in order to determine whether a given intervention in autism has any real effect, randomized controlled trials are required. Indeed, it’s not so difficult to see why, if you take into account the widespread belief that autistic children do not improve, along with parents’ imperfect human memories riddled with confirmation bias, confusing correlation with causation, and other confounders like regression to the mean, so many parents believe that “autism biomed” treatments have actually helped their children. Moreover, improvements observed in autistic children tend to be uneven, with periods of little change interspersed with periods of rapid development. Should such a period of rapid development appear after a “biomed” intervention, guess what gets the credit for the improvement?
But how much improvement is possible? Do autistic children “recover,” and, if they do, how much can they recover? The autism biomed movement is rife with stories of “recovered” children, but often, if you investigate these stories, they turn out to be less than convincing, not unlike the way that alternative medicine cancer “cure” testimonials tend not to be so impressive when examined closely. However, in the case of autism, this isn’t always the case. There are clearly children who lose their diagnosis of autism or ASD, with observations published as far back as 1970, when Rutter reported that 1.5% of adults who had been diagnosed with autism were functioning normally, while 30 years later Sigman et al reported that 17% of autistic children in their group lost their diagnosis and 10 years after that Kleinman et al reported that up to 19% of autistic children “lose their diagnosis.” The reason for this observation is hotly debated, and until fairly recently it was often assumed that these children’s recoveries were in fact not true recoveries but children who were either misdiagnosed or overdiagnosed. Such an assumption made intuitive sense because such an outcome is more likely with children diagnosed with Asperger’s disorder or pervasive developmental disorder, not otherwise specified (PDD-NOS), both of which are categories that resulted from the expansion of the diagnostic criteria for autism. Be that as it may, when you boil it all down, it is estimated that between 3% and 25% of autistic children “lose their diagnosis.” However, few of these studies explicitly address whether the social and communication abilities of these children are fully typical.
A recent study might help clarify what degree of recovery is and is not possible. Most of the previous studies before this have been small and did not look specifically at the outcomes people are curious about. Published in The Journal of Child Psychology and Psychiatry by Fein et al and entitled Optimal outcome in individuals with a history of autism, this study got some news coverage last week under titles such as Some With Autism Diagnosis Can Overcome Symptoms, Study Finds; Scientists seek clues in kids who outgrow autism symptoms; Some children outgrow autism: study; Health Buzz: Can Autism Fade Over Time?; and Children ‘may grow out of autism’. The authors set the stage in their introduction after surveying the literature, some of which I’ve touched on above: