Hidden in plain sight.

This issue of IJE includes a theme on autism edited by two of us (Ezra Susser and Michealine Bresnahan). Our colleague Guohua Li joins us to write this introduction. In choosing this topic we venture into territory marked by controversy and conflict. The heat of the conflict is reflected in two recently published books, reviewed here by Stein and by Grinker. One of the flashpoints of controversy pertains to time trends in autism and autism spectrum disorders (ASDs). ASDs are variously defined in epidemiological studies, but usually include ICD-10 Childhood Autism, Atypical Autism, Other Pervasive Developmental Disorder, Pervasive Developmental Disorder Unspecified and Asperger’s Syndrome. Recent reports suggest that 0.4% of children are affected with autism, and 1% affected with ASD. This is about a 10-fold higher prevalence for autism than was found in the earliest research such as the classic 1966 study by Lotter, and 5-fold higher than was found for autism in most studies as recently as the 1990s. A similarly dramatic time trend is found when attention is extended to the broader category of ASD. These figures have generated intense debate over whether the observed increase is mainly ‘real’ (increasing incidence) or ‘artefactual’ (increasing ascertainment). The original research piece presented by King and Bearman, commentaries on this article by leaders in the field and the paper by Naaser et al., contribute important data and thoughtful perspectives to this debate. Naaser and colleagues draw on the registries of Western Australia to illustrate the concurrence of changes in diagnostic criteria and available services with periods of increase in ascertainment, providing a type of ecological evidence for factors other than increasing incidence playing an important role in the autism and ASD time trends. King and Bearman take a quite different approach. They draw on the California Department of Developmental Services, a source that has figured large in the controversy over time trends in autism. They use the individual-level data, available from this source, to estimate the proportion of the increased prevalence of autism that could be due to diagnostic substitution and accretion. Based on observed changes in diagnosis, the paper infers that a significant part of the increase can be attributed to a shift away from mental retardation alone towards autism (with or without mental retardation). The commentary by Charman et al. also presents intriguing data that illustrate the impact of varying diagnostic criteria. In this commentary they include a remarkable comparison of autism diagnoses made by the old Lotter criteria, which focus on social aloofness and rituals, and newer ICD-10 criteria for the same children. The experts applying Lotter criteria included Lorna Wing and Susan Bryson who were familiar with the use of the criteria from the early time period. A significantly lower proportion of children were diagnosed with autism by Lotter criteria than by ICD-10 criteria. Although all of these articles present evidence that some of the time trend is related to increasing ascertainment, they also concur that a significant portion of the time trend remains unexplained. If any part of this represents an increase in incidence—as seems likely—we are pressed to intensify the search for environmental contributions to aetiology. As Rutter points out, this investment must include studies with the means to identify these factors in the relevant time frame—prospective research beginning in gestation with careful case identification and ongoing followup. Such methods are already being pursued in some studies such as the Autism Birth Cohort nested in the Norwegian Mother and Child Study, * Corresponding author. Department of Epidemiology, Mailman School of Public Health, 722 West 168th Street, NY 10032, New York, USA. E-mail: [email protected] 1 Department of Epidemiology, Mailman School of Public Health, Columbia University, New York, NY, USA. 2 New York State Psychiatric Institute, New York, NY, USA. 3 Department of Psychiatry, College of Physicians and Surgeons, Columbia University, New York, NY, USA. 4 Department of Anesthesiology, College of Physicians and Surgeons, Columbia University, New York, NY, USA. Published by Oxford University Press on behalf of the International Epidemiological Association