Despite years of research efforts, there is little understanding of the etiology or pathophysiology of Autistic Disorder (AD) and no treatment. Three significant developments have raised the possibility that progress can be made in unraveling the mysteries of AD: (1) morphometric analyses using MERI have shown that the brains of AD children ar larger in volume with significant differences in white matter, (2) the POMC pathway compounds B-endorphin and ACTH respond and interact differently before and after challenge in some children with AD and (3) the Human Genome Project has provided ahead of schedule a dense map of the human genome including about half of the expressed genes (ESTs). The observation that the brains of AD children are larger than normal controls has been documented by several research teams. This enlargement is localized to white matter and to the association regions of the brain. Can these findings be replicated in a newly-recruited cohort of late childhood children with AD? Using sequential MRI scan, can the effects of puberty on brain development in AD be defined? Are these findings consistent across the entire sample of AD children? Project 2: MRI is addressing these points. The observation that some children lose skills, social and/or verbal in the early course of AD is an anecdotal but consistent finding. Do these children have normal development and regress? Does a Full Autistic Regression exist separate from a Language-Only Regression? Does the neurobiological train from the MRI study differentiate these AD children with regression? Project 3: Regression will investigate a group of very young AD children and answer these questions about Autistic Regression. The first two new developments are from studies by researchers at UCI. Both the MRI and opiate findings could represent genetics differences between AD and normal children and/or within this group of children diagnosed as AD. Using the Human Genome framework and the genetic analysis expertise at UCI we can ask specifically if neurobiological differences represent the action of genes contributing to AD (several studies estimate the number of AD loci is 3-6). Are some of these half-dozen loci represented by the phenotypes of brain size? Are some of these half-dozen loci represented by the phenotypic characteristic of Autistic Regression subtypes? Project 1: Genetics is designed to answer this question. For AD, where we known so little by the underlying biology, it is essential that we pursue clues in a systematic fashion, using the state of our genetic knowledge as a tool. Our research provides clues on where to begin the genetics search. This grant is designed to exploit those clues to the fullest. It is unlikely that all cases, or even a large proportion of cases, of AD are purely genetic. But some AD surely is genetic, and finding those genes constitutes a critical first step towards understanding this disastrous disorder.