The role of altered gene-dosage (copy-number variation) in adult-onset neurological disorders has recently been shown. An extra copy of the gene APP causes familial early-onset Alzheimer disease (AD) with involvement of the blood vessels, extra copies of the gene SNCA causes familial Parkinson disease (PD), and an extra copy of the gene LMNB1 causes a disease resembling a type of multiple sclerosis. The frequency of these copy-number changes in non-familial cases has not yet been studied. These extra copies might occur frequently as architectural features of the human genome predisposes to extra or missing copies of the genes. AD has a substantial genetic component: fraternal twins are more likely to have AD if their twin has it than maternal twins and about half of the patients with AD have a family member who is also affected. It is conceivable that a type of genetic change that appears to happen quite frequently may be contributing to a common disease with a strong genetic component. I propose an exploratory pilot study to look for these copy number variations in 200 cases of AD by a new method called array CGH. I propose to look at 3 sets of genes: i) genes that have been shown to have copy number changes (APP, SNCA), ii) all genes and chromosomal regions which have been associated with AD and iii)genes that contribute to the neurovascular unit(the special interaction between brain vessels and surrounding other cell types). The neurovascular unit has been implicated in the pathogenesis of AD with multiple lines of evidence. Looking at certain genes (candidate genes) i) focuses the analysis and increases the likelihood of pattern recognition, ii) allows higher resolution of regions of interest, thus increasing sensitivity at reduced cost, and iii) is feasible with the ability to custom design/redesign arrays at no additional cost. During the design process, the candidate genes/regions were bioinformatically analyzed for architectural features that may predispose to increased or decreased copy numbers in databases of the human genome sequence. Boundaries for candidate genes and regions were chosen to include the region harboring the candidate genes/regions, the architectural features, and regions flanking the architectural features (0.7-6 Mb). I propose to design follow-up case-control studies to assess the significance of the identified copy number change in a given gene by a method called RT-qPCR. While the array CGH allows the concomitant detection of changes in many genes in one individual, the RT-qPCR allows the detection of the specific one change in many individuals at the same time. The sample size of the case-control studies will be calculated based on the data from the exploratory study.