Eric B. Larson, MD, MPH, MACP - US grants

The goal of this proposal is to establish a population based registry of patients with Alzheimer's type dementia (DAT) and other dementias and to collect epidemiologic, genetic (pedigree), and clinical information for the investigation of etiologic factors related to DAT. Group Health Cooperative (GHC) central membership (approx. 25,000 over age 60) will form the population base. Incident cases of probable dementia presenting to GHC facilities will be identified, screened, and diagnostically evaluated. Approximately 150 of the projected 230 new dementia cases occurring yearly are expected to DAT. Since the registry is population based, incidence rates can be calculated. Sensitivity and specificity of the screening device will be evaluated; eventually accuracy of the clinical diagnosis also will be evaluable because neuropathologic data will be collected for validation of the clinical diagnosis of DAT. This registry will serve as the primary resource for future population based studies of the etiology of DAT and other types of dementia in conjunction with the University of Washington Alzheimer's Disease Research Center.

Existing studies suggest that, while the overall prevalence rates of dementia in Japan and the U.S. are similar, the relative prevalence of Alzheimer's disease (AD) to vascular dementia (VaD) is much greater in the U.S. It is not known whether these differences result from differing study methods or whether they reflect underlying differences in the incidence of dementia subtypes in the two countries with concomitant differences in genetic susceptibility and/or exposure to environmental risk factors. Results from a prevalence survey of a total population of Japanese-Americans aged 65 and over living in King County, WA show that prevalence rates more closely resemble those of Caucasian populations than native Japanese. We propose to (1) follow this cohort over the next five years to (2) determine age- and sex-specific incidence rates of dementia and its subtypes; (3) follow existing prevalent and developing incident cases to monitor the clinical and neuropsychological progression and describe survival in AD and VaD cases, (4) conduct prospective risk factor analyses, including survival analyses, of genetic (apolipoprotein E and family history) and environmental factors for AD and VaD, particularly those affected by lifestyle and diet, and examine potential interactions between genes and environment on the risk of dementia and its subtypes; (5) compare prevalence, incidence as well as relative and attributable risks with methodologically standardized studies in Honolulu, Hiroshima and Machida City. The baseline examination (1992- 1994) contained nearly 2,000 individuals; to cohort of cognitively intact persons contains approximately 1,650. The cognitively intact cohort will be screened every two years, the persons with cognitive impairment are followed annually. Pending the results of an analysis of our screening strategy during the currently funded grant period, we plan to screen the cognitively intact cohort every two years using a cutoff score of 81/100 on the Cognitive Abilities Screening Instrument (CASI) and/or a decline of 7 points at any CASI level, to estimate age and sex-specific incidence rates. Patients scoring below the cut scores will have routine, standardized neurologic and neuropsychological evaluations. The followup of the cohort to the year 2001 will yield approximately 127 new cases of dementia. Risk factors for dementia subtypes will be examined in this group, including possible protective factors which show preliminary associations with cognitive impairment and prevalent disease, such as estrogen therapy replacement, head circumference and the apolipoprotein E-e2 (apoE) allele. Changes over time in CASI scores in the cognitively intact cohort also will be analyzed with regard to apoE genotype and other host factors. Finally, standardization efforts between the study sites will focus on the enhancement of the uniformity of diagnostic methods, and cross-site data analysis to compare rates and risk factors.

Literature from Japan suggests that the prevalence rate of vascular dementia far exceeds that of Alzheimer's disease in that country. In the U.S., the opposite trend is observed. However, direct comparison of prevalence rates is confounded by the lack of standardized diagnostic criteria. A population-based study of dementia in Japanese-Americans over age 65 residing in King County, WA is proposed. The project has five main aims: (1) to establish age- and gender-specific prevalence rates for dementia subtypes in this population, (2) to gather information on potential risk factors for dementia, specifically for Alzheimer's disease, (3) to follow prevalent cases longitudinally to describe disease progression and survival in dementia subtypes, and to obtain autopsy diagnosis on cases dying during the course of the study, (4) to develop a cognitively intact fixed cohort, and (5) to compare our assessments of Japanese-American elderly to parallel studies being conducted in Honolulu, Hawaii and Japan. The base population will be censused to identify persons of Japanese-American origin over age 65. A screening instrument (CAST) will be administered to all identified persons. A validation study will determine the CAST cut-off score, and this score will determine whether or not a person continues the diagnostic process or becomes a member of the cognitively intact cohort. Persons screening positive will be referred to an algorithmic evaluation, which will determine whether cognitive impairment was acquired and progressive, and if so, whether associated with cerebrovascular disease or not. Persons determined to have such a dementia will be further referred to a clinical evaluation for conventional diagnosis. The first two phases are designed in parallel with methods used in Honolulu, Hawaii, where the Honolulu Heart Watch cohort is being evaluated for prevalence and incidence of dementia. Methods in Japan will also be standardized with those in Honolulu. Cross-national comparison of prevalence rates of dementia associated with cerebrovascular disease and primary degenerative dementia will use diagnostic classifications from the algorithmic evaluation. Clinical evaluations in the proposed study will be based on the CERAD battery (Consortium to Establish a. Registry for Alzheimer's Disease). A case-control study will be conducted to examine risk factors using both prevalent and incident cases. Prevalent cases will be followed to study disease progression and survival among dementia subtypes, and to obtain neuropathologic diagnoses for cases who die during the course of the study. During the study period, a fixed cohort of cognitively intact persons will bc formed and monitored bi-annually. A renewal application will be submitted for continued followup of the cohort, to study incidence and risk factors associated with dementia subtypes in this ethnic population.

The University of Washington, Group Health Cooperative Alzheimer's Disease patient Registry (ADPR) is a model incident case registry for dementia and Alzheimer~s disease developed in 1986 in response to a National Institute of Aging request for proposals. This application is a competing continuation of the ADPR. The application proposes to continue follow-up of cases enrolled between January 1987 and may 1996, during which approximately 1,000 patients were enrolled. These cases have been and will continue to be part of ongoing studies of diagnostic markers, natural history and have served as sources of cases for other investigators at the University of Washington and throughout the country. During the last funding cycle, we established a cohort of 2,58 persons over age 65 from the Group Health enrollment. Based on biennial follow-up examinations, we are detecting all incident cases of Alzheimer~s disease and related dementias from this cohort. This application proposes to continue our follow-up and maintenance of this cohort. Our goal is to estimate age-group specific incident rates of Alzheimer~s disease and related dementias and to test environmental and genetic risk factors which have been previously identified in case-control studies. In addition, the methods of the cohort study have been standardized to companion cross-cultural studies of Japanese-Americans in Honolulu and Seattle, and Japanese in Hiroshima. Since these studies have similar design and data collection protocols, we propose to compare our results wit the results from these affiliated cross-cultural studies, comparing incidence rates, distribution of dementia subtypes, and distribution of risk factors in different sites. The cohort study design affords the opportunity to reduce bias in measurement of exposure and allows us to obtain truly incident dementia cases. The parallel cross- cultural studies offer a unique opportunity to take advantage of the natural variation in environmental exposures that occurs with population migration. Migration studies have traditionally been helpful in determining whether there are modifiable risk factors for prevention of important chronic diseases.

DESCRIPTION (provided by applicant): Since 1987, the UW (University of Washington)and GHC/CHS(Group Health Cooperative/Center for Health Studies) Alzheimer's Disease Patient Registry (ADPR) and Adult Changes in Thought (ACT) Study has established itself as a model incident case-finding registry as well as developed a prospective cohort of over 3,300 persons, initially without dementia. The ACT inception cohort, established 1994-96, was designed to detect new cases of dementia (especially AD), and to collect prospective data to identify potentially modifiable risk factors for AD and related conditions. To date, 309 persons have developed dementia, with an additional 291 cases projected in the near future, for a total of 600 cases. The combination of prospectively collected data over time (baseline & biennial follow-ups), plus comprehensive medical and pharmacy historic records for 10 to 20 years (median GHC membership is 19 years prior to enrollment), creates unparalleled opportunity for analyses using the ACT cohort data in the coming 5 years. The specific aims for this next cycle are: 1. To undertake innovative analyses of the pharmacoepidemiology of dementia and AD; specifically antihypertensive agents, post-menopausal hormone replacement therapy, HMC-CoA reductase inhibitors ("statins") and other antihyperlipidemic agents, non-steroidal anti-inflammatory drugs (NSAIDs), histamine H2 receptor antagonists, and antioxidant vitamins. 2. To evaluate other potential risk factors for AD and dementia, with emphasis on vascular risk factors, capitalizing on the rich ACT and GHC data sources to investigate: blood pressure, plasma lipids, exercise history, obesity, diabetes, and tobacco use. 3. To continue to serve as a vital source of long-term, comprehensive data and well-characterized research subjects for other investigators. The results of these studies will be invaluable to the development and testing of strategies to reduce the risk of AD and related conditions, and to reduce the burden of cognitive decline in the growing population of older persons.

[unreadable] DESCRIPTION (provided by applicant): Linking biorepositories of patients in healthcare delivery systems with electronic medical records (EMRs) is an efficient strategy for high-throughput genome wide association (GWA) studies, as phenotype, covariable and exposure data of public health importance can be economically abstracted and pooled across delivery systems to facilitate the large numbers of subjects needed for GWA studies of each phenotype. Key obstacles to the success of this strategy remain. In this project, which will use population-based genomic and phenotype data from a well characterized population served by a delivery system which captures virtually all health care encounters in its data bases. Researchers from Group Health Cooperative's Center for Health Studies, the University of Washington, and the Fred Hutchinson Cancer Research Center will address these obstacles by pursuing the following specific aims: [unreadable] [unreadable] 1. Informed by results from targeted focus groups, implement a consensus process with key stakeholders to develop recommendations concerning consent, data sharing, and return of research results to subjects. [unreadable] 2. Work together with other network sites to develop a virtual data warehouse (VDW) analogous to that used in the Cancer Research Network, and extend natural language processing (NLP) to pathology, radiology, and clinical chart notes. [unreadable] 3. Develop and test strategies to determine whether each candidate EMR-based phenotype is sufficiently valid to pursue analyses of GWA data, and develop statistical methods that explicitly account for heterogeneous phenotype validity within and between sites. [unreadable] 4. Perform a series of GWA analyses in the GHC biorepository and linked biorepositories. 4a: Alzheimer's disease (AD). 4b: Carotid artery atherosclerotic disease (CAAD). 4c: Complications of statin use, including elevations of CPK and muscle pain. [unreadable] [unreadable] Through cooperation with other investigators and the NHGRI, this work will facilitate development of policies and procedures to realize the incredible potential of EMR-linked biorepositories for GWA studies to improve understanding, prevention and treatment of chronic diseases and illnesses. Specific GWA research will allow us to explore both etiologic research (AD and CAAD progression) and pharmacogenetics (statin therapy). The implications of this portfolio of research extend far beyond the specific phenotypes we have chosen to emphasize; we expect this work represents the beginning of a large and productive enterprise. [unreadable] [unreadable] [unreadable] [unreadable]

DESCRIPTION (provided by applicant): The Seattle eMERGE project aims to bring personal genomics to practice settings by taking advantage of the extensive electronic medical record (EMR) and biorepository of Group Health Cooperative (GH), including a 33-year pharmacy database and longitudinal data on an aging population. Algorithms developed in eMERGE I will be used to combine genome-wide association studies with phenotypes mined from EMRs to discover new polymorphism-phenotype relationships. Target phenotypes are infectious disease susceptibility, specifically to Clostridium difficile diarrhea, shingles from varicella zoster virus, and fungal nail infection, responses to antihypertensive drugs, serotonin-specific reuptake inhibitors, and statins, including adverse events. A new algorithm will follow longitudinal glycemia and hematocrit trajectories, and a novel automated method will detect karyotype abnormalities for assessing correlation to myelodysplasia and leukemia. Data will also support phenotypes investigated at other eMERGE sites. To create a model for introducing genomics into clinical practice, successful needs assessment methods from eMERGE I will engage stakeholders in guiding development of prototype EMR user interfaces in a clinical decision support format. The test case will be human leukocyte antigen-typing for an adverse drug reaction and the setting will be the patient-centered medical home care model developed at GH. This proposal provides the eMERGE network and its collaborators with the Seattle team's unique expertise in using natural language processing (NLP) to extract information from EMRs, and assisting in adoption of NLP methods. To disseminate eMERGE results and foster collaborations, it takes advantage of leadership positions of the investigators, including partners within eMERGE, other consortia and the HMO Research network, especially the potential for developments supported by the NIH Director's Common Fund in biobanking and megaepidemiology. Completion of the aims will reveal new, medically useful markers, improve the linking of high-throughput genomic methods to EMR data, and develop policies and practices for bringing individualized evidence-based medicine to communities. RELEVANCE (See instrucfions): To advance personalized medicine-treatment and preventive care based on individual traits; this project matches small differences in DNA to infectious disease susceptibility and response to statins, serotonin- specific reuptake inhibitors (SSRIs) and blood pressure medications. Methods to use these results in clinical care will be guided by focus groups of patients and caregivers in the patient-centered Group Health system.

DESCRIPTION (provided by applicant): The ability to pay attention to sound sources of interest is critical for verbal communication in many situations. When multiple talkers are speaking simultaneously, being able to direct and redirect attention is necessary to extract relevant information. However, the neural processing involved in directing auditory attention is not fully understood. Moreover, establishing how top-down control of attention interacts with the bottom-up differentiability of sound sources provides a critical first step in understanding how hearing deficits can impact effective cognitive control; understanding this issue could inform the design of improved coding algorithms for hearing aids and cochlear implants. Previous neuroimaging studies in both audition and vision have implicated the involvement of several cortical structures in the voluntary switching of attention using spatial information, but the mechanisms and dynamics of switching attention using non-spatial information are unknown. We will study orientation and switching of auditory attention based on non-spatial stimulus features in behavioral and neuroimaging studies of human subjects performing psychoacoustical tasks. To capture cortical dynamics, we will employ a multimodal imaging approach, combining the temporal resolution provided by magneto- and electro-encephalography (M-EEG) while co-constraining the localization of cortical activation using anatomical magnetic resonance imaging (MRI) scans. Our first aim is to perform a set of psychophysical experiments designed to examine the cost of switching attention based on non-spatial stimulus features as a function of the peripheral separability of competing sounds. Our second aim is to use neuroimaging techniques to examine the cortical dynamics of switching attention based on non-spatial features. By combining psychophysics and neuroimaging, we will determine the temporal dynamics of the cortical network involved in orientation and switching of auditory attention. This research will allow us to better understand how the auditory system enables us to communicate effectively in challenging acoustic environments.

? DESCRIPTION (provided by applicant): This application from the Group Health (GH)/University of Washington (UW) eMERGE team proposes specific aims designed to advance integration of genomic data into clinical practice with a focus on clinical discovery and implementation on Mendelian forms of colorectal cancer and/or polyposis (CRC/P) and incidental findings in other actionable genes. Our aims will also allow us to address challenges involved in bringing genomic medicine into standard medical care. Our focus on CRC/P, and quantitative traits and incidental findings (IF) in other actionable genes represents a unique opportunity to move the field forward towards the goal of bringing genomic medicine into effective, standard medical practice in an everyday community practice setting. We have 3 Aims. Aim 1: Genomic medicine discovery and implementation focused on CRC/P, Triglycerides (TG), and neutrophil count (NPC). We proposed sequencing of 1000 CRC and 1000 Asian ancestry participants, to achieve sub- aims of understanding the genetic basis of CRC, TG, and NPC. Aim 2: Integrate genomic information into GH-wide clinical care and the EMR. We will develop intuitive, comprehensive reports to return CRC and other genes deemed actionable by the American College of Medical Genetics and Genomics (ACMG). We will incorporate stakeholder input and then to implement integrated processes and tools into an integrated delivery system with a focus on CRC/P and Long QT syndrome. We will develop and evaluate educational outreach and online resources. Aim 3: Evaluate the effectiveness and economic impact of result return to patients and their families. We will implement a novel tool to increase family communication of CRC genetic results and evaluate the economic impact and cost effectiveness of this tool as well as of returning IFs. Completion of the work in this eMERGE III proposal will guarantee that the Seattle site remains an engaged and effective leader in the eMERGE network in support of NHGRI's mission to ensure that barriers to successful integration of genomic medicine in clinical care are overcome.

?DESCRIPTION (provided by applicant): The overarching goal of this project is to deepen our understanding of the aging brain in a well characterized community-based longitudinal prospective cohort study. The Adult Changes of Thought (ACT) study enrolled a cohort of 2,581 Group Health participants in 1994-1996, an expansion cohort of 811 in 2000-2002, and in 2005 began continuous enrollment to keep 2,000 people alive and at risk for dementia outcomes; total enrollment to date is nearly 5,000. Extensive and unmatched clinical and research study data are available for study participants. Autopsy consent is sought; 560 autopsies have been completed to date, and more than a third of active participants have signed autopsy consents. Investigators propose three aims. Aim 1 addresses MULTIMORBIDITY: the single and joint effects of blood pressure, glucose, and cholesterol, and their treatments, on the aging brain. Outcomes include time to dementia and Alzheimer's disease, cognitive functioning, standard neuropathological indices, and newly developed quantitative measures of regional Aß, tau, and synaptophysin. This aim leverages important methodological innovations, including hierarchical Bayesian modeling of exposure trajectories to incorporate sporadically collected clinical data, modern psychometric approaches to cognitive data, use of marginal structural models to account for selection bias for autopsy studies, new criteria for neuropathological assessment of AD, and the newly developed quantitative measures of neuropathology. Aim 2 addresses RESILIENCE: why some people do well despite factors associated with doing poorly. Aim 2a addresses the role of physical activity on trajectories of cognition and physical performance, using accelerometers to address the amount and type of activity associated with protection. Aim 2b addresses factors associated with living to age 85 while preserving cognition and mobility, specifically focusing on physical activity and on medical comorbidity. Aim 2c addresses factors associated with avoiding dementia despite accumulation of neuropathology. In particular, Aim 2c will focus on synaptic contents, applying flow cytometry of a synaptosome preparation. Aim 3 addresses a LIVING LABORATORY: continuing to serve as a resource to researchers around the world.

ABSTRACT Genetic risk for AD now clearly highlights the potential for multiple molecular drivers and perhaps multiple pathogenic pathways, including forms of AD that derive from disease causing mutations in PSEN1 or PSEN2, increased risk from APOE ?4, and sporadic disease that does not have identified genetic risk. Regardless of genetic risk, AD is a chronic illness whose ultimate clinical expression as dementia follows years if not decades of injury, response to injury, consumption of reserve, and compensation. Moreover, as highlighted at the 2013 AD Related Dementias summit, longitudinal population-based cohort studies have repeatedly observed that AD most commonly is co-morbid with vascular brain injury (VBI) and less commonly with Lewy body disease (LBD). Finally, these same longitudinal cohort studies have revealed individuals who had high levels of AD neuropathologic change but no significant clinical expression ? a state of apparent resilience to AD. Here we propose to enable progress in precision medicine for AD by vastly improving the molecular characterization of disease and sharing this unique resource with the community of scientists. Indeed, much of our knowledge about injury/response to injury in AD is based on histopathologic assessments rooted in technology that is about 140 years old. Emerging technologies now permit a depth of molecular phenotyping that until recently was difficult even to imagine. We hypothesize that determining quantitative, high dimensional protein phenotypes from carefully clinically characterized individuals from longitudinal cohorts who have donated their brains for research will illuminate components of AD that currently are obscured by limited standard neuropathologic assessments. We will use this novel approach to test our hypothesis through three Specific Aims. In Aim 1, we will collect proteomics data on post-mortem brain samples from the University of Washington Alzheimer's Disease Research Center and the Adult Changes in Thought Study using a liquid chromatography-tandem mass spectrometry (LC-MS/MS) strategy known as data independent acquisition (DIA). DIA enables the comprehensive and systematic sampling of protein digests. This data acquisition will create a permanent digital molecular archive of this unique and highly valuable sample collection. In Aim 2, we will analyze the molecular phenotype of three groups that will be critical to precision medicine for AD: (i) different genetic risk, (ii) common co-morbidities, and (iii) resilience to AD neuropathologic change. We anticipate identifying a molecular signature that is predictive of cognitive impairment as a replacement for traditional histopathological assessment. In Aim 3, we will make our data available through a novel cloud based solution, called the Chorus Project (http://chorusproject.org), engineered to enable big data reanalysis by the community of scientists. We will develop a novel query engine that will enable informatics experts without knowledge of the complexities of mass spectrometry signal processing to perform reanalysis of our data.

Project Summary/Abstract Alzheimer's disease (AD) diagnoses and deaths are on the rise ? 5.4 million are currently living with the disease and 700,000 will die in 2016 alone. There is no sure-fire method for prevention, early detection, or cure. The Adult Changes in Thought (ACT) study has recruited over 5,000 adults over the age of 65 since 1994, and follows them every 2 years for new dementia and AD diagnoses. ACT data contribute to a living, learning co-laboratory, which serves the research community as a valuable scientific resource for AD and brain aging research. For the past four years, we have held an annual ACT Symposium to highlight new ACT resources, innovative research findings, and cross-disciplinary collaborations. We have done this with minimal support from the ACT parent U01. This conference grant will provide new support for our growing symposia for the next three years. The specific objectives of our conference grant are to: 1) promote appropriate scientific use and understanding of ACT data; 2) promote collaborations across scientific disciplines, institutions, and studies; and 3) provide opportunities for feedback on ongoing and planned work including analyses in progress and new grant ideas particularly for junior investigators. This new funding source will provide support for us to move our conference location to a larger, more technologically advanced facility. It will also provide travel stipends for investigators traveling from out of town so that we may extend our reach to new external investigators. We anticipate that future conferences (in late summer of 2017, 2018, and 2019 in Seattle, Washington) will cover hot topics in aging and dementia research such as resilience, imaging, dementia subtypes, molecular phenotypes, and traumatic brain injury ? topic to which ACT is making significant contributions. Long term, expanding our ACT symposia will increase use of the ACT living laboratory, and advance AD and brain aging research.

Alzheimer's disease prevalence is expected to triple by 2050, and current medications do not change the disease course. However, recent studies have projected that up to 30% of Alzheimer's disease may be preventable by targeting modifiable risk factors. Indeed, population-based studies are finding that, although the total number of dementia cases is growing, age-specific incidence rates are declining in parallel with population-level reductions in risk factors such as low education, smoking and cardiovascular disease. Furthermore, studies in Europe have demonstrated slower cognitive decline and less cognitive impairment with multi-domain risk reduction interventions. Given the importance of Alzheimer's prevention and the potential benefits of risk reduction for general as well as brain health, even relatively small reductions in Alzheimer's incidence would have a dramatic public health impact. Despite the tremendous promise of the multi-domain risk reduction approach, no primary prevention trials have been performed in the U.S. We propose to conduct a pilot study of a multi-domain Alzheimer's disease risk reduction intervention in the U.S. To maximize feasibility and the potential for implementation, we propose to perform our trial within an integrated healthcare delivery system. In addition, we plan to utilize a personalized intervention, which has proven successful in the healthcare setting for managing conditions such as depression and hypertension, and which should yield greatest change on risk factors. Specifically, we propose to randomize 200 higher-risk older adults (age ?70 with subjective cognitive complaints, performance in the lower end of normal on a cognitive test and ? 2 modifiable risk factors that will be targeted by our intervention) to a two year Multi-domain Alzheimer's Risk Reduction Study (MARRS) intervention or a Health Education (HE) control. In the MARRS intervention group, the team will work with participants to develop a personalized action plan to address risk reduction including cardiovascular risk management; smoking cessation; physical, mental and social activity; healthy diet; improved sleep quality and medication management. The goals of our study are to collect pilot data on the effect of MARRS on 2-year rate of cognitive decline compared to HE (Aim 1); compare changes in Alzheimer's modifiable risk factors (Aim 2); and gather preliminary data on the effect of MARRS versus HE on other important health-related outcomes (Aim 3). We will also compare strategies for identifying higher-risk patients for this and future studies. This innovative pilot trial will provide critically needed information to support a future multi-site trial, with the ultimate goal of delaying or preventing cognitive decline leading to Alzheimer's disease.

The overarching goals of this R01 proposal are to leverage extensive existing clinical imaging data from a well- characterized cohort of older adults to improve scientific understanding of potential mechanisms by which anticholinergic medications (Aim 1) and glucose metabolism (Aim 2) lead to dementia and Alzheimer?s disease risk, and to provide a valuable resource that characterizes longitudinal changes in brain structure over time for this cohort of older adults (Aim 3). The proposal builds on the resources of the Adult Changes in Thought (ACT) study. ACT is situated within Kaiser Permanente Washington, known until February 2017 as Group Health, which has extensive clinical data resources including laboratory data stretching back to 1988 and pharmacy data stretching back to 1977. These data have enabled study investigators to develop longitudinal exposure models to identify risk factors for Alzheimer?s disease and dementia, including strong anticholinergic drugs (Aim 1) and glucose levels (Aim 2). Little is known about imaging correlates of these exposures. Some 2300+ clinical MRI scans have been performed on 1432 ACT participants but they currently are stored in records departments at dozens of hospitals and are not available for research. The investigators propose to reclaim the scan data for research and to obtain follow-up scans on living participants with a single existing scan. This economical approach will enable the investigators to obtain longitudinal (at least 2 year and in most cases much longer interval) imaging data on over 600 well-characterized study participants. The investigators propose to have trained neuroradiologists apply NIH Neuroimaging Common Data Elements (CDEs) to extract data from these scans, and will use image thresholding software to robustly measure volumes of large structures such as ventricles and total brain. The investigators will use these data in Bayesian modeling approaches to test hypotheses regarding the effects of anticholinergic medications on changes in ventricular enlargement and brain atrophy (Aim 1) and regarding the effects of high glucose levels on white matter hyperintensities (Aim 2). The investigators propose to store de-identified raw scan data, imaging CDEs, and data gleaned from brain volume analyses at the Laboratory of Neuro Imaging (LONI) to ensure promulgation to the broader research community (Aim 3).