Richard J. Kryscio - US grants

The specific aims of The Biostatistics and Data Management Core (BDMC) are: 1. To maintain data sets for all information derived from the Clinical Core, 2. To maintain data sets for all information derived from the Neuropathology Core, 3. To maintain data sets for all information derived from other ADRC components and from ADRC affiliated projects, 4. To maintain accuracy and confidentiality of all data entered into the centralized database, 5. To assist in the development of data collection forms, 6. To provide statistical consultation to projects, cores, and other investigators affiliated with the ADRC, 7. To routinely provide data tables to monitor patient and control subject enrollment and flow, 8. To provide periodic summaries of data to the Administrative Core, and portions of the centralized data to investigators when needed, and 9. To perform statistical analyses for ADRC cores and projects. The ADRC has purchased a SUN 4/490 computer and a site license for the data management and statistical package SAS, to replace the previous system which used M.U.M.P.S. on a DEC-10 computer. A computer program was written to archive data from the DEC-10 to the SUN 4/490. The BDMC will create a centralized data management system using SAS on the SUN 4/490 to receive, store, retrieve, and analyze data obtained from the various components of the ADRC. This system will provide a mechanism for ADRC investigators to access a variety of data relevant to their research endeavors, and to insure confidentiality and protection of intellectual property. Quality control procedures including standardized forms, double data entry, and regular audits will be put in place in order to insure accuracy of all data. Standardized forms have been developed for some data. The approval process by the Executive Committee for all manuscripts, grant submissions, and new projects using ADRC resources will include a review by the BDMC to insure the proper use and interpretation of statistical methods. Biostatistical consultation will be available from the BDMC for all projects using ADRC resources.

The purpose of this proposal is to propose a conference to bring together for the first time biostatisticians, neuroepidemiologists, neuropsychologists, data managers, interested clinicians and basic scientists for the sole purpose of discussing the statistical/analytical methodology associated with key issues arising in AD research. Invited sessions will be given on each of the following specific aims of this conference: l. reliability and validity of measurements used to identify and monitor AD and non-AD subjects in the clinic, at autopsy, and in general population surveys, 2. methodology for longitudinal measurements used to monitor the cognitive and behavioral decline experienced by AD subjects, 3. methods for promoting interactions among basic scientists, neuropathologists and clinicians and biostatisticians on issues of importance in AD research, 4. epidemiological methodology for estimating the incidence and prevalence of AD and for identifying environmental and genetic factors which alter the risks of developing and the age of onset of the disease. These sessions have implications for data collection and standardization of measurements made at various Alzheimer's Disease Centers (ADCs). They are particularly relevant for data sharing at a national level; this is a topic which will be discussed at the conference. These sessions also have implications for the design and conduct of clinical trials on dementia as well as diagnostic methods (e.g. neuroimaging). Finally, because of time constraints these sessions are not inclusive of all the methodological problems encountered in AD research, Up to four poster sessions on these additional topics are proposed for this conference. Graduate students and post doctoral students will be actively recruited for poster presentations.

Dr. Richard Kryscio will direct the data analysis Core. This Core will be in charge of assembling and analyzing all behavioral, neurochemical, MRI, fMRI, tract tracing, immunohistochemical and other data collected from the unilateral MPTP-lesioned monkeys that have received chronic infusions of GDNF.

DESCRIPTION (Provided by applicant): The purpose of this proposal is to bring together biostatisticians, neuroepidemiologists, neuropsychologists, demographers, social scientists, data managers, and interested clinicians and basic scientists for the purpose of discussing statistical/analytical methodology associated with key issues arising in AD research. This is the second such conference; the first, held in May, 1998, was very successful in attracting a mix of quantitatively oriented scientists working in this area. The invited sessions of this second conference will build on that success, and will cover some issues not previously addressed. The specific aims of the proposed conference will cover the following topic areas: 1. epidemiological methodology associated with estimating the incidence and prevalence of AD and other forms of dementia; 2. probabilistic and statistical problems encountered in the systematic identification of genes associated with late onset AD and multivariate methods for the analysis of gene expression data on single neurons; 3. longitudinal data analysis used to monitor the cognitive and behavioral decline experienced by AD patients; 4. methodological issues involved in the design of prospective clinical trials with an emphasis on early detection/prevention of the disease process; and 5. methodological issues involved in the interpretation of neuroimages of the brain both volumetric and functional. In addition, the conference will host a half day workshop on methodologic issues in AD research with an emphasis on issues involved in the planning and implementing of clinical trials for the prevention of AD. The conference will include two contributed poster sessions on related topics with one of these sessions concentrating on student and/or post doctoral research. The conference will have implications for data collection and standardization of measurements made at ADCCs and for data sharing at the national level.

This subproject is one of many research subprojects utilizing the resources provided by a Center grant funded by NIH/NCRR. The subproject and investigator (PI) may have received primary funding from another NIH source, and thus could be represented in other CRISP entries. The institution listed is for the Center, which is not necessarily the institution for the investigator. This core provides expert advice on experimental design, data management, subject recruitment, and data analysis to the clinical projects in this COBRE and it provides expertise on the analysis of the bioinformatics data generated in the basic science projects. This advice promotes efficiency in terms of the minimal number of patients to be recruited to a clinical study or the minimum number of chips to be processed in a microarray experiment so that information is maximized. Advice rendered at an early stage in the design of the project makes a difference. It also significantly improves the chances that the project will be successful as evidenced by the role this core played in one success in each area. Inherent in this core is a training component since its personnel supervise graduate students trainees in their own discipline and work on a one to one basis with young investigators in the dental sciences. The younger faculty in this core are also mentored by the senior faculty. A journal paper summarizing a pilot project supporting a current clinical project and several abstracts based on related dental graduate student research have been submitted. An R01 application in the basic science area, in which this core provided key technical support, has been awarded and another application is in progress. As the projects in this COBRE mature personnel in this core will analyze the data generated and when needed will create new methodology to help interpret the results.

The Biostatistics and Data Management Core (BDMC) has become a major part of the infrastructure of the University of Kentucky Alzheimer's Disease Center (ADC). It plays an increasingly important role in data management, especially in reporting multiple data sets to NACC. This Core also provides an important service in experimental design and statistical analysis to the ADC and affiliated AD research. The BDMC has two specific aims, one related to quality data management and the other related to effective statistical consultation. The first specific aim is to maintain a centralized database of the information collected by all ADC cores, pilot projects, and affiliated AD and aging studies in an integrated manner. Relevant activities that support this specific aim include: ensuring confidentiality at the subject level and security at the database level with appropriate backup;assisting in the development of data collection forms and entering all data into the centralized, relational database in an accurate manner;training ADC personnel to practice quality data collection procedures and to meet BDMC standards;providing expertise on data management procedures to all ADC personnel and affiliated investigators;monitoring the flow of data in the longitudinal normal control and AD cohorts and the minority satellite clinic;providing monthly summaries to the ADC Executive Committee on subject recruitment, retention, and follow up;creating appropriate subsets of the database for approved investigators and collaborative projects and archiving all such requests;and reporting accurate and complete datasets to NACC. The second specific aim is to provide expertise on experimental design and statistical analysis to ADC cores, pilot projects, and affiliated AD projects. Relevant activities that support this specific aim include: consultation for investigators in the early stages of all projects, grant submissions, presentations, and manuscripts;reviewing all ADC pilot projects for statistical content;collaborating with other Alzheimer's Disease Centers on special projects of mutual interest, such as periodic conferences on quantitative methodologies in AD research;and incorporating advances in statistical methodology for interpreting and/or analyzing data collected. In addition, this Core continues to contribute to the literature with innovative studies on statistical methodology in AD research. It also participates in the training of graduate students, postdoctoral researchers, and young investigators.

DESCRIPTION (provided by applicant): Population demographics suggest that with the expected dramatic increase in age-associated dementias a public health crisis is looming. Current emphasis is on disease prevention with a focus on elderly individuals who express some cognitive impairment. We propose to identify authoritatively the risk and protective factors for cognitive decline in older persons. We have shown how to define these impaired states retrospectively, how to account for reverse transitions, how to distinguish prevalence from incidence, and how to account for competing risks by using a unique statistical (Markov) model. But sufficient longitudinal and neuropathological data is currently not available to distinguish among different types of dementia. No single Alzheimer's Disease Center (ADC) or cooperative study has an adequate sample size to reliably track transitions to dementia and differentiate Alzheimer's disease (AD) from other prevalent brain diseases that include vascular dementia (VaD) and Lewy body disease (LBD). This project will pool data from six well established longitudinal cohorts to identify risk factors for (1) preservation of intact cognition in those meeting neuropathological criteria for varying types of dementia and MCI and (2) specific forms of dementia (clinical and neuropathological). This will improve our understanding of intervening impaired states and factors that promote resistance to clinical symptoms despite the presence of neuropathology. These considerations lead to the specific aims below. Specific Aim 1: To merge databases from six large projects that follow cohorts of cognitively intact subjects to dementia, for the purpose of rigorous, statistical, biologically-informed analyses that accentuate longitudinal follow-up: BRAiNS (University of Kentucky), Nun Study (University of Minnesota), Memory and Aging Project (Washington U), Kuakini Honolulu-Asia Aging Study, Religious Orders Study (ROS, Rush Medical University), and the OHSC ADC (Oregon Health &Science University). The database would be made publicly accessible. Specific Aim 2: To identify appropriate intervening states between intact cognition and dementia based on periodic assessments of cognition and functional skills from data collected at these centers. Specific Aim 3: To study transitions and associated risk factors (e.g., genetic, medical, time in an impaired state) using novel statistical methods. Specific Aim 4: To standardize the neuropathological findings across databases (including quantitative neuropathological assessments) to enable the analysis of novel pathogenetic determinants of outcomes. This aim allows us to evaluate how actual brain pathology (e.g., microinfarcts, Lewy bodies, hippocampal sclerosis, and mixed pathologies) relates to antemortem states in the subset of participants coming to autopsy. This aim could support proposed revisions of current neuropathological and clinical research diagnostic criteria in dementia and preclinical dementia conditions. PUBLIC HEALTH RELEVANCE: With the graying of America the cost of caring for demented elderly will rise substantially in the next few decades. Current emphasis is on preventing this disease. This project will identify risk factors for various forms of dementia as well as impaired states that precede this disease.

DESCRIPTION (provided by applicant): Population demographics suggest that with the expected dramatic increase in age-associated dementias a public health crisis is looming. Current emphasis is on disease prevention with a focus on elderly individuals who express some cognitive impairment. We propose to identify authoritatively the risk and protective factors for cognitive decline in older persons. We have shown how to define these impaired states retrospectively, how to account for reverse transitions, how to distinguish prevalence from incidence, and how to account for competing risks by using a unique statistical (Markov) model. But sufficient longitudinal and neuropathological data is currently not available to distinguish among different types of dementia. No single Alzheimer's Disease Center (ADC) or cooperative study has an adequate sample size to reliably track transitions to dementia and differentiate Alzheimer's disease (AD) from other prevalent brain diseases that include vascular dementia (VaD) and Lewy body disease (LBD). This project will pool data from six well established longitudinal cohorts to identify risk factors for (1) preservation of intact cognition in those meeting neuropathological criteria for varying types of dementia and MCI and (2) specific forms of dementia (clinical and neuropathological). This will improve our understanding of intervening impaired states and factors that promote resistance to clinical symptoms despite the presence of neuropathology. These considerations lead to the specific aims below. Specific Aim 1: To merge databases from six large projects that follow cohorts of cognitively intact subjects to dementia, for the purpose of rigorous, statistical, biologically-informed analyses that accentuate longitudinal follow-up: BRAiNS (University of Kentucky), Nun Study (University of Minnesota), Memory and Aging Project (Washington U), Kuakini Honolulu-Asia Aging Study, Religious Orders Study (ROS, Rush Medical University), and the OHSC ADC (Oregon Health & Science University). The database would be made publicly accessible. Specific Aim 2: To identify appropriate intervening states between intact cognition and dementia based on periodic assessments of cognition and functional skills from data collected at these centers. Specific Aim 3: To study transitions and associated risk factors (e.g., genetic, medical, time in an impaired state) using novel statistical methods. Specific Aim 4: To standardize the neuropathological findings across databases (including quantitative neuropathological assessments) to enable the analysis of novel pathogenetic determinants of outcomes. This aim allows us to evaluate how actual brain pathology (e.g., microinfarcts, Lewy bodies, hippocampal sclerosis, and mixed pathologies) relates to antemortem states in the subset of participants coming to autopsy. This aim could support proposed revisions of current neuropathological and clinical research diagnostic criteria in dementia and preclinical dementia conditions.

Program Summary/Abstract: Data Management and Statistical Core The Data Management and Statistical (DMS) Core can be conceptualized as the lynchpin for integration of the ADC Cores and is vital to the success of the UK-ADC. Its critical roles include managing a large centralized database, consulting with ADC-affiliated researchers, and working to develop and apply innovative statistical methodology for data analysis. Data management efforts focus on collecting and storing high quality data. This focus begins with the leadership and vision and attention to detail provided by the DMS Core. This core has an enviable track record of timely and accurate reporting of a high volume of data to NACC. Further, weekly Core meetings are popular with UK-ADC investigators who find that the expert advice provided by our seasoned investigators improves their success in pilot studies, grant applications, and publications. The Core also participates as a full partner to the research mission of the UK-ADC emphasizing transitions and translations. One such partnership with the Clinical and Neuropathology Cores relates to clinico-pathological models of mixed dementias. A key element of this DMS Core is the well-established track record of developing novel methodology to analyze data collected at the UK-ADC and from other cohorts with a focus on elderly subjects? transitions to MCI and eventually dementia. The Core also provides training for students enrolled in the graduate programs in Gerontology, Public Health, Epidemiology and Biostatistics, Psychology, and Statistics. It is committed to developing younger faculty to become productive ADC investigators. In keeping with the mission of the UK-ADC, faculty in this Core also contribute to the dementia research community at large through service on external advisory committees, study sections, manuscript reviews, and data safety monitoring boards. The DMS Core will continue these critical responsibilities through the following specific aims. 1: Create and maintain a web-based, integrated data warehouse for ADC Cores and projects. 2: Provide expertise on experimental design and statistical analysis. 3: Support the clinical, neuropathological, and educational/outreach missions of the ADC. 4: Contribute to the activities of the dementia community at large through service and research.

Project Summary/Abstract: Data Management and Statistics Core The Data Management and Statistics (DMS) Core can be conceptualized as the lynchpin for integration of the ADRC Cores and is vital to the success of the UK-ADRC. Its critical roles include managing a large centralized database, consulting with ADRC-affiliated researchers, and working to develop and apply innovative statistical methodology for data analysis. Data management efforts focus on collecting, storing, and sharing high quality data using a web-based platform. This focus begins with the leadership and vision and attention to detail provided by the DMS Core. This core has an enviable track record of timely and accurate reporting of a high volume of data to NACC. Further, weekly Core meetings are popular with UK-ADRC investigators who find that the expert advice provided by our seasoned investigators improves their success in pilot studies, grant applications, and publications. The Core also participates as a full partner to the research mission of the UK- ADRC emphasizing transitions in late life from normal to multiple etiology dementia. A key element of this Core is the well-established track record of developing novel methodology to analyze data collected at the UK- ADRC and from other cohorts with a focus on elderly subjects? transitions to MCI and eventually dementia. The Core also provides training for students enrolled in the graduate programs in Gerontology, Public Health, Epidemiology and Biostatistics, Psychology, and Statistics. It is committed to developing younger faculty to become productive ADRC investigators. In keeping with the mission of the UK-ADRC, faculty in this Core also contribute to the dementia research community at large through service on external advisory committees, study sections, manuscript reviews, and data safety monitoring boards. The DMS Core will continue these critical responsibilities through the following specific aims. Aim 1: Maintain and expand an integrated data warehouse and suite of web-based data collection applications and reporting platforms for ADRC Cores and projects. Aim 2: Provide expertise on experimental design and statistical analysis. Aim 3: Support the clinical, biomarker, neuropathological, educational, and outreach missions of the ADRC. Aim 4: Contribute to the activities of the dementia community at large through service and research.