REGARDS: A Case Study in Aging and Disparities Research, Mentoring, and Data Sharing

Abstract Investigators in the NIH-funded REGARDS (REasons for Geographic and Racial Differences in Stroke) project have taken a novel approach to break the paradigm of epidemiologic studies limited to clinic-based convenience samples, by developing a national cohort of 30,239 black and white participants recruited from communities across all lower 48 US states, including 1,855 of the 3,033 counties. Mean age at enrollment (Jan 2003-Oct 2007) was 65.3 years. The four initial aims were to further understanding of: 1) geographic and racial differences in stroke risk factors; 2) geographic and racial differences in stroke incidence and mortality; 3) association of stroke risk factors and stroke risk (incidence and mortality) focusing on effect modification by race or region; and 4) establishment of a repository of serum, plasma, urine and DNA for use in future studies. When the grant was awarded, the study goals were broadened to include longitudinal remote assessment of cognitive function. A second in-home visit was completed May 2013-Dec 2016 including measures of functional status. The cohort is in its 17th year of follow-up. We will detail recruitment and enrollment methods, characteristics of the cohort and status, with brief overview of the biological, medical, psychosocial, environmental, and contextual data collected in the parent study. Speakers will discuss in more detail the stroke and cognitive data, ancillary studies focused on caregiver and heart disease outcomes, and provide examples of national and international mentoring that has leveraged REGARDS data. Finally, we will describe opportunities for additional data sharing and new ancillary studies.


BEHAVIORAL AND SOCIAL CONSIDERATIONS Lisa Barry, UConn Center on Aging, Farmington, Connecticut, United States
Cognitive, behavioral and social dimensions also demonstrate increasing heterogeneity with aging. For example, a longitudinal study of over 1,000 clergy revealed increasing heterogeneity in cognitive function and rate of decline with aging. Moreover, studies of individuals with probable Alzheimer's disease have shown heterogeneity in terms of clinical manifestations and rates of cognitive decline. Older adults also demonstrate greater heterogeneity in mood, anxiety, and the nature and patterns of symptoms over time. Heterogeneity of overall health status increases with aging, as does reported quality of life. Health and Retirement Study (HRS) data have shown that low socioeconomic status or being an underrepresented minority are both associated with greater intra-individual variability in health status in old age, with greatest differences seen in Hispanics. Finally, early life adversity can contribute to heterogeneity of multidimensional health trajectories even in late life.

CLINICAL CONSIDERATIONS George Kuchel, University of Connecticut, Farmington, Connecticut, United States
Varied physiological functions demonstrate increased heterogeneity with aging. Variability in force exertion and motor performance is higher in old age, with increased stepto-step gait variability indicating greater risk of falls and cognitive decline. Even in healthy older adults, renal function may show no change, slight decline, or marked decline. In contrast, heart rate variability declines with age, with decreased complexity and a higher risk of cardiac events. The risk of death, disease and disability varies among individuals with increasing heterogeneity with aging. As a result, frailty has been conceptualized as both as a phenotype and an accumulation deficit index, offering strong predictive validity when seeking to understand the heterogeneity of aging from the perspective of risk of mortality and physiologic dysregulation across different systems. Physical resilience defined as ability to maintain or restore function following exposure to stressors also demonstrates increased heterogeneity with aging.

POPULATION AND HEALTH POLICY CONSIDERATIONS
Julie Robison, University of Connecticut, Farmington, Connecticut, United States The risk of death, disease, disability, hospitalization, institutionalization and high health care costs varies among individuals with increasing heterogeneity associated with aging. Frailty, physical performance measures, self-reported measures and multimorbidity all represent measures that are useful in helping to better define such heterogeneity at the level of populations and to ultimately define such risk in individuals. These higher risk individuals account for a growing proportion of this nation's health care costs, with continued increases over time that appear unsustainable in the long term. Therefore, efforts to better define the nature of such heterogeneity of risk and improved targeting, with the goals of improving outcomes and reducing costs, are essential. A closely related challenge is to effectively translate proven clinical and health system interventions from the world of research to that of health policy and real-world clinical practice via pragmatic trials.

REGARDS: A CASE STUDY IN AGING AND DISPARITIES RESEARCH, MENTORING, AND DATA SHARING
Chair: Virginia Howard Co-Chair: Jennifer Manly Discussant: Maria Glymour Investigators in the NIH-funded REGARDS (REasons for Geographic and Racial Differences in Stroke) project have taken a novel approach to break the paradigm of epidemiologic studies limited to clinic-based convenience samples, by developing a national cohort of 30,239 black and white participants recruited from communities across all lower 48 US states, including 1,855 of the 3,033 counties. Mean age at enrollment (Jan 2003-Oct 2007) was 65.3 years. The four initial aims were to further understanding of: 1) geographic and racial differences in stroke risk factors; 2) geographic and racial differences in stroke incidence and mortality; 3) association of stroke risk factors and stroke risk (incidence and mortality) focusing on effect modification by race or region; and 4) establishment of a repository of serum, plasma, urine and DNA for use in future studies. When the grant was awarded, the study goals were broadened to include longitudinal remote assessment of cognitive function. A second in-home visit was completed May 2013-Dec 2016 including measures of functional status. The cohort is in its 17th year of follow-up. We will detail recruitment and enrollment methods, characteristics of the cohort and status, with brief overview of the biological, medical, psychosocial, environmental, and contextual data collected in the parent study. Speakers will discuss in more detail the stroke and GSA 2020 Annual Scientific Meeting Innovation in Aging, 2020, Vol. 4, No. S1 cognitive data, ancillary studies focused on caregiver and heart disease outcomes, and provide examples of national and international mentoring that has leveraged REGARDS data. Finally, we will describe opportunities for additional data sharing and new ancillary studies. The REGARDS study enrolled 30,239 whites and blacks aged >45 from 2003 -2007, with oversampling of blacks and residents of the Stroke Belt. Potential participants were mailed a letter/brochure followed by telephone call. After verbal consent, telephone interview assessed cardiovascular health and cognitive function. In a home visit, measurements of risk factors, biological samples, EKG, written consent were obtained; during the in-home visit, self-administrated questionnaires were left to be completed and returned. Participants are followed for hospitalizations via telephone at 6-month intervals. Annually and biennially, brief and more comprehensive assessments of global cognitive function are conducted. Medical records for suspected strokes are collected with adjudication by stroke experts. A 2nd in-home and telephone assessment was conducted 2013-2016, approximately 10 years after baseline. This presentation will describe the methodological details of REGARDS, progress on the specific aims of the current grant, and establish the context for the remaining presentations. Since 2003, REGARDS participants have taken part in telephone-based cognitive assessments. Global cognitive status is assessed annually with the Six-item Screener. Between 2006 and 2009, measures of learning and memory (CERAD Word List) and language/executive function (Animal and Letter Fluency) were implemented, and are administered biennially. A Brain Health Substudy, conducting in-home clinical examinations of neuropsychological, neurological, and functional status among 1000 participants, is underway to validate telephone assessments and estimate prevalence of VCID in REGARDS. Approaches to defining incident cognitive impairment and cognitive change, including definitions employed for case/cohort studies using stored blood samples, will be described. We will discuss psychometric and methodological considerations for characterization of risks for cognitive impairment across race and region, as well as longitudinal trajectories of cognitive function. David Roth,1 and William Haley,2 ,1. Johns Hopkins University,Baltimore,Maryland,United States,2. University of South Florida,Tampa,Florida,United States The REGARDS study has provided a unique opportunity to study both disease-specific (stroke) and broader samples of family caregivers, and to examine the effect of transitions to caregiving over time. Using REGARDS has afforded many advantages over conventional caregiving research, including the availability of biomarker and mortality data, a large sample of non-caregiving controls who can be carefully matched to caregivers, and ability to track onset of caregiving over time. Our findings illustrate the complex nature of caregiving-related effects. While caregiving leads to worse psychological well-being, we have found minimal physical health decreases and reduced mortality rates compared to matched non-caregiving samples. These findings have policy implications and have challenged the conventional beliefs about caregiving based on previous studies of convenience samples. Diverse students and junior faculty members from multiple universities have also gained experience and contributed to high impact papers from this work. The REGARDS-MI ancillary study provided new outcomes of heart disease events and adjudicated cause of death. A primary focus has been disparities in and risk factors for coronary artery disease. We demonstrated that compared to White men, Black men have a higher risk of fatal coronary heart disease (CHD) but a lower risk of non-fatal CHD. Ongoing work is investigating potential reasons for this. We have investigated the role of CHD in aging including the relationship between heart failure and cognitive function and the association of MI with functional status. The REGARDS-MI study has served as a platform for mentoring trainees and early stage investigators, many from underrepresented groups, and provided data to a large number of investigators to purse research in CHD. To date, REGARDS-MI has contributed to nearly 200 publications and spawned additional ancillary studies. This presentation will highlight some of these publications and other research in progress.

OPPORTUNITIES FOR MORE AGING AND DISPARITIES RESEARCH, MENTORING, AND DATA SHARING WITH REGARDS
Suzanne Judd, 1 Virginia Howard, 1 Mary Cushman, 2 Jennifer Manly, 3 and George Howard, 1 , 1. University of