CHANGES IN SOCIAL SUPPORT AND THE INFLUENCE OF DRIVING CESSATION

Abstract Driving cessation has been shown to be a potentially isolating transition in life, with important implications for mental health, social isolation, and social support. Older adults who live alone are vulnerable to social isolation in the context of driving cessation. Although some research has examined the association between driving cessation and certain kinds of social engagement activities, no research has specifically examined changes in social support, particularly among older adults most vulnerable to social isolation – those who live alone . The present study addresses this gap, using data drawn from the 2006-2014 waves of the Health and Retirement Study (HRS) to examine how social support changes in the context of driving cessation among older adults who live alone (N=412). This study specifically focuses on instrumental and emotional social support, and how different sources of the support (children, friends, and other family) are influenced by loss of driving. I use a series of ordinary least squares regression (OLS) to examine four-year changes in various forms of social support among those who live alone, comparing those who lose the ability to drive relative to their continuously driving counterparts. Preliminary results indicate that driving cessation is associated with decline in perceived instrumental support of friends (-0.984, p<01) for older adults who live alone. However, these effects did not extend to children or other family members. These results suggest that loss of driving may perpetuate vulnerabilities facing individuals who live alone by leading to lower levels of perceived support from non-family members.

well-being of individuals by influencing behavioral and biological processes. This symposium contains a collection of papers seeking to address the influence of resilience and coping on health outcomes in middle to late life adults. Tobin and Thorpe identified profiles of psychosocial resilience and examined their association with allostatic load (AL) among 283 Black men in the Nashville Stress and Health Study. Using Latent class analysis (LCA), individuals in the high resilience class had the greatest odds of high AL; high resilience worsened physical health for older but not younger Black men. Tan and colleagues explored satisfaction across life domains and correlates of satisfaction across domains in 93 Black adults. The authors report that higher satisfaction was associated with less education, less financial strain, lower depressive symptoms, and better self-rated physical health. Nguyen examined the association between everyday discrimination and generalized anxiety disorder (GAD) and whether church-based relationships buffer the negative effects of everyday discrimination on GAD among older African Americans. Using data from 670 African American respondents age 55 and older from the NSAL, the author reports, that significant interactions indicated that frequent contact with church members and high levels of subjective closeness to church members buffered against the negative effects of discrimination on GAD. These presentations collectively will bolster our knowledge of how stress and resilience impacts health disparities.

DO PSYCHOSOCIAL RESOURCES CONFER PHYSIOLOGICAL RISK? EXAMINING ALLOSTATIC LOAD BLACK MEN ACROSS THE LIFE COURSE
Courtney S. Thomas Tobin, 1 and Roland J. Thorpe Jr 2 , 1. UCLA Fielding School of Public Health,Los Angeles,California,United States,2. Department of Health,Behavior,and Society,Baltimore,Maryland,United States Research suggests positive psychosocial resources promote resilience, although this has been underexplored among Black men. The present study identified profiles of psychosocial resilience and examined their association with allostatic load (AL) among young, middle-aged, and older Black men. Data come from 283 Black men in the Nashville Stress and Health Study. Latent class analysis (LCA) identified resource profiles comprised of eight psychosocial resources across four categories (coping strategies, sense of control, racial identity, social support). Logistic regression was used to estimate the odds of high AL (5+ high-risk indicators) across classes. LCA indicated three latent classes: low (33%), moderate (26%), and high (41%) psychosocial resources. Unexpectedly, individuals in the high resilience class had the greatest relative odds of high AL; high resilience worsened health for older but not younger Black men. Findings suggest elevated levels of resources resilience may undermine physical health in this population. The African American church has played a major role in African American communities, and church relationships represent an important stress-coping resource for older African Americans. This study examined 1) the association between everyday discrimination and generalized anxiety disorder (GAD) and 2) whether church-based relationships buffer the negative effects of everyday discrimination on GAD among older African Americans. Logistic regression analyses were conducted using data from 670 African American respondents age 55 and older from the National Survey of American Life: Coping with Stress in the 21st Century. More frequent experiences of everyday discrimination was associated with higher odds of meeting criteria for GAD. Significant interactions indicated that frequent contact with church members and high levels of subjective closeness to church members buffered against the negative effects of discrimination on GAD. Interventions that focus on the use of church members for support capitalize on a major strength among older African Americans. This study explored satisfaction across life domains (e.g., family, daily life, health, finances, city of residence) and correlates of satisfaction across domains. Black adults (n=93, age range=55-80) completed the domains of life satisfaction scale and measures of sociodemographic factors, personality, and mental/physical health. Participants' satisfaction was highest for home condition, but lowest for health. Univariate analyses of variance demonstrated better life satisfaction in the oldest-old (80+) than the youngest-old (55-64; p<.05), particularly in the domains of daily life/leisure, current financial situation, and total household income. Linear regression models suggested that higher satisfaction was associated with less education, less financial strain, lower depressive symptoms, and better self-rated physical health, although the pattern of results varied by domain. Satisfaction may increase with advancing old age, at least in some life domains. It can also vary across life domains and unique factors likely relate to satisfaction in each life domain. Driving cessation has been shown to be a potentially isolating transition in life, with important implications for mental health, social isolation, and social support. Older adults who live alone are vulnerable to social isolation in Innovation in Aging, 2019, Vol. 3, No. S1 the context of driving cessation. Although some research has examined the association between driving cessation and certain kinds of social engagement activities, no research has specifically examined changes in social support, particularly among older adults most vulnerable to social isolation -those who live alone . The present study addresses this gap, using data drawn from the 2006-2014 waves of the Health and Retirement Study (HRS) to examine how social support changes in the context of driving cessation among older adults who live alone (N=412). This study specifically focuses on instrumental and emotional social support, and how different sources of the support (children, friends, and other family) are influenced by loss of driving. I use a series of ordinary least squares regression (OLS) to examine four-year changes in various forms of social support among those who live alone, comparing those who lose the ability to drive relative to their continuously driving counterparts. Preliminary results indicate that driving cessation is associated with decline in perceived instrumental support of friends (-0.984, p<01) for older adults who live alone. However, these effects did not extend to children or other family members. These results suggest that loss of driving may perpetuate vulnerabilities facing individuals who live alone by leading to lower levels of perceived support from non-family members.

INVESTIGATING SOCIAL NETWORKS OF OLDER SINGAPOREAN LEARNERS: THE MIXED-METHODS SOCIAL NETWORK APPROACH
Pei-Chun Ko 1 , 1. National University of Singapore, Singapore, Singapore Lifelong learning has been regarded as an important factor of promoting active engagement in later life for researchers and policy makers. Most of the studies tend to illustrate old learners as a homogeneous and self-resilient group of people to engage in lifelong learning. Few studies address older learners' social capital in affecting their decision of engagement and in sustaining their motivation. The study documented the existing social networks of older Singaporeans in lifelong learning programs and illustrated how social networks contributed their participation in learning. The mixed methods consist of in-depth interviews and two network instruments (Name Generator and Position Generator) based on 30 older Singaporeans (between 50 and 79 years old) who attended lifelong learning courses between 2016 and 2018. Interviews are transcribed and analyzed. The network instruments of are quantified and visualized. The findings show that older learners' networks included a mixture of social ties from family and friends. Learners' closeness with network members and their living arrangement with them influenced learners' involvement in learning and future planning. Single respondents who had more non-kin members in the networks reported to be more active due to their weak ties. Overlapping networks among couple learners increase the spousal support for learning. Learners who had wider ranges of social resources are associated with their interest in learning activities. The study suggests that advocating lifelong learning needs to take older adults' networks into considerations as networks represent the social forces that influence their decisions and motivations.

IS SOCIAL ENGAGEMENT RELATED TO FEWER HOSPITALIZATIONS IN COMMUNITY-DWELLING OLDER ADULTS?
Bryan D. James, 1 Raj C. Shah, 2 Melissa Lamar, 3 Lisa L. Barnes, 3 Robert S. Wilson, 3 David A. Bennett, 3 and Julie A. Schneider 3 , 1. Rush Alzheimer's Disease Center,Chicago,Illinois,United States,2. Rush University,Chicago,Illinois,United States,3. Rush Alzheimer's Disease Center,Chicago,Illinois,United States We tested the hypothesis that more socially engaged older adults experience fewer hospitalizations. Data came from 1,153 older adults (72.4% female, mean age 80.8, 14.6 years education at baseline), enrolled in the Rush Memory and Aging Project, with survey data linked to Medicare claims records (mean 5.0 [SD=3.1] years of Medicare coverage after study baseline). Linear regression models were fit with annual rate of hospitalization as outcome with terms for age, sex, and education. Engaging in more social activities (est=-0.16, SE=0.05, p=0.002) and larger life space (est=-0.08, SE=0.03, p=0.005) were associated with a lower rate of hospitalization, while a higher level of loneliness (est=0.18, SE=0.06, p=0.002) was associated with greater rate of hospitalization; size of social networks (est=-0.01, SE=0.01, p=0.069) was not associated with hospitalization. When examined separately by admission type, the same significant associations were found for nonelective (emergency and urgent) hospitalizations, but not for elective hospitalizations. After further adjusting for marital status, baseline levels of depressive symptoms, chronic medical conditions, physical activity, and ADL disabilities, only social activities were significantly related to hospitalization rate (total and nonelective). Adjusting for disability attenuated these associations the most, indicating that functional status may confound the relationship between social engagement and hospitalization. More research is necessary to determine if social engagement in older age can proactively help to keep older adults out of the hospital, or alternatively whether level of social engagement is a marker for functional status or other health factor that is directly related to risk of hospitalization.

SOCIAL ENGAGEMENT AND DEPRESSIVE SYMPTOMS IN OLDER AFRICAN AMERICANS: A 6-YEAR CROSS-LAGGED PANEL ANALYSIS
James Muruthi, 1 and J. Tina Savla 2 , 1. University of Oregon,Eugene,Oregon,United States,2. Virginia Tech,Blacksburg,Virginia,United States Although previous studies have extensively investigated the cross-sectional relationship between social engagement and depressive symptoms in late life, longitudinal studies have produced mixed results. Furthermore, studies on the associations between these two concepts among aging African Americans are few. Using a sample of 1688 older African Americans adults from waves 1 and 7 of the National Health and Aging Trends Study (60% women; Average age = 77 years), the present study investigates the longitudinal associations between social engagement (an index from scores on visiting friends and family, attending religious services, attending religious services, participating in group activities, and going out for enjoyment) and depressive symptoms across seven years. Structural equation modeling was used to test cross-lagged relationships between the variables.