THE DOCTOR-PATIENT RELATIONSHIP, PERSONALITY, MOOD, AND FUNCTIONING IN OLDER ADULTS

Abstract Personality pathology has been tied to mental and physical health in older adulthood. Less is known regarding the combined impact of personality and the doctor-patient relationship on mental health outcomes. This study examined relationships between personality, mood, and trust in physicians. Participants (N=170) were a sample of primary care older adults ages 60-99 (M = 70.73, SD = 7.054) who completed self-report measures of personality traits (NEO-FFI), processes (IIP-PD-25), depression (GDS-30; PHQ-9), social adjustment (SAS-SR) and trust in one’s physician (GTIP). Medical burden data (CIRS) were retrieved from medical records. After adjusting for relevant covariates such as age, perceived health, cumulative illness burden, and income security there were several significant predictive relationships. In combined models more neuroticism (NEO-N, ß = .082, p < .000) and lower trust (GTIP, ß = -.025, p = .014) but not agreeableness (NEO-A, ß = -.006) or interpersonal problems (IIP-25, ß = .254) predicted depression. In combined models, higher neuroticism (NEO-N, ß = .018, p < .000) and interpersonal problems (IIP-25, ß = .186, p = .002) but not agreeableness (NEO-A, ß = -.003) or trust (GTIP, ß = -.002) predicted social adjustment. The results are consistent with previous findings that neuroticism predicts both depression and social adjustment in older adults. In addition, lower trust augmented neuroticism to predict depression. Results suggest that apart from general personality risk factors, situational personality processes such as trust in physicians may affect mood state, whereas personality processes such as interpersonal problems contribute to longer term functional impairment.

disabilities in seven of Oregon's seventeen Area Agencies on Aging (AAAs). This poster describes the evaluation of the expansion using three data sources: administrative data about consumer characteristics compiled by the AAAs and State of Oregon from 2015-2017 (N=3,824 traditional consumers, N= 581 OPI-E consumers), qualitative interviews conducted with AAA directors and OPI-E case managers (N=23), and a survey of current OPI-E consumers (N=126). Compared to traditional OPI consumers, OPI-E consumers were somewhat more likely to be men and people of color. Interviews with AAA staff highlighted the need for outreach, lack of service provider capacity, unique characteristics of younger consumers, and issues related to data management and rural access. Staff reported valuing the program, noting how "even low levels of service go a long way." Qualitative and quantitative consumer responses showed consumers found OPI-E services invaluable. The majority stressed their appreciation for the program, with several describing it as "lifesaving." These three sources informed recommendations for expanding the OPI-E program statewide.

FEDERAL POLICY SUPPORTS AND GAPS IN ADDRESSING RACIAL-ETHNIC HEALTH DISPARITIES IN U.S. LONG-TERM CARE FACILITIES
Rebecca L. Mauldin, 1 Kathy Lee, 1 and Antwan Williams 1 ,

University of Texas at Arlinton, Arlington, Texas, United States
Older adults from racial and ethnic minority groups face health inequities in long-term care facilities such as nursing homes and assisted living facilities just as they do in the United States as a whole. In spite of federal policy to support minority health and ensure the well-being of long-term care facility residents, disparities persist in residents' quality of care and quality of life. This poster presents current federal policy in the United States to reduce racial and ethnic health disparities and to support long-term care facility residents' health and well-being. It includes legislation enacted by the Patient Protection and Affordable Care Act of 2010 (ACA), regulations of the U.S. Department of Health and Human Services (DHHS) for health care facilities receiving Medicare or Medicare funds, and policies of the Long-term Care Ombudsman Program. Recommendations to address threats to or gaps in these policies include monitoring congressional efforts to revise portions of the ACA, revising DHHS requirements for long-term care facilities staff training and oversight, and amending requirements for the Long-term Care Ombudsman Program to mandate collection, analysis, and reporting of resident complaint data by race and ethnicity.

SERVICE ENVIRONMENTS AND PSYCHOLOGICAL WELL-BEING AMONG RESIDENTIAL CARE FACILITY AND NURSING HOME RESIDENTS
Yeon Jin Choi 1 , 1. University of Southern California, Los Angeles, California, United States One's competence, or ability to cope, becomes lower as people age due to multiple factors, such as decreased physical and cognitive functioning, social isolation, and reduced income. Considering this change, providing appropriate services may help older adults to maintain their independence and improve their psychological well-being and quality of life. Therefore, this study aimed to investigate services that are positively associated with psychological well-being (i.e., mood, psychological health, self-efficacy) of older adults. For the analysis, observations were derived from the National Health and Aging Trends Study (NHATS), which includes a nationally representative sample of Medicare beneficiaries ages 65 and older. Service environments were assessed by (1) reflect availability and usage of 13 services (i.e., service unavailable, service available/ not being used, service available/ being used by residents) and (2) number of services available within three categories (i.e., help with ADL/IADL, transportation services, social and health-related services) Both having social events and activities available and participating in it were positively associated with older residents' mood, while using housekeeping services and walking areas (e.g., outdoor walking path) were associated with better psychological health and self-efficacy. When models were estimated with service categories, having more social and/or healthrelated services available were associated with better psychological health and self-efficacy. Findings of this study suggests that social events and activities, housekeeping services, and areas to walk for pleasure or exercise would improve older residents' psychological well-being. For Not only providing those services, residential facilities should encourage older residents to participation in or use the services. Personality pathology has been tied to mental and physical health in older adulthood. Less is known regarding the combined impact of personality and the doctor-patient relationship on mental health outcomes. This study examined relationships between personality, mood, and trust in physicians. Participants (N=170) were a sample of primary care older adults ages 60-99 (M = 70.73, SD = 7.054) who completed self-report measures of personality traits (NEO-FFI), processes (IIP-PD-25), depression (GDS-30; PHQ-9), social adjustment (SAS-SR) and trust in one's physician (GTIP). Medical burden data (CIRS) were retrieved from medical records. After adjusting for relevant covariates such as age, perceived health, cumulative illness burden, and income security there were several significant predictive relationships. In combined models more neuroticism (NEO-N, ß = .082, p < .000) and lower trust (GTIP, ß = -.025, p = .014) but not agreeableness (NEO-A, ß = -.006) or interpersonal problems (IIP-25, ß = .254) predicted depression. In combined models, higher neuroticism (NEO-N, ß = .018, p < .000) and interpersonal problems (IIP-25, ß = .186, p = .002) but not agreeableness (NEO-A, ß = -.003) or trust (GTIP, ß = -.002) predicted social adjustment. The results are consistent with previous findings that neuroticism predicts both depression and social adjustment in older adults. In addition, lower trust augmented neuroticism to predict depression. Results suggest that apart from general personality risk factors, situational personality processes such as trust in physicians may affect mood state, whereas personality processes such as interpersonal problems contribute to longer term functional impairment. The literature on health locus of control (HLC) suggests that individuals who believe that their health is internally determined are more likely to use active coping strategies than those who believe their health is determined by chance or powerful others (Brosschot, Gebhardt, & Godaert, 1994;Gibson & Helme, 2000). Coping strategies (Klapow et al., 1995) and HLC (Campbell, Hope, & Dunn, 2017) have been found to influence the relation between chronic pain and depression. We hypothesized that the relation between osteoarthritis pain and depression would be serially mediated by HLC and coping. Self-report measures of osteoarthritis pain (Meenan, Mason, Anderson, Guccione, & Kazis, 1992;Parmelee, Katz, & Lawton), HLC (Wallston, Wallston, & DeVellis, 1978), coping strategies (Felton & Revenson, 1984;Rosenstiel & Keefe, 1983), and depression (Radloff, 1977) were examined in 367 older adults with osteoarthritis of the knee. Hayes' (2013) PROCESS macro was used to test the hypothesized serial multiple mediation for three subscales of HLC: internality (IHLC), chance (CHLC), and powerful others (PHLC). After controlling for age, the hypothesized serial mediation was statistically significant for IHLC and CHLC but not PHLC. More specifically, osteoarthritis pain significantly increased CHLC, which increased negative coping and depression in turn. Osteoarthritis pain significantly decreased IHLC, which was associated with both positive and negative coping strategies in a complex serial mediation. These findings suggest that interventions targeting HLC and/or coping strategies may be able to alter the pain-depression pathway for older adults with chronic osteoarthritis pain. (Supported by R01-MH51800, P. Parmelee, PI).

POWER OF SOCIAL SUPPORT AGAINST DEPRESSIVE SYMPTOMS OF OLDER AFRICAN AMERICANS Seungjong Cho 1 , 1. Case Western Reserve University, Cleveland, Ohio, United States
Depressive symptomatology is one of the most prevalent mental health problems. About 20% of Americans experience depressive symptoms in their lives (Gotlib & Hammen, 2014). Social support, on the other hand, was proved to be a powerful buffer against depressive symptoms of older adults (Kim & Ross, 2009;Sheiman & Meersman, 2004). Few studies have explored this association exclusively among older African Americans who had a culture of powerful social support. Therefore, the purpose of this study is to compare the effects of different sources of social support (from spouse/partner, children, relatives, and friends) against depressive symptoms among older African Americans. This study analyzed the 2014 Health and Retirement Study (HRS; N = 187). Depressive symptomatology was operationalized as a count outcome (number of having symptoms; CES-D8 scale). A negative binomial regression model of depressive symptoms showed that higher levels of spousal support were associated with lower levels of depressive symptoms (coefficient = -.179, p < .001). For each additional score in spousal support, the expected log count of the number of depressive symptoms was decreased by .179. Other sources of social support were not significant predictors of depressive symptoms among older African Americans in this study. Among the covariates, self-rated health (coefficient = .358, p < .001) and household income (logged; coefficient = -.275, p = .014) were significant. The current study supported the results of previous studies showing the power of positive spousal interactions against depressive symptomatology, especially among a nationally representative sample of older African Americans. Introduction: The Personality Inventory for DSM-5 (PID-5) is a measure of the alternative model of personality disorders with limited evidence of validity among older adults. This study examined validity of the model through associations with the Horney-Coolidge Tridimensional Inventory (HCTI). Method: Older adults (N=125) completed the PID-5 and the HCTI. Results: Zero-order correlations were computed between the PID-5's five domains (Negative Affect, Detachment, Antagonism, Disinhibition, and Psychoticism) and the HCTI's three domains (Compliance, Aggression, and Detachment). Compliance was moderately negatively correlated with Detachment (r = -.27), as expected. Aggression was significantly positively related to all five PID-5 domains and was most strongly correlated with Antagonism (r = .56), Psychoticism (r = .48), and Disinhibition (r = .32). As predicted, Horney's Detachment was most strongly related to the PID-5's Detachment (r = .48). Regression analyses were also conducted with PID-5 domains predicting each HCTI type. The Compliance model was significant, with PID-5 domains predicting 13% of variability in Compliance. Negative Affect (positive) and Detachment (negative) were significant predictors. The Aggression model was also significant, with the PID-5 domains accounting for 40% of variability. Antagonism was the only significant positive predictor. Lastly, the Detachment model was significant, with the PID-5 domains predicting 29% of variance in Detachment scores. Negative Affect (negative) and Detachment (positive) were significant predictors. Discussion: Results indicate that the two measures of personality pathology generally converge regarding theoretically similar constructs and diverge around dissimilar domains, providing evidence of validity of the PID-5 for its ability to capture personality traits.