SOCIAL PARTICIPATION AND GENDER DISPARITY ON DEPRESSIVE SYMPTOMS AMONG INDONESIAN ELDERLY

Abstract A voluminous literature empirically proves the significant relationship between social participation and elderly mental health. Further analysis discovers the different behavior between rural and urban elderly. Despite the importance of taking into account regional inequality, gender disparity which explains the inconsistent empirical results across countries, is sometimes negligible from the discussion. This study aims to investigate the link of social participation in voluntary activity and community regular meeting to depressive symptom among the elderly in Indonesia. Separated analysis based on gender is also conducted to examine the extent to which social activities could explain the depressive symptom disparity between male and female elderly. Data from Indonesia Family Life Survey 2007 and 2014 were analyzed using logistic regression. Sample of 2994 and 2917 respondents aged 60 and over in 2007 and 2014 respectively are combined as pooled cross-sectional data instead of panel data to deal with the large reduction of sample size due to mortality. To minimize the potential endogeneity, covariates are included in the model such as residence location, living arrangement, socio-economic status, and health condition. The results indicate that economic condition, chronic disease, and difficulty on instrumental activity daily living (IADL) play a significant role in depressive symptom among Indonesian elderly, regardless of the gender. Surprisingly, social participation that is widely believed in strengthening mental health is statistically significant for female sample only. Moreover, the contrast sign of voluntary participation and community regular meeting coefficients indicate a special behavior between these two activities.

that cognitive functioning is positively related to changes in conscientiousness, agreeableness, openness, and extraversion and negatively related to changes in neuroticism. The findings advance our understanding of the association between changes in personality and cognition. The impact of cognitive change on personality stability and the role of personality traits for maintaining cognitive function in later life are discussed. The results have implications for developing interventions to maintain or enhance cognitive functioning and personality in middle and later adulthood.

WHEN LIFE DOESN'T GO AS PLANNED: UNMET EXPECTATIONS ABOUT RETIREMENT TIMING AND SUBSEQUENT DEPRESSIVE SYMPTOMS
Leah R. Abrams, 1 and Neil Mehta 1 , 1. University of Michigan School of Public Health, Ann Arbor, Michigan, United States The 2008 Great Recession affected American's retirement timing, but it remains unclear how unfulfilled expectations about retirement timing influence psychological well-being. This study examines how unmet expectations about working at age 62 relate to subsequent depressive symptoms, with special attention to sociodemographic differences in unmet expectations and their consequences. We use longitudinal data from 10,557 adults ages 51+ in the Health and Retirement Study . Mean expected probability of working full time at age 62 (ranging 0-100) was 40.5 (SD=54.65). We created quartiles: no chance (0 probability, 35% of sample), unlikely (1-30, 16%), unsure (33-80, 28%), and very likely (85-100, 21%). Expected probability and the association between expectations and reality were significantly lower for racial minorities compared to whites, low education compared to high, and pre-baby boomers compared to baby boomers. Those who were working at age 62 but expected to be retired did not experience elevated depressive symptoms compared to those who correctly expected to be working. In contrast, those who were unexpectedly not working experienced significantly higher depressive symptoms compared to those who correctly expected to be retired (Unsure: IRR=1.16 p=0.024, Very likely: IRR=1.19, p=0.010). This association was slightly attenuated after adjusting for declines in functioning, suggesting partial, but not complete, mediation by health status. The association was consistent by race, education, and birth cohort, but was larger in men than women. Taken together, our findings indicate that unexpected continued employment does not harm psychological well-being, but earlier than expected retirement may result in higher depressive symptoms.

SOCIAL PARTICIPATION AND GENDER DISPARITY ON DEPRESSIVE SYMPTOMS AMONG INDONESIAN ELDERLY
Riska D. Astuti, 1 and Bondan Sikoki 2 , 1. SurveyMETER institute, Yogyakarta, Indonesia, Indonesia, 2. SurveyMETER, Yogyakarta, Indonesia, Indonesia A voluminous literature empirically proves the significant relationship between social participation and elderly mental health. Further analysis discovers the different behavior between rural and urban elderly. Despite the importance of taking into account regional inequality, gender disparity which explains the inconsistent empirical results across countries, is Innovation in Aging, 2019, Vol. 3, No. S1 sometimes negligible from the discussion. This study aims to investigate the link of social participation in voluntary activity and community regular meeting to depressive symptom among the elderly in Indonesia. Separated analysis based on gender is also conducted to examine the extent to which social activities could explain the depressive symptom disparity between male and female elderly. Data from Indonesia Family Life Survey 2007 and 2014 were analyzed using logistic regression. Sample of 2994 and 2917 respondents aged 60 and over in 2007 and 2014 respectively are combined as pooled cross-sectional data instead of panel data to deal with the large reduction of sample size due to mortality. To minimize the potential endogeneity, covariates are included in the model such as residence location, living arrangement, socio-economic status, and health condition. The results indicate that economic condition, chronic disease, and difficulty on instrumental activity daily living (IADL) play a significant role in depressive symptom among Indonesian elderly, regardless of the gender. Surprisingly, social participation that is widely believed in strengthening mental health is statistically significant for female sample only. Moreover, the contrast sign of voluntary participation and community regular meeting coefficients indicate a special behavior between these two activities.

University of Massachusetts Boston, Boston, Massachusetts, United States
Personality traits, such as those identified in the Big Five Personality Model (i.e., openness to experience, conscientiousness, extraversion, agreeableness, and neuroticism), may be associated with different aspects of friendship among older adults. Additionally, men and women form and maintain their friendships differently, which may result in gender differences in their friendships. This study examined the relationship between specific personality traits and friendship characteristics, including friendship quantity, frequency of social interactions with friends, positive and negative friendship quality. The study also explored whether gender moderates the relationship between personality traits and friendships in later life. This study used data from 7,250 community-dwelling older adults, aged 65 years and above (M=75.4 years old, SD=6.91), who participated in 2012 and 2014 Leave-Behind Questionnaire of the Health and Retirement Study. Results from the linear regression analysis indicated significant main effects for the different personality traits and friendship quantity, quality and social interaction frequency, but no main effect for gender was found. However, the moderating effect of gender was significant for the relationship between specific personality traits (i.e. openness to experience, agreeableness and extraversion) and social interaction with friends, as well as for positive and negative friendship quality. For example, older women who scored high on openness to experience reported significantly lower social contact frequency with friends (B=-.16, p<.05) and higher negative friendship quality (B=.08, p<.05) than men who scored high on openness to experience. Study results provide insights for understanding better how personality traits and gender play a role in friendships in later life.

ASSOCIATION BETWEEN SOCIAL SUPPORT AND ANXIETY IN OLDER ADULTS NEW TO LONG-TERM SERVICES AND SUPPORTS
Yeji Hwang, 1 Nancy Hodgson, 1 Justine Sefcik, 1 Augustine Boateng, 1 Anjali Rajpara, 1 Kierra Foley, 1 and Karen Hirschman 1 , 1. University of Pennsylvania, Philadelphia, Pennsylvania, United States Anxiety symptoms are common among older adults and are often associated with adverse outcomes. Thus, it is important to examine modifiable factors and manage anxiety symptoms in this population. While many biological and psychological factors related to anxiety symptoms in older adults have been found, little is known about social factors which are essential in one's mental health. The purpose of this study was to examine the prevalence of anxiety symptoms among older adults new to long-term services and supports (LTSS) and to investigate the relationship between social support and presence of anxiety. This was a secondary data analysis from a study funded to examine health related quality of life in older adults new to LTSS. Anxiety was assessed using a single item, "Recently, how often have you felt anxious?" and the answers were dichotomized into "anxiety" (ratings: 'very often', 'often', 'sometimes', and 'seldom') and "no anxiety" (rating: 'never'). Social support was measured by Medical Outcomes Study Social Support Scale. Prevalence of anxiety symptoms in this sample was 82.7% (n=225). In multivariate logistic regression, adjusting for age, gender, LTSS type, cognitive status, physical and emotional health, and depressive symptoms, older adults with more tangible social supports had lower odds of having anxiety symptoms (Odds ratio=0.515; 95% CI: 0.289-0.919, p=0.025). Improving access to tangible social supports for older adults at the start of LTSS may impact anxiety in older adults. Implications for