154P
Depression Among Community-Dwelling Older Adults in China: A Multilevel Modeling Perspective

Schedule:
Friday, January 16, 2015
Bissonet, Third Floor (New Orleans Marriott)
* noted as presenting author
Yu-Chih Chen, MSW, Doctoral student, Washington University in Saint Louis, St. Louis, MO
Yi Wang, MSW, Doctoral student, Washington University in Saint Louis, St. Louis, MO
Nancy Morrow-Howell, PhD, Bettie Bofinger Brown Distinguished Professor of Social Policy Director, Harvey A. Friedman Center for Aging, Washington University in Saint Louis, St. Louis, MO
Background and Purpose. Depression is one of the most common psychiatric disorders among older populations. There are estimates that the prevalence of depressive symptoms among Chinese older adults is 29.2% in rural regions and 20.5% in urban areas. Depression is associated with individual-level factors, including low self-rated health, poor economic security, and lower levels of social activities. The person-in-environment perspective as well as previous studies suggest a relationship between built/social environment and individuals’ physical and cognitive functioning. Thus, this study examines community-level characteristics (transportation, housing, social services, health services, and community SES) and depression. The objectives of this study include:

(1) to explore the amount of variance in depression associated with communities within which older Chinese live; (2) to examine the relationships between individual-level factors and depression; (3) to investigate the influence of community-level factors on depression; (4) and to explore possible cross-level interactions and their effects on depression.

Methods. Data were drawn from the baseline (2011-2012) of the China Health and Retirement Longitudinal Survey (CHARLS) of older adults aged 60 and above (N= 6,674) in 447 communities. The dependent variable depression was measured by 10-item CES-D. Individual-level independent variables include economic security, social engagement, and self-rated health. Community-level predictors include transportation (road passing though, degree of road tidiness), housing (sewer system, and toilet type), social services (employment center, senior association), and health services (health centers, health posts).Controls included age, gender, marital status, and education. A two-level hierarchical linear regression model (HLM) was conducted using the statistical package R. HLM methods were used to handle clustered data by building a model with individual-level and community-level variables in one model to predict values of depression and to test cross-level interaction hypotheses.  

Results. The Intraclass Correlation Coefficient (ICC) was calculated as result of a null model. The ICC of 0.14 indicates 14% variance in depression was explained by community-level variables. Results showed that at the individual level, economic security (b = −.34, t = −2.60), health (b = −3.10, t = −26.28), and social engagement (b = −.55, t = −2.49) were negatively associated with depression. On the community level, toilet type in the “housing” domain (b = −.84, t = −3.00), senior associations in “social services” domain (b = −2.45, t = −2.40), and community SES (b = −.33, t = −3.15) were negatively associated with older adult's depression. In addition, senior associations (yes/no) on the community level moderated two individual-level relationships: social engagement and depression (b = −.72, t = −1.77), as well as health and depression (b = .47, t= 2.17).

Discussion. Both individual and community factors are associated with older adult's depression. On the individual level, improving older adult's economic security, increasing health status, and encouraging older adults' social engagement by offering opportunities in the community are recommended. On community level, making the built environment more age-friendly, providing more social and health services, and advancing community socioeconomic development will alleviate depression among community-dwelling older adults.