Korean elderly immigrants in the U.S. have been found to suffer from a higher level of depression than other elderly Asian immigrants. Despite the higher incidence, few studies have examined depression and its risk factors among this population. Thus, the overarching objective of this study was to empirically investigate principal bio-psycho-social risk factors associated with depression among elderly Korean immigrants. Data were collected from a survey of 234 community-dwelling cognitively competent Korean immigrants (age ³ 55) residing in the New York metropolitan area. Symptoms of depression were measured by the 30-item GDS scale. Due to the lack of a theoretical model to assess depression, the study utilized an adaptation of George's Social Antecedent model of depression (1994) from which six sets of exogenous variables were selected. They included 1) demographic risk factors, 2) early and later events and achievements, 3) acculturation, 4) vulnerability and protective factors (physical illness, IADL, and social integration), 5) provoking agents (stressful life events and migratory grief), and 6) coping efforts.

The statistical methodology employed robust path analysis, combined with the two different modeling strategies (hierarchical inclusion and general to specific), to obtain reliable results. Robust path analyses were used; algorithms were applied to down-weight outliers (OD) and to estimate variances that were heteroskedastically consistent (HC). Therefore, robust path analyses were performed using four main-effect path models (OD hierarchical, HC hierarchical, OD backward elimination, and HC backward elimination regression models) and two interaction-effect path models (OD hierarchical and HC hierarchical regression models). A reliability analysis of the findings from the 6 models confirmed that the results were not an artifact of any particular statistical algorithm applied.

The results indicated a high prevalence rate of depression (40%) in the sample of Korean immigrant elders studied. From the robust findings, significant risk and buffering factors for depression emerged. Risk factors were found to include age, prior treatment for depression, stressful life events, and migratory grief, while buffering factors were found to include perceived health status and perceived satisfaction with social support. Also noteworthy were two significant interactions including perceived health status by life events and perceived health status by satisfaction with social support. The former interaction implied that the effects of life events on depression were alleviated by positive perceived health status, and the latter suggested that positive perception of health did not suppress depression when there is dissatisfaction with social support.

The results have important implications for social work practice, policy and research. With respect to practice, psychoeducation about late-life depression and its risk factors, and culturally relevant mental health treatment (e.g. intergenerational family therapy and health education) implemented in community programs can de-stigmatize "mental illness" and improve access to appropriate treatment. With regard to policy, welfare reform, Medicaid, and Medicare programs are discussed to propose ways to reduce financial barriers to mental health treatment for newly arrived elderly immigrants. For directions for future research, longitudinal research is warranted to investigate the effects of the length of exposure to various risk factors on depression over time.