Relationship of depression, chronic disease, self-rated health, and gender with health care utilization among community-living elderly

Background: We investigated the interactive effects of depressive symptoms and chronic diseases on health care utilization among elderly people and explored the potential moderating effect of gender and the mediating effect of self-rated health (SRH) on the association between depressive symptoms and health care utilization. Method: We analyzed data from 5223 people aged 60 years or older living in the community from the Korea Welfare Panel Study in 2015. Depressive symptoms were measured using an 11-item version of the Center for Epidemiologic Studies Depression (CES-D-11) Scale and morbidity within 28 disease categories was assessed. Health care utilization was evaluated as the number of outpatient visits (OV), number of hospitalizations (NH), and number of days spent in the hospital (DH) during past year. Hierarchical moderated regression analyses were applied to investigate the interactive effects. We also adopted the mediation analysis method by Hayes and Preacher. Results: A significant interactive effect of CES-D-11 score and chronic disease on OV was found. A positive correlation between CES-D-11 score and OV was only observed in those with chronic disease. Gender had a moderating effect on the association of depression symptoms with OV, NH, and DH among elderly people with chronic disease. SRH had mediating effects on the association of CES-D-11 with OV, NH, and DH only among those with chronic disease. Limitations: The severity or multimorbidity of chronic diseases, which could affect health care utilization among elderly were not considered. Conclusions: We elucidated the complex aspects of the relationship between depressive symptoms and chronic disease and their interactive effects on health care utilization among elderly people, and identified important roles of gender and SRH.