The association between dual sensory loss and healthcare expenditure: Mediating effect of depression.

Background: Previous studies have suggested the dual sensory loss (DSL) is linked to depression, and that they are associated with higher healthcare expenditures, respectively. However, the association between DSL, depression and healthcare expenditures remains ambiguous.

Objectives: The current study aims to examine the association between DSL, depression and healthcare expenditures as well as catastrophic health expenditures (CHE) among Chinese people aged 45 and above.

Methods: We first utilized the China Health and Retirement Longitudinal Survey (CHARLS) 2018 to obtain data from a total of 13,412 Chinese individuals aged 45 and above to conduct a cross-sectional study. DSL was defined as a combined variable of self-reported vision loss and hearing loss. Depression was measured using The Center for Epidemiologic Studies Depression Scale (CESD-10). The healthcare expenditures, including outpatient out-of-pocket cost and inpatient out-of-pocket cost, were obtained from the Harmonized CHARLS section. CHE were defined as out-of-pocket (OOP) health spending equal to or higher than 40 % of a household's capacity to pay. A Tobit linear regression with three models and a path analysis were conducted to estimate the association between DSL, depression and healthcare expenditures and CHE. Then we utilized 2011CHARLS and 2018CHARLS to present a longitudinal analysis. A path analysis was conducted to estimate the association between 2011DSL, 2018depression and 2018healthcare expenditures and CHE.

Results: Depression has a significant mediating effect between DSL and healthcare expenditures. (For outpatient OOP cost: a = 0.453, b = 23.559, c = 25.257, the proportion of mediating effect in total effect = 29.71 %; for inpatient OOP cost: a = 0.453, b = 13.606, c = 15.463, the proportion of mediating effect in total effect = 28.50 %; all P < 0.05). The mediating effect of depression also exists in the association between DSL and CHE (a = 0.453, b = 0.018, c = 0.043, the proportion of mediating effect in total effect = 15.90 %; P < 0.05). The mediation effect of depression on healthcare expenditures and CHE also exists in the longitudinal analysis using CHARLS 2011 and CHARLS 2018 (all P < 0.05).

Limitations: The DSL status were based on self-report and we used 2018CHARLS to conduct the study, which may cause some bias.

Conclusion: Significant mediating effect of depression exists between DSL and higher healthcare expenditures and CHE. The mental health of elder people with DSL should be focused on, and we should have an overall viewpoint on the topic of healthcare expenditures and CHE.