Mediating and moderating role of health and social care utilisation in the association between social frailty and depression

Frailty describes the condition that the age-related deteriorations in the various systems within the human body, which lead to subsequent loss of ability to carry out necessary daily activities. Recent literature expanded the concept of physical frailty to include multidimensional domains. Among them, social frailty refers to the loss of social and general resources, activities or abilities that are essential to fulfil basic social needs. Despite the international interest of assessing the relationship between social frailty and depression, there is a lack of longitudinal evidence from primary prospective cohort studies. Furthermore, as social frailty refers to the vulnerability of an individual in meeting social needs, the social determinants of health may play a crucial role to explain its negative health consequence. The Andersen’s Healthcare Utilisation Model might provide a coherent framework to delineate how social frailty affect care seeking, which in turn affect health outcome. According to the Bunt’s social frailty model, the resources limitations underpinning social frailty tie in with the enabling resources as stated in the Andersen’s model. The proposed study aimed to quantify the longitudinal relationship between social frailty and depression; to investigate the moderating and/or mediating role of health and social services utilisation in such relationship; and to explore the bidirectional relationships of social frailty, depression and health and social services utilisation among older adults. A two-year prospective cohort study will be conducted with 1000 community-dwelling older Chinese adults in Hong Kong. Purposive sampling will be employed to recruit participants with diverse levels of social frailty and health and social service utilistaion. Potential participants will be recruited from community centres, public areas, health care centres/clinics, elder academy and social media, and older households and hidden elders lists of the non-governmental organisations and the PI’s working contacts. In addition, a random sample will be recruited to explore the composition of the different subgroups. A structured questionnaire will be used, with the 8-item social frailty scale, a modified local adaptation of Client Service Receipt Inventory, the Geriatric Depression Scale, and other covariates. Baseline and four half-yearly assessments will be performed. The integrated model based on Andersen’s Healthcare Utilisation Model and Bunt’s social frailty model will be used to guide the whole study and interpretation of results. Linear mixed models and structural equation models will be used in the statistical analysis. The findings would inform service providers on the provision of services to reduce social frailty and depression.