Patient preference and adherence
Direct and Indirect Pathways Associated with Self-Rated Health Among Older Adults with Diabetes in China: A Path Analysis of CHARLS 2020
By Yu X, Sun Q, Zhang D and 2 others.
Abstract
Objective
Using data from the fifth wave of CHARLS, this paper applies path analysis to examine the determinants and mechanisms of self-rated health among elderly diabetic patients, aiming to inform individualized clinical health management and improve patients' health and quality of life.
Method
Analyses were conducted using SPSS 27.0 and Mplus 8.3. A path analysis model was developed based on the Andersen behavioral model, and parameters were estimated with the WLSMV estimator given the categorical outcome variables.
Result
A cross-sectional study design was adopted, and the final analysis sample included 1440 elderly people diagnosed with diabetes. Among them, 835 cases (accounting for 58.0%) rated their self-health as good or average, and 605 cases (accounting for 42.0%) rated their self-health as poor. The path analytic model showed an adequate fit to the empirical data. The path coefficient indicates that the degree of pain is relatively mild (β=0.212, P<0.001), the number of comorbid chronic diseases is relatively small (β=0.400, P<0.001), there is no experience of falling (β=0.291, P=0.001), and the ability of daily activities is relatively strong (β=-0.624, P < 0.001). Elderly diabetic patients who had fewer hospitalizations in the past year (β=0.204, P<0.001), consumed alcohol (β=-0.347, P<0.001), and had longer sleep duration (β=-0.257, P<0.001) had better self-rated health conditions. Beyond their direct effects, pain, comorbidity burden, daily activity capacity, and fall history also exert indirect influences on self-rated health, mediated through three variables-hospitalization frequency, sleep duration, and alcohol consumption.