Evaluation of self-perceived oral function among older adults using linear regression and structural equation modeling

With global population aging, oral hypofunction in older adults has become increasingly prominent. Self-perception is a key factor for early identification. Establishing effective predictive models is essential for improving oral health management. This study aimed to identify factors associated with self-perceived oral function and validate the model structure.

Data were obtained from a large-scale cross-sectional survey conducted in China in 2021. Dental examination and questionnaire data from 423 adults aged 65-74 years were analyzed using SPSS. Spearman correlation, Mann-Whitney U test, and Kruskal-Wallis test were used to screen variables for a multiple linear regression model. A structural equation model (SEM) was constructed using AMOS, and confirmatory factor analysis (CFA) was performed to assess model fit.

Predictors were identified from four domains: I) demographics, II) dental service utilization, III) oral health knowledge, self-esteem, and practices, and IV) health status. The regression model explained 21.6 % of the variance in self-perceived oral function scores (p < 0.001) without multicollinearity (VIF=1.047-2.193). Significant predictors included gender, reason for last dental visit, desserts or candy consumption frequency, tooth brushing frequency, decayed teeth index (dt), number of teeth with dental calculus, number of teeth with attachment loss (AL) ≥4 mm, and systemic disease (p < 0.05). SEM showed acceptable fit (χ²/df=1.865, RMSEA=0.045, GFI=0.982, CFI=0.823).

Oral practices emerged as the most important predictor, among other factors identified across all domains. SEM confirmed the structural validity of the regression model, enhancing the interpretability of predictors and providing a reliable method for identifying the risk of oral hypofunction in older adults.

Since oral hypofunction often manifests subtly and varies by individual perception, incorporating self-perceived measures provides critical early signals and enhances personalized preventive strategies.