Estimating the causal impact of chewing disability on depressive symptoms mediated by loneliness: a longitudinal marginal structural model study of older adults in Singapore.

Background and Objectives: Chewing disability may contribute to depressive symptoms in older adults, but causal pathways, accounting for time-varying confounding factors, remain unexplored. Previous research shows an association between chewing disability, loneliness, and depression. This study examines the causal relationship between chewing disability and clinically significant depressive symptoms (CSDS) and whether loneliness mediates this association among older adults. Research Design and Methods: In total, 1,277 participants aged ≥60 years, without CSDS at baseline, were selected from a nationally representative study of older adults in Singapore (2009-2015, three waves). Marginal structural models were utilized to estimate total and indirect effects of chewing disability and CSDS over time, where loneliness was treated as a mediator. Results: Across the study period, 10.3% developed CSDS, 40.7% experienced loneliness, and 33.6% had chewing disability. Individuals with chewing disability had a 48% higher risk of CSDS (RR: 1.48, 95% CI: 1.15-1.82), and the indirect effect through loneliness was 26% (RR: 1.26, 95% CI: 0.99-1.53; 17.3% of the total effect). Nonetheless, the estimates varied by the operationalization of chewing disability and loneliness. A broader definition of chewing disability showed stronger total effects (RR: 1.57, 95% CI: 1.24-1.91), while a stricter loneliness threshold had a greater indirect effect (RR: 1.70, 95% CI: 1.30-2.09; 21.8% of the total effect). Discussion and Implications: Chewing disability increases the risk of CSDS among older adults, with partial mediation by loneliness. Further research on oral rehabilitative interventions that improve chewing function and mitigate depressive symptoms in older adults is needed.