Cumulative social disadvantage and incident complete tooth loss among middle-aged and older adults: evidence from CHARLS and HRS

Complete tooth loss (edentulism) is linked to multiple adverse health outcomes, yet the role of multidimensional social disadvantage in its onset remains poorly understood. Existing evidence has largely focused on single socioeconomic indicators. We used data from two nationally representative cohorts: CHARLS (Wave 3 to Wave 4, N = 3,906) and HRS (Wave 8 to Wave 11, N = 6,135), including participants aged ≥ 50 years and dentate at baseline. A cumulative social disadvantage score (range 0–6) was constructed from six indicators: low education, low wealth, non-employment, rural residence, no social activity, and digital exclusion. Incident complete tooth loss was defined as transition from having at least one natural tooth at baseline to none at follow-up. Logistic regression estimated odds ratios (ORs) with sequential covariate adjustment. Restricted cubic splines and subgroup analyses were performed to examine dose-response patterns and effect modification. In fully adjusted models, each one-point score increment was associated with 15% higher odds in CHARLS (OR = 1.15, 95% CI: 1.00–1.32) and 36% higher odds in HRS (OR = 1.36, 95% CI: 1.23–1.51). Dose-response relationships were linear in both cohorts (CHARLS P-nonlinear = 0.510; HRS P-nonlinear = 0.702). In CHARLS, the point estimate was highest among participants aged 50–64 years (OR = 1.29), although the age interaction was not significant ( P = 0.317); HRS showed consistent effects across all age strata. Cumulative social disadvantage independently predicts incident complete tooth loss in both Chinese and American older adults, with cross-national differences in effect magnitude but similar linear dose-response patterns. Structural policies reducing multidimensional social disadvantage may help prevent edentulism and narrow oral health disparities in aging populations.