Clinical oral health indicators and self-rated oral health in Chile: effect modification by educational attainment in a cross-sectional study of the 2016– 2017 National Health Survey

Abstract Background Self-rated oral health (SROH) is widely used in population surveillance, but its correspondence with clinical oral status may differ across socioeconomic groups. We examined whether educational attainment modifies the association between clinically assessed oral health indicators and SROH in Chile. Methods Cross-sectional analysis of 5520 participants aged ≥ 15 years who underwent a standardized oral examination in the 2016–2017 Chilean National Health Survey. The indicators were the number of missing teeth, the number of teeth with cavitated carious lesions, maxillary and mandibular anterior tooth loss, and removable denture use. SROH was dichotomized as good (very good/good) or not good (fair/poor/very poor). Separate logistic regression models, adjusted for sex, age, time since the last dental visit, anxiety/depression symptoms and, in three models, smoking, estimated odds ratios (ORs). Effect modification by educational attainment (< 8, 8–12 and ≥ 13 years of schooling) was assessed using stratum-specific ORs, Mantel–Haenszel tests of homogeneity and product terms; additive interaction was quantified with the relative excess risk due to interaction (RERI). Results Not-good SROH was reported by 55.1% of participants. All indicators were associated with not-good SROH, most strongly maxillary anterior tooth loss (adjusted OR 2.02; 95% CI 1.76–2.33), followed by removable denture use (1.28; 1.11–1.48), mandibular anterior tooth loss (1.25; 1.09–1.44), each tooth with cavitated caries (1.16; 1.12–1.20) and each missing tooth (1.03; 1.02–1.04). For all five indicators, stratum-specific ORs increased with educational attainment. Effect modification was clearest for removable denture use (ORs 0.87, 1.36 and 1.76 for low, intermediate and high attainment; Mantel–Haenszel p = 0.020). It was statistically significant but of negligible magnitude for missing teeth, and inconsistent for mandibular anterior tooth loss (RERI 0.34; 95% CI − 0.01 to 0.70). No modification was observed for cavitated caries or maxillary anterior tooth loss. Conclusions Clinical oral conditions were reflected less strongly in the self-ratings of people with fewer years of schooling, particularly for conditions whose appraisal depends on social norms and expectations rather than on visibility or symptoms. Surveillance that relies exclusively on SROH may therefore underestimate educational inequalities in oral health; self-reported and clinical indicators should be monitored together.