Association Between Dental Caries and Concentration of Selected Minerals in Drinking Water Among Mongolian Adolescents: A Cross-Sectional Study With Nationwide Calcium, Magnesium, and Chloride Analyses and Regional Fluoride Analysis

Introduction: Dental caries remains common among Mongolian children, but the relationship between the mineral composition of local drinking water and caries in adolescents has not been well characterized.Methods: This nationwide cross-sectional study linked oral health screening records for Mongolian adolescents aged 13-17 years with area-level water quality data.The primary national analysis included 92,776 participants from 323 geographic areas with calcium, magnesium, and chloride data.Generalized estimating equations with exchangeable working correlation, clustered by geographic area, estimated adjusted associations with decayed, missing, and filled teeth (DMFT) and caries experience (DMFT >0), controlling for age, sex, and macroregion.A pre-specified supplementary fluoride analysis included 75,448 participants from 217 areas in 16 provinces/municipality where fluoride data were available.Results: In the primary analysis, mean age was 14.59 (SD 1.31) years, mean DMFT was 1.87 (SD 2.47), and 59.24% had caries experience.Chloride was significantly associated with DMFT (B (unstandardized regression coefficient) per 10 mg/L = 0.014, 95% CI 0.004 to 0.024; p=0.006) and caries experience (adjusted OR=1.015, 95% CI 1.007 to 1.022; p<0.001) but with a small magnitude.Calcium and magnesium were not significantly associated with either outcome.In the restricted fluoride analysis, fluoride was weakly inversely associated with DMFT (B per 0.1 mg/L = -0.009,95% CI -0.017 to -0.001; p=0.027) and caries experience (adjusted OR=0.991, 95% CI 0.983 to 0.999; p=0.021).Conclusions: Area-level drinking water mineral concentrations were associated with dental caries outcomes among Mongolian adolescents.The fluoride result is limited to areas with available fluoride measurements, and the cross-sectional, ecological exposure design precludes causal inference.