Development and Validation of a Nomogram for Predicting Dental Caries Risk in Adolescents Aged 12 to 18 Years

Introduction and aims To develop and internally validate a nomogram for estimating the probability of dental caries among adolescents aged 12 to 18 years using readily available demographic and behavioural factors, with the aim of supporting early risk assessment and individualised preventive strategies. Methods A cross-sectional study was conducted using a multistage, stratified, cluster sampling method to survey 3359 adolescents from 21 secondary schools in Yinchuan, China. Caries status was assessed using the International Caries Detection and Assessment System-II (ICDAS-II), and the primary outcome was defined according to the DMFT index. Participants were randomly assigned to either a training group (70%) or a validation group (30%). LASSO regression identified non-zero coefficient predictors, followed by multivariable logistic regression to determine optimal predictors and construct a nomogram. Model discrimination was evaluated using ROC curves (AUC), calibration curves assessed prediction accuracy and decision curve analysis (DCA) evaluated clinical utility. Sensitivity analyses were performed using an alternative outcome definition and an alternative data-splitting strategy to evaluate model robustness. Results The final model included 7 predictors: older age (OR = 1.14, 95% CI:1.06-1.21), female sex (OR = 1.81, 1.45-2.25), no post-meal brushing/rinsing (OR = 1.58, 1.27-1.97), poor self-rated oral health (OR = 1.60, 1.35-1.91) and tooth sensitivity to stimuli (OR = 1.28, 1.02-1.60) increased caries risk; fewer sugary milk beverages (OR = 0.66, 0.58-0.76) and thorough tooth brushing (OR = 0.54, 0.42-0.69) were associated with reduced risk. The AUCs were 0.708 (training, 95% CI:0.682-0.733) and 0.715 (validation, 95% CI:0.674-0.755). Hosmer-Lemeshow tests indicated good calibration ( P = .220/.746), and potential clinical utility. Sensitivity analyses yielded comparable results, supporting the robustness of the model. Conclusion This internally validated nomogram demonstrated acceptable performance for estimating the probability of dental caries among adolescents within the study population. The model may assist risk stratification and support preventive decision-making in school-based oral health programs and routine clinical practice. Further external validation is warranted before broader implementation. Clinical significance This non-invasive, easily applicable nomogram uses routinely available demographic and behavioural information to support dental caries risk assessment among adolescents and may facilitate targeted preventive interventions in school and clinical settings.