Comparison of Salivary Statherin and Beta-Defensin-2 Levels, Oral Health Behaviors, and Demographic Factors in Children With and Without Early Childhood Caries Publisher Pubmed



M Koopaie MARYAM ; Fk Shahri Faezeh KHAJEHREZA ; R Montazeri ROSHANAK ; S Kolahdooz SAJAD ; Mm Shahri Majid MARDANI ; E Moshkbouy ELHAM
Source: BMC Oral Health Published:2025

Abstract

Background: Early childhood caries (ECC) is a widespread pediatric dental condition that is influenced by a combination of biological, behavioral, and demographic factors. Salivary biomarkers, including beta-defensin-2 (BD-2) and statherin (STATH), offer potential as non-invasive tools for detecting and assessing the risk of ECC. This study aims to compare the levels of salivary statherin and beta-defensin-2, alongside oral health behaviors and demographic factors, in children both with and without early childhood caries. Methods: This case-control study involved 75 children diagnosed with ECC and 75 age- and gender-matched caries-free controls. Unstimulated saliva samples were obtained and analyzed via ELISA to quantify the levels of beta-defensin-2 and statherin. Demographic and behavioral data were gathered through structured interviews with parents. Statistical analyses included t-tests, logistic regression, and machine learning models to predict the risk of ECC. Results: Salivary beta-defensin-2 levels were significantly higher in children with ECC (9.25 ± 2.89 ng/mL) compared to caries-free controls (6.41 ± 2.45 ng/mL, p = 0.003), indicating its potential as a diagnostic biomarker. Statherin levels, although lower in the ECC group, did not differ significantly between groups (p = 0.08). Behavioral factors such as regular dental visits and parental education levels were strongly associated with ECC prevalence. Machine learning models demonstrated high accuracy in predicting ECC, with the Gradient Boosting and CatBoost achieving the highest performance. Conclusions: Salivary beta-defensin-2 is a promising ECC risk assessment biomarker, while statherin is less effective as an independent predictor. Behavioral and demographic factors significantly influence ECC prevalence. Machine learning models integrating clinical, demographic, and salivary data provide a robust tool for detection and targeted prevention strategies. Comprehensive approaches combining salivary biomarkers and behavioral interventions are critical to managing ECC, particularly in resource-limited settings. © 2025 Elsevier B.V., All rights reserved.