Breastfeeding and Early Childhood Dental Caries: Evidence from Birth Cohort Studies in Australia and Brazil

While breastfeeding is strongly recommended for infant health, uncertainty remains regarding its independent association with early childhood caries after accounting for sugar exposure. <b>Objective:</b> This study aimed to evaluate the controlled direct effect of breastfeeding on dental caries. <b>Methods:</b> Data were drawn from two population-based birth cohort studies: the 2015 Pelotas Birth Cohort Study (Brazil) and the SMILE Study (Australia). The exposure was any breastfeeding at 3, 6, and 12 months, with sugar consumption at 12 and 24 months as the mediator. The outcome was dental caries at ages 4/5, assessed as early childhood caries (ECC), severe ECC, any disease experience, or any decayed teeth. Marginal Structural Models with inverse probability of treatment weight were used to estimate the controlled direct effect of breastfeeding on dental caries, accounting for sociodemographic confounders and sugar consumption. <b>Results:</b> A total of 751 Australian and 3545 Brazilian children were included in at least one sub-analysis. Findings indicate a contextual difference: in Australia, breastfeeding had no effect on dental caries after considering sugar consumption as mediator, whereas in Brazil, children not breastfed had a lower risk of dental caries. For instance, adjusted Brazilian estimates showed a reduced relative risk ranging from 0.63 (95% CI 0.55; 0.72) for ECC to 0.43 (95% CI 0.34; 0.55) for severe ECC. <b>Conclusions:</b> The association between breastfeeding and dental caries appears to vary across socio-environmental contexts. In settings with high caries burden, prolonged breastfeeding may increase caries risk independently of sugar consumption; however, breastfeeding remains strongly recommended given its substantial overall health benefits. These findings highlight the importance of integrating early oral health guidance such as oral health education into breastfeeding support programs.