Association between breastfeeding and lower risk of behavioral problems in children: a systematic review and...

To systematically evaluate the association between breastfeeding and the risk of behavioral problems in children, and to provide evidence-based support for early prevention strategies.

PubMed, the Cochrane Library, Embase, Web of Science, SinoMed, CNKI, WanFang Data, and VIP were systematically searched from database inception to November 2025 for observational studies examining the association between breastfeeding and behavioral problems in children. Two reviewers independently screened the literature, extracted data, and assessed study quality. Meta-analysis was performed using Stata MP 18.0, and pooled odds ratios (ORs) with 95% confidence intervals (CIs) were calculated.

Ten studies involving 69,621 children were included, comprising four cohort studies and six cross-sectional studies. Breastfeeding was significantly associated with a lower risk of overall behavioral problems (OR = 0.66, 95% CI: 0.51-0.86, p = 0.002), internalizing behavioral problems (OR = 0.79, 95% CI: 0.69-0.89, p < 0.001), and externalizing behavioral problems (OR = 0.78, 95% CI: 0.69-0.89, p < 0.001). Substantial heterogeneity was observed for overall behavioral problems, whereas the pooled results for internalizing and externalizing behavioral problems were relatively stable. Subgroup analyses indicated that the association between breastfeeding and overall behavioral problems varied by study design and region. No significant association was observed in cohort studies or in studies conducted in China, whereas significant inverse associations were found in cross-sectional studies and in studies conducted in other countries. Longer breastfeeding duration (≥6 months) showed a stronger trend toward a lower risk, although the between-subgroup difference was not statistically significant.

The findings suggest a potential association between breastfeeding and a lower risk of behavioral problems in children. However, the available evidence is limited, and residual confounding and publication bias cannot be excluded. Further high-quality prospective studies are needed to clarify the nature, magnitude, and potential causal mechanisms underlying these associations.

https://www.crd.york.ac.uk/prospero/, identifier CRD420251233415.