Breastfeeding’s link to teen cognition emerges after researchers account for family background

by · News-Medical

In Chinese youths, longer breastfeeding showed no clear cognitive link until researchers accounted for socioeconomic disadvantage, revealing modest gains in math and word recognition that highlight how family context can reshape child development findings.

Study: Breastfeeding Duration and Cognitive Performance Among Youths. Image Credit: Asian foto / Shutterstock

Breastfeeding Duration and Cognitive Performance Background

Does breastfeeding for longer actually make children smarter, or are family circumstances the real driver? The World Health Organization (WHO) recommends breastfeeding as the best source of nutrition for infants and recommends exclusive breastfeeding for the first 6 months, followed by continued breastfeeding for at least 2 years, but evidence remains inconsistent on whether breastfeeding has long-term cognitive benefits.

The main challenge is socioeconomic status (SES), which strongly shapes both feeding practices and learning environments. In many countries, wealthier families breastfeed longer, making it difficult to isolate true biological effects. Conversely, in some populations, the opposite trend exists.

Understanding how the duration of breastfeeding is associated with a child’s cognitive development in diverse settings will provide the basis for developing appropriate health and education policies.

Further research is needed to provide clarity on the causal relationship between breastfeeding and cognition.

Breastfeeding and Youth Cognition Study Design

This cross-sectional study used data from the China Family Panel Studies (CFPS), a nationally representative longitudinal survey conducted by the Institute of Social Science Survey at Peking University. The study included youth aged 10-15 years that participated in waves of data collection in 2010, 2014, or 2018 and who were born between 1995-2009. Institutional ethics approval for the study was obtained from the Peking University Biomedical Ethics Review Committee, and informed consent was collected.

Breastfeeding duration was assessed retrospectively from caregiver reports and classified as less than or equal to 6 months or greater than 6 months, per WHO recommendations. This measure captured any breastfeeding, including exclusive and mixed feeding. Standard tests of cognitive performance were completed, including mathematics and word recognition tests based on school curricula. Standardized scores were used, and all scores were converted to age-adjusted z-scores to control for differences in developmental level. Poor cognitive performance was defined by a score at or below the 15th percentile.

Covariates included child characteristics (sex, ethnicity, gestational age, birth order, birth year), parental factors (age at birth, educational level), and household indicators (residence, income, cooking fuel type). Survey-weighted generalized linear models were used to estimate associations. Three models were applied: unadjusted, non-SES adjusted, and fully SES-adjusted. Sensitivity analyses, including adjustment for parental cognition and birth weight, were performed to ensure robustness.

Socioeconomic Adjustment and Cognitive Findings

The final sample included 5,436 youths with a mean age of 12.03 years, of whom 52.5% were male. Among them, 13.8% were breastfed for 6 months or less, while 86.2% were breastfed for a longer period of time. It was noted that longer breastfeeding was more common among families with lower SES, including lower parental education and household income, and these children were also more likely to live in rural areas and use unclean cooking fuels, reflecting broader environmental disadvantages.

Time trends showed an overall decline in breastfeeding duration across birth cohorts from 1995 to 2009. However, even as overall duration decreased, longer breastfeeding remained more prevalent in lower-income households and among parents with less education. This pattern of association contrasts with what is observed in many high-income countries, where higher SES is usually associated with longer breastfeeding duration.

Cognitive function was not significantly associated with how long a child was breastfed in unadjusted analyses. In this study, children who were breastfed for more than six months had a slightly higher, non-significant point estimate for math tests (β = 0.03; 95% confidence interval (CI), −0.06 to 0.13) and a slightly lower, non-significant point estimate for word recognition tests (β = −0.02; 95% CI, −0.13 to 0.09) than those who were breastfed for six months or less. However, after adjustment for non-SES covariates, then for SES, the estimates became positive, and in the fully SES-adjusted model, they were statistically significant. The children who had been breastfed for over six months also had significantly better math scores (β = 0.14; 95% CI, 0.05 to 0.22) and word recognition scores (β = 0.12; 95% CI, 0.02 to 0.21).

In the fully SES-adjusted analyses of binary outcomes, there was a 35% decrease in the odds of poor mathematics performance (odds ratio (OR), 0.65; 95% CI, 0.48 to 0.88) and a 36% decrease in the odds of poor word recognition performance (OR, 0.64; 95% CI, 0.47 to 0.87) associated with longer periods of breastfeeding. This study provides evidence that breastfeeding duration is associated with small improvements in cognitive ability, even after controlling for SES.

Nonlinear analyses using restricted cubic splines suggested that the lowest odds of poor performance occurred around 6 to 12 months of breastfeeding, with attenuation at longer durations. This suggests that associations may not be strictly linear, but does not establish an optimal breastfeeding duration, and may reflect the timing of infant development, changes in milk composition, or the introduction and quality of complementary feeding during this period.

Sensitivity analyses supported the robustness of these findings across different analytic specifications, adjustments, and analyses. These findings also highlight how an individual's socioeconomic context can either enhance or obscure an association with breastfeeding.

Breastfeeding Policy and Research Implications

In this cross-sectional analysis, breastfeeding duration was not associated with cognitive performance in unadjusted models but showed a positive association after accounting for SES. Longer breastfeeding was more common among lower socioeconomic groups, which initially obscured its relationship with cognitive outcomes. After adjustment, modest improvements in mathematics and word recognition were observed, along with reduced odds of poor performance.

These results indicate that breastfeeding may be associated with adolescent cognitive development, although causality cannot be inferred. Supporting breastfeeding through policy, along with decreasing SES inequalities, may help improve child development opportunities, but further longitudinal research is needed.

Journal reference: