Association Between Breastfeeding and Autism Spectrum Disorder in Preschool Children: An Analysis of Data from the Study to Explore Early Development (SEED)

Background: Most causes of autism spectrum disorder (ASD) are not well understood. The potential effect of breastfeeding is of interest because breastmilk provides key nutrients for physical and cognitive development of the child, and breastfeeding contributes to maternal-child bonding. However, breastmilk may also contain environmental pollutants or other substances with potential negative effects on cognitive development. Conflicting results have been reported in past studies on the relationship between breastfeeding and ASD, and few studies have examined a possible association with its duration. Methodological limitations noted in past studies include small samples, use of parental report to document ASD diagnosis, lack of a comparison group, and incomplete adjustment for confounding factors.

Objectives: To evaluate possible associations between ASD and (1) breastfeeding and (2) breastfeeding duration, and to identify factors associated with breastfeeding and its duration in cases and typically-developing controls.

Methods: We used data from the Study to Explore Early Development (SEED), a multi-site case-control study that includes children aged 2-5 years ascertained through birth certificates (controls) and clinical and educational sources providing services to children with developmental disabilities (cases). Enrolled children underwent an extensive assessment by expert clinicians using standardized instruments. The diagnosis of ASD was based on the Autism Diagnostic Observation Schedule and the Autism Diagnostic Interview-Revised. Mothers participated in a telephone interview about their pregnancy, including whether the child was ever breastfed and if “yes,” then how old the child was when breastfeeding stopped. We compared cases and controls using logistic regression to assess any association with breastfeeding and multinomial logistic regression to compare child age (tertiles) at breastfeeding cessation. These analyses were adjusted for a variety of child and maternal characteristics.

Results: The sample included 707 cases and 1223 controls. Overall, 75% of children were “ever breastfed.” The mean child age (in months) at breastfeeding cessation was 7.3 in cases and 9.2 in controls. There was no independent association between breastfeeding and ASD (adjusted odds ratio [aOR]: 0.92 (95 CI: 0.65, 1.30). Compared to controls, cases were younger at the time of breastfeeding cessation: aOR comparing highest age tertile (> 12 months) to lowest age tertile (<=3 months) in cases versus controls was 0.60 (95%CI 0.42, 0.87). In both cases and controls, higher maternal education attainment was associated with breastfeeding and its longer duration. In cases, higher family income and having a mother born outside of the United States were each associated with breastfeeding, and older maternal age was associated with longer breastfeeding duration. In controls, white race was associated with breastfeeding and longer breastfeeding duration; cesarean birth, mother with a psychiatric condition, and lower family income were each associated with shorter breastfeeding duration (p-values <0.05).

Conclusions: Although initiation of breastfeeding was not significantly different between groups, its duration was shorter among ASD cases compared to controls. These results document the importance of assessing duration of breastfeeding rather than initiation alone. Multiple factors were associated with both the initiation and duration of breastfeeding, and should be taken in account when assessing the relationship between breastfeeding and ASD.