The Association Between Developmental Risk Status and Early Feeding Patterns

Background: Feeding practices: a vital physical and psychological aspect of the parent-infant interaction are influenced by an array of factors.  Although the presence of a child with Autism Spectrum Disorder (ASD) contributes to a markedly different family milieu than the presence of an unaffected child (Davis & Carter, 2008), there has been little research on the effect of having an older child with Autism Spectrum Disorder (ASD) on the feeding practices employed with younger offspring. 

Objectives: This descriptive and exploratory study examines whether a risk status for ASD influences early feeding patterns.

Methods: We analyzed questionnaire data submitted by the parents of 81 infants classified as high risk (HR) or low risk (LR) for ASD.  HR infants (N=47) had an older sibling with ASD, whereas LR infants (N=34) did not.  Participants were enrolled in a longitudinal study examining early signs of ASD.  Data collected at 12 months included questions on the feeding method (breast, bottle-fed, or both), feeding difficulties, and parental education level.  At 24 months, infants were classified based on Mullen, ADOS, Vineland, and CSBS data as likely having ASD (N=12), Broader Autism Phenotype (BAP; N=14), or other specific delays (DD; N=16), or no concerns (N=39). 

A chi-square test was used to examine a relationship between risk status (HR or LR) and feeding method (breast/both or bottle).  The HR sample was further examined for an association between feeding method and clinical concerns. 

Results: HR infants were more likely to be fed exclusively with a bottle (23%) than LR controls (3%), x²(1)= 6.55, p=.01.  However, neither group reported feeding difficulties (13% and 21% for HR and LR groups, respectively).  No relationship to education level was found.

A comparison within the HR group suggests a significant relationship between feeding method and the presence of clinical concern at 24 months, x²(1)=5.49, p=.02.  When separated into 24 month diagnostic categories, HR infants who raised clinical concerns at 24 months were more likely to be exclusively bottle-fed than infants with no clinical concerns, x²(6)=10.85, p=.09 (30%, 36%, 30%, and 0% for the ASD, BAP, DD, and No Concerns groups, respectively). 

Conclusions: Results suggest a significant association between risk status and feeding method.  HR infants who later exhibited clinical concerns were less likely to be breast-fed (exclusively or in combination with bottle) than LR and HR infants without clinical concerns at 24 months. 

While the exact nature of this relationship remains unclear, this study raises questions related to feeding practices in high-risk populations.  Some hypotheses potentially explaining this phenomenon include: high stress levels associated with raising an older child with ASD or the influence of previous breast feeding experience with an older child with ASD on parental choices regarding breastfeeding of the younger sibling.  In the group that demonstrated a varied degree of developmental difficulties at 24 months, feeding practices may have been influenced by oral-motor or behavioral regulatory difficulties during the neonatal period. We are examining some of these hypotheses both in high-risk and high-density samples; this data will be available in May 2011.