ABSTRACT
Précis: Our goal is to assess both the direct and indirect effects of diet during the postpartum period in
breastfeeding mother-infant dyads. In addition to accomplishing the primary aims of a randomized controlled
trial (RCT) designed to characterize the effects of diet-transmitted vegetable flavors in mother’s milk during
sensitive period in flavor learning, we have a comprehensive data set of repeated and diverse maternal and
infant phenotypes collected during the course of the RCT. The trial lasted from 2 weeks postpartum until the
infants were weaned to solid foods, with post-intervention visits at 10.5 months and 1 year. Therefore, we have
a unique opportunity that goes beyond the goals of the original RCT on the highly pervasive phenomenon of
breastfeeding and nutritional programming.
The secondary outcome phenotypes collected throughout the intervention and post-intervention periods (10.5
mos, 1 y) included measures of anthropometry (from which we determined weight gain trajectories [g/d], a risk
factor for later obesity) and dietary intake of both members of the dyad; infant developmental milestones; fat
and caloric content of breast milk; and maternal feeding styles and goals. The secondary analyses we propose
here will test several hypotheses to compare the secondary outcomes with primary outcomes of lab-based
measures of infant vegetable acceptance, maternal vegetable taste ratings, and taste genotype. Breastfeeding
(BF) was robust and persistent for our RCT cohort, lasting for 10.5 mos or longer for 81% of the infants, and
only one-third ever fed infant formula. Thus, for two aims (AIMS 1, 2) we will also compare data collected from
this well-characterized BF cohort with data collected using identical methods on a longitudinal cohort of
exclusively formula-fed (FF) infants and mothers. At the heart of the proposal are four aims, which upon
completion will yield at least six publications.
AIM 1 will use data from repeated measures of maternal feeding practices and breast milk composition to test
the primary hypothesis that the fat content of breast milk and maternal responsive feeding contributes to infant
weight gain trajectories over time. Because breastfeeding protected against obesity at 1 y compared to the FF
cohort, we will also determine whether there are differences between BF and FF groups in maternal feeding
styles over time and the type of complementary diet they feed their infants. AIM 2 will determine how early risk
factors—rapid weight gain trajectories and breast milk composition,—interact together or act independently to
influence primary tooth eruption, a hypothesized biomarker for later obesity risk. AIM 3 will characterize
nutrient and energy content of the diets of mothers over time and concordance with their children’s diets. While
genetic variation did not contribute to initial vegetable acceptance, AIM 4 will test the hypothesis that TAS2R38
genotype explains variation in post-intervention diet measures such that those with insensitive alleles will show
greater vegetable intakes than those homozygous for the bitter-sensitive allele.
The knowledge gained from these secondary analyses could be translated to behavioral interventions to
promote healthy maternal diet and feeding practices that in turn support healthy weight gain and diet in infants.
