Development of infant feeding algorithms
Dietary factors in early life (infant feeding practices and timing of introduction of solid foods) are the most potentially modifiable early life exposures associated with childhood growth, as compared to genetic determinants, co-morbidity, and other environmental influences. Yet, studies assessing the association of infant feeding with growth may be limited by out-of-date data and unable to compare results due to inconsistent definitions of infant feeding practices. Mixed feeding (i.e. breast and bottle feeding) calls for special attention due to the reality of mothers returning to work after childbirth in the US. This report used data from the National Children’s Study Formative Research in Physical Measurements. A discovery set of 300 participants were selected by ethnicity from the sample available when this report was developed. This report emphasized statistical methods as well as data pre-processing, which are critical but typically under-studied. This report is intended to contribute towards closing this gap by describing a study from design, data pre-processing, to analysis. Results showed that non-Hispanic Black children had the lowest rates of ever and exclusively breastfeeding, compared to Hispanics and non-Hispanic Whites. Mothers aged 30 years and over, married, and educated above the high school level exclusively breastfed more than other mothers. Mixed feeding was categorized into three and five subgroups according to maternal recalls of the extent or frequency of breast/formula-feeding and compared by mean durations of breast/formula-feeding. Mixed feeding groups may provide unique opportunities to assess the relationship between mixed feeding versus exclusively breast/formula-feeding and childhood linear growth in the author’s dissertation. The percentage of children who were breastfed less than 6 months differed from those breastfed more than 6 months by ethnicity, child’s birthweight, gestational age, maternal age at childbirth, education level, and marital status, which suggests 6 months as a reasonable cut-off for breastfeeding categorization. Children of low birthweight and born preterm were introduced to solid foods later than those of normal/high birthweight and those born on time/postterm, even after adjusting for ethnicity. Analyses on a re-test set will be performed and compared to this discovery set in the author’s dissertation.