Early Feeding of High Sugar-containing Foods/Drinks: A Birth Cohort Study

Objectives: Early introduction of sugar to young children can increase the preference for sugar and the risk of consuming a cariogenic high-sugar diet later in life. This study aimed to investigate early life factors influencing feeding of foods/drinks containing added sugar in the first 6-months of life.
Methods: Data from an ongoing population-based birth cohort study in Adelaide, Australia, were used. Mothers of newborn children were invited to participate and complete questionnaires at birth, and subsequently at periodic ages. The outcome was reported feeding (Yes/No) by age 6-month of common foods/drinks with high levels of added sugar (hereafter referred as sugary foods/drinks). Exposures were household income (Q1(lowest)/Q2/Q3/Q4) and mother’s sugar-sweetened beverage (SSB) consumption (Everyday/Sometimes/Never). Other socioeconomic (SES) factors and health behaviours were treated as confounders or covariates. Analysis was conducted progressively from bivariate to multivariable log-binomial regression with robust standard error estimation to calculate prevalence ratios (PR) of having early sugar feeding. Models for complete cases and with multiple imputations (MI) were generated. Confounding effect of SES and interaction between income and mother’s SSB consumption were further controlled for in a marginal structural model (MSM) with stabilised inverse probability weights to estimate controlled direct effect of income on the outcome.
Results: Of 1222 mother/child dyads, 21% of children had been fed sugary foods/drinks by 6-months. There were strong income gradient and significant association with mother’s SSB consumption. In the complete-case model, income Q1 and Q2 had PRs of 1.94(1.23–3.05) and 1.75(1.17–2.62) against Q4, respectively. The PR among mothers ingesting SSB everyday was 1.64(1.15–2.33). Breastfeeding at 3-month was protective. Similar findings were observed in the MI model. MSM estimates confirmed direct effect of income (the PRs of Q1 and Q2: 1.69(1.08–2.66) and 1.59(1.09–2.32), respectively.
Conclusions: Household income at birth and maternal behaviours were significant determinants of early feeding of sugary foods/drinks.