Externalizing behavior is prospectively associated with intake of added sugar and sodium among low socioeconomic status preschoolers in a sex-specific manner

Erica C. Jansen, Alison L. Miller, Julie C. Lumeng, Niko Kaciroti, Holly E. Brophy Herb, Mildred A. Horodynski, Dawn Contreras, Karen E. Peterson

High intake of added sugar and sodium is a public health concern for preschool-aged children living in the US. Externalizing behavior may predict higher consumption of added sugar and/or sodium; however, previous studies have mostly been cross-sectional. The aim was to evaluate whether externalizing behavior is prospectively related to added sugar and intake in a sex-specific manner among preschoolers. This was a secondary analysis of 524 preschool children (48% male) from Michigan who participated in an obesity prevention trial that occurred during one school year from 2011 to 2015. Teacher-assessed externalizing behaviors and three 24-h dietary recalls were completed at baseline and follow-up. We used linear mixed effects regression to evaluate the association between externalizing behavior at baseline and added sugar (% of total Calories) and sodium intake (mg/1000 Calories) at follow-up. In adjusted analysis, we included baseline income-to-needs ratio, child race/ethnicity, and baseline overweight status. All models were adjusted for total energy intake and accounted for clustering by classroom. Baseline externalizing behavior was positively associated with added sugar intake at follow-up among boys; after adjustment for confounders, every 5 points lower externalizing T-score (corresponding to higher externalizing behavior) was associated with a 0.6 higher percentage of added sugar per total Calories (95% CI 0.2 to 1.1; P value = 0.004). In contrast, girls with higher levels of externalizing behavior had lower consumption of added sugars; after confounder adjustment, every 5 points lower externalizing T-score was related to 0.6 lower percentage intake (95% CI -1.0 to -0.1; P value = 0.01). Baseline externalizing behavior was inversely associated with sodium intake at follow-up among boys. After potential confounder adjustment, for every 5 points lower externalizing behavior T-score, there was a 22 mg/1000 Cal lower sodium intake (95% CI -45 to 1; P value = 0.06). In contrast, after adjustment for confounders, every 5 points lower externalizing T-score among girls was related to 24 mg/1000 Cal higher sodium intake (95% CI 1 to 46; P value = 0.04). Externalizing behavior among preschool-aged children was prospectively related to added sugar and sodium intake in a sex-dependent manner. NCT01398358 Registered 19 July 2011.