Cross-lagged Associations Between Maternal Feeding to Soothe and Infant Sleep: Implications for Infant Weight Outcomes

• In Event: Infant sleep and the development of obesity: Direct and indirect associations in four longitudinal cohorts

Abstract

Poor sleep in infancy has been linked with subsequent obesity (Hagger et al., 2016; Li et al., 2022) and may contribute to maternal use of more obesogenic feeding practices, such as using feeding to calm a distressed infant or to settle an infant to sleep (Leerkes et al., 2022). In turn or alternatively, these feeding practices may undermine infants’ development of more adaptive self-regulation contributing to poorer infant sleep outcomes (Paul et al., 2016). In the current study, we examine cross-lag associations between two infant sleep characteristics, nighttime sleep duration and frequency of night wakings, and two obesogenic parenting practices, using food to soothe and feeding the infant to sleep, across 3 waves during infancy and the implications of these pathways for later weight outcomes.

Participants were 176 mothers (64% non-white, 45% first-time mothers) and their infants (49% female). At 2, 6, and 14 months postpartum, mothers completed questionnaires to assess the focal variables (Brief Infant Sleep Questionnaire, Food to Soothe Scale) and covariates (CES-Depression, Pittsburgh Sleep Quality Index, Infant Feeding Practices). Infants were measured and weighed at 14 months, and these measures are currently being collected at 24 months and will be complete prior to the conference.

Four cross-lagged path models were computed using MPLUS. Maternal education, race, and income were entered as time invariant covariates specified on wave 1 variables; and maternal breastfeeding status (1 = any breastfeeding), depressive symptoms, and sleep quality were specified as time varying covariates at each wave. As illustrated in Figure 1, Panel A, low infant nighttime sleep duration at 2 months predicted mothers’ greater use of feeding to sleep at 6 months which in turn predicted lower infant sleep duration at 14 months. Moreover, the indirect effect from sleep duration at 2 months to 14 months via 6 month feeding to sleep was statistically significant, B =.06, SE =.04, CI[ .01,.15]. Frequency of night wakings at 6 months predicted mothers’ greater likelihood of feeding the infant to sleep at 14 months (Figure 1, Panel B). As illustrated in Figure 2, low infant nighttime sleep duration at 2 months predicted higher maternal feeding to soothe at 6 months. No significant cross-lagged effects were apparent in the model involving night wakings and feeding to soothe. Generally, maternal feeding behaviors were more stable over time than infant sleep behaviors. Notably, lower nighttime sleep duration at 14 months was associated with infants’ concurrent higher waist to length ratio (B = -.25, p =.018) over and above infant sex, gestational age, and birthweight.

Most significant cross-lagged effects were from infant sleep to parent behavior, suggesting problematic sleepers may elicit more obesogenic feeding behaviors over time. The indirect effect demonstrates a transaction whereby infant sleep duration altered parenting (feeding to sleep) which in turn reduced infant sleep duration. If longitudinal associations to infant weight outcomes at 24 months are apparent, this will be strong evidence for the role of infant sleep and maternal feeding to soothe in early weight problems.

Authors

• Esther M. Leerkes, UNC Greensboro
  Presenting Author

• Agona Lutolli, University of North Carolina at Greensboro

• Cheryl Buehler, University of North Carolina at Greensboro

• Laurie Wideman, UNC Greensboro

• Lenka Shriver, University of North Carolina at Greensboro