Maternal well-being is associated with child health and developmental outcomes, including obesity risk (Woo Baidal et al., 2016). Such associations may be due to factors such as child feeding practices, though findings are inconsistent (McPhie, Skouteris, Daniels and Jansen, 2014). Being of low socio-economic status is associated with higher risk of poor child feeding and dietary intake (Zarnowiecki, Dollman and Parletta, 2014; Darmon and Drewnowski, 2008), which in turn is an important predictor of childhood obesity (Perry et al., 2015). There is mixed evidence suggesting that maternal stress and depression impact on weight outcomes (Tate, Wood, Liao and Dunton, 2015) and dietary intake (O’Connor et al., 2017), possibly through impacting the parent-child feeding relationship (Webb, Zimmer-Gembeck, Scuffham, Scott and Barber, 2018). Maternal self-esteem influences the parent-child relationship but has not been investigated in relation to weight outcomes (Benton, Skouteris and Hayden, 2015).
The aim of this study is to examine the effect of stress, depression and self-esteem on child dietary intake in a low socio-economic group.
The Preparing for Life (PFL) study was a longitudinal, randomised controlled trial aimed at improving the school-readiness of children in several disadvantaged communities in North Dublin. We used data from the PFL study (Northside Partnership, Doyle and UCD Geary Institute PFL Evaluation Team, 2018) that were collected when study children were eighteen months old. The following data were collected by trained interviewers using Computer Assisted Personal Interviewing software: parenting stress (Parenting Daily Hassles Scale (PDH; Crnic and Greenberg, 1990) using Frequency and Intensity subscales); parenting stress (Parenting Daily Hassles Scale (PDH; Crnic and Greenberg, 1990) using Frequency and Intensity subscales); depression (Edinburgh Postnatal Depression Scale (EPDS; Cox, Holden and Sagovsky, 1987) with scores of nine or over representing clinically elevated depressive symptoms); depression (Edinburgh Postnatal Depression Scale (EPDS; Cox, Holden and Sagovsky, 1987) with scores of nine or over representing clinically elevated depressive symptoms); and self-esteem (six-item Rosenberg Self-Esteem Scale (RSE; Rosenberg, 1965)); as well as demographic information. Child dietary intake was assessed using a Food Frequency questionnaire (FFQ) adapted from the Lifeways Cross-Generation Cohort Study (Shrivastava et al., 2013). Binary ‘healthy’ or ‘unhealthy’ categorical variables were created to represent whether children were or were not meeting Irish dietary recommendations for toddlers (Health Promotion Unit of the Department of Health and Children, 2004) for each food group (protein, fruit and vegetables, dairy, protein, and sugars/fats - see Table 1). A composite ‘Overall Healthy Diet’ binary variable was also created to quantify the number of children meeting all dietary recommendations listed above.
Descriptive statistics were used to explore associations between variables of interest and each dietary outcome variable using t-tests, chi square tests and Mann Whitney tests. Multivariable logistic regressions were conducted to further explore relationships between exposure and outcome variables and were adjusted for relationship status, household income and maternal age. Analyses were carried out using SPSS Version 24.
A total of 225 participants provided data at eighteen months, representing a response rate of sixty-eight per cent of the initial sample of 322. The key results are presented in Table 2. Seventy participants reported having maternal depression. Mean PDH Frequency score was 33.83, and median PDH Intensity score was 26. Median RSES score was 14. Overall, only two children (0.9%) were reported to have a diet meeting all dietary requirements at eighteen months. None of the psychological variables significantly predicted any dietary intake outcomes.
Dietary recommendations were met by less than one per cent of participants. This corroborates research on the poorer diet quality of people in low socio-economic brackets (Harrington et al., 2011; Kunaratnam, Halaki, Wen, Baur and Flood, 2018). As this analysis did not produce significant associations between maternal well-being and dietary intake, socio-economic disadvantage may be a more salient factor in child dietary intake than maternal psychological factors and may explain the linkages between maternal well-being and infant feeding within the literature. This may occur due to financial constraints on the type of food available and consumed in the home, or differences in feeding practices and beliefs among mothers of different socio-economic groups (Zarnowiecki et al., 2014). Maternal stress, depression and self-esteem may also influence weight outcomes through other weight-related behaviours, but not dietary intake.