UT SON Dissertations (Open Access)

Date of Award

Spring 5-2018

Degree Name

Doctor of Philosophy in Nursing (PhD)


Janet Meininger, PhD - Chair

Second Advisor

Sheryl Hughes, PhD

Third Advisor

Wenyaw Chan, PhD

Fourth Advisor

Cathy Rozmus, PhD


Background: Low-income Hispanic preschoolers face disproportionately high prevalence of food insecurity (FI) and obesity. Consumption of low-cost, energy-dense foods to compensate for FI leads to excess body-mass. FI parents may adopt feeding styles that contribute to decline in children’s dietary quality. Feeding style describes the amount of demandingness (i.e., control of children’s eating) and responsiveness (i.e., warmth used to express demandingness). FI may indirectly contribute to obesity through dietary quality and feeding style.

Purpose: This study investigated: 1. if dietary quality mediated the relationship between food security status (FSS) at Time 1 (T1) and child body-mass at Time 2 (T2), 2. if feeding demandingness (PFD) and/or responsiveness (PFR) mediated the relationship between FSS at T1 and child dietary quality at T2, 3. explored if gender and/or parental acculturation moderated the mediation.

Method: The current study was a secondary analysis of an observational study (R01 HD06257, PI: Hughes). Hispanic parent-preschooler dyads (n=137) provided data through the 6-item Household Food Security Survey, Healthy Eating Index-2015 (HEI), Caregiver’s Feeding Style Questionnaire, Bidmensional Acculturation Scale, and body-mass-index z-score (BMIz) at two timepoints 18 months apart. Mediation and moderated mediation analyses were conducted using regression models while controlling co-variates. Bias-corrected bootstrap confidence intervals estimated indirect effects.

Outcomes: FSST1 did not indirectly influence child BMIzT2 through HEIT1 (ab= -0.00, bootstrap CI [-0.15, 0.03]). FSST1 also did not indirectly influence HEIT2 through PFDT1 (ab= -0.01, bootstrap CI [-0.15, 0.03]) or PFRT1 (ab= 0.01, bootstrap CI [-0.04, 0.14]). However, as FSST1 worsened, HEI-2015T2 improved (c= 1.06, 95% CI [0.43, 1.69]). As a co-variate, higher baseline English acculturationT1 predicted lower HEI-2015T2 (β= -3.44, 95% CI [-5.62, -1.26]) and higher BMIzT2 (β= 0.13, 95% CI [0.05, 0.21]); however, it did not have significant conditional effects in moderated mediation models. Gender (pFSSxGender= .04) moderated the direct effect of FSST1 on BMIzT2; however, effect size ((β= 0.05, 95% CI [0.002, 0.09]) was too small to be clinically relevant.

Conclusion: FI did not affect body-mass through dietary quality, nor did it affect dietary quality through PFD or PFR. However, an unexpected positive direct relationship between FI and subsequent dietary quality warrants further exploration.


Food security, Child obesity, Parenting, Diet

Included in

Nursing Commons