| Purpose. Children who are overweight or obese experience many negative health and social consequences, and weight problems disproportionately affect socioeconomically disadvantaged populations. With data from the Fragile Families and Child Wellbeing Study (FFCWS), this study used life course theory and cumulative risk approaches to examine the relationship between family risk factors, food insecurity, receipt of food assistance benefits, and child overweight or obesity.;Informed by prior research, hypotheses were tested to examine the mechanisms that explain the connection between family risks and child overweight or obesity. Specific research questions were: To what extent are family risk factors, food insecurity, and food benefits associated with child overweight or obesity?; To what extent are aggregated risk factors (rather than specific risk factors), food insecurity, and food benefits associated with child overweight or obesity?; Does exposure at earlier or later ages have more impact on child overweight or obesity?; Do federal food policy benefits moderate the relationship between food insecurity and child overweight or obesity?;Methods. Four waves of publicly available data from the FFCWS were used, when children were 1, 3, 5 and 9 years of age (n= 1,092). BMI percentile was used to create a dichotomous measure, with cut-off at 85% of BMI for overweight or obesity. Family risks were operationalized and analyzed in 3 ways: (i) separate dichotomous measures of financial risk, maternal depression, and parenting stress at each wave, (ii) an aggregated number of these specific risks from infancy through age five, and (iii) a 6-level family risk categorical variable (no risks, low risks during years 1 to 5, high risks in year 1 only, high risks in year 3 only, high risks in year 5 only, and always high risks). These three approaches complement each other. Separate family risk factors provide specific information for potential interventions, whereas a single family risk score is statistically parsimonious. Also, the categorized family risk variable allows for the examination of the role of timing of risk exposure. Food program participation was dichotomized with 1 indicating any of program participation and 0 indication non-participation. Food insecurity was based on an 18-item scale from U.S Department of Agriculture (USDA) and measured as an ordinal variable with four levels (food security, marginally food security, food insecurity, and very high food insecurity). Univariate analysis displayed sample characteristics, and the Pearson correlation and polychoric correlation applied to examine binary relationships. The current study included dichotomous and ordinal variables, but Pearson correlation assumes the variables are continuous and follow a multivariate normal distribution. Thus, both correlation analyses were performed. Hypotheses were tested using generalized structural equation modeling (GSEM) techniques. Results were adjusted by including a weight available in the FFCWS data set at wave 9. All GSEM models included controls for low birth weight status, breastfeeding, children overweight or obesity status (BMI ≥85th percentile) at previous ages, mother's BMI, and number adults in households. Data were analyzed with STATA 14 (StataCorp, 2015).;Results. Thirty-six percent of children were either overweight or obese at age 9, and being overweight or obese at age 9 was positively correlated with family risks and food assistance benefits at the bivariate level. Results of GSEM partially supported the research hypotheses. Food benefits were a significant predictor of BMI, but not as hypothesized. According to models that analyzed specific family risks, financial strain had a strong consistent positive impact on BMI through the receipt of food assistance benefits. Also, the models that analyzed risk as a composite measure was positively associated with child BMI at age 9, but indirectly through food assistance benefits. Food insecurity did not have a significant direct or indirect impact on BMI. In addition, the moderating effect of food supports on the relationship between food insecurity and child overweight or obesity was not statistically supported.;Conclusions. This dissertation extends our understanding about the consequences of family risks on child overweight or obesity in several ways. First the pathways through which family factors matter for children were considered, as little is known about the factors that mediate or moderate this association, especially from ages 1 to 9 years. Second, I use a nationally representative sample of non-marital births to parents residing in cities with populations over 200,000 from multiple waves, which allows consideration of the importance of long-term consequences of family risks on child weight. The study findings should be interpreted with caution due to limitations of the design and measures in the available dataset. Nevertheless, the findings set the stage for future research and have implications for social welfare policy. (Abstract shortened by UMI.). |
