Food Insecurity & BMI outcomes among pre-school and school aged children in an inner city setting

Abstract

Introduction: While the paradoxical association between overweight and household food insecurity (HFI) is well established amongst low-income women, findings remain inconclusive amongst children. The purpose of this study was to determine the relationship between household food insecurity (HFI) and child overweight outcomes in an inner-city, pre-school and school aged population. Methods: This study used a cross-sectional study design augmented by validated Early Pregnancy Study (EPS) data collected during a pregnancy 6 years prior. A random subset of mothers of child-bearing age (23-44 years) and the child resulting from that pregnancy (between 4-7 years of age at time of re-enrollment) were tracked and re-enrolled in this follow-up study. The primary exposure, HFI, was captured using an adaptation of the standardized US Food Security Scale -- embedded within a larger self-report questionnaire addressing selected maternal and child influences (such as health, exercise, and dietary risk factors) on child weight patterns. Maternal and the child BMI outcome was determined through in-home clinical assessment of height and weight and using standard CDC based cutoffs to classify weight status. Self-report and biologically confirmed prenatal EPS data were also obtained to establish useful baseline data and to enhance study results. Exploratory hypotheses examined: (1) Interrelatedness between select maternal and child health, social-environmental, and socio-demographic exposures and (2) Household food insecurity (HFI), prenatal and current maternal and child exposures were explored for their relationship with child BMI. Results: Thirty-six mother-child dyads participated in this follow-up study. Twenty-eight percent of mothers reported HFI; 83% of mothers were classified as overweight or obese and 58% of children were categorized as at-risk for overweight or overweight. Most bi-variate analyses yielded non-significant results, though in-utero drug use was significantly correlated, and mothers who tested positive for tobacco use during pregnancy were also likely to test positive for marijuana use, as determined through biological confirmation (p<0.05). Additionally, current maternal BMI as well as current maternal hip-to-waist ratio were each significantly associated with current child BMI (p<0.05). One unexpected finding included a significant association between current maternal BMI and presence of a co-morbidity in the child (p<0.05). No significant relationships were found between HFI or additional exploratory multivariate models looking at independent effects of prenatal and current maternal-child exposures in predicting child overweight, though current maternal BMI was predictive of child overweight in most analyses. Access to electronic medical record information was highly predictive of participation in this follow-up study (p<0.001). Discussion: This study provided an opportunity to better characterize an attrition-prone population, and, while it is possible to track and identify these women through electronic medical record databases, recruitment efforts were labor-intensive and attrition rates particularly high (~22% of mothers re-enrolled). Overall, the influence of household food insecurity on child overweight remains unclear. However, study findings also demonstrated that maternal weight alone was highly predictive of child overweight as early as pre-school, mirroring empirical evidence in this area. Ultimately, childhood obesity is more effectively prevented when treated as a familial concern. Future studies should continue to explore cumulative influences catalyzing obesity in young children in order to better inform understanding of early obesity onset.