While the prevalence of childhood obesity seems to have reached a plateau recently, substantial disparities in
child obesity by race/ethnicity and family socio-economic status (hereafter, social disparity in child obesity)
continue to widen; the rates decreased in non-Hispanic white or high-socioeconomic status (SES) children but
increased among minority or low-SES children. Given the adverse health outcomes that result from obesity, it
is critical to examine which factors affect childhood obesity and its social disparity. Changes in neighborhood
SES over the last 2 decades correspond to the recent increases in social disparity in child obesity, thus
neighborhood SES is potentially one important contributor to the greater social disparity in child obesity. Past
studies have established a scientific premise for research in this area; however, the majority of evidence has
(1) been based on cross-sectional studies, (2) not accounted for residential mobility of individuals, (3) utilized
less valid self-reported heights and weights, and/or (4) not examined mediating mechanisms by which
neighborhood SES affects child obesity. Further, little research has examined whether neighborhood SES has
differential effects on child obesity by important subgroups (e.g., based on race/ethnicity and family SES). This
proposed study is a secondary data analysis of a large, longitudinal study, namely the Fragile Families & Child
Wellbeing Study (FFCWS), to examine the longitudinal association between neighborhood SES changes and
childhood obesity and its social disparity. The FFCWS is a $40 million study of 4,898 children born in 1998-
2000 and their families. This proposed study will address the gaps described above by using data from
FFCWS. The specific aims are to examine from young childhood to adolescence: 1) the contribution of
neighborhood SES changes on child obesity; 2) the extent by which neighborhood SES changes explain social
disparity in obesity; 3) the mediating effects of individual obesity-related behaviors and neighborhood social
environments on the association between neighborhood SES changes and child obesity; and 4) the interacting
effects of individual sociodemographic factors on the association between neighborhood SES changes and
child obesity. Using the FFCWS is advantageous because data from the FFCWS (1) include objectively-
measured body mass index of children when they were three, five, nine, and fifteen; (2) have information about
multiple levels of exposures ranging from the individual and family levels to the neighborhood and social levels,
which can be utilized as potential moderators, mediators, or covariates; and (3) were collected in 2000s during
the drastically changing time period in neighborhood SES, which would allow for better causal inference. This
study builds on important preliminary studies and the extensive relevant experience of two principal
investigators. This study is very cost-effective and will elucidate the role of neighborhood socioeconomic status
on child obesity including mediating and moderating relationships, allowing for recommendations that will
inform community-level interventions to reduce child obesity.
