PI: Kathleen Keller
NIH R01 DK144201
Administered in: College of Health and Human Development
Abstract:
Excess food consumption is the primary cause of pediatric obesity, but not all children are equally susceptible to overeating in the current environment. This suggests there are child-level (i.e., neural, cognitive, and behavioral) and family- and societal-level (i.e., home food environment, socioeconomic status- SES) characteristics that can either increase or decrease risk for excessive pre-adolescent increases in adiposity. Identifying these characteristics so they can be targeted to develop precision nutrition interventions is critical for prevention of the disease during a critical developmental window. In this project, we are proposing to test a novel eating behavior phenotype called PACE. This phenotype combines children’s Portion size susceptibility, Appetitive traits, loss of Control eating, and Eating speed. We hypothesize that the PACE phenotype has a consistent trajectory that predicts development of obesity in children, with underlying mechanisms rooted in interoception and executive (i.e., self-regulation) abilities. In our preliminary studies, PACE predicted baseline and prospective adiposity gains in a cohort of 7–8 year-old children with primarily healthy weight; however, the relationship between PACE and adiposity was stronger in lower relative to higher SES families, suggesting there may be conditions under which the effects of high PACE could be ameliorated. To confirm and extend our prior findings in a sample with greater SES diversity, we plan to enroll 210, 7–9 year-old children with healthy weight status separated into three groups by familial risk status: 70 high-risk (child & both parents with obesity), 70 medium-risk (child with healthy weight, parents with obesity), and 70 low-risk (child & parents both with healthy weight) for a 12 month prospective study to characterize the relationship between PACE and developmental changes in adiposity in rural Appalachian children. Aim 1 will test whether PACE is predictive of baseline adiposity in a new cohort, including children who have already developed obesity, and test the extent to which family SES moderates this relationship. Aim 2 will identify the neural underpinnings of this phenotype by testing the hypothesis that PACE will be associated with hypoactivation in regions implicated in interoception and executive function and reduced functional connectivity between these regions. Aim 3 will follow children for 12 months to test the hypothesis that baseline PACE will predict amount of adiposity gain over this timeframe. In the final aim, we will use machine learning to identify positive outliers—children with high PACE scores (and other risk factors) who exhibit less adiposity gain than predicted–to characterize the individual- and family-level traits and behaviors that protect these children from pre-adolescent weight gain. We expect the proposed studies to inform the science of obesity prevention by identifying specific modifiable eating phenotypes that can be targeted by interventions to prevent and treat this disease. In addition, data- driven approaches to characterize “positive outliers” will contribute knowledge of resilience phenotypes to the literature, thus advancing understanding of primary prevention of obesity.