Trajectories leading to childhood obesity result from multifactorial Gene x Environment x Behavior interactions during a period of high developmental plasticity. Defining the causal mechanisms that mediate risk for obesity development in early life and how risk can be modified by psychosocial, contextual, and/or environmental factors, requires systems biology integration, which is currently lacking. The objective of this application is to develop predictive algorithms for childhood obesity risk by integrating data on growth trajectories, body composition, and dietary intake with multi omic (genetic, epigenetic and microbiome) data sets and assessments of temperament-self-regulation and child eating behavior across early childhood. We will leverage a rich biobank of longitudinal biological samples, questionnaire, and observational data from the ongoing Synergistic Theory and Research on Nutrition and Growth Kids 2 (SK2) cohort (n=468)11. Our central hypothesis is that computational and systems biology approaches will illuminate interactions between modifiable and non-modifiable obesity-related risk factors to provide predictive indices that identify at-risk individuals in early life. The three specific aims are: Aim 1. Determine how microbial profiles implicated in rapid growth and childhood obesity differentially influence children’s weight gain and body composition and determine whether there are differential associations based on genetic and or epigenetic risk factors; Aim 2. Determine how microbial profiles implicated in executive and emotion processes map onto profiles of child temperament and indicators of self-regulation. Establish the impact of these interacting systems on assessments of children’s eating-related behavior and weight trajectories, and whether there are differential associations based on genetic/epigenetic risk; and Aim 3. Uncover and characterize interactions between dietary components, genetics, fecal microbiota, temperament, and cognitive control on obesity risk using Bayesian and neural networks. These outcomes will enable development of algorithms to predict responses to diet within the context of genetic variation, microbiome composition, temperament and other environmental factors.
