Intergenerational & Environmental Drivers of Obesity: an Agent Based Model Approach
Obesity prevalence remains high and exhibits pronounced social patterning in the U.S., despite widespread intervention efforts. Complex individual, environmental, and intergenerational influences on obesity are difficult to study using traditional methods. Agent Based Models (ABM) are increasingly utilized to study complex determinants of population health by simulating experimental conditions that would be infeasible to conduct on human populations.
We present the motivating complexity and design of our recently developed multigenerational ABM, which simulates development of obesity over the life course and across generations within multiple sub-populations. The ABM incorporates interdependence of multiple obesity drivers that occur throughout the life course. It simulates autonomous, interacting individuals (agents) representative of the US population in age, gender, Body Mass Index (BMI), reproduction rate, and mortality rate. Agents are experimentally assigned to live in a high or low socioeconomic status (SES) area. Agents grow as a function of age, gender, local environment, and maternal BMI, in a dynamic simulation over time. Differentials in offspring BMI trajectory as a function of maternal BMI, assignment to residential areas as a function of individual SES, obesogenicity of local areas as a function of neighborhood, and other key factors are experimentally modified. Obesity prevalence and within SES subgroups, over multiple generations, are calculated under pre-specified simulated experimental conditions to study the relative and combined influences of possible prevention strategies.
The goals of this modeling effort are first, to quantify the theoretical influence of maternal obesity-induced intrauterine effects on (a) obesity prevalence and (b) SES disparities in obesity prevalence over multiple generations; and second, to quantify and compare the theoretical contributions of intergenerational transmission of (a) maternal obesity-induced biological susceptibility, (b) environmental factors, and (c) socioeconomic status on obesity prevalence and disparities over multiple generations.
Janne Boone-Heinonen, PhD, MPH
Dr. Boone-Heinonen is an Associate Professor in the OHSU-PSU School of Public Health. As an obesity epidemiologist, her research interests focus on how social, environmental, and behavioral factors in early life influence development of obesity and related health conditions throughout the life cycle. She uses interdisciplinary quantitative methods with large population-based cohorts to address methodological challenges in obesity and life course epidemiology. This seminar presents an application of systems science methods to investigate the influence of prenatal conditions on obesity over multiple generations.
Dale Frakes is a Ph.D. candidate in the Systems Science Program at Portland State University studying under Wayne Wakeland. He has completed a graduate certificate in Computer Modeling & Simulation and has since worked as a graduate teaching assistant and adjunct instructor for Modeling & Simulation courses at PSU. He received an MBA with a concentration in Global Business from the University of Portland and worked for more than 10 years in Analytics and Supply Chain Performance Management at Nike, Inc. developing several tools to enhance reporting and analytics. His current research is in applying agent-based modeling to understanding and mitigating the challenge of the spread of false news stories and intergenerational obesity.
Computer Sciences | Medicine and Health Sciences
Frakes, Dale, "Intergenerational & Environmental Drivers of Obesity: an Agent Based Model Approach" (2019). Systems Science Friday Noon Seminar Series. 2.