Virtual Population Obesity Prevention (VPOP) Labs
Project Lead: Bruce Y. Lee, MD, MBA
In the Global Obesity Prevention Center (GOPC), scientists use computational simulation models to understand complex systems by evaluating the multiple factors at play. Using information from different sources and disciplines—such as statistics, epidemiology, nutrition, anthropology, sociology, health communications and economics—simulation models examine multifaceted problems like obesity by considering the interactions and effects of multiple factors, processes and influences. Agent-Based Models (ABMs), which treat each person as an individually simulated “agent,” appear particularly promising when it comes to developing systems-focused solutions to the obesity epidemic.
First of its Kind: The Virtual Population Obesity Prevention (VPOP) model is the first obesity-related agent-based model to include detailed representations of multiple levels, including individual behavior (e.g., energy intake and expenditure), social context, physical environment (including food and physical activity sources), and policy. Grounded on collected data and the obesity literature, the VPOP model simulates the food and physical activity behaviors of child and adult agents, allowing users and decision makers to design, test, and evaluate potential or existing obesity programs and policies, within a virtual environment.
Breaking New Grounds: “The challenge in building this kind of model is not only being sure to represent reality as much as possible, but to capture the most important pieces that are relevant to your question of interest” says Yeeli Mui, PhD student and systems science trainee. “A great advantage, though, is the potential time and resource savings. To actually implement a particular intervention or food-related policy takes a huge amount of time and effort. Agent-based models like VPOP allow us to test things out in a simulated environment and make more informed decisions related to specific interventions and policies.”
The VPOP team collaborates with experts from a range of disciplines and sectors:
- Behavioral sciences
- Medical anthropology
- Environmental health
- Weight management
- Health policy
“While historically this type of modeling has been used in the engineering and business fields, we’re excited to be bringing agent-based modeling into the public health arena,” Mui says. “We’ve introduced our model to various stakeholders and have received a lot of positive reactions, including from policymakers here in Baltimore City. Ultimately, this is something we hope to make transferable to other settings facing similar complex challenges.”
Example of a Current VPOP
Baltimore Low Income Food Environment (BLIFE) involves the use of mixed methods formative work to enhance the ABM of child agents navigating through a virtual representation of the low income Baltimore food environment.
Click here to find out further information about BLIFE.