Using Agent-Based Models to Develop Public Policy about Food Behaviours: Future Directions and Recommendations

Area of Science:

• Public Health
• Computational Modeling
• Behavioral Science

Background:

• Overweight and obesity are prevalent in Western countries, necessitating population-level interventions.
• Interventions often target environmental factors and behaviors, including food choices.
• Agent-based models (ABMs) offer a systems perspective for evaluating these interventions.

Purpose of the Study:

• To address limitations in current ABM procedures for food behavior interventions.
• To identify readily available datasets and techniques to improve ABMs.
• To propose an integrated process for developing more robust food behavior ABMs.

Main Methods:

• A scoping review was conducted to identify datasets and techniques.
• A three-step process was proposed to integrate data and methods.
• The proposed process aims to overcome limitations in current ABM development.

Main Results:

• Highlighted limitations in existing ABM approaches for food behaviors.
• Identified independent datasets and techniques to address these limitations.
• Proposed a novel, integrated three-step process for future ABM development.

Conclusions:

• Current ABMs for food behaviors have limitations regarding data utilization and environmental scope.
• The proposed integrated process represents a significant advancement for developing more effective ABMs.
• This approach can generate robust findings to inform the design of impactful food behavior interventions for public health.