Agent-based modelling in active commuting research: a scoping review

Agent-based modelling (ABM) is a promising research tool for studying complex healthy behaviours such as active commuting. This scoping review systematically examined the current literature on the use of ABM in active commuting intervention research, identifying key model components, methodological strengths and limitations. A comprehensive search identified 43 studies for analysis. Findings revealed a growing utilisation of ABM in recent years with considerable variations in model structures, agent attributes, behaviours, environmental representations, and types of interventions evaluated. Overall, ABM has been utilised to study active commuting behaviours across diverse populations and geographical contexts by incorporating individual-level heterogeneity and simulating interactions between agents and their environments. It also facilitated the in-silico evaluation of interventions such as infrastructure modifications, policy changes, financial incentives, and health promotion initiatives on their potential impact on active commuting behaviours. Nevertheless, challenges remain such as the limited theoretical grounding for agent behaviours, translation of findings to real-world settings and lack of standardised model reporting and validation methods. Addressing these gaps would enhance ABM's credibility and applicability as a decision-support tool for public health practitioners and policymakers to promote active commuting to combat physical inactivity and improve population health.