The role of whole network analysis in understanding health behaviours: a systematic review of smoking, alcohol use, physical activity, and diet.

BACKGROUND

Health behaviours such as smoking, alcohol use, physical activity, and dietary habits are not only individual choices but are influenced by social interactions and relational contexts. Recently, Whole Network Analysis (WNA) has become a powerful methodological tool to explore how these behaviours spread and become established within social systems. Unlike egocentric approaches, which focus on individuals' personal networks, WNA captures the entire structure of social relations within a defined population.

METHODS

We conducted a systematic review of empirical studies applying WNA to four key health behaviours: smoking, alcohol consumption, physical activity, and diet. A thorough search was performed across six major academic databases for peer-reviewed articles published in English. We adhered to PRISMA 2020 guidelines when selecting and analysing the studies. Inclusion criteria included sociocentric data collection, a focus on behavioural processes, and empirical design.

RESULTS

A total of 27 studies met the inclusion criteria. Findings highlighted consistent structural mechanisms across behaviours, including peer influence, homophily, network cohesion, and centrality. WNA proved particularly effective in identifying behaviour clusters, co-evolutionary processes between networks and behaviours, and structural barriers or facilitators of behavioural diffusion. Longitudinal studies employing stochastic actor-oriented models provided especially valuable insights into disentangling selection and influence.

CONCLUSION

WNA provides a distinctive perspective for understanding how health behaviours emerge and spread within social systems. This review underscores the importance of integrating WNA more systematically into public health research and intervention planning. By mapping entire social systems, WNA facilitates the identification of key actors, influential subgroups, and potential entry points for effective behavioural interventions. Future research should expand WNA applications to more diverse populations and settings and critically examine how social inequalities, such as socioeconomic status, gender, and contextual factors, interact with network dynamics to influence health behaviour trajectories.