INVESTIGATING INTERVENTION PATHWAYS BETWEEN 24-HOUR MOVEMENT BEHAVIOR AND SOCIODEMOGRAPHIC FACTORS IN CHILDREN: APPLICATION OF THE DIJKSTRA COMPLEX NETWORK ALGORITHM

Abstract

The aim of the study was to analyze the network structure and investigate intervention pathways between 24-hour movement behaviors and sociodemographic factors by applying the Dijkstra algorithm. Sixty-one parents and/or guardians of preschool children aged 3-6 years enrolled in public schools participated. The 24-hour movement behaviors were analyzed using a questionnaire adapted by Google Forms. Descriptive statistics were applied for the categorical variables, then an associative network analysis was conducted in Rstudio, as well as the expected influence centrality metric to verify the most influential nodes in the network. Subsequently, the Dijkstra algorithm was applied to identify the shortest intervention path between the study variables, starting from the variables that had the highest expected influence value. The results indicated that children from families with higher income and higher education tended not to adhere to physical activity, screen time on weekdays, on weekends and sleep on weekdays, with relationships varying between (-0.009 and -0.19). Similarly, younger children, in addition to the factors mentioned above, also do not adhere to sleep at the weekend (0.31). The djisktra algorithm showed that income would be the mediating variable for possible interventions, along with sleep at the week. Therefore, the study points to new possibilities for planning interventions in physical activity and related through the dynamics of complex systems.