Answers are grounded in this paper and cite their sources.
Longitudinal observational study uncovers direct and upstream socioeconomic and environmental determinants of physical inactivity across U.S. counties, highlighting complex multi-year pathways.
This study introduces a time-constrained multiple mediation framework for modeling longitudinal pathways in observational data, addressing prior research that often overlooks temporal ordering and interdependence among mediators. Using county-level data from 3,109 continental U.S. counties, we analyzed 31 social, economic, and environmental variables measured in 2015 and 2016, with leisure-time physical inactivity (LTPI) measured in 2017. The framework combines chained LASSO regression with explicit temporal constraints to distinguish direct predictors from upstream, time-ordered pathways and reduce spurious associations due to confounding and mediator interdependence. In the LTPI application, the model identified five direct predictors in 2016: prior-year LTPI, child poverty, food insecurity, median household income, and air pollution, explaining 76% of LTPI variation. Walkable park access was not retained as a direct predictor but appeared as an upstream condition linked to LTPI through child poverty, illustrating how the framework distinguishes direct LTPI predictors from earlier conditions operating through time-ordered pathways.
No takes yet. Share an insight, caveat, or question.
Lee et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: