Skip to main content
. 2024 Dec 6;132(12):127002. doi: 10.1289/EHP14585

Figure 2.

Figure 2 is conceptual directed acyclic graph that was used for pre-baseline survival bias simulations. The graph shows that the effect of traffic-related air pollution on the probability of being alive at the age of study recruitment was categorized into three levels and modeled with two parameters corresponding to medium exposure [lowercase gamma begin subscript 1 end subscript] and high exposure [lowercase gamma begin subscript 2 end subscript] in reference to low exposure. Further, the graph shows that an unmeasured confounder (U) is hypothesized to affect the probability of being alive at the age of study recruitment through the model parameter lowercase gamma begin subscript 3 end subscript. Finally, the graph shows that for some simulations, traffic-related air pollution and U were hypothesized to jointly affect the probability of being alive at the age of study recruitment, with the relevant interaction terms being lowercase gamma begin subscript 4 end subscript and lowercase gamma begin subscript 5 end subscript.

Conceptual directed acyclic graph underlying a simulation-based bias analysis for TRAP exposure on cognitive decline. Parameters next to arrows correspond to parameters in the prebaseline selection bias simulation. Note: TRAP, traffic-related air pollution.