Robust Statistical Approaches to Understanding the Causal Effect of Air Pollution Mixtures

This report, available for downloading below, presents a study led by Joseph Antonelli at the University of Florida. Antonelli and colleagues describe several new approaches to support causal inference, focusing on exposure to multiple pollutants, addressing bias resulting from confounding, and enhancing exposure assessment by incorporating people’s mobility patterns.

Key takeaways:

• This study is among the first to address a series of common challenges faced by researchers assessing the health effects of exposures to air pollution mixtures, from exposure assessment to causal inference.
• The approaches are applied to hypothetical scenarios and to data from a US Medicare cohort to demonstrate the effectiveness of the practice methods.
• Overall, the approaches presented here provide new solutions for overcoming challenges to assessing causal links between air pollution and health.