Abstract: Environmental epidemiologic studies often seek to estimate the joint health effects of environmental mixtures, where multiple correlated exposures may act together through complex relationships. However, policymakers and researchers frequently need to apply these findings to target populations that differ from the original study population. We address this problem by developing statistical methods for transporting the effect of an environmental mixture across populations. Motivated by the Strong Heart Study (SHS), a prospective cardiovascular disease cohort among three American Indian communities with measurements of multiple urinary metal exposures, we extend causal inference-based transportability methods to settings involving multiple continuous and correlated exposures. Specifically, we estimate the joint effect of six metals on coronary artery calcification (CAC) by transporting associations observed in the Multi-Ethnic Study of Atherosclerosis (MESA) to the SHS population, where CAC was not directly measured. This framework provides a principled approach for extending environmental mixture effect estimates to external populations and may improve the relevance of environmental epidemiologic findings for public health decision-making.
About the speaker: Melanie Mayer, Ph.D. is a postdoctoral fellow in Biostatistics at the Perelman School of Medicine. Her research develops causal inference and statistical learning methods for complex observational health data, with applications in environmental epidemiology, oncology, electronic health records, and real-world evidence generation.
.png){kind=logo}
