Advances in Epidemiological Approaches to Assessing Health Effects of Environmental Mixtures

Advances in Epidemiological Approaches to Assessing Health Effects of Environmental Mixtures

Workshop ChairsBonnie Joubert and Caroline Dilworth
Additional Faculty: Jessie Buckley, Matthew Strickland, Aimin Chen, Birgit Claus-Henn

*NOTE: Participants are encouraged to bring a laptop with SAS or R installed.

A well-known issue in environmental epidemiology is that real-world exposures occur in mixtures; commonly used statistical approaches evaluating independent exposure-outcome relationships may not be appropriate. Consequently, researchers are implementing Bayesian methods, smoothing, shrinkage, weighting, classification, variable selection, prediction strategies, and other approaches to determine which exposures within a mixture are most influential as well as how the mixture as a whole influences health outcomes. The NIEHS recently hosted a workshop, “Statistical Approaches for Assessing Health Effects of Environmental Chemical Mixtures in Epidemiology Studies,” to compare statistical approaches to chemical mixtures in epidemiological data. To continue the momentum, this workshop will showcase applications of various statistical approaches in large, real-world epidemiological datasets. The workshop will include presentations from researchers working with mixtures in the real-world context followed by examples of implementing the code on the presented data. The workshop will discuss broad issues in epidemiology and toxicology related to mixtures as well as logistics of how to determine which method to use for your data or hypotheses of interest. The workshop will motivate scientists to think outside the box of their current approaches by giving them a stepping board to work from to try new approaches in existing projects. Time for group discussion as well as individual support from experts during a hands-on code implementation exercise will be provided. Opportunities for networking and transdisciplinary collaborations are also anticipated.

Schedule: 8:30 am to 12:30 pm

8:30 – 8:45 am
Speaker: Bonnie Joubert, PhD, NIEHS
Topic: Introduction of the symposium and speakers (5 min);
Recent progress in mixtures research in epidemiology and goals of this symposium (10 min)

8:45 – 9:15 am
Speaker: Birgit Claus Henn, SD, Boston University
Topic: Associations between exposure to metal mixtures and child neurodevelopment: application of novel statistical approaches (20 min)
Code run through (10 min)

9:15 – 9:45 am
Speaker: Aimin Chen, MD, PhD, University of Cincinnati College of Medicine
Topic: LASSO application in the study of prenatal exposure to chemical mixtures and child reading skills at school age (20 min)
Code run through (10 min)

9:45 – 10:00 am
Break

10:00 – 10:30 am
Speaker: Jessie Buckley, PhD, University of North Carolina at Chapel Hill
Topic: Estimating effects of interventions on exposure mixtures using the parametric g-formula (20 min)
Code run through (10 min)

10:30 – 11:00 am
Speaker: Matthew Strickland, PhD, MPH, University of Nevada, Reno
Topic: A method to partially correct for residual confounding in a time-series study of ambient air pollution mixtures (20 min)
Code run through (10 min)

11:00 – 11:30 am
Discussion: Summary and context for the methods and general discussion with the audience.
Coordinators: Bonnie Joubert and Caroline Dilworth

11:30 – 12:15 am
Hands-on exercise and Q&A.
Time will be available for individual support related to optional code implementation or questions that were not specifically addressed in the presentations and discussion. This time can also be used to run the code on real datasets of interest to the workshop participants. Participants should have some working knowledge of either SAS or R programming languages for this portion of the workshop and have SAS or R installed on their lap tops. Working in partnerships or small groups is encouraged.

12:15 – 12:30 am
Wrap up and future directions
Bonnie Joubert and Caroline Dilworth