GECC - Gateway Exposome Coordinating Center

Methods

Overview

The Methods team focuses on key statistical and methodological issues that pose important challenges to research on Alzheimer’s disease and related dementias (AD/ADRD) and the exposome across the six substantive GECC exposome domains.

Research in this area is cross-cutting, multidisciplinary, and collaborative, focusing on topics impacting multiple domains, including the appropriate measurement of exposome domains, and the development and use of appropriate models to describe the associations between exposures and AD/ADRD outcomes.

The Methods core has three overarching aims:

Improve and refine exposure estimates over time and space

Estimates of various exposures, such as air quality or access to healthcare services, are often based on imperfect data. The methods team will focus on comparing, assessing, and developing a range of approaches to reduce bias and increase precision of estimates in the presence of many of the commonly occurring data challenges across substantive domains.

Model the impacts of multiple exposures simultaneously

Though research often focuses on one exposure at a time, individuals exist in a complicated world and are constantly faced with multiple, potentially interacting, exposures. Understanding and developing the optimal approaches for capturing and modeling these interactions is an important methodological topic.

Minimize bias in analyses estimating the effect of exposome features

There are many possible sources of bias or reasons why a study may get the wrong answer. Issues around who is in the study (selection bias) or the quality of measurement (measurement error) are important topics that the Methods team will explore. Additionally, understanding how to synthesize evidence and triangulate findings across different types of studies can help in reducing risk of bias.

Community Insights

In the fall of 2024, the GECC hosted a series of town hall meetings with hundreds of unique participants. These meetings yielded critical insights for the Methods domain, including highlighting key themes and gaps in research. Even among conversations that focused on specific exposome domains, such as the social environment, questions related to methodological approaches were pervasive throughout discussions.

Key Themes

Measurement of exposures and outcomes
Lifecourse research and cumulative exposures
Interactions between exposures
Mixture methods
Harmonization

Gaps in Research

Comparisons and documentation of methods for assessing lifecourse exposures and interactions or mixtures
Best practices for conducting harmonized research across cohorts or contexts
Consideration of mixtures across and within domains

Steps to Improve Methods in Exposome Research

Create documentation regarding best practices
Discuss comparisons of different existing methodologic approaches
Conduct simulation studies to evaluate the performance of different methods for different questions or types of data
Evaluate the performance of innovative approaches to address the limitations of existing methods

Priorities

Approaches to high-dimensional and multi-modal data: rich data from a variety of sources, such as novel wearable sensors or linked administrative data provide important exposure to improve exposure measurement, but appropriate methods are needed to handle high-dimensional and multi-modal data.
Measurement error: better methods to quantify measurement error, evaluate potential bias due to measurement error, and incorporate measurement error into substantive analyses can improve the quality of inferences.
Harmonization: research combining data sources needs to consider the harmonization of measures, and potentially the harmonization of methods and models as well to ensure comparisons or pooled analyses are valid.
Mixtures: existing approaches to evaluate the effects of chemical mixtures may be generalizable to broader exposome research, but theoretical considerations and extensions of existing methods to accommodate diverse data types are needed.
Effect heterogeneity: the utility of data-driven machine learning approaches to identify heterogeneous effects in exposome research should be explored, along with best practices in pairing more data-driven approaches with subsequent theory-based modeling efforts.
Exposure timing: existing methods developed for environmental exposures (e.g., distributed lag models) may be more broadly useful, though considerations around methods for imputing or modeling life-course exposures, and handling multicollinearity or differential measurement error over time deserve attention.
Cumulative exposures: careful thought is needed in matching approaches for calculating summary measures of cumulative exposures with theoretical reasoning or data-driven evidence on lifecourse exposures.
Reverse causation: though analytic methods largely cannot correct reverse causation, greater awareness of the potential source of bias is needed and study design choices or sensitivity analyses can evaluate or reduce the risk of reverse causation.
Sample selection: the integration of approaches to account for selection bias, or to apply generalizability and transportability tools, with other methods for dementia exposome research should be considered, and further efforts are needed to evaluate the performance of transportability tools under different plausible violations of assumptions.