Page 430 - Contributed Paper Session (CPS) - Volume 2
P. 430
CPS1916 Erica P. et al.
Bayesian measurement error models in air
pollution and health studies
1
2
2
Erica Ponzi , Marta Blangiardo , Paolo Vineis
1 University of Zurich, Hirschengraben 84, CH-8001 Zurich, Switzerland
2 Imperial College London, London, United Kingdom
Abstract
Measurement error can bias relevant parameters in studies on the effects of
air pollution and more generally environmental exposures on health and can
lead to inaccurate conclusions when evaluating associations among
pollutants, disease risk and biomarkers. Although the presence of
measurement error in such studies has been recognized as a potential
problem, it is rarely considered in applications and practical solutions are still
lacking. In particular, taking into account measurement error in environmental
exposures can become quite challenging when considering their association
with high-throughput molecular data, which are nowadays becoming more
and more popular in biomedical research. In this work, we formulate Bayesian
measurement error models and apply them to study the link between air
pollution and chronic cardiovascular diseases, focusing on how this link is
mediated by omics measurements. The data stem from the “Oxford Street II
Study”, a randomized crossover trial in which 60 volunteers walked for two
hours in a traffic-free area (Hyde Park) and in a busy shopping street (Oxford
Street) of London. Omics measurements are taken on each individual as well
as air pollution measurements, in order to investigate the association between
short-term exposure to traffic related air pollution and perturbation of
metabolic pathways.
Keywords
Bayesian hierarchical models; bias correction; environmental exposure; omic
signals
1. Introduction
The reliable estimation of associations between environmental exposures
and health conditions requires the collection of considerable amounts of
exposure data, which is often subject to several sources of error or imprecision.
This can lead not only to bias in the estimation of parameters relevant to the
study but also to inaccurate conclusions when evaluating associations among
pollutants, disease risk and biomarkers. Although the presence of
measurement error in such studies has been recently discussed (Rhomberg et
al. 2011; Edwards and Keil 2017; Shaw et al. 2018), it is often not accounted for
419 | I S I W S C 2 0 1 9