A Bayesian localized conditional autoregressive model for estimating the health effects of air pollution

Duncan Lee, Alastair Rushworth, Sujit K. Sahu

Research output: Contribution to journalArticle

24 Citations (Scopus)

Abstract

Estimation of the long-term health effects of air pollution is a challenging task, especially when modeling spatial small-area disease incidence data in an ecological study design. The challenge comes from the unobserved underlying spatial autocorrelation structure in these data, which is accounted for using random effects modeled by a globally smooth conditional autoregressive model. These smooth random effects confound the effects of air pollution, which are also globally smooth. To avoid this collinearity a Bayesian localized conditional autoregressive model is developed for the random effects. This localized model is flexible spatially, in the sense that it is not only able to model areas of spatial smoothness, but also it is able to capture step changes in the random effects surface. This methodological development allows us to improve the estimation performance of the covariate effects, compared to using traditional conditional auto-regressive models. These results are established using a simulation study, and are then illustrated with our motivating study on air pollution and respiratory ill health in Greater Glasgow, Scotland in 2011. The model shows substantial health effects of particulate matter air pollution and nitrogen dioxide, whose effects have been consistently attenuated by the currently available globally smooth models.

LanguageEnglish
Pages419-429
Number of pages11
JournalBiometrics
Volume70
Issue number2
Early online date24 Feb 2014
DOIs
Publication statusPublished - Jun 2014

Fingerprint

Conditional Model
Air Pollution
air pollution
Autoregressive Model
Air pollution
Health
Random Effects
Nitrogen Dioxide
Spatial Analysis
Particulate Matter
Scotland
Spatial Autocorrelation
Spatial Modeling
Collinearity
Model
Nitrogen
Covariates
Smoothness
Incidence
Simulation Study

Keywords

  • air pollution and health
  • conditional autoregressive models
  • spatial autocorrelation

Cite this

Lee, Duncan ; Rushworth, Alastair ; Sahu, Sujit K. / A Bayesian localized conditional autoregressive model for estimating the health effects of air pollution. In: Biometrics. 2014 ; Vol. 70, No. 2. pp. 419-429.
@article{60ecdca63008412088b10506bc213a52,
title = "A Bayesian localized conditional autoregressive model for estimating the health effects of air pollution",
abstract = "Estimation of the long-term health effects of air pollution is a challenging task, especially when modeling spatial small-area disease incidence data in an ecological study design. The challenge comes from the unobserved underlying spatial autocorrelation structure in these data, which is accounted for using random effects modeled by a globally smooth conditional autoregressive model. These smooth random effects confound the effects of air pollution, which are also globally smooth. To avoid this collinearity a Bayesian localized conditional autoregressive model is developed for the random effects. This localized model is flexible spatially, in the sense that it is not only able to model areas of spatial smoothness, but also it is able to capture step changes in the random effects surface. This methodological development allows us to improve the estimation performance of the covariate effects, compared to using traditional conditional auto-regressive models. These results are established using a simulation study, and are then illustrated with our motivating study on air pollution and respiratory ill health in Greater Glasgow, Scotland in 2011. The model shows substantial health effects of particulate matter air pollution and nitrogen dioxide, whose effects have been consistently attenuated by the currently available globally smooth models.",
keywords = "air pollution and health, conditional autoregressive models, spatial autocorrelation",
author = "Duncan Lee and Alastair Rushworth and Sahu, {Sujit K.}",
year = "2014",
month = "6",
doi = "10.1111/biom.12156",
language = "English",
volume = "70",
pages = "419--429",
journal = "Biometrics",
issn = "0006-341X",
number = "2",

}

A Bayesian localized conditional autoregressive model for estimating the health effects of air pollution. / Lee, Duncan; Rushworth, Alastair; Sahu, Sujit K.

In: Biometrics, Vol. 70, No. 2, 06.2014, p. 419-429.

Research output: Contribution to journalArticle

TY - JOUR

T1 - A Bayesian localized conditional autoregressive model for estimating the health effects of air pollution

AU - Lee, Duncan

AU - Rushworth, Alastair

AU - Sahu, Sujit K.

PY - 2014/6

Y1 - 2014/6

N2 - Estimation of the long-term health effects of air pollution is a challenging task, especially when modeling spatial small-area disease incidence data in an ecological study design. The challenge comes from the unobserved underlying spatial autocorrelation structure in these data, which is accounted for using random effects modeled by a globally smooth conditional autoregressive model. These smooth random effects confound the effects of air pollution, which are also globally smooth. To avoid this collinearity a Bayesian localized conditional autoregressive model is developed for the random effects. This localized model is flexible spatially, in the sense that it is not only able to model areas of spatial smoothness, but also it is able to capture step changes in the random effects surface. This methodological development allows us to improve the estimation performance of the covariate effects, compared to using traditional conditional auto-regressive models. These results are established using a simulation study, and are then illustrated with our motivating study on air pollution and respiratory ill health in Greater Glasgow, Scotland in 2011. The model shows substantial health effects of particulate matter air pollution and nitrogen dioxide, whose effects have been consistently attenuated by the currently available globally smooth models.

AB - Estimation of the long-term health effects of air pollution is a challenging task, especially when modeling spatial small-area disease incidence data in an ecological study design. The challenge comes from the unobserved underlying spatial autocorrelation structure in these data, which is accounted for using random effects modeled by a globally smooth conditional autoregressive model. These smooth random effects confound the effects of air pollution, which are also globally smooth. To avoid this collinearity a Bayesian localized conditional autoregressive model is developed for the random effects. This localized model is flexible spatially, in the sense that it is not only able to model areas of spatial smoothness, but also it is able to capture step changes in the random effects surface. This methodological development allows us to improve the estimation performance of the covariate effects, compared to using traditional conditional auto-regressive models. These results are established using a simulation study, and are then illustrated with our motivating study on air pollution and respiratory ill health in Greater Glasgow, Scotland in 2011. The model shows substantial health effects of particulate matter air pollution and nitrogen dioxide, whose effects have been consistently attenuated by the currently available globally smooth models.

KW - air pollution and health

KW - conditional autoregressive models

KW - spatial autocorrelation

UR - http://www.scopus.com/inward/record.url?scp=84902305284&partnerID=8YFLogxK

UR - http://onlinelibrary.wiley.com/journal/10.1111/(ISSN)1541-0420

U2 - 10.1111/biom.12156

DO - 10.1111/biom.12156

M3 - Article

VL - 70

SP - 419

EP - 429

JO - Biometrics

T2 - Biometrics

JF - Biometrics

SN - 0006-341X

IS - 2

ER -