Format

Send to

Choose Destination
J Expo Sci Environ Epidemiol. 2015 Mar-Apr;25(2):138-44. doi: 10.1038/jes.2014.40. Epub 2014 Jun 4.

Consequences of kriging and land use regression for PM2.5 predictions in epidemiologic analyses: insights into spatial variability using high-resolution satellite data.

Author information

1
1] Department of Biostatistics, Harvard School of Public Health, Boston, Massachusetts, USA [2] National Center for Atmospheric Research, Institute for Mathematics Applied to Geosciences, Boulder, Colorado, USA.
2
Department of Environmental Health, Harvard School of Public Health, Boston, Massachusetts, USA.
3
1] Department of Environmental Health, Harvard School of Public Health, Boston, Massachusetts, USA [2] Department of Geography and Environmental Development, Ben-Gurion University of the Negev, Be'er Sheva, Israel.
4
Department of Biostatistics, Harvard School of Public Health, Boston, Massachusetts, USA.

Abstract

Many epidemiological studies use predicted air pollution exposures as surrogates for true air pollution levels. These predicted exposures contain exposure measurement error, yet simulation studies have typically found negligible bias in resulting health effect estimates. However, previous studies typically assumed a statistical spatial model for air pollution exposure, which may be oversimplified. We address this shortcoming by assuming a realistic, complex exposure surface derived from fine-scale (1 km × 1 km) remote-sensing satellite data. Using simulation, we evaluate the accuracy of epidemiological health effect estimates in linear and logistic regression when using spatial air pollution predictions from kriging and land use regression models. We examined chronic (long-term) and acute (short-term) exposure to air pollution. Results varied substantially across different scenarios. Exposure models with low out-of-sample R(2) yielded severe biases in the health effect estimates of some models, ranging from 60% upward bias to 70% downward bias. One land use regression exposure model with >0.9 out-of-sample R(2) yielded upward biases up to 13% for acute health effect estimates. Almost all models drastically underestimated the SEs. Land use regression models performed better in chronic effect simulations. These results can help researchers when interpreting health effect estimates in these types of studies.

PMID:
24896768
PMCID:
PMC4758216
DOI:
10.1038/jes.2014.40
[Indexed for MEDLINE]
Free PMC Article

Supplemental Content

Full text links

Icon for Nature Publishing Group Icon for PubMed Central
Loading ...
Support Center