Air Pollution Epidemiology: Can Information Be Obtained from the Variations in Significance and Risk as a Function of Days After Exposure (Lag Structure)?
Archived files are provided for reference purposes only. The file was current when produced, but is no longer maintained and may now be outdated. Persons with disabilities having difficulty accessing archived files may contact the NCEA Webmaster for assistance. Please visit http://epa.gov/ncea to access current information.
Determine if analysis of lag structure from time series epidemiology, using gases, particles, and source factor time series, can contribute to understanding the relationships among various air pollution indicators. Methods: Analyze lag structure from an epidemiologic study of cardiovascular mortality in Phoenix including CO, NO2, and SO2; and PM2.5 (based on both filter and TEOM measurements); and source contributions from source apportionment analysis. Results: Statistically significant (SS) results (increase in risk for a increase in a air pollution component) on lag day 1 for CO, NO2, both PM1 indicators, and TEOM PM2.5 can be explained in terms of a vehicular emissions factor. SS results on lag day 3 for both PM1 indicators, TEOM PM2.5 can be explained by removal of a coarse-mode components (soil plus an unknown). SS results on lag day 4 for TEOM PM1 and PM2.5 are explained by retention of nitrate or other semivolatile PM component on the TEOM but not the gravimetric measure (filter PM2.5). Lack of a SS result on lag day 0 for PM1 and PM2.5 are likely due to the very low concentration of sulfate in Phoenix. The regional sulfate factor is SS on lag day 0 but sulfate alone is not SS although an increase in risk is observed. Conclusions: Analysis off lag structure helps explains the SS relationships between risk of cardiovascular mortality and air pollution parameters. It further suggests that different components of air pollution may cause mortality with different delay times and that the total risk due to air pollutants from various sources should be estimated from the sum of SS risks on several lag days.