Estimating cumulative ambient concentrations of air pollutants: description and precision of methods used for an epidemiological study

Abstract

Methods for estimating cumulative ambient air pollution concentrations for individuals enrolled in an epidemiological cohort study are described and studied. Monthly interpolations from fixed-site monitoring stations in California to zip code centroids were used. The precision of the interpolation methods for total suspended particulates and ozone was assessed, and fixed-site monitoring stations were used in turn as receptor sites. Actual versus interpolated 2-y mean concentrations did not differ significantly and were correlated with a Pearson correlation coefficient of .78 for total suspended particulates and .87 for ozone. The impact of the change from monitoring total oxidants to ozone on oxidant/ozone cumulations was evaluated, and monthly mean concentrations for total oxidants were compared with ozone for 435 station months during which both pollutants were measured simultaneously. Average concentrations did not differ significantly and were highly correlated (r = 0.94). Exceedance frequency statistics, which averaged slightly higher for ozone than for total oxidants, were also highly correlated.