Breathing Zone Exposure Assessment - Aerosols: Measurement, Dosimetry, and Health Effects - page 32

The most appropriate location to monitor personal-level exposures is in the breathing zone (BZ) in
relatively close proximity to the nose and mouth to minimize the inluence of strong gradients that
may exist near localized sources and near the body.
This chapter explores questions such as the following: (a) What information content is added by
conducting BZ exposures instead of simplistic, ixed location monitors? (b) does the monitoring
location on the body matter? (c) how are exposure misclassiication bias and confounding issues best
handled during BZ exposure assessments? (d) how do burden issues from BZ exposure monitors
impact the representativeness (and bias) of the resultant data? (e) does the application of BZ expo-
sures rather than applying exposure surrogate methods strengthen the epidemiologic associations
between elevated toxicant levels and adverse health outcomes? and (f) can potential doses also be
estimated from BZ exposure concentrations?
While concentrations are produced by emissions from sources, exposures only occur if an indi-
vidual is close enough (proximal) for a suficiently extended period to result in a signiicant expo-
sure level. Fixed-location monitors do not incorporate the element of proximity, nor do they account
for periods of time the person is actually nearby. The proximity to point, area, and line contaminant
sources produces concentration gradients that can be substantial. Personal activity sources such
as particle resuspension while walking (Oberoi et al., 2010), cooking (Koistinen et al., 2004), and
workplace operations (Maynard and Jensen, 2001) often comprise the single strongest source cat-
egory for many exposures. Since proximity changes as a person moves, personal exposure assess-
ment provides the most complete and integrated picture of exposure. The process of capturing
personal exposure data is more time consuming and expensive, and can be very burdensome to the
individuals being studied, compared with less representative surrogate monitoring locations. Thus,
apply BZ exposures involves trade-offs balancing the values of enhanced accuracy and representa-
tiveness of the collected data against the cost, burden, and complexity beneits of simpler surrogate
estimating methods and locations. The importance of conducting BZ exposure studies for children
is addressed in a review paper by Ashmore and Dimitroulopoulou (2009) for both gases and par-
ticles. They stress the importance of a holistic approach to managing aspects of indoor air pollution
by utilizing personal exposure methods to robustly deine the spatial and temporal ranges of their
exposure distributions and their impacts on children's health.
While BZ exposures are routinely collected for both gas- and particle-phase contaminants,
the inertial, diffusional, and transformational characteristics of airborne particles tend to add
an additional level of complexity to the issues inherent in conducting accurate and representa-
tive assessments. This chapter intentionally focuses on aerosol sampling issues, given their greater
complexity, often larger and more burdensome monitors, and potential for exhibiting the greatest
biases between locations imposed by either differences in monitoring methods or strong spatial and
temporal gradients encountered by individuals.
2.2 MINIMIZING EXPOSURE MISCLASSIFICATION BIASES
2.2.1 bz e xPosures versus F ixed -l ocation c oncentrations
2.2.1.1 Rationale
The strategy of sampling of contaminants in or near the human BZ as most representative of inhala-
tion exposure has been acknowledged as the most accurate approach (e.g., Cohen et al., 1984; NRC,
2004; Sarnat et al., 2007). These researchers recognized that relatively simple, ixed-location met-
rics are still needed to estimate cohort exposures in a simple and cost-effective manner over long
study periods. BZ exposure metrics are inherently more complex and costly to administer. But they
also recognized that ixed-location monitors (and exposure models) are only surrogates for true—
“gold standard”—BZ exposures and cannot possibly estimate every nuance of the inherent spatial
and temporal gradients individuals in a cohort experience daily. The key is deining the bias levels
inherent in the surrogate method and conducting careful trade-off analyses to deine the costs and