studies, the investigators can obviously only study
air concentrations that are present and the extent
they can be characterized and then relate the ambient concentrations to one or more health indices
such as changes in lung function.
The predominant source of information on the
health effects of the common air pollutants has
been from epidemiological studies. It is beyond
the scope of this entry to discuss the conduct and
results of those studies in detail. The reader interested in more detail will find the textbook on
Environmental Medicine edited by Ayres et al.
[6] useful. Suffice it to note that most epidemiological studies of air pollution are a kind of cohort
study or a time-series study. None of the studies
are strict cohort studies in which a population is
randomly assigned to one or more treatment
groups. This is rarely possible in studying air
pollution. Investigators are constrained to studying populations under natural conditions of ambient exposure and other risk factors associated with
the natural course of life. The study that may come
closest to the ideal cohort study is the Six-City
Study [20, 23] which will be discussed later. In
such a study, the investigators may use regulatory
compliance monitoring data supplemented by
research monitoring to characterize ambient air
concentrations of the various pollutants. Health
outcomes information may be obtained by contact
with the enrollees supplemented by information
acquired from Administrative databases such as
records of deaths. Other studies such as those
using the ACS populations, for example, Pope
et al. [54] are totally dependent upon regulatory
compliance modeling and administrative databases for health outcomes. As might be expected,
the quality of these studies is dependent upon the
quantity and quality of the monitoring data, the
availability of data on other risk factors, the duration of follow-up, the quality of the health outcome data, and the rigor of the analyses.
A second type of study design is that used in
the time-series studies as pioneered by the Johns
Hopkins University investigators in the National
Morbidity and Mortality Air Pollution Studies
(NMMAPS). These studies utilize regulatory
compliance monitoring data to characterize ambient concentrations of pollutants and administrative databases for morbidity and mortality. These
studies seek to characterize the relationship
between ambient concentrations on a given day
or series of days and excess morbidity or mortality
occurring concurrently or on subsequent days. An
example is the study of Bell et al. [8, 9] and Smith
et al. [58] that will be discussed later. An advantage is that it is assumed that many other risk
factors that may be influencing the health of the
population is not varying on a day-to-day basis,
for example, socioeconomic status. Other factors
such as temperature may vary along with pollution levels.
In the next section, information on the health
effects of particulate matter and ozone will be
briefly reviewed as examples of the kind of information used in setting NAAQS. In this section
I draw heavily on a previous review of risk assessment/risk management [40].
Particulate Matter as an Example
The Particulate Matter (PM) NAAQS was initially
set in 1971 using Total Suspended Particulate
(TSP) Matter as an indicator [36, 37, 39, 42,
64]. Later, the standard was revised with the TSP
indicator replaced by PM10. PM 10 is defined as
the size fraction collected using a device that
includes 50% of the particle with an aerodynamic
diameter of 10 mm. A progressively smaller fraction of PM mass is collected as the particle size
increases above 10 pm and a progressively larger
fraction of PM mass is sampled as particle size
decreases below 10 mm. Later, the standard was
again changed with PM 2.5 introduced as an indicator to complement the PM 10 indicator. PM 2.5 ,
the fine particle fraction indicator is defined as for
PM 10 indicator with a 50% cut point at 2.5 mm
aerodynamic diameter. To state the obvious, the
PM 2.5 fraction is contained within the PM 10 fraction which is contained within the TSP sample. In
recent years, increased attention has been focused
on the ultrafine particle number which is
contained within the PM 2.5 size range. There has
also been interest in characterizing health effects
of the PM 10–2.5 fraction which has been called the
course particle fraction. Consideration is given to
regulating this fraction with a separate NAAQS.
Air Quality Guidelines and Standards
27
air concentrations that are present and the extent
they can be characterized and then relate the ambient concentrations to one or more health indices
such as changes in lung function.
The predominant source of information on the
health effects of the common air pollutants has
been from epidemiological studies. It is beyond
the scope of this entry to discuss the conduct and
results of those studies in detail. The reader interested in more detail will find the textbook on
Environmental Medicine edited by Ayres et al.
[6] useful. Suffice it to note that most epidemiological studies of air pollution are a kind of cohort
study or a time-series study. None of the studies
are strict cohort studies in which a population is
randomly assigned to one or more treatment
groups. This is rarely possible in studying air
pollution. Investigators are constrained to studying populations under natural conditions of ambient exposure and other risk factors associated with
the natural course of life. The study that may come
closest to the ideal cohort study is the Six-City
Study [20, 23] which will be discussed later. In
such a study, the investigators may use regulatory
compliance monitoring data supplemented by
research monitoring to characterize ambient air
concentrations of the various pollutants. Health
outcomes information may be obtained by contact
with the enrollees supplemented by information
acquired from Administrative databases such as
records of deaths. Other studies such as those
using the ACS populations, for example, Pope
et al. [54] are totally dependent upon regulatory
compliance modeling and administrative databases for health outcomes. As might be expected,
the quality of these studies is dependent upon the
quantity and quality of the monitoring data, the
availability of data on other risk factors, the duration of follow-up, the quality of the health outcome data, and the rigor of the analyses.
A second type of study design is that used in
the time-series studies as pioneered by the Johns
Hopkins University investigators in the National
Morbidity and Mortality Air Pollution Studies
(NMMAPS). These studies utilize regulatory
compliance monitoring data to characterize ambient concentrations of pollutants and administrative databases for morbidity and mortality. These
studies seek to characterize the relationship
between ambient concentrations on a given day
or series of days and excess morbidity or mortality
occurring concurrently or on subsequent days. An
example is the study of Bell et al. [8, 9] and Smith
et al. [58] that will be discussed later. An advantage is that it is assumed that many other risk
factors that may be influencing the health of the
population is not varying on a day-to-day basis,
for example, socioeconomic status. Other factors
such as temperature may vary along with pollution levels.
In the next section, information on the health
effects of particulate matter and ozone will be
briefly reviewed as examples of the kind of information used in setting NAAQS. In this section
I draw heavily on a previous review of risk assessment/risk management [40].
Particulate Matter as an Example
The Particulate Matter (PM) NAAQS was initially
set in 1971 using Total Suspended Particulate
(TSP) Matter as an indicator [36, 37, 39, 42,
64]. Later, the standard was revised with the TSP
indicator replaced by PM10. PM 10 is defined as
the size fraction collected using a device that
includes 50% of the particle with an aerodynamic
diameter of 10 mm. A progressively smaller fraction of PM mass is collected as the particle size
increases above 10 pm and a progressively larger
fraction of PM mass is sampled as particle size
decreases below 10 mm. Later, the standard was
again changed with PM 2.5 introduced as an indicator to complement the PM 10 indicator. PM 2.5 ,
the fine particle fraction indicator is defined as for
PM 10 indicator with a 50% cut point at 2.5 mm
aerodynamic diameter. To state the obvious, the
PM 2.5 fraction is contained within the PM 10 fraction which is contained within the TSP sample. In
recent years, increased attention has been focused
on the ultrafine particle number which is
contained within the PM 2.5 size range. There has
also been interest in characterizing health effects
of the PM 10–2.5 fraction which has been called the
course particle fraction. Consideration is given to
regulating this fraction with a separate NAAQS.
Air Quality Guidelines and Standards
27