Environmental Engineering Reference
In-Depth Information
consistent with the general literature in showing positive associations between
increased concentrations of constituents of PM
2.5
and increased risk of respiratory
and cardiovascular morbidity; mortality lagged 1-3 days following the change in
ambient PM. The majority of the studies have the advantage of having included
large populations. None of the epidemiology studies covered in our review conclu-
sively implicated specifi c transition metals as being the cause of the increased
cardio-pulmonary morbidity and mortality observed following incremental
increases in levels of PM. However, the major fi ndings presented in the eleven epi-
demiology studies included in our review (Bell et al.
2009
; Burnett et al.
2000
;
Hirshon et al.
2008
; Laden et al.
2000
; Lipfert et al.
2006
; Lippmann et al.
2006
;
Ostro et al.
2007
,
2008
,
2009
; Patel et al.
2009
; Suh et al.
2011
) tended to be inter-
nally consistent in identifying some metals (Fe, Ni, V, Zn) more often than others
(As, Cu, Mn, Se) as potentially affecting health.
The fact that Fe and Zn are the two most abundant transition metals in ambient
PM may have contributed to these fi ndings. As discussed previously, the studies
conducted by Bell et al. (
2009
) and Lippmann et al. (
2006
) included New York City,
wherein levels of V and Ni were relatively high during the time period of data col-
lection, and the Dominici study (
2007
) showed that the associations disappeared
when New York City was removed from the other NMMAPS cities. The multiple
studies by Ostro did not report signifi cant effects associated with Ni, and signifi cant
associations between V and health outcomes were reported in only one study. In
those studies, in which transition metals in addition to Ni and V were evaluated, the
majority showed no effects of these other metals on health outcomes. Furthermore,
the metals identifi ed by the investigators as being more toxic were chosen
a priori
as targets of these investigations (e.g., in the many ROFA studies) far more often
than the others, biasing the ability to uncover potential associations.
The studies used to defi ne the role of metals in causing the respiratory and cardio-
vascular health effects had limitations that included: the characterization of exposure
from PM monitoring data, the presence of co-pollutants in ambient PM, and indi-
vidual risk factors for the same outcomes in members of the subject populations.
All of the studies used data from “centrally” located monitors to quantify the
concentrations of PM
2.5
, PM
10
, and individual transition metals to which the study
populations were exposed. In the California studies (Ostro et al.
2007
,
2008
,
2009
)
characterization of exposure was based on relatively few measurements, the moni-
tors were not co-located in most counties and data were reported every third or sixth
day. The study conducted by Hirshon et al. (
2008
) in Greater Baltimore inferred
exposure from a single sampling location as being representative of the entire 2.6
million population. Relying upon measurements from a single monitoring location
can be particularly problematic for constituents such as Zn, Ni and V that have high
spatial variability (Patel et al.
2009
; Bell et al.
2009
), even when the model accounted
for random effects. For example, when New York County and Queens County were
excluded from the analysis by Bell et al. (
2009
), the association between Ni and
respiratory hospitalizations was no longer signifi cant. It is also important to recog-
nize that the concentrations of Ni and V in the ambient PM in certain areas of
Search WWH ::
Custom Search