Health Effects of Aerosols - Aerosols: Measurement, Dosimetry, and Health Effects - page 612

1.00
1.00
0.92

1.00
0.88
0.96

CVDRESP
Total
Others

Lag 1
Lag 0
Lag 2

−0.2

0.0 0.2 0.4 0.6 0.8 1.0 −0.2 0.0 0.2 0.4 0.6 0.8 1.0

% Change in mortality per 10 µg/m 3 increase in PM 10 % Change in mortality per 10 µg/m 3 increase in PM 10 (a) (b)

PM 10
PM 10 +O 3

1.00
1.00
0.96
0.97
0.98

PM 10 +O 3 +NO 2

PM 10 +O 3 + SO 2

PM 10 +O 3 +CO

−0.25 0.0 0.25 0.5 0.75

(c)

% Change in mortality per 10 µg/m 3 increase in PM 10

FIGURE 23.24 Results from the revised NMMAPS analyses [310]. (a) Marginal posterior distribution of the
national average effect of 10 μg/m 3 increase in PM 10 on total, nonaccidental mortality based on different lags.
(b) Comparative marginal posterior distribution of the national average effect of 10 μg/m 3 increase in PM 10
at lag 1 on total, cardiorespiratory (CVDRESP), and “other” mortality. (c) Marginal posterior distribution of
effect of 10 μg/m 3 increase in PM 10 at lag 1 in models with other pollutants. Box at upper right includes pos-
terior probabilities that PM 10 effects are greater than 0, based on pooled city-speciic estimates in a two-stage
hierarchical model. a Revised pooled estimates. (From Dominici, F. et al., Mortality among residents of 90
cities, Special Report: Revised Analyses of Time-Series Studies of Air Pollution and Health, Health Effects
Institute, Ed., Health Effects Institute, Boston, MA, pp. 9-24, 2003.)

evidence of harvesting at the 60-day averaging window, and pneumonia showed evidence of short-
term harvesting, followed by increased percentage increases (Figure 23.31). These data imply that
COPD deaths are being brought forward by about 2 months, while deaths from pneumonia and IHD
are not due to harvesting and may relect “enrichment” of the at-risk pool as a consequence of per-
sistent exposure to increased average levels of PM [242]. An analysis of the APHEA project, based
on distributed lag models applied to ten cities, failed to ind strong evidence for mortality displace-
ment for total daily mortality (Figure 23.32) [244]. Moreover, the effect estimate for mortality for
exposures 11-60 days before death was more than two-thirds as large as that for 10 days just prior
to death (0.688 ± 0.261 vs. 0.922 ± 0.184, respectively—both estimates are ×10 3 ). Similar results