Dr.
Les
Grant,
RTP
Division
Director
National
Center
for
Environmental
Assessment
Office
of
Research
and
Development
Briefing
for
Congressional
Staff
January
16,
2004
Revised
Air
Quality
Criteria
for
Particulate
Matter
National
Ambient
Air
Quality
Standards:

Congressional
Requirements
and
EPA
Review
Process
°
National
Ambient
Air
Quality
Standards
(
NAAQS)

 
Required
by
Clean
Air
Act
to
protect
against
health
and
welfare
effects
of
widespread
major
air
pollutants
 
NAAQS
and
scientific
bases
("
criteria")
must
be
reviewed/
revised,

as
appropriate,
every
5
years.

°
NAAQS
Review
Process
 
Air
Quality
Criteria
Documents
are
prepared
by
the
National
Center
for
Environmental
Assessment
in
EPA's
Office
of
Research
and
Development.

 
Staff
Papers
are
prepared
by
Office
of
Air
Quality
Planning
and
Standards
(
OAQPS)
in
EPA's
Office
of
Air
and
Radiation.
Current
PM
NAAQS
Review:

PM
Criteria
Document
Revision
°
Status
of
PM
Air
Quality
Criteria
Document
(
PM
AQCD)

Revision
 
1997
decisions
to
retain
PM
10
and
set
new
PM
2.5
NAAQS
based
on
1996
PM
AQCD
findings.

 
New
scientific
research
from
EPA,
private
sector,
academia,
and
other
federal
agencies
is
incorporated
into
revised
ACQD
 
New
information
reduces
uncertainty;
improving
scientific
foundation
of
NAAQS
 
Four
PM
AQCD
External
Review
Drafts
(
Oct
1999;
March
2001;
April
2002;
June
2003)
have
undergone
public
comment
and
CASAC
review.
Major
revisions
still
needed
for
Chapters
7­
9.

 
Final
PM
AQCD
to
be
completed
by
April
30,
2004,
under
terms
of
Consent
Decree
Agreement.
Key
Findings/
Conclusions
of
Revised
Draft
PM
Criteria
Document:
New
Epidemiology
Evidence
°
Short­
Term
PM
Exposure
Effects
 
Many
new
PM
time­
series
analyses
studies
show
positive,
often
statistically
significant
associations
between
24­
hr
ambient
PM
levels
and
increased
risk
of
mortality
and/
or
morbidity
(
indexed
by
hospital
admissions,
respiratory
symptoms,
etc.).

 
Reanalyses
to
address
GAM­
related
statistical
issues
generally
confirmed
original
results;
reduced
PM
effect
sizes
for
some.

 
Inclusion
of
gaseous
co­
pollutants
(
e.
g.,
O
3,
CO,
NO
2,
SO
2)
in
multipollutant
models
generally
did
not
markedly
change
PM
effect
seen
in
single
pollutant
models.

 
Overall,
new
short­
term
PM
epidemiology
studies
appear
to
substantiate
likely
associations
of
human
health
outcomes
(
mortality,

morbidity)
with
ambient
PM
acting
alone
and/
or
in
combination
with
other
gaseous
co­
pollutants.
°
Long­
Term
PM
Exposure
Effects
 
HEI­
sponsored
reanalyses
of
Harvard­­
Six
Cities
and
American
Cancer
Society
(
ACS)
Studies
of
long­
term
PM
exposure
effects
on
mortality
confirm
original
findings.

 
New
ACS
Study
extension
(
doubling
follow­
up
time
from
8
to
16
years)

confirms
original
findings
of
increased
PM
2.5­
related
total
and
causespecific
cardiovascular
mortality
and
adds
strongest
epi
evidence
yet
for
increased
lung
cancer
association
with
fine
(
PM
2.5)
but
not
coarse
(
PM
15­
2.5)
fraction
particles.

 
New
"
intervention"
studies
indicate
decreased
risks
of
mortality
associated
with
reductions
in
airborne
PM
or
SO
2
after
implementation
of
air
pollution
controls.

Key
Findings/
Conclusions
of
Revised
Draft
PM
Criteria
Document:
New
Epidemiology
Evidence
Key
Findings/
Conclusions
of
Revised
Draft
PM
Criteria
Document:

New
Dosimetry/
Toxicology
Evidence
°
Toxicological
Evidence
of
Health
Effects
Related
to
Ambient
PM
 
Utah
Valley
PM
collected
before,
during,
and
after
steel
mill
closure
provided
natural
experiment;
coherence
seen
between
epidemiologic
results
and
PM
effects
shown
experimentally
in­
vitro,
in
animals,
and
in
human
clinical
studies.

 
Exposure
of
humans
and
animals
to
concentrated
ambient
particles
(
CAPs)
result
in
physiological
changes
and
dose­
related
impacts
on
respiratory
and
cardiovascular
functions.

 
In­
vivo
and
in­
vitro
studies
provide
evidence
for
health
effects
associated
with
PM
constituents/
sources
(
e.
g.,
metals,
sulfates,
coal/
oil
combustion).

 
Newly
available
findings
on
biologic
mechanisms
suggest
PM­
related
cardiovascular
effects
mediated
via
direct,
neurogenic,
and
systemic
mechanisms.
Attachments
to
January
16,
2004
Briefing
for
Congressional
Staff
°
Figure
8­
15
(
from
June
2003
Fourth
External
Review
Draft
of
PM
Criteria
Document),
showing
results
of
reanalyses
addressing
GAM­
related
statistical
issues
for
a
number
of
key
PM
time­
series
studies.

°
Figures
8­
18
to
8­
21
(
from
June
2003
Fourth
Draft
PM
Criteria
Document),
comparing
results
of
single
pollutant
models
(
PM
alone)
versus
multi­
pollutant
models
(
PM
plus
O
3;
PM
plus
CO;

etc.).
Figure
8­
15.
PM
10
excess
risk
estimates
for
total
non­
accidental
mortality
for
numerous
locations
(
and
for
cardiovascular
mortality[*]
for
Coachella
Valley,
CA
and
Phoenix,
AZ),
using:
(
1)
GAM
with
default
convergence
criteria
(
white
circle);
(
2)
GAM
with
stringent
convergence
criteria
(
black
circle);
and,
(
3)
GLM/
natural
splines
(
x)
that
approximate
the
original
GAM
model
from
the
GAM
reanalysis
studies.
The
numbers
in
parenthesis
indicate
lag
days
used
("
01"
is
average
of
0
and
1
day
lags).
Figure
8­
18.
Excess
risk
estimates
for
total
non­
accidental
mortality
in
single­
pollutant
(
PM
only)
and
multi­
pollutant
models.
PM
increments:
50

g/
m3
for
PM
10
and
25

g/
m3
for
PM
2.5
and
PM
10­
2.5.
Results
presented
from
time­
series
studies
that
did
not
use
GAM
or
were
reanalyzed
using
GLM.

*
Estimates
from
multi­
pollutant
models
in
Ito
(
2003)
obtained
from
the
author
via
personal
communication
(
November
25,
2003).
Figure
8­
19.
Excess
risk
estimates
for
cardiovascular­
related
effects,
including
mortality,
hospital
admissions,
and
changes
in
biomarkers
(
e.
g.,
increases
in
blood
parameters
or
decreases
in
heart
rate
variability
measures)
in
single­
pollutant
(
PM
only)
and
multi­
pollutant
models
.
PM
increments:
50

g/
m3
for
PM
10
and
25

g/
m3
for
PM
2.5
and
PM
10­
2.5.
Results
presented
from
time­
series
studies
that
did
not
use
GAM
or
were
reanalyzed
using
GLM.
IH
=
ischemic
heart
disease;
HF
=
heart
failure;
HR
=
heart
rate;
HRV
=
heart
rate
variability.

*
Estimates
from
multi­
pollutant
models
in
Ito
(
2003)
obtained
from
the
author
via
personal
communication
(
November
25,
2003).
Figure
8­
20.
Excess
risk
estimates
for
respiratory­
related
effects,
including
mortality,
hospital
admissions
and
medical
visits
in
single­
pollutant
(
PM
only)
and
multi­
pollutant
models.
PM
increments:
50

g/
m3
for
PM
10
and
25

g/
m3
for
PM
2.5
and
PM
10­
2.5.
Results
presented
from
time­
series
studies
that
did
not
use
GAM
or
were
reanalyzed
using
GLM.
Mort
=
mortality;
Pneu
=
pneumonia;
COPD
=
chronic
obstructive
pulmonary
disease.

*
Estimates
from
multi­
pollutant
models
in
Ito
(
2003)
obtained
from
the
author
via
personal
communication
(
November
25,
2003).
Figure
8­
21.
Excess
risk
estimates
for
increases
in
respiratory
symptoms
or
decreases
in
lung
function
measures
in
singlepollutant
(
PM
only)
and
multi­
pollutant
models.
PM
increments:
50

g/
m3
for
PM
10
and
25

g/
m3
for
PM
2.5
and
PM
10­
2.5.
Results
presented
from
time­
series
studies
that
did
not
use
GAM
or
were
reanalyzed
using
GLM.
