"
Grahame,
Thomas"
<
Thomas.
Grahame@
HQ.
DOE.
GOV>

06/
19/
2006
06:
26
PM
T
o
Jason
Burnett/
DC/
USEPA/
US@
EPA
c
c
"
Daniels,
Jarad"
<
Jarad.
Daniels@
hq.
doe.
gov>
S
u
b
j
e
c
t
is
it
fine
PM
or
traffic
emissions?
the
need
to
use
multi­
pollutant
models
to
ensure
robust
findings
Jason,
I
would
like
to
add
to
the
discussion
about
whether
it
is
fine
PM
generally,
or
traffic
emissions,
which
are
most
likely
to
be
associated
with
morbidity
and
mortality,
especially
in
daily
health
effects
studies.
A
new
2006
study
(
not
previously
brought
to
the
attention
of
EPA,
at
least
by
me,
and
attached
below
the
signature
block)
has
come
to
my
attention.
I
have
included
a
short
analysis
of
it
in
the
attached
MS
Word
document,
which
also
includes
a
fuller
discussion
of
the
Stieb
et
al
(
2000)
study,
now
that
I
have
got
hold
of
a
copy
of
it.

<<
ADDITIONAL
DAILY
MORTALITY
STUDY
ANALYSIS.
doc>>

Long
story
short,
a
summary
of
both
this
new
study,
as
well
as
of
the
seven
studies
(
including
Stieb
et
al)
in
the
R&
L
memo
examining
24
hour
mortality
and
morbidity
with
98th
percentile
concentrations
below
that
of
the
Lipfert
et
al
(
2000)
study
of
Philadelphia
(
e.
g.,
below
44.2
µ
g/
m3),
is
this:
fine
PM
is
never
significantly
associated
with
morbidity
or
mortality
in
such
studies
in
a
multi­
pollutant
model.
In
several
cases
fine
PM
wasn't
significantly
associated
even
in
single
pollutant
models.

Further,
authors
of
studies
included
in
the
R&
L
memo
state
that
single
pollution
model
results
are
not
robust,
compared
to
the
multi­
pollutant
model
results
which
find
no
PM
associations.
Here
is
a
quote
from
a
1998
Delfino
et
al
study
examined
and
referenced
in
the
attachment
and
which
I
believe
is
also
included
in
the
CD:
"
Although
there
were
adverse
effects
of
estimated
PM2.5
on
ER
visits
for
respiratory
illnesses
among
the
elderly,
the
association
was
unstable
and
completely
confounded
by
both
temperature
and
O3...
it
is
reasonable
to
accept
only
the
most
robust
findings
as
meaningful
to
community
health 
therefore
the
present
discussion
will
focus
on
the
most
notable
finding
of
adverse
affects
of
O3
among
the
elderly "
(
italics
ours)

In
summary,
then,
a
careful
reading
of
the
seven
studies
in
the
R&
L
memo
as
well
as
supporting
evidence
suggests
that:

(
1)
There
is
no
robust
evidence
that
fine
PM
mass
per
se
is
associated
with
daily
adverse
health
effects,
certainly
in
studies
with
concentrations
at
levels
below
44.2
µ
g/
m3,
in
the
studies
cited
in
the
R&
L
memo.
(
I
haven't
examined
the
Lipfert
et
al
(
2000)
study
which
is
said
to
have
made
findings
at
that
98th
percentile
concentration.)

(
2)
The
studies
consistently
find,
in
robust
multi­
pollutant
modeling,
that
it
is
vehicular
emissions
and
not
fine
PM
per
se,
which
is
consistently
linked
with
daily
health
effects
(
the
Stieb
et
al
study
is
of
a
town
with
a
refinery,
and
some
emissions
associated
with
health
effects
might
reflect
refinery
activities
there).

(
3)
The
findings
 
that
in
multi­
pollutant
modeling
it
is
not
fine
PM
but
rather
other
emissions,
usually
vehicular
emission,
which
are
associated
with
adverse
health
effects
 
are
the
same
findings
as
made
in
the
new
2006
study
which
is
attached.
This
new
study
shows
that
although
the
traffic
volume
density
variable
is
most
strongly
related
to
health
effects
in
single
pollutant
models,
several
pollution
variables
(
including
fine
PM)
are
also
significantly
associated.
But
in
multi­
pollutant
models,
while
the
traffic
volume
density
variable
remains
significant
and
barely
changes
in
size,
none
of
the
pollution
variables
retain
significance
except
peak
ozone.

(
4)
These
findings
are
consistent
with
other
studies
examined
in
the
lengthier
DOE
comments
that
have
been
provided,
e.
g.
studies
using
the
newest
epidemiology.
Such
studies,
such
as
those
of
Schwartz
et
al
(
2005)
and
Gold
et
al
(
2005),
use
vehicular
emissions
data
(
black
carbon
fine
PM)
which
accurately
reflects
the
exposure
of
subjects
to
these
emissions,
and
these
studies
 
although
of
different
design
that
the
24
hour
studies
using
gases
with
fine
PM
in
multi­
pollutant
analysis
 
find
that
it
is
the
vehicular
PM
but
not
the
other
types
of
PM
which
are
associated
with
adverse
health
effects.

In
short,
the
evidence
appears
to
be
strong
that
a
national
24
hour
standard
should
not
be
based
upon
findings
in
any
of
the
seven
studies
in
the
R&
L
memo
with
98th
percentile
concentrations
below
44.2
µ
g/
m3,
including
especially
Burnett
and
Goldberg
(
2003),
because
(
1)
these
findings
are
from
non­
robust
single
pollutant
models;
(
2)
three
of
the
seven
have
insignificant
results
even
in
single
pollutant
models;
(
3)
significant
single
pollutant
findings
are
contradicted
by
later,
robust
multi­
pollutant
studies
by
the
same
authors
(
or
within
the
same
study,
e.
g.,
Stieb
et
al,
2000);
and
(
4)
are
also
contradicted
by
the
strong
weight
of
evidence
in
the
newer,
more
sophisticated
studies,
including
the
2006
study
attached.
Such
weak
and
contradictory
evidence
does
not
appear
to
have
the
analytical
strength
upon
which
to
base
new
a
national
ambient
air
quality
standard
for
fine
PM
at
35
µ
g/
m3,
a
level
which
seems
to
be
based
upon
the
98th
percentile
concentration
in
Burnett
and
Goldberg
(
2003),
with
an
added
"
margin
of
safety."

Thanks,

Tom
Thomas
J.
Grahame
U.
S.
Department
of
Energy
1000
Independence
Ave.,
SW
Washington,
DC
20585
202
586
7149
(
voice)

202
586
7085
(
fax)
<<
Lipfert
et
al
2006
Atm
Env
traffic
density
article.
pdf>>
