Content uploaded by Meei-Maan Wu
Author content
All content in this area was uploaded by Meei-Maan Wu on May 15, 2017
Content may be subject to copyright.
Articles
Risk
of
Internal
Cancers
from
Arsenic
in
Drinking
Water
Knashawn
H.
Morales,1
Louise
Ryan,1'2
Tsung-Li
Kuo,3
Meei-Maan
Wu,4
and
Chien-Jen
Chen5
1Department
of
Biostatistics,
Harvard
School
of
Public
Health,
Boston,
Massachusetts,
USA;
2Dana-Farber
Cancer
Institute,
Boston,
Massachusetts,
USA;
3Department
of
Forensic
Medicine, College
of
Medicine,
National
Taiwan
University,
Taipei,
Taiwan;
4Institute
of
Biomedical
Sciences,
Academia
Sinica,
Taipei,
Taiwan;
5Graduate
Institute
of
Epidemiology,
National
Taiwan
University,
Taipei,
Taiwan
The
U.S.
Environmental
Protection
Agency
is
under
a
congressional
mandate
to
revise
its
current
standard
for
arsenic
in
drinking
water.
We
present
a
risk
assessment
for
cancers
of
the
bladder,
liver,
and
lung
from
exposure
to
arsenic
in
water,
based
on
data
from
42
villages
in
an
arseniasis-
endemic
region
of
Taiwan.
We
calculate
excess
lifetime
risk
estimates
for
several
variations
of
the
generalized
linear
model
and
for
the
multistage-Weibull
model.
Risk
estimates
are
sensitve
to
the
model
choice,
to
whether
or
not
a
comparison
population
is
used
to
define
the
unexposed
diease
mortality
rates,
and
to
whether
the
comparison
population
is
all
of
Taiwan
or
just
the
southwest-
em
region.
Some
factors
that
may
affect
risk
could
not
be
evaluated
quantitativel.
the
ecologic
nature
of
the
data,
the
nutritional
status
of
the
study
population,
and
the
dietary
intake
of
arsenic.
Despite
all
of
these
sources
of
uncertainty,
however,
our
analysis
suggests
that
the
current
standard
of
50
pg/L
is
asoatted
with
a
substantial
increased
risk
of
cancer
and
is
not
sufficiently
protective
of
public
health.
Key
wordr
bladder
cancer,
generalized
linear
model,
lifetime
death
risk,
lung
cancer,
margin
of
exposure,
multistage-Weibull.
Environ
Health
Perspect
108:655-661(2000).
[Online
5
June
2000]
htp:icepnntl.
niebs.
nib.gov/docs/2000/108p655-6616morakslbs/tracthl
A
metal
found
in
rocks
and
mineral
forma-
tions
in
the
earth's
crust,
arsenic
has
long
been
associated
with
the
development
of
can-
cer
in
humans.
Exposure
can
occur
via
inhalation,
primarily
in
industrial
settings,
or
through
ingestion.
Because
drinking
water
is
one
of
the
primary
routes
of
exposure,
stan-
dards
set
in
1942
established
a
maximum
contaminant
level
(MCL)
of
50
pg/L
in
drinking
water.
In
1975,
50
pg/L
was
adopt-
ed
as
the
interim
standard
in
response
to
the
1974
Safe
Drinking
Water
Act
(1).
In
a
1984
health
assessment,
the
U.S.
Environmental
Protection
Agency
(EPA)
dassified
arsenic
as
a
dass
A
human
carcinogen,
based
primarily
on
epidemiologic
evidence,
and
produced
quantitative
risk
estimates
for
both
ingestion
and
inhalation
routes
of
exposure
(2).
Although
the
EPA
assessment
for
the
inhala-
tion
route
is
well
accepted,
the
risk
assess-
ment
for
ingestion
remains
controversial.
The
1984
risk
assessment
for
arsenic
in
drinking
water
was
based
on
an
epidemio-
logic
study
in
Taiwan
that
examined
an
association
between
arsenic
exposure
via
drinking
water
and
skin
cancer
(non-
melanoma)
(3).
EPA
investigators
estimated
that
the
lifetime
risk
of
skin
cancer
for
indi-
viduals
who
consumed
2
L
water
per
day
at
50
psg/L
could
be
as
high
as
2
in
1,000.
This
high
value
prompted
questions
about
the
1984
risk
assessment,
including
applicability
of
the
risk
assessment
to
the
U.S.
popula-
tion,
the
role
of
arsenic
as
an
essential
nutri-
ent,
the
relevance
of
skin
lesions
as
the
basis
for
the
risk
assessment,
and
the
role
of
arsenic
intake
via
food.
In
1988,
the
EPA
Risk
Assessment
Forum
published
a
revised
skin
cancer
risk
assessment
and
focused
attention
on
these
questions
(4).
The
EPA
is
currently
under
a
congressional
mandate
to
finalize
a
new
rule
for
arsenic
in
drinking
water
by
1
January
2001
(5).
There
has
been
substantial
focus
on
the
association
between
arsenic
and
skin
cancer,
and
there
is
also
substantial
evidence
that
exposure
to
arsenic
in
drinking
water
increas-
es
the
mortality
risk
for several
internal
can-
cers.
Increases
in
bladder
and
lung
cancer
mortality
were
found
in
a
region
of
northern
Chile
(6).
An
association
was
also
found
between
bladder
cancer
mortality
and
arsenic
in
drinking
water
in
Argentina
(7).
Significant
increased
mortality
was
observed
for
males
and
females
in
Taiwan
due
to
lung,
liver,
skin,
kidney,
and
bladder
cancer
(8).
The
National
Research
Council
presents
a
more
detailed
summary
of
the
evidence
link-
ing
arsenic
exposure
to
internal
cancer
(1).
The
purpose
of
this
article
is
to
present
a
risk
assessment
for
mortality
due
to
several
internal
cancers
based
on
a
reanalysis
of
the
data
reported
by
Chen
et
al.
(8).
Brown
(9)
discussed
the
limitations
of
the
data
available
for
analysis
when
the
current
EPA
risk
assessment
(4)
was
prepared.
For
several
rea-
sons,
it
can
be
argued
that
the
risk
assess-
ment
of
internal
cancers
presented
in
this
paper
yields
more
convincing
results
than
the
previous
EPA
assessment
based
on
skin
cancer.
First,
the
current
study
focuses
on
mortality
from
bladder,
lung,
and
liver
can-
cers
identified
through
national
death
records.
In
addition,
unlike
the
Tseng
et
al.
(3)
study
that
was
used
in
the
EPA
analysis,
which
grouped
data
into
three
broad
exposure
intervals
[low
(<
300
pg/L),
medi-
um
(300-600
pg/L),
and
high
(>
600
pg/L)],
data
now
available
provide
exposure
at
the
individual
village
level.
This
paper
is
a
follow-up
to a
prelimi-
nary
study
that
focused
only
on
bladder
cancer
and
examined
model
sensitivity
(10).
The
current
analysis
is
expanded
to
include
lung
and
liver
cancers
and
examines
issues
of
dose-response
modeling
by
Poisson
regression,
in
addition
to
applica-
tion
of
the
multistage-Weibull
(MSW)
model,
in
more
detail.
Materials
and
Methods
Internal
cancer
data.
Data
used
in
this
analysis
were
derived
from
a
study
in
an
arse-
niasis-endemic
area
of
Taiwan
(11-13).
Cancer
mortality
data
were
collected
from
death
certificates
of
residents
of
42
villages
during
1973-1986.
These
data
were
original-
ly
collected
in
1987,
so
only
records
up
to
1986
were
available.
Causes
of
death
were
classified
according
to
the
Eighth
Revision
of
International
Classification
of
Diseases,
1965
Revision
(ICD)
(14).
The
data
consisted
of
person-years
at
risk
and
the
number
of
deaths
due
to
bladder
(ICD
code
188),
lung
(ICD
code
162),
and
liver
(ICD
code
155)
cancer
in
5-year
age
increments
for
both
males
and
females.
Table
1
summarizes
the
internal
cancer
data
and
provides
person-years
at
risk
and
observed
number
of
cancer
deaths
by
age,
sex,
and
arsenic
level.
Although
individ-
ual
village
arsenic
levels
are
available
and
will
be
used
in
subsequent
analyses,
exposure
lev-
els
are
grouped
in
Table
1
for
convenience of
presentation.
The
numbers
of
bladder,
liver,
and
lung
cancers
are
given,
along
with
the
number
of
person-years
at
risk.
For
example,
males
between
the
ages
of
50
and
69
con-
tributed
21,040
person-years
at
risk
and
6,
17,
and
12
deaths
were
observed
from
blad-
der,
liver,
and
lung
cancer,
respectively.
Address
correspondence
to
L.
Ryan,
Department
of
Biostatistics,
Dana-Farber
Cancer
Institute,
44
Binney
Street,
Boston,
MA
02115
USA.
Telephone:
(617)
632-3602.
Fax:
(617)
632-2444.
E-mail:
ryan@jimmy.harvard.edu
Support
was
received
from
the
National
Institutes
of
Health
(grants
SF3
1GM
18906,
ES0002,
and
CA48061),
the
David
and
Lucile
Packard
Foundation,
and
the
Department
of
Health,
Executive
Yuan,
ROC
(DOH88-HR-503).
Received
29
December
1999;
accepted
14
March
2000.
Environmental
Health
Perspectives
*
VOLUME
108
1
NUMBER
7
1
July
2000
655
Articles
*
Morales
et
al.
Exposure
data.
Drinking
water
samples
were
collected
from
wells
of
42
villages
in
1964-1966
(12).
The
artesian
wells
were
gradually
closed;
the
last
one
dosed
in
mid-
1970.
Although
mortality
data
were
collect-
ed
for
a
later
time
period
(1973-1986),
it
is
likely
that
arsenic
levels
in
well
water
remained
relatively
unchanged
over
this
time
period.
It
could
also
be
argued
that
because
of
the
long
latency
of
the
cancers
of
interest,
it
is
appropriate
for
exposure
to
be
based
on
a
time
period
10
to
20
years
before
death.
A
strength
of
the
currently
available
exposure
data
is
that
individual
well
concentration
levels
are
available
for
each
village.
Physical
and
chemical
characteristics
of
drinking
water
such
as
pH
value
and
levels
of
arsenic,
sodium,
calcium,
magnesium,
manganese,
iron,
mercury,
chromium,
lead,
nitrite
and
nitrate
nitrogen,
fluoride,
and
bicarbonate
have
been
intensively
studied
in
both
Blackfoot
disease-endemic
and
-nonendemic
areas
(15,16).
Arsenic
level
was
the
only
level
that
was
significantly
higher
than
the
maxi-
mal
allowable
limit
and
strikingly
different
in
water
from
shallow
wells
and
artesian
wells.
The
data
also
have
some
limitations.
The
drinking
water
was
not
tested
for
levels
of
dissolved
radon
and
other
a-emitters.
Fluorescent
compounds,
especially
humic
acids,
have
been
found
in
the
well
water.
These
fluorescent
substances
result
from
the
decomposition
of
organic
matter,
particular-
ly
dead
plants.
However,
it is
unlikely
that
their
presence
causes
confounding
in
this
analysis
because
widespread
contamination
is
not
confined
to
the
arseniais-endemic
area.
Standardized
mortality
ratio.
We
used
standardized
mortality
ratios
(SMRs)
to
sum-
marize
the
observed
patterns
of
mortality
in
data.
SMRs
provide
a
popular
approach
to
comparing
mortality
in
a
specific
population
with
mortality
from
a
suitable
comparison
population
(17).
SMRs
correspond
to
ratios
of
observed
and
expected
number
of
events
and
are
calculated
by
10/XEi,
where
0,
is
the
observed
number
of
deaths
in
the
ih
age
group
and
Ei
is
the
corresponding
expected
number
of
deaths,
calculated
by
multiplying
the
study
population
size
(P,)
by
the
age-spe-
cific
cancer
death
rate
(M)
in
a
comparison
population
(i.e.,
Ei
=
P.
x
M).
Usually
SMRs
are
expressed
as
a
percentage
so
that
the
value
100
x
12Oi
l/Ei
is
the
number
reported.
There
are
concerns
with
using
all
of
Taiwan
as
a
comparison
population
because
of
the
potential
for
bias
associated
with
differences
in
the
populations
(e.g.,
rural
vs.
urban).
For
this
reason,
we
considered
two
comparison
populations
in
this
analysis:
all
of
Taiwan
and
the
southwestern
region
of
Taiwan
(18).
The
latter
is
expected
to
provide
a
more
suit-
able
comparison
basis
for
the
study
popula-
tion,
which
is
largely
rural
and
fairly
poor.
Table
2
contains
the
data
from
the
two
com-
parison
populations.
The
number
of
deaths
due
to
bladder,
lung,
and
liver
cancers
and
person
years
at
risk
(PYR)
were
extracted
by
age
group
and
sex
for
1973-1986.
Generalized
linear
model,
Poisson
mod-
eling
is
often
used
in
epidemiologic
analysis,
particularly
for
rare
events
such
as
cancer
deaths.
In
fact,
SMRs
can
correspond
to
maximum
likelihood
estimates
of
risk
ratios
from
a
Poisson
model
(17).
In
our
analysis,
we
assumed
that
the
number
of
deaths
due
to
cancer
follows
a
Poisson
distribution
with
parameter
equal
to
the
person-years
at
risk
multiplied
by
the
hazard
of
dying
of
cancer.
The
hazard
is
often
modeled
as
a
function
of
age
(t)
and
exposure
(x).
As
described
by
Breslow
and
Day
(17),
a
broad
class
of
mod-
els
can
be
characterized
using
the
following
general
form,
h(x,t)
=
ho(t)
X
g(x),
[1]
where
h0(4
denotes
the
baseline
hazard
func-
tion
that
only
depends
on
age,
t,
and
describes
the
instantaneous hazard
of
dying
of
cancer
for
the
unexposed
population.
The
risk
ratio
attributed
to
exposure
level
x
is
denoted
by
g(x).
Of
course,
it
is
likely
that
a
variety
of
factors,
including
cigarette
smoking,
use
of
bottled
water,
and
dietary
intake
of
inorganic
arsenic,
could
influence
or
even
confound
the
model.
The
model
described
in
Equation
1
will
allow
consideration
of
other
covariates.
Unfortunately,
measurements
for
these
and
other
potentially
important
factors
were
not
available
for
our
study.
Rather,
this
is
an
eco-
logic
study
wherein
only
relatively
simple
exposure
and
population
characteristics
could
be
measured.
It
will
be
important
to
consider
this
and
other
sources
of
uncertainty
when
interpreting
the
results.
Although
not
dis-
cussed
extensively
here,
it is
possible
for
the
risk
ratio
g(x)
to
also
depend
on
age,
t.
For
example,
older
people
may
be
more
suscepti-
ble
to
exposure.
We
did
in
fact
explore
such
age-dependent
risk
models
and
found
that
in
general,
it
was
adequate
to
model
the
relative
risk
as
a
function
of
exposure
only.
A
wide
range
of
models
was
obtained
by
varying
a)
the
use
of
comparison
populations;
b)
the
way
age
is
modeled
in
ho(4,
e.g.,
linear,
quadratic,
or
the
use
of
regression
splines;
c)
transforma-
tions
of
exposure
concentrations;
and
a)
the
way
exposure
is
modeled.
Table
3
summarizes
the
various
modeling
options
considered
in
this
analysis.
Each
model
corresponds
to
choosing
one
option
from
each
column.
For
example,
the
model
excluding
the
comparison
population,
with
a
linear
age
effect,
an
expo-
nential
linear
dose
effect,
and
no
transforma-
tion
on
dose,
is
characterized
by
ho(t)
=
exp(ao
+
clt)
and
g(x)
=
exp(,lx).
Note
that
the
linear
and
quadratic
dose-effect
models
(generally
referred
to
as
additive
models)
do
not
fit
into
the
usual
dass
of
generalized
linear
models
(GLMs)
and
require
special
program-
ming.
Exponential
linear
and
exponential
quadratic
models
fall
under
the
general
dass
of
multiplicative
models.
The
spline
age
effect
was
modeled
using
natural
splines
because
of
the
ease
of
obtaining
predicted
values
(19).
There
are
three
options
for
the
baseline
haz-
ard:
model
the
hazard
without
including
a
comparison
population,
treat
the
comparison
population
as
an
unexposed
group,
or
replace
the
baseline
hazard
function
with
empirical
Table
1.
Person-years
at
risk
by
age,
sex,
and
arsenic
level
with
observed
number
of
deaths
from
cancer
(bladder,
liver,
and
lung).
Sex,
arsenic
Age
(years)a
level
(pg/I)
20-30
30-49
50-69
>70
Total
Male
<
100
35,818
34,196 21,040
4,401
95,455
(0,0,0)
(1,10,2)
(6,17,12)
(10,4,14)
(17,31,28)
100-299
18,578
16,301
10,223
2,166
47,268
(0,
0,
0)
(0,
4,
3)
(7,
15,14)
(2,
4,13)
(9,
23,
30)
300-599
27,556
25,544
15,747
3,221
72,068
(0,
3,
0)
(5,7,
9)
(15,
23,
30)
(12,
6,14)
(32,
39,
53)
.
600
16,609
15,773
8,573
1,224
42,179
(0,
0,
1)
(4,
12,
3)
(15,
15,
23)
(8,
2,
6)
(27,
29,
33)
Total
98,561
91,814
55,583 11,012
256,970
(0,
3,
1)
(10,
33,
17)
(43,
70,
79)
(32,
16,
47)
(85,
122,
144)
Female
<
100
27,901
32,471
21,556
5,047
86,975
(0,
0,
0)
(3,1,
5)
(9,
6,18)
(9,
5,
5)
(21,
12,
29)
100-299
13,381
15,514
11,357
2,960
43,212
(0,
0,
0)
(0,
3,
4)
(9,
6,10)
(2,
5,
5)
(11,
14,19)
300-599
19,831
24,343
16,881
3,848
64,903
(0,
0,
0)
(0,
5,
6)
(19,
6,
20)
(11,
2,10)
(30,13,
36)
.600
12,988
15,540
9,084
1,257
38,869
(0,
0,
0)
(0,
4,
6)
(21,
7,
28)
(7,
1,
4)
(28,
12,
38)
Total
74,101
87,868
58,878
13,112
233,959
(0,
0,
1)
(3,
13,
21)
(58,
25,
76)
(29,
13,
24)
(90,
51,
122)
"Values
in
parentheses
are
number
of
deaths
from
bladder,
liver,
and
lung
cancer,
respectively.
VOLUME
108
1
NUMBER
7
1
July
2000
*
Environmental
Health
Perspectives
656
Articles
*
Internal
cancers
from
arsenic
in
drinking
water
estimates
based
on
the
comparison
popula-
tion
(not
included
in
Table
1).
The
third
option
can
be
accomplished
by
fitting
a
Poisson
model
containing
indicators
corre-
sponding
to
the
age
categories
observed
in
the
comparison
population.
This
approach
essen-
tially
corresponds
to
the
traditional
SMR
approach.
Because
there
were
no
villages
with
zero
concentration
levels,
the
method
used
to
model
the
baseline
hazard
had
a
fairly
strong
influence
on
the
results.
In
particular,
the
choice
of
whether
to
include
a
comparison
population
had
a
strong
influence.
The
use
of
an
unexposed
comparison
population
has
the
potential
to
provide
more
information
about
the
shape
of
the
model
at
low
concentrations.
Although
not
a
member
of
the
usual
GLM
class,
the
MSW
model
was
also
considered
because
it
was
used
in
the
previ-
ous
risk
assessment
(4).
The
MSW
corre-
sponds
to
letting
g(x)
=
PO
+
IX
+
2x2
and
ho(t)
=
C(t-
To)+
(10),
where
tdenotes
age
and
x
denotes
exposure
concentration.
The
plus
sign
(+)
indicates
a
truncation
on
the
(t
-
To)
term
(i.e.,
if
To
>
t
then
the
term
is
set
to
zero).
Results
based
on
the
MSW
model
are
only
presented
for
comparison.
The
GLM
approach
has
several
advantages
over
the
MSW
model.
First,
the
MSW
model
appears
to
be
more
sensitive
to
outliers
than
the
GLM
model
(1J).
Also,
the
hazard
func-
tion
for
the
MSW
model
involves
a
trunca-
tion
in
t
that
complicates
estimation.
Finally,
the
inclusion
of
the
power
parameter
k
(for
our
purposes,
k
=
2)
tends
to
give
the
fitted
model
a
relatively
sublinear
shape
that
leads,
Table
2.
Comparison
population
data,
1973-1986.
Sex,
All-Taiwan
Southwestern
region
age
Deaths
(n)
Deaths
(n)
(years)
PYR
Bladder
Lung
Liver
PYR
Bladder
Lung
Liver
Male
20-25
13,271,386
3
45
206
2,956,638
2
14
43
25-30
11,054,191
4
86
426
2,175,046
3
26
88
30-35
8,628,516
8
144
782
1,580,019
2
33
140
35-40
6,793,545
20
217
1,351
1,320,637
6
38
245
40-45
6,375,466
50
447
2,030
1,327,866
18
89
403
45-50
6,384,052
91
951
3,145
1,334,769
34
181
565
50-55
6,062,515
164
1,852
4,140
1,214,443
52
323
716
55-60
5,018,542
213
2,882
4,562
977,820
61
478
832
60-65
3,666,535
345
3,557
4,030
739,460
103
595
722
65-70
2,443,367
413
3,569
3,259
520,965
126
607
704
70-75
1,480,126
418
2,658
2,107
320,158
130
465
463
75-80
720,375
305
1,318
1,170
158,750
88
230 246
80-85
287,294
146
512
436
63,236
32
80
103
85+
105,411
66
152 188
22,651
15
22
33
Female
20-25
12,612,276
0
39
81
2,595,529
0
7
15
25-30
10,548,089
2
70
134
1,846,189
2
19
34
30-35
8,210,507
2
102
168
1,402,764
0
17
39
35-40
6,458,620
5
205
247
1,215,899
2
41
53
40-45
5,802,856
20
365 396
1,191,615
8
75
75
45-50
5,157,821
41
525 590
1,111,810
14
112
138
50-55
4,335,755
76
730
763
957,985
36
160
169
55-60
3,517,193
124
1,018
1,018
774,836
52
200
255
60-65
2,776,622
153
1,224
1,039
634,758
77
258 243
65-70
2,106,715
173
1,280
1,039
492,203
68
230 235
70-75
1,490,659
185
1,062
875
342,767
70
190
199
75-80
888,468
157
707
602
199,630
43
108
127
80-85
433,245
81
330
300
96,293
21
45
59
85+
217,590
41
136 153
46,089
9
10
31
PYR,
person-years.
Data
from
the
Department
of
Health
(18).
Table
3.
Poisson
modeling
options.
Comparison
population
Age
effect
hoat)
Dose
transformation
Dose
effect
g(x)
None
Linear
Linear Linear
exp(aO
+
x
X,
x=
ppba
PJx
Southwestern
Taiwan
Quadratic
Logarithmic
Quadratic
exp(ot,
+
act+
2f2)
x=log(l
+ppb)
l1Xx+l32
All
of
Taiwan
Regression
spline
Square
root
Exponential
linear
exp[xo
+
(xins(4lb
x
=ppb
exp(,l
x)
Exponential
quadratic
exp(p1x
+
02xk2)
&Represents
exposure
concentration
in
parts
per
billion,
which
is
equivalent
to
micrograms
per
liter.
bns~ft
represents
a
natural
spline
applied
to
t.
in
general,
to
higher
benchmark
doses
than
the
GLM
models.
Quantitative
risk
assessment.
Because
the
risk
of
dying
from
cancer
is
age
dependent,
it
is
common
to
base
risk
assessment
on
the
excess
risk
of
dying
from
cancer
over
the
course
of
a
typical
lifetime.
The
adjusted
life-
time
death
risk
can
be
calculated
by
integrat-
ing
the
death
hazard
over
the
typical
lifetime
in
the
population
of
interest,
Idr(x)
=
f1oh(xt)S(t)dt,
where
S(t)
is
the
probability
of
surviving
until
age
t
and
h(x,t)
is
the
hazard
for
dying
of
the
cancer
of
interest
at
age
t
for
someone
exposed
at
level
x.
Applying
integration
by
parts,
Idr(x)
can
also
be
written
as
Idr(x)
=
1-
JO,
k(t)S(t)dt,
where
X(t)
denotes
the
hazard
of
dying
at
age
t
from
causes
other
than
the
cancer
of
inter-
est.
This
function
can
be
approximated
by
ldr(x)
7,1
Sq,
exp[-5
Ih(x,
s)
j
t
S<t
where
Idenotes
the
sum
over
all
5-year
age
groups
in
the
study
and
qt
is
the
probability
of
dying
during
the
5-year
time
interval
indi-
cated
by
t.
Values
for
qt
were
taken
from
the
1996
U.S.
population
lifetable
for
males
and
females
(Table
4)
(20).
Traditionally,
standards
for
carcinogenic
compounds
have
been
set
by
finding
the
exposure
level
that
yields
a
rate
of
10-
over
background.
As
suggested
by
Brown
(9)
and
discussed
by
Smith
and
Sharp
(21),
this
esti-
mate
is
probably
unreliable
for
epidemiolog-
ic
data,
where
exposure
is
not
typically
mea-
sured
accurately
enough
to
extrapolate
to
such
low
risk
levels.
The
new
EPA
guidelines
for
cancer
risk
assessments
(22)
suggest
the
use
of
a
point-of-departure
analysis
for
set-
tings
where
the
mode
of
action
is
supportive
of
linearity
or
there
is
insufficient
support
for
a
nonlinear
mode
of
action.
The
idea
is
to
Table
4.
U.S.
death
probabilities
by
age
and
sex,
1996.
Probability
of
death
(q)
Age
(years)
Male
Female
20-25
0.00742
0.00239
25-30
0.00755
0.00307
30-35
0.00962
0.00423
35-40
0.01227
0.00595
40-45
0.01621
0.00834
45-50
0.02182 0.01224
50-55
0.03144
0.01938
55-60
0.04622
0.02938
60-65
0.06966
0.04577
65-70
0.09278
0.06417
70-75
0.12183
0.09207
75-80
0.14149 0.12267
80-85
0.15457
0.16036
85+
0.24949
0.41813
Data
from
Vital
Statistics
of
the
United
States,
1996120).
Environmental
Health
Perspectives
*
VOLUME
108
1
NUMBER
7
July
2000
657
Articles
*
Morales
et
al.
estimate
a
point
on
the
exposure
response
curve
within
the
observed
range
of
the
data
and
then
extrapolate
linearly
to
lower
doses.
The
dose
associated
with
10%
excess
risk
(EDIO)
is
the
standard
point
of
departure,
but
often
in
epidemiologic
studies,
an
excess
risk
of
10%
is
fairly
large
and
occurs
only
at
rela-
tively
high
doses.
We
will
use
both
1%
and
5%
excess
risks
for
the
point
of
departure
(EDO0
and
ED05,
respectively).
We
computed
confidence
intervals
for
excess
lifetime
risk
using
the
Delta
method
(23).
Bootstrap
methods
were
also
used
for
models
with
non-
parametric
age
effects,
yielding
similar
results
(24).
The
new
guidelines
also
suggest
a
mar-
gin-of-exposure
analysis
(MOE),
defined
as
the
point
of
departure
divided
by
the
environmental
exposure
of
interest.
This
approach
is
the
proposed
default
mode
of
action
when
linearity
is
not
the
most
reason-
able
assumption
(22).
For
subsequent
discus-
sion
we
will
use
MOE01(50)
to
represent
the
margin
of
exposure
using
the
EDO1
point
of
departure
and
50
1ig/L
as
the
environmental
exposure
of
interest.
Results
Table
5
contains
a
descriptive
summary
of
the
internal
cancer
data,
showing
person-years
at
risk,
observed
number
of
cancers,
and
the
SMRs
for
age,
sex,
and
exposure
grouped
into
the
same
intervals
used
by
the
EPA
in
the
skin
cancer
risk
assessment.
As
in
Wu
et
al.
(12),
the
analysis
is
limited
to
persons
.
20
years
of
age
because
there
were
essen-
tially
no
cancer
deaths
observed
in
those
younger
than
20
years
of
age.
Note
that
the
entire
Taiwanese
population
was
used
to
cal-
culate
the
expected
deaths
used
in
the
com-
putation
of
SMRs
in
Table
5.
Although
the
computed
SMRs
display
a
large
amount
of
noise,
there
appear
to
be
higher
SMRs
at
high
exposure
levels
compared
to
exposures
in
the
lower
range,
especially
for
bladder
and
lung
cancer.
There
is
no
observed
tendency
in
SMRs
with
respect
to
age,
which
suggests
no
age
dependency
on
the
risk
ratio,
g(x),
defined
in
Equation
1.
Overall,
females
have
higher
SMRs
than
males.
Liver
cancer
mor-
tality
is
generally
higher
than
expected,
although
there
is
no
particularly
strong
exposure-response
relationship.
The
GLM
analysis
began
by
fitting
all
possible
models
and
comparing
the
Akaike
information
criterion
(AIC)
to
narrow
the
model
choice
(25).
The
AIC
is
a
commonly
used
criterion
for
comparing
nonnested
models.
It
penalizes
the
model
deviance
by
Table
5.
Summary
statistics.
Observed
no.
of
cancer
deaths
(SMRa
x
100)
Villages
(n)
PYRb
Bladder
Lung
Liver
Combined
Overall
42
490,929
175
(1,327)
266
(266)
173
(134)
614
(254)
Exposure
(pg/L)
0-50
8
92,920
26
(1,002)
30
(156)
29
(118)
85
(183)
50-100
6
102,797
12
(415)
31
(143)
18
(65)
61
(116)
100-200
4
40,679
12(1,047)
21
(243)
19
(174)
52
(251)
20-300
3
36,514
8
(766)
24
(308)
14
(144)
46
(247)
300-400
4
28,870
6
(744)
12
(197)
6
(77)
24
(163)
400-500
3
28,655
22
(2,968)
21
(365)
12
(160)
55
(393)
500-600
7
79,446
34
(1,490)
56
(332)
34
(159)
124
(306)
600+
7
81,048
55
(3,271)
71
(514)
41
(217)
167
(486)
Age
(years)
20-40
42
258,789
2
(1,446)
5
(178)
18
(169)
25
(184)
40-60
42
164,549
21
(730)
116
(365)
76
(130)
213
(228)
60+
42
67,591
121
(1,189)
145
(222)
79
(133)
345
(256)
Sex
Male
42
256,970
85
(1,005)
144
(220)
122
(127)
351
(206)
Female
42
233,959
90
(1,904)
122
(354)
51
(158)
263
(368)
&Definition
from
Breslow
and
Day
(17).
b1973.1986.
adding
twice
the
number
of
parameters
to
it.
Thus
a
model
with
a
low
AIC
will
be
one
that
describes
the
observed
data
well
(a
low
deviance)
yet
with
relatively
few
parameters
(small
penalty).
The
models
with
the
lowest
AIC
provide
the
best
fit.
Because
it
was
not
appropriate
to
compare
AIC
for
the
models
based
on
different
data
sets,
we
provide
sepa-
rate
analyses
with
and
without
comparison
populations.
Table
6
identifies
models
1-9
and
the
MSW
model
that
appear
in
Tables
7-10
and
Figures
1-3.
For
simplicity
we
will
refer
to
the
model
numbers.
Table
7
compares
the
four
top
performing
models
based
on
AMC
for
the
models
with
and
without
comparison
populations
for
male
bladder
cancer.
Several
other
models
fit
reasonably
well,
but
we
chose
to
present
only
four
(see
"Discussion").
It
is
important
to
note,
however,
that
models
induding
exposure
concentration
were
highly
significant
compared
to
models
excluding
-concentration.
We
also
present
the
MSW
model.
Although
detailed
results
are
shown
here
only
for
male
bladder
cancer,
the
same
general
patterns
apply
to
females
and
to
all
cancer
outcomes
except
for
the
combined
analysis
(see
"Discussion").
In
general,
mod-
els
with
no
transformation
on
dose
and
an
exponential
linear
dose
effect
fit
well
when
we
used
no
comparison
population.
When
we
used
population
data
from
the
southwest-
ern
region
of
Taiwan
or
the
entire
Taiwanese
population,
models
with
the
square
root
and
log
transformation
fit
well.
This
is
most
likely
Table
6.
Model
description.
Model
1
2
3
4
5
6
7
8
9
MSW
Dose
transformation
Identity
Identity
Identity
Log
Log
Log
Sqrt
Sqrt
Sqrt
Identity
Dose
effecta
Linear
linear
Quadratic
Linear
Quadratic
Quadratic
Linear
Quadratic
Quadratic
Quadratic
Age
effect
Quadratic
Spline
Spline
Quadratic
Quadratic
Spline
Quadratic
Quadratic
Spline
Truncated
aExponential
linear
or
exponential
quadratic.
0.10
.m
0.08
L..
a
0.08
0.04
0
500
1,000
1,5W
2000
U.S.
equivalent
concentration
(igl/L)
Figure
1.
Estimated
lifetime
death
risk
for
male
bladder
cancer
without
comparison
population.
For
a
description
of
models,
see
Table
6.
0W
o
500
1,000
1,500
2,000
U.S.
equivalent
concentration
(jg/L)
U.S.
equuivalent
concentration
(jig/I)
Figure
3.
Estimated
lifetime
death
risk
for
male
Figure
2.
Estimated
lifetime
death
risk
for
male
blad-
bladder
cancer
with
southwestern
Taiwanese
der
cancer
with
Taiwanese-wide
comparison
pop-
region
comparison
population.
For
a
description
ulation.
For
a
description
of
models,
see
Table
6.
of
models,
see
Table
6.
VOLUME
108
1
NUMBER
71
July
2000
*
Environmental
Health
Perspectives
658
Articles
*
Internal
cancers
from
arsenic
in
drinking
water
due
to
the
relatively
low
cancer
death
rates
in
the
comparison
population.
The
log-transfor-
mation
allowed
the
fitted
curve
to
rise
more
quickly
from
zero
to
accommodate
this
dif-
ference.
Using
the
log-transformation
with-
out
the
comparison
population
gave
a
good
model
fit,
according
to
AIC,
but
risk
esti-
mates
were
not
easy
to
interpret
because
of
instability
of
the
fitted
model
at
low
dose.
For
this
reason,
we
chose
not
to
pursue
the
log-transformation
without
the
comparison
population
any
further.
A
few
additive
mod-
els
gave
a
good
fit,
but
in
most
cases,
the
multiplicative
models
did
a
better
job,
so
we
chose
not
to
continue
with
the
additive
mod-
els.
Also
note
that
the
MSW
model
fit
rea-
sonably
well
(Figures
1-3
show
graphical
representations).
Each
dot
in
Figures
1-3
corresponds
to
the
estimated
lifetime
risk
of
dying
of
bladder
cancer
for
villages,
grouped
by
50-pg/L
exposure
levels
(0-50,
50-100,
etc.).
The
grouping
is
for
presentation
pur-
poses
only
because
village-specific
estimates
were
highly
variable.
The
idea
of
grouping
for
the
purpose
of
graphical
presentation
of
a
fitted
model
has
been
widely
used
in
the
logistic
regression
context
as
well
(26).
Fitted
curves
for
the
models
without
the
compari-
son
population
are
very
similar
in
shape,
whereas
there
is
a
considerable
amount
of
variability
in
the
models
with
a
comparison
population.
Tables
8-10
contain
risk
statistics
for
the
best-fitting
GLM
models
and
the
MSW
model
with
and
without
comparison
popula-
tion
data.
Concentrations
are
reported
in
U.S.
equivalent
concentrations
of
arsenic
in
drinking
water,
based
on
conversions
that
account
for
the
average
weight
and
average
water
intake
for
a
male
living
in
the
United
States
compared
to a
male
living
in
Taiwan.
For
models
1
and
2,
which
have
no
transfor-
mation
on
dose,
EDOI
estimates
equal
395
and
351
pg/L,
respectively,
for
male
bladder
cancer.
For
models
3,
4,
and
5,
which
have
a
log-transformed
dose
effect,
EDO0
estimates
for
male
bladder
cancer
range
from
21
to
54
pg/L.
Models
7
and
8,
which
have
a
square
root
dose
effect,
give
higher
estimates
(156
and
108
pg/L,
respectively).
Results
for
model
9
are
similar
to
models
7
and
8.
When
Table
7.
AIC
for
best-fitting
models.
All
of
Southwestern
Model
None
Taiwan
area
1
302.1655
2
302.5547
3
334.8289
326.9948
4
302.9700
326.6287
5
330.0863
6
303.3353
330.9968
7
326.1207
8
327.1098
9
333.8307
a
comparison
population
is
used
(Tables
9
Estimates
of
ED
and
ED05
based
on
using
and
10),
there
is
more
variability
in
the
pre-
the
southwestern
region
of
Taiwan
tended
to
dicted
lifetime
risk
from
model
to
model.
It
be
much
lower
than
those
based
on
using
appears
that
the
inclusion
of
a
large
unex-
the
Taiwanese-wide
population.
The
MSW
posed
comparison
population
had
a
relative-
model
implies
a
lower
risk
when
no
compar-
ly
strong
influence
on
estimation
of
risk.
ison
population
was
used
(ED01
=
633
pg/L
Table
8.
Concentrations
(pg/I)
for
different
measures
of
risk
(without
comparison
population).
Bladder
Lung
Liver
Combined
Model
no."
M
F
M
F
M
F
M
F
lb
ED01
395
252
364
258
573
673
169
121
LED01
326
211
294
213
437
410
148
105
MOE01(50)
7.9
5.0
7.3
5.2
11.5
13.5
3.4
2.4
ED05
1,277
813
1,345
885
-c
-
720
493
LEDo5
1,076
690
1,086
733
-
-
629
430
MOE05(50)
25.54
16.3
26.9
17.7
--
14.4
9.9
2d
ED01
351
244
343
256
585
657
164
120
LED01
296
209
279
215
451
405
144
106
MOE01(50)
7.0
4.9
6.9
5.1
11.7
13.1
3.3
2.4
ED05
1,181
796
1,288
879
-
-
703 492
LED05
1,005
683
1,045
735
-
-
617
433
MOE05(50)
23.6
15.9
25.8
17.6
-
-
14.1
9.8
Mswe
ED01
633
365
227
396
864
824
163
267
MOE01(50)
12.7
7.3
4.5
7.9
17.3
16.5
3.3
5.3
ED05
1,439
828
1,171
898
-
-
706
605
MOE05(50)
28.8
16.6
23.4
18.0
-
-
14.1
12.1
aDose
transformation,
dose
effect,
and
age
effect,
respectively.
bidentity,
linear,
and
quadratic.
CED,5
outside
the
observ-
able
range
of
data.
'Identity,
linear,
and
spline.
'Identity,
quadratic,
and
truncated.
Table
9.
Concentrations
(pg/L)
for
different
measures
of
risk
(with
Taiwanese
comparison
population).
Bladder
Lung
Liver
Combined
Model
no.a
M
F
M
F
M
F
M
F
3b
ED01
22
21
11
8
254
331 3
2
LED01
18
17
8
6
54
63
3
2
MOE01(50)
0.4
0.4
0.2
0.2
5.1
6.6
0.1
0.0
ED05
504
330
1,145
448
-C
-
111
54
LED05
355
248
514
260
-
-
76
42
MOE05(50)
10.1
6.6
22.9
9.0
-
-
2.2
1.1
5d
ED01
23
19
11
8
239
339
3
2
LED01
19
16
8
6
51
65
3
2
MOE01(50)
0.5
0.4
0.2
0.2
4.8
6.8
0.1
0.0
ED05
539
304
1,276
476
--
113
56
LED05
380
231
564
274
-
-
77
43
MOE05(50)
10.8
6.1
25.5
9.5
-
-
2.3
1.1
6e
ED01
41
17
128
33
608
404
86
9
LED01
18
9
42
10
337
87
35
3
MOE01(50)
0.8
0.3
2.6
0.7
12.2
8.1
1.7
0.2
ED05
611
293
925
491
-
-
389
125
LED05
416
185
684
346
-
-
278
75
MOE05(50)
12.2
5.9
18.5
9.8
-
-
7.8
2.5
9f
ED01
100
72
76
68
895
511
45
17
LED01
65
52
32
34
542
148
19
10
MOE01(50)
2.0
1.4
1.5
1.4
17.9
10.2
0.9
0.3
ED05
708
407
978
579
-
-
499
228
LED
516
309
659
433
-
-
337
160
MOj5(50)
14.2
8.1
19.6
11.6
-
-
10.0
4.6
MSwg
ED01
164
88
196
116
480
551
106
53
MOE01(50)
3.3
1.8
3.9
2.3
9.6
11.0
2.1
1.1
ED05
852
455
1,014
579
1,089
-
544
273
MOE
5(50)
17.0
9.1
20.3
11.6
21.8
-
10.9
5.5
Dose
transformation,
dose
effect,
and
age
effect,
respectively.
hLog,
linear,
and
quadratic.
cED,5
outside
the
observable
range
of
data.
"Log,
linear,
and
spline.
"log,
quadratic,
and
spline.
fSqrt,
quadratic,
and
spline.
#Identity,
quadratic,
and
truncated.
Environmental
Health
Perspectives
*
VOLUME
1081
NUMBER
7
July
2000
MSW
302.0293
348.4275 334.3308
659
Articles
*
Morales
et
al.
Table
10.
Concentrations
(pg/L)
for
different
measures
of
risk
(southwestern
Taiwanese
comparison
pop-
ulation).
Bladder
Lung
Liver
Combined
Model
no.8
M
F
M
F
M
F
M
F
4b
ED01
21
19
10
10
119
467
32
LED01
17
16
8
8
37
76
2
2
MOE01(50)
0.4
0.4
0.2
0.2
2.4
9.3
0.1
0.0
ED05
649
452
768
522
-c
-
93
63
LED05
422 313
403
312
-
-
66
48
MOE05(50)
13.0
9.0
15.4 10.4
-
-
1.9
1.3
5d
ED01
54
25
76
27
503
455
62
9
LED01
21
12
22
9
247
110
22
3
MOE01(50)
1.1
0.5
1.5
0.5
10.1
9.1
1.2
0.2
ED05
723
464
780
520
-
-
330
132
LED05
508 315
558
362
__
226
79
MOE05(50)
14.5
9.3
15.6
10.4
-
-
6.6
2.6
7e
ED01
156
136
79
76
309
485
21
20
LED01
131
117
62 63
174
242
17
17
MOE01(50)
3.1
2.7
1.6
1.5
6.2
9.7
0.4 0.4
ED05
917
624
880
608
-
-
347
250
LED05
786
548
705
510
--
292
219
MOE05(50)
18.3
12.5
17.6 12.2
-
-
6.9
5.0
8f
ED01
108
85
50
63
779
559
31
18
LED01
65
56
25
35
400
168
14
10
MOE01(50)
2.2
1.7
1
1.3
15.6
11.2
0.6
0.4
ED05
817
536
778
582
-
-
416
238
LED05
594 406 489
431
--
275
167
MOE05(50)
16.3
10.7
15.6
11.6
-
-
8.3
4.8
MSW
EDoa
185
101
181
113
709
597
98
55
M8E01(50)
3.7
2.0
3.6
2.3
14.2
11.9
2.0
1.1
ED05
959
520 936
583
1,608
-
506
284
M0E05(50)
19.2
10.4
18.7
11.7
32.2
-
10.1
5.7
Dose
transformation,
dose
effect,
and
age
effect,
respectively.
bLog,
linear,
and
quadratic.
CED01
outside
the
observable
range
of
data.
dLog,
quadratic,
and
quadratic.
eSqrt,
linear,
and
quadratic.
'Sqrt,
quadratic,
and
quadratic.
for
male
bladder
cancer)
compared
to
esti-
mates
when
a
comparison
population
is
used
(164
and
185
jig/L).
Discussion
In
contrast
to
the
1988
EPA
risk
assessment
that
focused
on
skin
cancer
incidence
(4),
this
study
examines
cancer
mortality
in
a
set-
ting
where
exposure
is
measured
at
village
level.
Although
there
is
an
advantage
to
hav-
ing
individual
village
measurements,
there
also
appears
to
be
variability
in
the
exposure
assessment,
causing
high
variability
in
the
risk
estimates.
Depending
on
the
model
and
whether
or
not
a
comparison
population
is
used
in
the
analysis,
EDO1
estimates
range
in
value
from
21
to
633
jig/L
for
male
bladder
cancer.
For
males,
the
lung
cancer
risk
tends
to
be
slightly
higher
than
the
risk
for
bladder
cancer,
with
EDOI
values
ranging
from
10
to
364
pg/L.
Although
this
result
seems
in
con-
trast
to
the
high
SMRs
for
bladder
cancer
in
Table
5,
the
risk
estimates
are
calculated
on
an
additive
scale
and
are
influenced
by
back-
ground
cancer
rates.
Hence,
even
though
bladder
cancer
has
high
SMRs,
the
number
of
excess
bladder
cancer
deaths
associated
with
exposure
is
only
moderate
because
of
the
low
bladder
cancer
death
rate
in
the
gen-
eral
population.
In
contrast,
because
lung
cancer
is
more
prevalent
in
the
general
pop-
ulation,
even
a
moderate
SMR
can
lead
to
high
numbers
of
excess
deaths.
There
does
not
appear
to
be
high
risk
associated
with
liver
cancer
in
males
with
the
exception
of
estimates
based
on
three
models
that
used
a
log-transformation
of
exposure
(models
3,
4,
and
5).
EDOI
estimates
range
from
309
to
895
pg/L
for
models
apart
from
the
latter,
which
yields
values
that
range
from
199
to
254
pg/L.
The
risk
associated
with
female
cancers
tends
to
be
higher
than
that
of
males
for
each
cancer
type.
For
bladder
cancer,
EDOI
estimates
for
females
range
from
17
to
365
gg/L.
For
lung
and
liver
cancer,
female
EDOI
estimates
range
from
8
to
396
pg/L
and
331
to
824
gg/L,
respectively.
The
best
models
according
to
AIC
for
bladder,
lung,
and
liver
cancer
combined
did
not
exactly
correspond
to
the
models
presented
in
Tables
8-10.
For
males,
the
best
model
with
no
comparison
population
is
model
1,
which
has
a
linear
untransformed
dose
effect
and
a
quadratic
age
effect
(Table
8).
For
females
the
best
model
for
combined
cancer
has
a
square
root
transformation
on
dose
with
a
quadratic
dose
and
age
effect.
The
EDO1
esti-
mate
based
on
this
model
equals
844
pg/L.
When
a
comparison
population
is
used
(either
all
of
Taiwan
or
the
southwestern
region
of
Taiwan),
the
best
model
for
both
males
and
females
has
a
square
root
transfor-
mation
on
dose
with
a
linear
dose
effect
and
spline
age
effect.
EDOI
estimates
based
on
this
model
with
the
entire
Taiwanese
population
equal
22
and
18
pg/L
for
males
and
females,
respectively.
When
the
southwestern
region
was
used,
EDOI
estimates
equal
21
and
20
pg/L
for
males
and
females,
respectively.
Our
results
show
that
exposure-response
assessments
depend
highly
on
the
choice
of
model,
as
well
as
whether
or
not
a
compari-
son
population
is
used
in
the
analysis.
As
dis-
cussed
by
Morales
et
al.
(10),
one
possible
explanation
is
the
uncertainty
associated
with
an
ecologic
study
design.
We
assumed
the
same
arsenic
concentration
for
all
per-
sons
in
the
same
village
and
individual
expo-
sures
can
vary
widely
in
a
village.
Mortality
records
are
available
for
individuals,
but
their
individual
exposures
are
not.
The
National
Academy
of
Sciences
(1)
provides
a
good
dis-
cussion
on
this
subject.
Although
one
might
argue
that
the
appropriate
strategy
would
be
to
select
the
best
model
based
on
accepted
statistical
cri-
teria,
several
models
gave
essentially
the
same
quality
fit
(as
measured
by
AIC),
yet
yielded
substantial
differences
in
risk
estimates.
For
example,
for
the
models
without
a
compari-
son
population,
the
MSW
model
gave
a
fit
comparable
to
some
of
the
GLM
models,
but
produced
EDOI
estimates
almost
twice
as
high.
Despite
the
comparably
good
fit,
we
preferred
the
GLM
models
to
the
MSW
model.
For
example,
sensitivity
analysis
revealed
that
the
MSW
model
was
influ-
enced
strongly
by
the
removal
of
various
subsets
of
villages,
whereas
the
GLM
was
not
(10).
The
poor
nutritional
status
of
the
Taiwanese
in
the
Blackfoot
disease
region
could
be
another
contributing
factor
of
uncertainty.
We
could
not
account
for
dietary
intake
of
inorganic
arsenic
in
food
for
either
population,
or
for
other
con-
founders
in
this
analysis.
Differences
in
EDOI
estimates
were
par-
ticularly
affected
by
whether
or
not
a
com-
parison
population
was
used.
There
is
reason
to
believe
that
the
urban
Taiwanese
popula-
tion
is
not
a
comparable
population
for
the
poor
rural
population
used
in
this
study.
Thus,
risk
estimates
using
the
Taiwanese
population
may
be
biased.
As
an
alternative,
we
used
the
southwestern
region
of
Taiwan;
we
found
very
different
risk
estimates
based
on
the
two
different
comparison
populations
(Tables
9
and
10).
We
could
have
done
other
analyses.
For
example,
we
could
have
calculat-
ed
lifetime
death
rates
for
the
unexposed
VOLUME
108
1
NUMBER
7
1
July
2000
*
Environmental
Health
Perspectives
660
Articles
*
Internal
cancers
from
arsenic
in
drinking
water
group
[Idr(0)]
using
U.S.
population
data.
It
would
be of
interest
to
see
how
the
unex-
posed
death
rates
in
the
United
States
com-
pare
to
the
death
rates
in
Taiwan.
Despite
the
considerable
variation
in
estimated
EDO0,
the
results
are
sobering
and
indicate
that
current
standards
are
not
ade-
quately
protective
against
cancer.
For
the
combined
analysis
with
no
comparison
pop-
ulation
and
identity
transformation
on
dose,
the
MOE01(50)
values
range
from
0.4
to
16.9
for
both
males
and
females.
When
we
include
a
comparison
population,
the
MOE01(50)
values
range
from
0.2
to
3.4.
The
current
arsenic
standard
of
50
j.g/L
(4)
is
actually
below
the
estimated
EDO0,
which
suggests
that
the
risk
at
the
current
standard
is
higher
than
1
in
100.
Note,
however,
this
estimate
is
likely
to
be
overly
conservative
because
the
data
suggest
that
the
log-trans-
formations
lead
to
somewhat
unstable
results.
Even
considering
the
identity
trans-
formation,
which
tended
to
give
less
extreme
results,
the
risk
associated
with
a
concentra-
tion
of
50
pg/L
is
approximately
1
in
300,
based
on
linear
extrapolation
from
the
point
of
departure.
Risks
of
a
similar
magnitude
were
reported
by
Smith
et
al.
(27).
This
is
an
extremely
high
value.
We
could
argue
that
if
indeed
the
risk
were
this
high,
we
would
expect
to
find
epidemiologic
evidence
even
within
the
U.S.
population.
The
SEER
Cancer
Statistics
Review
(28)
estimates
that
the
age-adjusted
U.S.
mortality
rates
for
bladder,
lung,
and
liver
cancer
are
3.2,
49.5,
and
2.8
per
100,000,
respectively.
It
is
also
estimated
that
approximately
5%
of
large
and
small
regulated
water
supply
systems
in
the
United
States
have
arsenic
concentra-
tions
>
20
)ig/L
(291).
Thus,
if
the
excess
can-
cer
risk
associated
with
50
pg/L
arsenic
is
on
the
order
of
magnitude
1
in
1,000,
we
would
expect
an
increase
of
approximately
0.05
per
1,000
or
5
per
100,000
in
the
pop-
ulation.
It
is
not
surprising
that
epidemio-
logic
studies
in
the
United
States
have
not
so
far
been
able
to
identify
clear
associations.
Thus,
we
conclude
that
arsenic in
drinking
water
may
indeed
be
contributing
to
excess
cancer
mortality
in
the
United
States.
REFERENCES
AND
NOTES
1.
National
Research
Council.
Arsenic
in
Drinking
Water.
Washington,
DC:National
Academy
Press,
1999.
2.
U.S.
EPA.
Health
Assessment
Document
for
Inorganic
Arsenic.
EPA
600/8-83/021F.
Cincinnati,
OH:U.S.
Environmental
Protection
Agency,
1984.
3.
Tseng
WP,
Chu
HM,
How
SW,
Fong
JM,
Lin
CS,
Yeh
S.
Prevalence
of
skin
cancer
in
an
endemic
area
of
chronic
arsenicism
in
Taiwan.
J
Natl
Cancer
lnst
40:453-463
(1968).
4.
U.S.
EPA.
Special
Report
of
Inorganic
Arsenic:
Skin
Cancer;
Nutritional
Essentiality.
EPA
625/3-87/013.
Washington,
DC:U.S.
Environmental
Protection
Agency,
1988.
5.
Safe
Drinking
Water
Act
Amendments.
42
U.S.C.
300g-1
(b)
(12)
(A)
(iv),
1996.
6.
Smith
AH,
Goycolea
M,
Haque
R,
Biggs
ML.
Marked
increase
in
bladder
and
lung
cancer
mortality
in
a
region
of
northern
Chile
due
to
arsenic
in
drinking
water.
Am
J
Epidemiol
147:660-669
(1998).
7.
Hopenhayn-Rich
C,
Biggs
ML,
Fuchs
A,
Bergoglio
R,
Tello
EE,
Nicolli
H,
Smith
AH.
Bladder
cancer
mortality
associated
with
arsenic
in
drinking
water
in
Argentina.
Epidemiology
7:117-124
(1996).
8.
Chen
CJ,
Chuang
YC,
Lin
TM,
Wu
HY.
Malignant
neo-
plasms
among
residents
of
a
Blackfoot
disease-endemic
area
in
Taiwan:
high-arsenic
artesian
well
water
and
cancers.
Cancer
Res
45:5895-5899
(1985).
9.
Brown
KG.
Assessing
risk
of
inorganic
arsenic
in
drink-
ing
water
in
the
United
States.
Hum
Ecol
Risk
Assess
4:1061-1070
(1998).
10.
Morales
KH,
Ryan
L,
Brown
KG,
Kuo
TL,
Chen
CJ,
Wu
MM.
Model
sensitivity
in
an
analysis
of
arsenic
exposure
and
bladder
cancer
in
southwestern
Taiwan.
In:
Proceedings
of
the
Third
International
Conference
on
Arsenic
Exposure
and
Health
Effects,
12-15
July
1998,
San
Diego,
California.
New
York:Elsevier,
1999;207-215.
11.
Chen
CJ,
Kuo
TL,
Wu
MM.
Arsenic
and
cancers
[Letter].
Lancet
ii:414-415
(1988).
12.
Wu
MM,
Kuo
TL,
Hwang
YH,
Chen
CJ.
Dose-response
relation
between
arsenic
concentration
in
well
water
and
mortality
from
cancers
and
vascular
diseases.
Am
J
Epidemiol
130:
1123-1132(1989).
13.
Chen
CJ,
Chen
CW,
Wu
MM,
Kuo
TL.
Cancer
potential
in
liver,
lung,
bladder
and
kidney
due
to
ingested
inorganic
arsenic
in
drinking
water.
Br
J
Cancer
66:888-892
(1992).
14.
WHO.
International
Classification
of
Diseases:
Manual
of
the
International
Statistical
Classification
of
Diseases,
Injuries,
and
Causes
of
Death,
1965
Revision,
Vol
1
and
2.
Geneva:World
Health
Organization,
1967.
15.
Chen
KP,
Wu
HY,
Wu
TC.
Epidemiologic
studies
on
Blackfoot
disease
in
Taiwan.
3:
physicochemical
charac-
teristics
of
drinking
water
in
endemic
Blackfoot
disease
areas.
Mem
Coil
Med
NatI
Taiwan
Univ
8:115-129
(1962).
16.
Yeh
SJ,
Yang
MH.
Preliminary
report
on
trace
element
content
in
water
samples
from
the
Blackfoot
disease
endemic
area
and
in
biopsy
samples
from
Blackfoot
dis-
ease
patients.
In:
Report
on
Blackfoot
Disease
Research,
Vol.
8.
Taichung,
Taiwan:Taiwan
Provincial
Department
of
Health,
1980;22-28.
17.
Breslow
NE,
Day
NE.
Statistical
Methods
in
Cancer
Research.
Vol
II:
The
Design
and
Analysis
of
Cohort
Studies.
New
York:Oxford
University
Press,
1987.
18.
Department
of
Health.
Health
Statistics.
Vol
Il:
Vital
Statistics,
1973-1986.
Taipei,
Taiwan:Department
of
Health,
1974-1987.
19.
Venebles
WN,
Ripey
BD.
Modern
Applied
Statistics
with
S-PLUS.
New
York:Springer-Verlag,
1994.
20.
Vital
Statistics
of
the
United
States,
1996.
Hyattsville,
MD:National
Center
for
Health
Statistics,
1996.
21.
Smith
AH,
Sharp
DS.
A
standardized
benchmark
approach
to
the
use
of
cancer
epidemiology
data
for
risk
assessment.
Toxicol
Ind
Health
1:205-212
(1985).
22.
U.S.
Environmental
Protection
Agency.
Proposed
guide-
lines
for
carcinogen
risk
assessment;
notice.
Fed
Reg
61:17959-18011
(1996).
23.
Stuart
A,
Ord
K.
Kendall's
Advanced
Theory
of
Statistics.
Vol
1:
Distribution
Theory.
London:Edward
Arnold,
1994.
24.
Efron
B,
Tibshirani
RJ.
An
Introduction
to
the
Bootstrap.
New
York:Chapman
&
Hall,
1993.
25.
Hastie
TJ,
Tibshirani
RJ.
Generalized
Additive
Models.
New
York:Chapman
&
Hall,
1990.
26.
Hosmer
DW,
Lemeshow
S.
Applied
Logistic
Regression.
New
York:John
Wiley
&
Sons,
1989.
27.
Smith
AH,
Hopenhayn-Rich
C,
Bates
MN,
Goeden
HM,
Hertz-Picciotto
I,
Duggan
HM,
Wood
R,
Kosnett
MJ,
Smith
MT.
Cancer
risks
from
arsenic
in
drinking
water.
Environ
Health
Perspect
97:259-267
(1992).
28.
Ries
LAG,
Kosary
CL,
Hankey
BF,
Miller
BA,
Clegg
L,
Edwards
BK,
eds.
SEER
Cancer
Statistics
Review,
1973-1996.
Bethesda,
MD:National
Cancer
Institute,
1999.
29.
Welch
AH,
Helsel
DR,
Focazio
MJ,
Watkins
SA.
Arsenic
in
ground
water
supplies
of
the
United
States.
In:
Proceedings
of
the
Third
International
Conference
on
Arsenic
Exposure
and
Health
Effects,
12-15
July
1998,
San
Diego,
CA.
New
York:Elsevier,
1999;9-17.
*LUNING
MORE
MAD
COWSJ:
The
Myth
ot
fCt4cy
Toxic
Waste
So
you'd
better
know
who
you
can
trust.
Environmental
Health
Perspectives
draws
on
more
than
25
years
of
scientific
publishing,
backed
by
the
authority
of
the
National
Institutes
of
Health-National
Institute
of
Environmental
Health
Sciences,
to
provide
you
with
the
most
current
and
credible
information
available
to
help
you
make
decisions
about
your
health
and
the
environment.
Each
month
we
delve
into
the
issues
you
care
about
such
as:
water,
air,
and
soil
pollution
urban
sprawl
population
and
world
health
toxic
pesticides
environmental
products
and
services
Call
today
to
subscribe
and
visit
us
online
at
S
-
S
Environmental
Health
Perspectives
*
VOLUME
108
1
NUMBER
7
1
July
2000
661
