Arsenic - Toxicology of Metals - page 371

is generally accepted as the most reliable indicator of
recent arsenic exposure, and this approach has proved
useful in identifying above-average exposures in pop-
ulations living near industrial point sources of arsenic
(Polissar et al., 1990). Normal levels of arsenic in the
urine of people with no known high exposure are
apparently in the range of 5-50

determined such values by means of hydride genera-
tion combined with AAS or AAF and reported a mean
value of approximately 7

µ g/g creatinine and a 95%

cut off of 18 g/g creatinine in a group from the gen-

eral population of Belgium. In a population group from
China not known to be exposed to inorganic arsenic,
there was a mean of approximately 25

µ

g As/L (Baker et al .,
1977; Bencko and Symon, 1977; Braman and Foreback,
1973; Buchet et al ., 1980; Smith et al ., 1977). Ingestion
of seafood may, however, increase the concentrations
to more than 1 mg As/L (Buchet et al ., 1980; Norin and
Vahter, 1981; Pinto et al ., 1976; Schrenk and Schreibeis,
1958).
In a study of As 2 O 3 exposure in a smelter, Pinto et
al . (1976) found an average urinary arsenic concentra-
tion of 0.053 mg/L among 200 workers who had not
been exposed to arsenic. The average arsenic concen-
tration in urine of 24 men who had been exposed to a
mean air concentration of 53

µ µ g/g creatinine

and a 95% cut off of 50 µ g/g creatinine (Buchet et al .,

2003).

6.3 Blood
By using NAA, Brune et al . (1966) found a mean
concentration of 0.004 mg/kg in whole blood of nor-
mal people and 0.035 mg/kg in uremic patients. Hey-
dorn (1970) reported a mean arsenic concentration in
whole blood of 0.022 mg/L in Taiwanese subjects. He
also used NAA.
In three studies of As-exposed populations, Vahter
et al . (1995), Concha et al . (1998a), and Concha et al .
(1998b) found the average blood total As to be 7.6, 9,
and 10

g/m 3 )
in the same factory ranged from 0.038-0.539 mg/L,
with an overall average of 0.174 mg/L. They found a
correlation between airborne arsenic concentrations and
arsenic in urine, but the values showed a wide scatter.
Watrous and McCaughey (1945) determined urinary
levels of arsenic among workers in a factory producing
arsphenamines from arsanilic acid. The mean arsenic
level was 0.44 mg/L with a range of 0.04-3.8 mg/L.
Calderon et al . (1999) found a correlation between
the log of the mean total urinary arsenic concentra-
tion/creatinine (TAs/c,

µ g/m 3 (range, 3-295 µ

g/L, respectively. In one of these studies, the
range of blood As concentration of an unexposed pop-
ulation was 1-2

µ

g/L (Concha et al ., 1998b). By use of
HPLC-ICP-MS, Suzuki et al . (2002) and Mandal et al .
(2004) were able to separate and quantify the As spe-
cies in the blood of an As-exposed population. They
found AsB, DMA, MMA, and inorganic As in the blood
samples by speciating the blood. They also found the
total As concentration of 10.1

µ

g/mg) in people living
in areas with arsenic-contaminated drinking water
sources and the log of the inorganic arsenic concentra-
tion in the drinking water (lnAs,

µ

g/L As in blood from an
As-exposed individual. Mandal et al . (2004) also sepa-
rated the blood into red blood cells (RBCs) and plasma
and determined the fraction of the various species in
these blood components.
As stated in Section 5.2, the clearance of arsenic from
blood is very rapid. The time lapse between exposure
and sampling will, therefore, be of importance if the
blood levels are being related to the exposure. Fur-
thermore, intake of seafood may greatly infl uence the
blood arsenic levels. For these reasons, blood arsenic
cannot be regarded as a useful indicator of exposure.
A review article by Taylor et al . (2004) tabulates both
the technologies used for the determination of As in
blood, hair, nails, and urine and summarizes the basic
fi ndings for the different biological matrices.

µ

g/L).
Arsenic concentration in urine may be used as an
index of exposure, but a number of factors such as a
diet containing mainly seafood and the time between
exposure and urine sampling have to be considered.
Because “fi sh arsenic” (arsenobetaine) is essentially
nontoxic, total urinary arsenic content may overesti-
mate exposures to arsenic species that are of health con-
cern. Experience has shown that total urinary arsenic
may lead to overestimation of exposure to inorganic
arsenic even if the subjects are requested to refrain
from eating seafood (Pinto et al ., 1976). Dietary habits
should be examined thoroughly when the total urinary
arsenic concentration is to be used as an index of expo-
sure. Analytical differentiation of the different forms
of arsenic in the urine is an appropriate way of deter-
mining how much and what form of arsenic has been
absorbed (Buchet et al ., 1980; Norin and Vahter, 1981).
The sum of urinary excretion of inorganic arsenic and
its monomethylated and dimethylated metabolites is a
good indicator of exposure to inorganic arsenic when
there is some fi sh consumption. Buchet et al . (2003)

µ

6.4 Hair
Attempts have been made to correlate normal con-
centrations of arsenic in hair to exposure to inorganic
arsenic. Smith (1964) found that 80% of 1000 peo-
ple tested had a concentration below 1 mg As/kg in
hair, with an average of 0.81 mg/kg and a median
of 0.51 mg/kg. Liebscher and Smith (1968) gave