Lead - Toxicology of Metals - page 614

However, in connection with a decrease of mean B-Pb
from 1.02

covariates. However, still overadjustment and/or
underadjustment, or incorrect inference, may easily
occur. Also, the limits of the precision of analytical and
psychometric measurements increase the uncertainty
of any estimate of effect, especially at B-Pb < 100-
150

µ mol/L (212 µ g/L) at age 2, to 0.38 µ mol/L

(79 g/L) at ages 11-13, there was no signifi cant rela-

tionship between the individual B-Pb decrease and the
IQ change. The data may indicate a long-lasting effect
of damage caused in the fetus and/or infant.
In the Yugoslavia Prospective Lead Study, IQ at 3,
4, 5, 7, or 10-12 years in 390 children to women living
near a smelter (mean, B-Pb 309

µ

g/L. Hence, if a threshold exists, it is unlikely to
be detected (U.S. CDC, 2005). Likewise, if there is no
threshold, that is also very diffi cult to prove.
There is limited information indicating a gene-
environment interaction. Hence, children with the ALAD 2
genotype had low dentine lead and consistently better
performance compared with children with ALAD 1 ,
although the effects were small (Bellinger et al ., 1994b).
Lead-associated declines in mental development
in children living in contaminated housing were not
reversed by chelation (Liu et al ., 2002; Rogan et al .,
2002). However, there are several caveats regarding
these results.
In the U.S., Attention Defi cit Hyperactivity Disorder
(ADHD) in children (age 4-15) was associated with B-Pb
(Braun et al ., 2007). Further, in children (age 11) from Pitts-
burgh, Pennsylvania, there was a slight, but statistically
signifi cant, association between tibia-lead and delin-
quent behavior (Needleman et al ., 1996). Furthermore,
arrested and adjudicated youths had higher bone lead
than controls (Needleman et al ., 2002). It may be men-
tioned, that in the United States, there is an ecological
relationship (at county level) between B-Pb and homicide
rate (Stretesky and Lynch, 2001). Needless to say, such
studies include large problems of residual confounding.
In summary, retardation of neurobehavioral devel-
opment and growth, as well as electrophysiological
and hearing changes, have in a long series of studies of
infants and children been shown to be associated with
B-Pb in the mother, newborn, infant, and child. It is not
known which period is most critical. For example, the
in utero exposure is often related to postnatal uptake
of lead. Despite the many methodological problems in
such studies, there were indications that effects may
occur even at B-Pbs in the pregnant woman and infant
of only 0.5

µ

µ g/L at age 10-12) and

a control (B-Pb, 61 g/L) town displayed a negative

association with prenatal and postnatal B-Pb (Wasser-
man et al ., 2003). A doubling of the average lifetime
B-Pb was associated with a decrease of full-scale IQ by
1.6 points. The association with tibia lead at age 10-12
was closer than with B-Pb.
In the Mexico City Prospective Lead Study, 112 chil-
dren were followed to 3-5 years of age (Gomaa et al .,
2002). B-Pb (geometric mean, approximately 100

µ

g/L)
measured in cord blood was negatively associated
with adjusted intellectual function, whereas postnatal
B-Pb was not.
In an important study of 172 children in Roches-
ter, New York, in which B-Pb was measured serially
between the ages 6 months and 5 years, 101 did not
have a recorded B-Pb >100

µ

g/L (Canfi eld et al ., 2003;
additional information in Koller et al ., 2004). Strong
and signifi cant negative associations between B-Pb and
adjusted IQ (0.74 points per 10

µ

g/L) were observed
at 3 and 5 years. There were indications that the effect
was larger at lower B-Pb than at higher ones.
In a thorough review of some of the recent prospec-
tive studies of children, it was concluded that even
at B-Pb consistently <100

µ

g/L, there was intellectual
impairment (Koller et al ., 2004). Some of the reviewed
studies indicate a supralinear exposure-response rela-
tionship (i.e., the loss of IQ at a B-Pb change from 0 to
100

µ

g/L would be larger than at a change from 100
to 200

µ

g/L) (Lanphear et al ., 2005). Although this is
under debate, it undermines the idea of a threshold.
However, lead exposure was believed to account for
only a small fraction (1-4%) of the total variation in
cognitive ability while social and parenting factors
account for 40% or more (Koller et al ., 2004).
Critics have questioned the importance of such small
decrements of IQs of individual children. However, the
tests are blunt instruments for detecting subtle changes
in brain function. Also, on a population basis, the effect
is considerable. Furthermore, even small changes of IQ
in large numbers of children will dramatically increase
the proportion of children with low (e.g., < 80) and
decrease the proportions with high (> 120) IQ.
It should be stressed that the confounding and effect-
modifi cation issues are complicated (Bellinger, 2000).
All the aforementioned studies adjust for different

µ

mol/L, perhaps even lower (Table 3; Sker-
fving, 2005). WHO (2000b) estimated that there was a
net decrease of 3.4 (95% confi dence interval 1.1-5.0) IQ
points at 150

µ

g/L. Although the disablement in the
individual is small, as is the fraction of the total vari-
ance of the CNS function explained by lead, the effect
is defi nitely adverse. The reversibility of the effects
is not adequately known but they seem to be at least
partly irreversible.
In an attempt to calculate the global burden of dis-
ease caused by lead-induced mild mental retardation,
the estimate was 9.8 million disability-adjusted life years
(DALYs), mostly in the West Pacifi c, Southeast Asia, and
Central and South America regions (Fewtrell et al ., 2004).

µ