Matern Child Health J (2012) 16:1518–1524
DOI 10.1007/s10995-011-0917-3
Relationship Between Prenatal Lead Exposure and Infant Blood
Lead Levels
Natalie P. Archer • Carrie M. Bradford •
David M. Klein • Jim Barnes • L. J. Smith •
John F. Villanacci
Published online: 8 December 2011
Ó Springer Science+Business Media, LLC 2011
Abstract Recent literature has shown that analyzing
newborn dried blood spots (DBS) may be effective in
assessing some prenatal environmental exposures, such as
exposure to lead. The purpose of this study was to evaluate
the relationship between prenatal exposure to lead (as
measured by newborn DBS results) and blood lead levels
(BLLs) in infants 6 months of age or younger, using public
health registry data for infants born in Texas from July 2002
through July 2006. The Texas Child Lead Registry (TCLR)
was used to identify infants with documented elevated
BLLs of 10 lg/dL or higher as well as infants with docu-
mented low BLLs. BLLs for these children were compared
to their corresponding newborn DBS results using Pearson
correlation coefficients and exact logistic regression mod-
els. Overall, a significant but weak positive correlation was
found between infant BLLs and corresponding newborn
DBS lead levels (r = 0.48). However, the odds of an infant
with an elevated newborn DBS lead level having an ele-
vated BLL at 6 months of age or younger were much
greater than for an infant with a low newborn DBS lead
level of \5 lg/dL (adjusted odds ratio 27.95, 95% CI:
5.52–277.28). Although an association was observed
between newborn DBS lead levels and BLLs in infants
tested between 0 to 6 months of age, our findings suggest
N. P. Archer (&)
C. M. Bradford
L. J. Smith

J. F. Villanacci
Environmental Epidemiology and Disease Registries Section,
Texas Department of State Health Services, P.O. Box 149347,
Austin, TX 78714-9347, USA
e-mail: natalie.archer@dshs.state.tx.us
D. M. Klein
J. Barnes
Chemical Threat Laboratory, Laboratory Services Section,
Texas Department of State Health Services, P.O. Box 149347,
Austin, TX 78714-9347, USA
123
that prenatal exposure may not be the only significant
source of lead exposure for infants B6 months of age.
Keywords Dried blood spots
Newborn screening

Prenatal lead exposure
Blood lead levels

Childhood lead poisoning
Introduction
Reducing childhood exposure to lead is a top environ-
mental public health priority among state and federal
agencies. Young children exposed to lead can suffer
damage to the brain and other organs and can experience
behavioral problems, developmental delays, and learning
disabilities [1–4].
The risk of exposure to lead in the environment
increases as children become more mobile and exhibit
more hand-to-mouth behaviors [5, 6], resulting in blood
lead levels (BLLs) peaking around 2 years of age [6–8].
Consistent with this age-dependent risk, the American
Academy of Pediatrics recommends children be tested at 1
and 2 years of age [9], and the Texas Department of State
Health Services (DSHS) recommends children at higher
risk of exposure be tested as early as 6 months of age [10].
In Texas, all child blood lead results are required by law to
be reported to DSHS.
Some Texas children below ages included in the rec-
ommendations occasionally receive blood lead tests. From
July 2002 through July 2006 Texas received 10,325 blood
lead reports for children less than or exactly 6 months of
age, 85 of which were considered elevated (BLL C 10
lg/dL) [11]. The sources of lead for these exposures are not
known; however, limited mobility at these ages would limit
behaviorally associated environmental exposures.
Matern Child Health J (2012) 16:1518–1524
Recent advances in analytical methods make it possible
to evaluate prenatal exposures to environmental contami-
nants, such as lead, using newborn dried blood spots (DBS)
[12–14]. The purpose of this study was to evaluate the
relationship between prenatal exposure to lead (as mea-
sured by newborn DBS results) and BLLs in infants exactly
6 months of age or younger, using public health registry
data for infants born in Texas from July 2002 through July
2006. This study was approved by the DSHS Institutional
Review Board.
Methods
Selection Criteria
All infant records were selected from the Texas Child Lead
Registry (TCLR), a legislatively-mandated registry main-
tained by the DSHS. The TCLR contains blood lead results
for children who have received a blood lead test in Texas
since 1996 [15].
Infants from the TCLR who were born in Texas from
July 5, 2002 through July 31, 2006 and whose BLLs were
tested at exactly 6 months of age or younger were eligible
for inclusion. Storage of newborn DBS in Texas began in
July 2002, thus DBS for children born on or after July 5,
2002 were the only specimens available for this study.
When the study was initiated, data required to link TCLR
infant records to the corresponding newborn DBS were
available through the end of 2006. To analyze 4 years of
data, the timeframe of July 2002 through July 2006 was
chosen for this study. Infants with BLL tests taken on or
before the day they turned 6 months of age were of interest
based on the hypothesis that at these young ages, lead
exposure would be more likely attributed to prenatal
exposures than environmental exposures.
Of those infants who met the study criteria, the study
population consisted of all children with elevated BLLs
(C10 lg/dL) and a subset of those with very low BLL
(B2 lg/dL). If a child was reported as having an elevated
BLL more than one time during the first 6 months of life,
the elevated BLL corresponding with the youngest age was
used in the analyses. Four infants with very low reported
BLLs were randomly sampled for each infant identified
with elevated BLLs (1:4 ratio). Infants identified with very
low BLLs were never reported as having a BLL of more
than 2 lg/dL in the first 6 months of life.
Data Linkage
All study infants’ information was obtained from pre-
existing public health data from the TCLR, birth certificate
records, and newborn screening (NBS) data, including
1519
infant DBS specimens. Data on all infants identified from
the TCLR first were linked to birth certificate information
to obtain more complete identifiers for associating the
study infants with their correct DBS specimens and to
obtain additional information on possible confounders or
effect modifiers. The linked TCLR—birth certificate data
set was linked to the NBS data records to identify corre-
sponding DBS specimens. As there was no unique identi-
fying field common to all datasets, both deterministic and
probabilistic linkage methods were performed, using
Microsoft Access 2007 and Link King, version 5.0.
The DSHS Chemical Threat Laboratory (CT Lab)
obtained a dried blood spot sample and a blank sample for
each DBS specimen by punching a 3/16-inch diameter
circle from the blood spot and white area surrounding the
spot, respectively.
Dried Blood Spot Laboratory Analysis
The method used to analyze lead in newborn DBS was
adapted from the Kansas Department of Health Laboratory
[16]. This method involves extracting the lead into a
solution containing acid and an internal standard (terbium)
followed by analysis with inductively coupled plasma-mass
spectrometry (ICP-MS). ICP-MS is used in the CT Lab to
analyze blood for mercury, lead, and cadmium and to
analyze urine for 14 toxic metals using the Centers for
Disease Control and Prevention (CDC) method.
Calibration curves were prepared in acidic solutions and
demonstrated linearity of the method from 0.05 to 1 lg/dL.
The dilution factor of the blood spots when analyzed then
represents a lead concentration of near 0 to 300 lg/dL in
the original spot. Over this range the coefficient of deter-
mination (R2) values for the calibration curves were
0.9999, with 1.0 representing perfect linearity.
Quality control (QC) samples were prepared at three
levels covering the range of the method (5, 20, and 50 lg/
dL), and were run with each set of specimens. QC samples
were prepared by spiking sheep’s blood with aqueous
solutions containing known amounts of lead. These spiked
blood pools were spotted onto blood collection cards pro-
vided by the newborn screening laboratory, which were
then punched and interspersed with the DBS specimens
during analysis as quality controls. Accuracy for QC blood
spots was within 28% at the 5 lg/dL level, within 6% at
the 20 lg/dL level, and within 9% at the 50 lg/dL level.
Precision, reported as the standard deviation from expected
values, was 0.78 at the 5 lg/dL level, 0.60 at the 20 lg/dL
level, and 0.68 at the 50 lg/dL level.
During the CT Lab’s development of the blood spot
method, tests were done to ensure the capability and
accuracy of the tests. Spiked sheep’s blood pools were also
analyzed directly by adding 6.7 lL of the blood, which is
123
Matern Child Health J (2012) 16:1518–1524
Recent advances in analytical methods make it possible
to evaluate prenatal exposures to environmental contami-
nants, such as lead, using newborn dried blood spots (DBS)
[12–14]. The purpose of this study was to evaluate the
relationship between prenatal exposure to lead (as mea-
sured by newborn DBS results) and BLLs in infants exactly
6 months of age or younger, using public health registry
data for infants born in Texas from July 2002 through July
2006. This study was approved by the DSHS Institutional
Review Board.
Methods
Selection Criteria
All infant records were selected from the Texas Child Lead
Registry (TCLR), a legislatively-mandated registry main-
tained by the DSHS. The TCLR contains blood lead results
for children who have received a blood lead test in Texas
since 1996 [15].
Infants from the TCLR who were born in Texas from
July 5, 2002 through July 31, 2006 and whose BLLs were
tested at exactly 6 months of age or younger were eligible
for inclusion. Storage of newborn DBS in Texas began in
July 2002, thus DBS for children born on or after July 5,
2002 were the only specimens available for this study.
When the study was initiated, data required to link TCLR
infant records to the corresponding newborn DBS were
available through the end of 2006. To analyze 4 years of
data, the timeframe of July 2002 through July 2006 was
chosen for this study. Infants with BLL tests taken on or
before the day they turned 6 months of age were of interest
based on the hypothesis that at these young ages, lead
exposure would be more likely attributed to prenatal
exposures than environmental exposures.
Of those infants who met the study criteria, the study
population consisted of all children with elevated BLLs
(C10 lg/dL) and a subset of those with very low BLL
(B2 lg/dL). If a child was reported as having an elevated
BLL more than one time during the first 6 months of life,
the elevated BLL corresponding with the youngest age was
used in the analyses. Four infants with very low reported
BLLs were randomly sampled for each infant identified
with elevated BLLs (1:4 ratio). Infants identified with very
low BLLs were never reported as having a BLL of more
than 2 lg/dL in the first 6 months of life.
Data Linkage
All study infants’ information was obtained from pre-
existing public health data from the TCLR, birth certificate
records, and newborn screening (NBS) data, including
1519
infant DBS specimens. Data on all infants identified from
the TCLR first were linked to birth certificate information
to obtain more complete identifiers for associating the
study infants with their correct DBS specimens and to
obtain additional information on possible confounders or
effect modifiers. The linked TCLR—birth certificate data
set was linked to the NBS data records to identify corre-
sponding DBS specimens. As there was no unique identi-
fying field common to all datasets, both deterministic and
probabilistic linkage methods were performed, using
Microsoft Access 2007 and Link King, version 5.0.
The DSHS Chemical Threat Laboratory (CT Lab)
obtained a dried blood spot sample and a blank sample for
each DBS specimen by punching a 3/16-inch diameter
circle from the blood spot and white area surrounding the
spot, respectively.
Dried Blood Spot Laboratory Analysis
The method used to analyze lead in newborn DBS was
adapted from the Kansas Department of Health Laboratory
[16]. This method involves extracting the lead into a
solution containing acid and an internal standard (terbium)
followed by analysis with inductively coupled plasma-mass
spectrometry (ICP-MS). ICP-MS is used in the CT Lab to
analyze blood for mercury, lead, and cadmium and to
analyze urine for 14 toxic metals using the Centers for
Disease Control and Prevention (CDC) method.
Calibration curves were prepared in acidic solutions and
demonstrated linearity of the method from 0.05 to 1 lg/dL.
The dilution factor of the blood spots when analyzed then
represents a lead concentration of near 0 to 300 lg/dL in
the original spot. Over this range the coefficient of deter-
mination (R2) values for the calibration curves were
0.9999, with 1.0 representing perfect linearity.
Quality control (QC) samples were prepared at three
levels covering the range of the method (5, 20, and 50 lg/
dL), and were run with each set of specimens. QC samples
were prepared by spiking sheep’s blood with aqueous
solutions containing known amounts of lead. These spiked
blood pools were spotted onto blood collection cards pro-
vided by the newborn screening laboratory, which were
then punched and interspersed with the DBS specimens
during analysis as quality controls. Accuracy for QC blood
spots was within 28% at the 5 lg/dL level, within 6% at
the 20 lg/dL level, and within 9% at the 50 lg/dL level.
Precision, reported as the standard deviation from expected
values, was 0.78 at the 5 lg/dL level, 0.60 at the 20 lg/dL
level, and 0.68 at the 50 lg/dL level.
During the CT Lab’s development of the blood spot
method, tests were done to ensure the capability and
accuracy of the tests. Spiked sheep’s blood pools were also
analyzed directly by adding 6.7 lL of the blood, which is
123
1520
the amount contained in a 3/16 inch blood spot [17], to the
diluent solution. To further test the method’s accuracy,
whole human blood containing known levels of lead (from
an earlier CDC-sponsored blood lead test) were analyzed
as whole blood directly and were spotted onto cards for
analysis as blood spots. The method was accurate within
10–15% of the known values for whole blood analyzed
directly, and was accurate within 10% when the samples
were analyzed as blood spots.
The DBS samples used in this study were analyzed
between October 30, 2009 and November 24, 2009. Total
DBS lead level and DBS lead level with the background
amount of lead found on the card (or filter paper) sub-
tracted out were obtained for each specimen. The DBS lead
results with the amount of lead found on the cards sub-
tracted out were used in the analyses and reported in the
same units as BLLs reported in the TCLR (lg/dL).
Statistical Analysis
Both descriptive statistical calculations (crosstabs) and
correlation analyses were performed. Pearson’s correlation
coefficient (q) was calculated to assess the relationship
between TCLR infant BLL results and newborn DBS lead
levels. This parametric measure was chosen because the
sample size was large and scatterplots showed that the
relationship between these two variables was reasonably
linear.
Unconditional logistic regression models were used to
further examine the association between BLL status (ele-
vated or non-elevated) in infants 6 months of age or
younger and their prenatal exposure to lead (as measured
by newborn DBS lead level), and to adjust for other
potential covariates. Odds ratios and corresponding 95%
confidence intervals (CI) were calculated. Because the use
of categorical variables resulted in a high level of stratifi-
cation, exact logistic regression analyses were performed
instead of standard logistic regression methods.
In the logistic regression analyses, the outcome variable
was infant BLL results, with 2 levels: elevated BLLs
(C10 lg/dL) and very low BLLs (B2 lg/dL). The CDC
considers BLLs C10 lg/dL to be elevated [11]; the same
criterion for elevated BLLs was used in this study. Infants
with very low BLLs were chosen in order to compare
elevated infants’ DBS levels with those of infants who had
little to no lead exposure. Newborn DBS lead values were
modeled as a categorical independent variable, with the
following three categories: \5 lg/dL, 5 to \10 lg/dL, and
C10 lg/dL. The \5 lg/dL category was the reference
category. Newborn DBS lead levels were modeled as a
categorical rather than a continuous variable to evaluate the
association between elevated infant BLLs and elevated
newborn DBS lead levels. The highest DBS level category
123
Matern Child Health J (2012) 16:1518–1524
(C10 lg/dL) was chosen to be consistent with the outcome
variable’s elevated BLL category. We divided the non-
elevated lead levels into 2 separate categories (\5 lg/dL
and 5 to \10 lg/dL) to compare results for infants with
elevated newborn DBS lead levels with those with low
DBS lead levels. Possible covariates adjusted for in the
model included infant’s age in months, maternal race/eth-
nicity, infant race/ethnicity, infant sex, maternal age,
maternal education and paternal education, maternal
tobacco use, gestational age at birth, and whether infant
resided in a metropolitan or nonmetropolitan area. These
variables were added to the model only if a univariable
regression analysis yielded a P-value of 0.25 or less [18].
Infant age and gestational age at birth were modeled as
continuous variables; all other potential covariates were
modeled as categorical variables. A manual backwards
elimination method was used to determine the final model.
Subgroup analyses also were performed, stratified by
age. Infants reported in the TCLR as being tested at 0
through 3 months and 4 to 6 months were analyzed as two
separate subgroups to see if there were different associa-
tions for these two age groups. The same statistical anal-
yses described above were used for these subgroups.
Statistical analyses were performed using SAS, version
9.1.3, and Stata/IC, version 10.1.
Results
A total of 85 infant records meeting our selection criteria
were identified from the TCLR as having elevated BLLs.
For each of these 85 records, four records of infants with
very low reported BLLs were randomly sampled from the
TCLR (340 records), for a total of 425 study infants.
Of the 85 infant records with an elevated BLL, 51 were
linked to corresponding newborn screening records (a 60%
success rate), and DBS were obtained for 50 of these
infants. Likewise, 204 of the 340 infant records with very
low reported BLLs were linked to their newborn screening
records (also a success rate of 60%), and DBS were pulled
for 199 of these infants. A total of 249 infant records (50
with elevated BLLs, 199 with very low BLLs) were used in
the analyses. Of these 249 DBS specimens, 12 had elevated
newborn DBS lead levels. Ten of these were infants with
elevated BLLs (20%), and two were infants with very low
BLLs (1%) (Table 1).
A correlation analysis of all study infants yielded a
Pearson correlation coefficient of 0.48 (P \ .0001), which
is indicative of a weak positive relationship between infant
BLLs and their corresponding newborn DBS lead levels
(Table 2).
The final exact logistic regression model consisted only
of the newborn DBS lead results (our variable of interest)
Matern Child Health J (2012) 16:1518–1524 1521

Table 1 Texas Child Lead Registry BLL status (elevated vs. low) by newborn DBS lead levels
Texas Child Lead Registry BLL at 0 to 6 months old Newborn DBS lead level Total
C10 lg/dL 5 to \10 lg/dL \5 lg/dL

Elevated BLL (C10 lg/dL) 10 (20%) 4 (8%) 36 (72%) 50

Low BLL (B2 lg/dL) 2 (1%) 4 (2%) 193 (97%) 199
Total 12 (5%) 8 (3%) 229 (92%) 249 (100%)

Table 2 Pearson product moment correlations between infant BLL
results and newborn DBS lead levels, stratified by infant age group
Infant age group Pearson correlation P-value
coefficient
All study infants (0 to 6 months) 0.48 \0.0001
0 through 3 months 0.65 \0.0001
4 to 6 months 0.31 \0.0001
and maternal age, modeled as a categorical variable with 2
levels (\25 years of age, C25 years of age). Maternal age
was positively associated with infant BLL status, although
this association was not statistically significant (P = 0.10).
We considered maternal age to be a biologically relevant
variable, and as it had a relatively small P-value, we left it
in the model. In the final model, the odds of an infant with
an elevated newborn DBS lead level having an elevated
BLL at 0 to 6 months of age were almost 28 times greater
than for an infant with a low newborn DBS lead level of
\5 lg/dL (95% CI 5.5–277.3) (Table 3). A test for trend
showed that overall, newborn DBS lead levels were sig-
nificantly associated with infant BLLs (P \ .0001).
Of the 249 study infants, 44 infants were tested at 0
through 3 months of age. Approximately 43% of these
younger infants with elevated BLLs had elevated newborn
DBS lead levels, compared to 3.3% of infants with very
low BLLs (Table 4). The remaining 205 study infants were
tested at 4 to 6 months of age. Approximately 11% of these
older infants with elevated BLLs had elevated newborn
DBS lead levels, compared to less than 1% of infants with
very low BLLs (Table 4).
Pearson’s test of correlation showed a slightly stronger
positive relationship between infant BLLs and their cor-
responding newborn DBS lead levels in the younger infant
subgroup (correlation coefficient = 0.65, P \ 0.0001) than
for all study infants combined (correlation coeffi-
cient = 0.48, P \ .0001), while infants in the older infant
subgroup showed a slightly weaker correlation between
infant BLLs and newborn DBS lead levels (correlation
coefficient = 0.31, P \ 0.0001) (Table 2). Exact logistic
regression analyses of the two infant age subgroups showed
a strong association between elevated DBS lead level and
elevated infant BLLs in both subgroups (Table 3). Again,
only maternal age was adjusted for in the final model.
Discussion
In this study, a weak positive linear correlation was
observed between BLLs in TCLR infants tested between 0
to 6 months of age and newborn DBS lead levels. This
relationship also held true both for infants tested at 0
through 3 months and infants tested at 4 to 6 months,
although the correlation seemed to be stronger for the
younger infant subgroup.
A significant association between elevated DBS lead
levels and elevated infant BLLs was also observed using
exact logistic regression models. Even though odds ratios
were quite large, because of small cell counts and the

Table 3 Adjusted odds ratios

Infant age group Newborn DBS N Adjusted odds
of elevated infant BLLs for
lead level ratio* (95% CI)
newborn DBS lead levels
relative to referent DBS lead All study infants (0 to 6 months) C10 lg/dL 12 27.95 (5.52–277.28)
level (\5 lg/dL), stratified by
infant age group 5 to \10 lg/dL 8 4.86 (0.85–27.73)
\5 lg/dL 229 1.00 (referent)
0 through 3 months C10 lg/dL 7 32.20 (2.53–2,152.83)
5 to \10 lg/dL 2 2.47 (0.03–215.69)
\5 lg/dL 35 1.00 (referent)
4 to 6 months C10 lg/dL 5 22.53 (2.12–1,151.99)
5 to \10 lg/dL 6 5.43 (0.69–42.71)
\5 lg/dL 194 1.00 (referent)
* Adjusted for maternal age

123
1522
Table 4 Texas Child Lead Registry BLL status (elevated vs. low) by newborn DBS lead levels, stratified by infant age group
Infant age group Texas Child Lead Registry BLL
0 through 3 months Elevated BLL (C10 lg/dL)
Low BLL (B2 lg/dL)
Subgroup total
4 to 6 months Elevated BLL (C10 lg/dL)
Low BLL (B2 lg/dL)
Subgroup total
method of analysis used (exact logistic regression), confi-
dence intervals around odds ratios were very wide, reduc-
ing the reliability of these estimates.
Although an association was seen between elevated
infant BLLs and elevated newborn DBS lead levels
(Table 1), the majority of infants with elevated BLLs
(80%) did not have elevated newborn DBS lead levels,
suggesting that these infants were somehow being exposed
to lead after birth. This observation held true for the 0
through 3 month subgroup, in which 57% of infants with
reported elevated BLLs did not have elevated newborn
DBS lead levels.
The half-life of lead in blood is approximately 1 month
[19, 20]; thus, we would expect that lead in a child’s blood
resulting from prenatal exposure would continue to be
present in the blood for at least 1 to 2 months after birth.
Infants with elevated newborn DBS lead levels who also
had elevated BLLs at 4 to 6 months of age could have had
continued exposure to lead in addition to prenatal lead
exposure [21]. Potential sources for continued exposure to
lead could include dietary sources such as formula pre-
pared with lead-contaminated water or lead in breast milk
as well as environmental sources such as lead dust in the air
from household renovations [22–24].
Study infants with elevated BLLs and those randomly
selected with very low BLLs were comparable to each
other based on demographic characteristics. Infants with
elevated BLLs tended to be younger than those with very
low BLLs. However, no significant differences were noted
between the two groups in terms of maternal age, metro-
politan versus nonmetropolitan residence, sex, infant race/
ethnicity, and mother’s Medicaid status.
Some studies debate the practice of subtracting back-
ground levels of contaminants from DBS results [25, 26].
While results using background-subtracted data were used
in this study, analyses conducted without removing the
background levels produced similar results.
One of the strengths of this study was the relatively large
number of infants available with known blood lead levels.
The TCLR data are the most complete source of blood lead
information for the state; each year, approximately 250,000
123
Matern Child Health J (2012) 16:1518–1524
Newborn DBS lead level Total
C10 lg/dL 5 to \10 lg/dL \5 lg/dL
6 (42.90%) 1 (7.10%) 7 (50.00%) 14
1 (3.33%) 1 (3.33%) 28 (93.33%) 30
7 (15.90%) 2 (4.55%) 35 (79.55%) 44 (100%)
4 (11.10%) 3 (8.30%) 29 (80.60%) 36
1 (0.60%) 3 (1.80%) 165 (97.60%) 169
5 (2.50%) 6 (2.90%) 194 (94.60%) 205 (100%)
children under the age of 6 are tested for lead. Also, to the
best of the authors’ knowledge, this study is the first to look
at the relationship between BLLs in the first few months of
life and corresponding lead levels in newborn DBS.
DBS offer a number of benefits for exposure analysis in
newborns; they are required to be collected from all new-
borns by law and techniques exist to screen them for
environmental contaminants such as lead. Early detection
of exposure could allow the implementation of actions to
reduce the potential for adverse health effects. Collection
of DBS also is less invasive than venous blood draws.
While lead levels in the mother’s blood are strongly cor-
related with levels in the child [27, 28], DBS provide more
precise information on the actual levels in the children.
Others have noted potential problems with using DBS to
assess environmental exposures. The inability to exactly
quantify the volume of blood within a DBS punch [13]
hinders a precise determination of the contaminant con-
centration in the blood. A single DBS punch contains a very
small amount of blood, making analytical detection difficult
[14]. The filter paper itself could contain some level of the
contaminant of interest, and may contain an uneven distri-
bution of the contaminant throughout the filter paper, mak-
ing determination of background lead levels difficult [25].
A large percentage (41%) of the infant records from the
TCLR could not be linked to their NBS record or matched
with their corresponding DBS, due largely to missing
information in the databases. These infants, who were
excluded from analysis, were comparable to infants who
were matched with their DBS specimens in terms of sex,
race/ethnicity, metro/nonmetro residence, blood sample
type (capillary vs. venous), year of test, and age of test in
months. Infants excluded from the analysis did have higher
reported BLLs than infants who were able to be analyzed.
The association seen between newborn DBS lead levels
and BLLs in infants 0 to 6 months of age might not hold
for infants excluded from analyses.
Vital statistics data indicated that mothers of 52.2% of
the infants in this study were enrolled in Medicaid at the
time of delivery, whereas mothers of only 36.7% of all
infants born in Texas from 2002 through 2006 were
Matern Child Health J (2012) 16:1518–1524
enrolled in Medicaid. Medicaid children are considered to
be at high risk for blood lead poisoning [29], and are dis-
proportionately represented in the TCLR database. Also, a
larger percentage of study infants were born to younger
mothers. This study had a higher percentage of Hispanic
infants and a lower percentage of non-Hispanic white
infants when compared to all live births in 2002 through
2006. The demographic differences between infants in this
study and infants in the Texas population may indicate that
these results might not be generalizable to all Texas
infants.
Lastly, the TCLR includes BLL results for both capil-
lary and venous lead tests. In this study, 91% of infants
with reported elevated BLLs and 74% of infants with
reported very low BLLs had capillary tests. While capillary
tests themselves are accurate, improper cleaning of the skin
surface prior to the test can lead to false-positive results.
However, because there is limited mobility in infants of the
young ages used in this study, and because it is recom-
mended that a heel stick rather than a finger stick be per-
formed for infants of these ages, contamination of capillary
specimens may not be as much of a concern as it is in older
children.
Conclusions
Although infant BLLs were statistically significantly cor-
related with prenatal lead exposure (measured by newborn
DBS lead results), they were only weakly positively cor-
related, and evidence of prenatal lead exposure was not
seen for a majority of infants with elevated BLLs at 0 to
6 months of age. These data suggest that young infants
may be exposed to lead after birth, which underscores the
importance of testing young children for lead poisoning.
Reducing prenatal exposure to lead also is important.
Women who are pregnant should avoid or minimize
exposure to lead through work and/or hobbies, and preg-
nant women who are or who have been exposed to lead
should be advised to have their BLLs tested. Adequate
calcium and vitamin D supplementation may be helpful in
reducing maternal release of lead stored in bone into cir-
culation, particularly during the third trimester and during
periods of breastfeeding, when the potential for release of
lead from the mothers’ bones is highest [30, 31].
While DBS analysis could be used as a screening tool to
identify newborns exposed to lead prenatally, the relatively
high costs of DBS lead analysis coupled with the small
number of infants likely to be exposed to elevated lead
levels in utero are barriers to implementing this practice
statewide. More cost-effective options could include test-
ing mothers with a history of exposure to lead during the
1523
third trimester of pregnancy or obtaining capillary blood
lead tests of infants at birth.
Acknowledgments The authors would like to thank Taren Land and
Lisa Marengo for their help with record linking, Dr. Peter Langlois
and Dr. Lucina Suarez for their epidemiological and statistical advice,
and Dr. Susan Tanksley, Dr. Grace Kubin, Teresa Willis, and Gene
Willard for their support. Lab work for this project was funded in part
by the Texas Environmental Health Institute.
References
1. Lyngbye, T., Hansen, O. N., Trillingsgaard, A., Beese, I., &
Grandjean, P. (1990). Learning disabilities in children: signifi-
cance of low-level lead-exposure and confounding factors. Acta
Paediatrica Scandinavica, 79, 352–360.
2. Mendelsohn, A. L., Dreyer, B. P., Fierman, A. H., Rosen, C. M.,
Legano, L. A., Kruger, H. A., et al. (1998). Low-level lead
exposure and behavior in early childhood. Pediatrics, 101, E10.
3. Ruff, H. A., Markowitz, M. E., Bijur, P. E., & Rosen, J. F. (1996).
Relationships among blood lead levels, iron deficiency, and
cognitive development in two-year-old children. Environmental
Health Perspectives, 104, 180–185.
4. Sciarillo, W. G., Alexander, G., & Farrell, K. P. (1992). Lead
exposure and child behavior. American Journal of Public Health,
82, 1356–1360.
5. Dietrich, K. N., Krafft, K. M., Pearson, D. T., Harris, L. C.,
Bornschein, R. L., Hammond, P. B., et al. (1985). Contribution of
social and developmental factors to lead exposure during the first
year of life. Pediatrics, 75, 1114–1119.
6. Dietrich, K. N., Berger, O. G., & Succop, P. A. (1993). Lead
exposure and the motor developmental status of urban six-year-
old children in the Cincinnati Prospective Study. Pediatrics, 91,
301–307.
7. Baghurst, P. A., Robertson, E. F., McMichael, A. J., Vimpani, G.
V., Wigg, N. R., & Roberts, R. R. (1987). The Port Pirie cohort
study: lead effects on pregnancy outcome and early childhood
development. Neurotoxicology, 8, 395–402.
8. Dietrich, K. N., Ris, M. D., Succop, P. A., Berger, O. G., &
Bornschein, R. L. (2001). Early exposure to lead and juvenile
delinquency. Neurotoxicology and Teratology, 23, 511–518.
9. American Academy of Pediatrics Committee on Environmental
Health. (2005). Lead exposure in children: Prevention, detection,
and management. Pediatrics, 116, 1036–1046.
10. Texas Department of State Health Services. (2010). Blood lead
screening and testing guidelines for Texas children: Quick ref-
erence guide. Available: http://www.dshs.state.tx.us/lead/pdf_files/
BL%20Screening%20Guidelines.05.05.09.pdf. Accessed Decem-
ber 2010.
11. Centers for Disease Control and Prevention. (2002). Managing
elevated blood lead levels among young children: Recommen-
dations from the Advisory Committee on Childhood Lead
Poisoning Prevention. Atlanta: Centers for Disease Control and
Prevention.
12. Burse, V. W., DeGuzman, M. R., Korver, M. P., Najam, A. R.,
Williams, C. C., Hannon, W. H., et al. (1997). Preliminary
investigation of the use of dried-blood spots for the assessment of
in utero exposure to environmental pollutants. Biochemical and
Molecular Medicine, 61, 236–239.
13. Olshan, A. F. (2007). Meeting report: The use of newborn blood
spots in environmental research: Opportunities and challenges.
Environmental Health Perspectives, 115, 1767–1769.
123
1524
14. Spliethoff, H. M., Tao, L., Shaver, S. M., Aldous, K. M., Pass, K.
A., Kannan, K., et al. (2008). Use of newborn screening program
blood spots for exposure assessment: Declining levels of per-
fluorinated compounds in New York State infants. Environmental
Science and Technology, 42, 5361–5367.
15. Texas Department of State Health Services. (2006). Texas
childhood lead poisoning prevention program 2006 annual
report. Austin: Texas Department of State Health Services.
16. Sandstrom, S., & Brady, J. (2003). Determination of lead in filter
paper blood spots by inductively coupled plasma mass spec-
trometry. Test method CM 202.01. Kansas: Kansas Department of
Health and Environment.
17. Wang, S. T., & Demshar, H. P. (1992). Determination of blood
lead in dried blood-spot specimens by Zeeman-effect background
corrected atomic absorption spectrometry. Analyst, 117, 959–961.
18. Hosmer, D. W., & Lemeshow, S. (2000). Applied logistic
regression. New York: John Wiley & Sons.
19. Anderson, A. C., Pueschel, S. M., & Linakis, J. G. (1996).
Pathophysiology of lead poisoning. In S. M. Pueschel, J. G.
Linakis, & A. C. Anderson (Eds.), Lead poisoning in childhood
(pp. 75–96). Baltimore: Paul H. Brookes Publishing.
20. Zenz, C., Dickerson, O., & Horvath, E. (1994). Occupational
medicine. St. Louis: Mosby.
21. Agency for Toxic Substances and Disease Registry. (2007).
Toxicological profile for lead (update). Atlanta: US Department
of Health and Human Services.
22. Baum, C. R., & Shannon, M. W. (1997). The lead concentration
of reconstituted infant formula. Journal of Toxicology. Clinical
Toxicology, 35, 371–375.
23. Ettinger, A. S., Tellez-Rojo, M. M., Amarasiriwardena, C.,
Bellinger, D., Peterson, K., Schwartz, J., et al. (2004). Effect of
breast milk lead on infant blood lead levels at 1 month of age.
Environmental Health Perspectives, 112, 1381–1385.
123
Matern Child Health J (2012) 16:1518–1524
24. Shannon, M. W., & Graef, J. W. (1992). Lead intoxication in
infancy. Pediatrics, 89, 87–90.
25. Chaudhuri, S. N., Butala, S. J. M., Ball, R. W., & Braniff, C. T.
(2009). Rocky Mountain Biomonitoring Consortium. Pilot study
for utilization of dried blood spots for screening of lead, mercury,
and cadmium in newborns. Journal of Exposure Science Envi-
ronmental Epidemiology, 19, 298–316.
26. Cizdziel, J. V. (2007). Determination of lead in blood by laser
ablation ICP-TOF-MS analysis of blood spotted and dried on
filter paper: A feasibility study. Analytical and Bioanalytical
Chemistry, 388, 603–611.
27. Gardella, C. (2001). Lead exposure in pregnancy: A review of the
literature and argument for routine prenatal screening. Obstetrical
and Gynecological Survey, 56, 231–238.
28. Nashashibi, N., Cardamakis, E., Bolbos, G., & Tzingounis, V.
(1999). Investigation of kinetic of lead during pregnancy and
lactation. Gynecologic and Obstetric Investigation, 48, 158–162.
29. Advisory Committee on Childhood Lead Poisoning Prevention.
(2000). Recommendations for blood lead screening of young
children enrolled in Medicaid: Targeting a group at high risk.
MMWR Recommendations and Reports, 49, 1–13.
30. Ettinger, A. S., Tellez-Rojo, M. M., Amarasiriwardena, C.,
Peterson, K. E., Schwartz, J., Aro, A., et al. (2006). Influence of
maternal bone lead burden and calcium intake on levels of lead in
breast milk over the course of lactation. American Journal of
Epidemiology, 163, 48–56.
31. Ettinger, A. S., Lamadrid-Figueroa, H., Tellez-Rojo, M. M.,
Mercado-Garcia, A., Peterson, K. E., Schwartz, J., et al. (2009).
Effect of calcium supplementation on blood lead levels in preg-
nancy: A randomized placebo-controlled trial. Environmental
Health Perspectives, 117, 26–31.