American Journal of Epidemiology
ª The Author 2010. Published by Oxford University Press on behalf of the Johns Hopkins Bloomberg School of
Public Health. All rights reserved. For permissions, please e-mail: firstname.lastname@example.org.
Vol. 171, No. 10
Advance Access publication:
April 20, 2010
Association Between Blood Lead and the Risk of Amyotrophic Lateral Sclerosis
Fang Fang, Lydia C. Kwee, Kelli D. Allen, David M. Umbach, Weimin Ye, Mary Watson, Jean Keller,
Eugene Z. Oddone, Dale P. Sandler, Silke Schmidt, and Freya Kamel*
* Correspondence to Dr. Freya Kamel, Epidemiology Branch, National Institute of Environmental Health Sciences, PO Box 12233,
Mail Drop A3-05, Research Triangle Park, NC 27709 (e-mail: email@example.com).
Initially submitted September 11, 2009; accepted for publication March 8, 2010.
The authors conducted a 2003–2007 case-control study including 184 cases and 194 controls to examine the
association between blood lead and the risk of amyotrophic lateral sclerosis (ALS) among US veterans and to
explore the influence on this association of bone turnover and genetic factors related to lead toxicokinetics. Blood
lead, plasma biomarkers of bone formation (procollagen type 1 amino-terminal peptide (PINP)) and resorption
(C-terminal telopeptides of type 1 collagen (CTX)), and the K59N polymorphism in the d-aminolevulinic acid
dehydratase gene, ALAD, were measured. Odds ratios and 95% confidence intervals for the association of blood
lead with ALS were estimated with unconditional logistic regression after adjustment for age and bone turnover.
Blood lead levels were higher among cases compared with controls (P < 0.0001, age adjusted). A doubling of
blood lead was associated with a 1.9-fold increased risk of ALS (95% confidence interval: 1.3, 2.7) after adjustment
for age and CTX. Additional adjustment for PINP did not alter the results. Significant lead-ALS associations were
observed in substrata of PINP and CTX levels. The K59N polymorphism in the ALAD gene did not modify the lead-
ALS association (P ¼ 0.32). These results extend earlier findings by accounting for bone turnover in confirming the
association between elevated blood lead level and higher risk of ALS.
amyotrophic lateral sclerosis; bone and bones; bone resorption; lead; odds ratio; osteogenesis
Abbreviations: ALAD, d-aminolevulinic acid dehydratase gene; ALS, amyotrophic lateral sclerosis; CTX, C-terminal telopeptides of
type 1 collagen; GENEVA, Genes and Environmental Exposures in Veterans with Amyotrophic Lateral Sclerosis; ICD-9, Inter-
national Classification of Diseases, Ninth Revision; PINP, procollagen type 1 amino-terminal peptide; VALE, Veterans with ALS
and Lead Exposure.
An association between lead exposure and amyotrophic
lateral sclerosis (ALS) is a long-standing hypothesis. Most
previous studies have supported this relation but in general
have relied on indirect measures of lead exposure (1–12).
Previously, we reported that increases in blood and bone
lead levels were associated with a higher risk of ALS
(1, 12). Another recent study reported a similarly strong
association for blood, albeit not bone, lead level (13).
Blood lead levels may reflect both current environmental
lead exposure and mobilization of lead from bone (14). The
distribution of lead between blood and bone may change
during ALS progression as a patient’s level of physical activ-
ity declines, but no known study has taken bone turnover into
account by using direct measurements. Lead toxicokinetics
may also modify the lead-ALS association. For example,
the K59N polymorphism of the d-aminolevulinic acid
dehydratase gene, ALAD, influences lead toxicokinetics,
leading to lower bone lead levels and sometimes to higher
blood lead levels in carriers of the variant allele (ALAD2)
(15), and should thus also be accounted for. In a previous
study, we found no evidence for such an effect modification
by ALAD genotype, but we had limited power to evaluate this
In the present study, Veterans with ALS and Lead Expo-
sure (VALE), we sought to corroborate our earlier findings
on the relation of lead to ALS risk in a different and larger
population. In addition, we evaluated the role of bone turn-
over and ALAD genotype in the lead-ALS relation.
1126Am J Epidemiol 2010;171:1126–1133
MATERIALS AND METHODS
Cases: the National Registry of Veterans with ALS
Cases for the VALE study were derived from the National
Registry of US Veterans with ALS. Details of this study
have been provided elsewhere (16). Briefly, between April
2003 and September 2007, nationwide publicity efforts and
a search of Veterans Administration databases were used to
identify living US veterans with motor neuron diseases.
Veterans or their proxies completed a screening question-
naire to verify eligibility, and veterans were invited to par-
ticipate in the registry DNA bank by donating a blood
sample. Subsequently, neurologists with expertise in ALS
reviewed medical records to determine motor neuron dis-
ease diagnosis in accordance with the original El Escorial
Criteria (17), including ALS (International Classification of
Diseases, Ninth Revision (ICD-9) code 335.20), progressive
muscular atrophy (ICD-9 code 335.21), progressive bulbar
palsy (ICD-9 code 335.22), pseudobulbar palsy (ICD-9 code
335.23), primary lateral sclerosis (ICD-9 code 335.24), and
other motor neuron diseases (ICD-9 code 335.29). Ulti-
mately, 1,998 veterans with motor neuron diseases enrolled
in the registry, of whom 1,167 donated a blood sample to the
DNA bank (16, 18).
VALE included a subset of registry cases comprising
motor neuron disease cases who donated a blood sample
between January 23, 2007, and September 30, 2007;
208 cases were eligible, of whom 200 were enrolled, in-
cluding 163 ALS cases, 30 progressive muscular atrophy
cases, and 7 primary lateral sclerosis cases (Table 1). The
remaining 8 cases were excluded because their final
diagnosis was not motor neuron disease. Cases included
in VALE were similar to those in the registry as a whole
with respect to age, gender, race, and distribution of
Controls: Genes and Environmental Exposures in
Veterans with Amyotrophic Lateral Sclerosis Study
Controls for the VALE study were obtained from the
Genes and Environmental Exposures in Veterans with
Amyotrophic Lateral Sclerosis (GENEVA) study; enroll-
ment procedures for GENEVA have been described previ-
ously (18). Briefly, in June 2005, a random sample of 10,000
US veterans were identified; control recruitment was initi-
ated in January 2006. Eligible controls, free of ALS or
other neurologic disorders per the telephone screener, were
frequency matched to the cases by age (within 5 years),
gender, race (white/nonwhite), and past use of the Veterans
Administration system for health care.
Between May 2007 and May 2008, VALE contacted
359 controls already enrolled in GENEVA for additional
informed consent and blood sample collection. A total
of 252 controls consented to participate in VALE, of
whom 229 ultimately donated a blood sample. There
were no differences in age, gender, race, smoking, or
education between those who enrolled and provided
a blood sample and those who did not (P > 0.41 for all
The National Registry of Veterans with ALS conducted
home visits with case enrollees and collected as many as 4
tubes of blood to provide DNA and plasma. For cases en-
rolled in VALE, blood collection procedures remained the
same except that the first whole-blood sample was collected
in a 6-mL BDVacutainer blue-top Trace Element metal-free
tube for lead measurement (Becton, Dickinson and Com-
pany, Franklin Lakes, New Jersey).
For GENEVA controls, saliva samples, collected by mail
using Oragene kits (DNA Genotek Inc., Kanata, Ontario,
Canada), were used as a source of DNA. VALE conducted
a home visit for controls, during which 2 blood samples
were collected: a 6-mL whole-blood sample in a metal-free
tube for lead measurement and a 9-mL plasma sample for
bone turnover biomarkers.
For both cases and controls, blood samples were chilled
immediately, shipped with cold packs, and processed and
frozen within approximately 48 hours after blood draw. All
samples were stored at ?80?C until assay. Samples were
collected and processed in the same way for cases and
Lead concentration in blood samples was determined by
inductively coupled plasma mass spectrometry. The testing
laboratory was blinded to case-control status and made ex-
tensiveefforts to preventmetal contamination, including use
of a class 100 plastic hood for sample preparation and
ultrex-grade acids and oxidants as well as approximately
18-MX-quality deionized water to eliminate contamination.
Before analysis, samples were digested in a digitally con-
trolled digestion block with high-purity acids and oxidants.
Quality control samples processed with each batch of study
samples to continuously monitor assay performance indi-
cated good precision: the relative standard deviation per-
centage was less than 10% for all and less than 5% for
96% of the batches. Method blanks and aliquots of digestion
reagents were carried through the analytical procedure to
monitor the analyte background contribution from the re-
agents and the procedure. Aliquots of a Standard Reference
Material (National Institute of Standards and Technology
(NIST) SRM 966 Toxic Elements in Bovine Blood) were
also processed as an accuracy check. In addition, approxi-
mately 5% of the study samples were prepared and analyzed
in duplicate to monitor precision.
Bone turnover measurement
Because bone formation and bone resorption are coupled
processes, we measured plasma biomarkers for both. We
conducted a pilot study to determine whether collecting
samples under field conditions affected biomarker stability.
We subjected blood samples collected from nonveteran vol-
unteers to 1 of 3 conditions: processed immediately, held at
4?C for 24 hours before processing, and stored at room
temperature for 24 hours before processing. Plasma was
then stored at ?80?C until assay. On the basis of stability
Blood Lead and the Risk of ALS 1127
Am J Epidemiol 2010;171:1126–1133
observed under these conditions, we assessed bone forma-
tion by measuring plasma procollagen type 1 amino-
terminal peptide (PINP) using the UniQ PINP RIA
radioimmunoassay (Orion Diagnostica Oy, Espoo, Finland;
intraassay coefficient of variation: 8.8%; interassay coeffi-
cient of variation: 5.1%) and assessed bone resorption by
measuring plasma C-terminal telopeptides of type 1 colla-
gen (CTX) using the Serum CrossLaps ELISA assay
(Nordic Bioscience Diagnostics, Herlev, Denmark; intraas-
say coefficient of variation: 5.1%; interassay coefficient of
variation: 6.7%). Both assays were run with negative and
positive controls and met specified assay requirements for
all kit-calibrated standards. These biomarkers are both
specific to bone relative to other connective tissues.
DNA was extracted from whole blood for cases and
Oragene (DNA Genotek Inc.) saliva collection kits by using
Puregene reagents (Gentra Systems, Inc., Minneapolis,
Minnesota) for controls. Previous studies have documented
that DNA extracted from Oragene kits generates both high
yields and high-quality DNA, as judged by polymerase
chain reaction and genotyping success rates (19, 20). Results
United States, 2003–2007
Characteristics of Participants in the Veterans With ALS and Lead Exposure Study,
All ParticipantsWhite Men Only
(n 5 200)
(n 5 229)
(n 5 184)
(n 5 194)
No.% No.% No.% No.%
63.3 (34–83)63.4 (34–84)63.4 (34–83)64.3 (34–84)
Men 19698 216 94184 194
White18794 205 90184194
Ever133 6615467 12367 13670
Never6231 75 3356 30 5830
Clinically definite ALS31 16 2916
Clinically probable ALS,
Clinically probable ALS 73 3667 36
Clinically possible ALS24 12 2212
30 15 2815
Site of onset
Bulbar 392035 19
Symptom onset to
102 51 96 52
98 4988 48
Diagnosis to sample
Abbreviation: ALS, amyotrophic lateral sclerosis.
aValues are expressed as mean (range).
bPercentages may not add to 100% because of rounding.
1128Fang et al.
Am J Epidemiol 2010;171:1126–1133
for our study were similar. The coding change (K59N,
rs1800435) intheALAD genewasgenotyped with aTaqMan
assay (Applied Biosystems Inc., Foster City, California) at
the Duke Center for Human Genetics Molecular Genetics
Core. We required 95% genotyping efficiency and that ge-
notypes of quality control samples match within and across
all plates before including samples in the statistical analysis.
The aims of VALE included examining whether 1) ele-
vated blood lead level is associated with a higher risk of
ALS among the veterans, 2) the lead-ALS association is
influenced by bone turnover, and 3) the lead-ALS associa-
tion is modified by ALAD genotype. We based our power
calculations on 200 cases and 200 controls and treated lead
as a continuous variable. For aim 1, on the basis of an ex-
pected mean lead level of 3.4 lg/dL (standard deviation,
2.5) (12) but conservatively assuming a standard deviation
of 1.0, we had greater than 80% power to detect a 1.4-fold
increase in ALS risk for each 1-lg/dL increase in blood
lead. For aim 2, on the basis of expected increases in bone
turnover biomarkers in ALS patients, we had greater than
80% power to detect an odds ratio of 1.9–2.0 for interaction
of blood lead with these biomarkers. For aim 3, on the basis
of an expected prevalence of 0.2 for the ALAD2 allele, we
had greater than 80% power to detect an odds ratio of 1.8 for
the main effect and an odds ratio of 2.2 for an interaction
with blood lead.
Most study participants were white men. To reduce sam-
ple heterogeneity, we excluded women and nonwhites from
the main analyses, leaving 184 cases and 194 controls
(Table 1). We compared the means of lead, PINP, and
CTX levels between cases and controls using analysis of
covariance (PROC GLM; SAS version 9.1.3 software,
SAS Institute, Inc., Cary, North Carolina); P values were
calculated after adjustment for age (as a continuous vari-
able; age at diagnosis for cases and age at interview for
controls). Since case samples with various lag times after
diagnosis were collected, changes in physical activity asso-
ciated with different disease stages might have influenced
blood lead and bone turnover levels. Consequently, we ex-
amined whether the means of lead, PINP, and CTX levels
varied with the time interval between diagnosis and sample
collection (<1 year, 1–2 years, and >2 years) among the
We used unconditional logistic regression models to
estimate odds ratios for ALS and their 95% confidence
intervals. We used log2-transformed blood lead level as
a continuous variable; this transformation was used to en-
sure linearity in model fitting and better interpretability. All
models included adjustment for age as a continuous vari-
able. In some models, we further adjusted for smoking,
which may be associated with ALS (21–23).
In additional models, we adjusted for PINP and CTX
separately and jointly, as log2-transformed continuous vari-
ables. Analyses were also conducted after stratification
by untransformed PINP (?34.45 lg/L and >34.45 lg/L,
the median for controls) or CTX (?0.32 ng/mL and >0.32
ng/mL, the median for controls) levels; log2-transformed
PINP and CTX were still included in these models as con-
tinuous variables to mitigate residual confounding.
To evaluate robustness of results, we conducted several
additional analyses. First, we repeated analyses after ex-
cluding progressive muscular atrophy and primary lateral
sclerosis cases (n ¼ 33) to ensure that results were pertinent
to ALS. Second, the time interval between diagnosis and
sample collection was 2 years or less for 144 cases (78.3%,
Table 1), and we repeated analyses after excluding cases for
whom this interval exceeded 2 years to allay potential con-
cern that survival bias would influence our results. Third,
diagnostic delay was 1 year or less for 96 cases (52.2%,
Table 1), and we repeated the analyses after stratification
by this factor (?1 year or >1 year) to assess whether di-
agnostic delay affected the lead-ALS relation. Finally, we
repeated the analyses including women and nonwhites and
adjusting for gender and race to evaluate the generalizability
of results from the main analyses.
Since the ALAD2 allele is rare, few individuals are
homozygous for that allele. Accordingly, we dichotomized
ALAD genotype into those homozygous for ALAD1 versus
thosewith at least one copy of ALAD2. To evaluate potential
effect modification of the lead-ALS relation, we conducted
an analysis stratified by ALAD genotype. In a separate anal-
ysis, a potential interaction between blood lead and ALAD
genotype was tested by adding an interaction (product) term
of these 2 variables in the logistic regression model.
The VALE study was approved by the institutional review
boards at the National Institute of Environmental Health
Sciences, the Durham VA Medical Center, Duke University,
and the Copernicus Group.
The unadjusted mean level of untransformed blood lead
was 1.76 lg/dL (range, 0.32–6.90) among the controls
and 2.41 lg/dL (range, 0.72–7.58) among the cases; the
difference was statistically significant after adjustment
for age (Figure 1). The mean levels of blood lead were
2.38 lg/dL (standard deviation, 1.31) among ALS cases,
2.67 lg/dL (standard deviation, 1.28) among progressive
muscular atrophy cases, and 1.93 lg/dL (standard deviation,
0.79) among primary lateral sclerosis cases; these values did
not differ (P ¼ 0.25). The interval between diagnosis and
sample collection did not influence the mean lead level of
cases (Figure 2), although lead was weakly correlated with
time since diagnosis among cases when time was used as
a continuous variable (r ¼ 0.18; P ¼ 0.01). After adjustment
for age, a 1-unit increment of log2-transformed lead (equiv-
alent to a doubling of blood lead) was associated with a 2.6-
fold higher odds of ALS (95% confidence interval: 1.9, 3.7),
indicating a dose response (Table 2). Adjustment for smok-
ing (ever/never) in addition to age did not change the results
(data not shown).
Compared with controls, cases had higher CTX but not
PINP levels (Figure 1). The interval between diagnosis and
sample collection did not influence mean levels of PINP or
CTX among cases (Figure 2). Lead was correlated with
CTX among cases (r ¼ 0.20; P ¼ 0.008) and controls
Blood Lead and the Risk of ALS 1129
Am J Epidemiol 2010;171:1126–1133
(r ¼ 0.23; P ¼ 0.002) but not with PINP (P > 0.20 for both
groups). Adjustment for CTX diminished the magnitude but
did not eliminate the association of lead with ALS (odds
ratio ¼ 1.9, 95% confidence interval: 1.3, 2.7). Further ad-
justment for PINP alone or jointly with CTX did not alter
results; thus, we present results adjusted for age and CTX
only. A dose response for the lead-ALS association was also
seen when blood lead was categorized in tertiles; after ad-
justment for age and CTX, the odds ratio for the highest
compared with the lowest tertile was 2.1 (95% confidence
interval: 1.1, 3.8; Ptrend¼ 0.008).
Models stratified by either PINP or CTX showed a signif-
icant association of blood lead with ALS in all strata
(Table 2). Slightly stronger associations of lead with ALS
were suggested among individuals with lower CTX or
higher PINP levels (i.e., less demineralization), but a statis-
tically significant interaction was not noted (P > 0.20).
Excluding cases with progressive muscular atrophy or
primary lateral sclerosis did not change results substantially
(Table 2). After we excluded cases with an interval of more
than 2 years between diagnosis and sample collection, the
odds ratio for the association of lead with ALS was 1.7 (95%
confidence interval: 1.2, 2.5). Odds ratios were 1.9 (95%
confidence interval: 1.3, 2.9) for cases with more than a
1-year diagnostic delay and 1.7 (95% confidence interval:
1.1, 2.7) for cases with a diagnostic delay of 1 year or less.
Blood Lead, µg/dL
Plasma PINP, µg/L
>2 1–2 <1
<1 1–2 >2 <1 1–2>2
Plasma CTX, ng/mL
(PINP), and C) plasma C-terminal telopeptides of type 1 collagen (CTX) levels among amyotrophic lateral sclerosis (ALS) cases by interval
between diagnosis and sample collection (years), the Veterans with ALS and Lead Exposure study, United States, 2003–2007. P values were
calculated after adjustment for age at diagnosis (as a continuous variable), A): P ¼ 0.38, B): P ¼ 0.45, C): P ¼ 0.76.
Unadjusted, untransformed means and standard deviations of A) blood lead, B) plasma procollagen type 1 amino-terminal peptide
Blood Lead, µg/dL
Plasma PINP, µg/L
Plasma CTX, ng/mL
(PINP), and C) plasma C-terminal telopeptides of type 1 collagen (CTX) levels among amyotrophic lateral sclerosis (ALS) cases and controls, the
Veterans with ALS and Lead Exposure study, United States, 2003–2007. P values were calculated after adjustment for age (as a continuous
variable; age at diagnosis for cases and age at interview for controls)—A): P < 0.0001, B): P ¼ 0.77, C): P < 0.0001.
Unadjusted, untransformed means and standard deviations of A) blood lead, B) plasma procollagen type 1 amino-terminal peptide
1130Fang et al.
Am J Epidemiol 2010;171:1126–1133
When women and nonwhites were included, the odds ratio
was 1.6 (95% confidence interval: 1.2, 2.2) after adjustment
for gender and race in addition to age and CTX.
Two cases and 6 controls had no ALAD genotype data
and were excluded from corresponding analyses (Table 2).
ALAD2 carriers did not have different odds of ALS
compared with ALAD1-1 homozygotes (age-adjusted odds
ratio ¼ 0.8, 95% confidence interval: 0.4, 1.4). Among
ALAD1-1 homozygotes, mean lead levels were 2.43 lg/dL
(standard deviation, 1.31) for cases and 1.77 lg/dL (stan-
dard deviation, 0.94) for controls; among ALAD2 carriers,
they were 2.26 lg/dL (standard deviation, 1.13) for cases
and 1.88 lg/dL (standard deviation, 0.89) for controls. Us-
ing log2-transformed lead as a continuous variable, we noted
a significant lead-ALS association among ALAD1-1 carriers
after adjustment for age and CTX, whereas the association
was weaker and not significant among ALAD2 carriers
(Table 2). The interaction between lead and ALAD genotype
was not statistically significant (P ¼ 0.32), however.
In the VALE study, we extended our previous findings to
show that elevated blood lead levels were associated with
a higher ALS risk for a new and larger population and after
taking bone turnover into account. ALAD genotype did not
seem to modify the lead-ALS association.
Association of lead exposure with ALS is a long-standing
hypothesis (1–12), although many previous studies used in-
direct measures of lead exposure. In a previous study, we
showed that measured blood and bone lead levels were both
associated with a higher risk of ALS; the association was
particularly marked for blood lead (odds ratio ¼ 1.9 for each
Am J Epidemiol 2010;171:1126–1133
1-lg/dL increment) (1, 12). Recently, a similar association
between blood lead and ALS was also shown in a Northern
California population (13). Findings from our previous and
present studies are similar both qualitatively and quantita-
tively. In the previous study, the possible contribution of
bone demineralization was takeninto account by controlling
for physical activity. The present study extended these find-
ings by measuring bone turnover directly.
Although blood lead is often considered an indicator of
current lead exposure, it may also reflect bone lead levels. In
older individuals with no obvious sources of external expo-
sure, bone lead is the largest source of blood lead (14),
suggesting that the latter may serve as an indirect indicator
of cumulative lifetime exposure. Alternatively, increased
blood lead level may be a consequence of the disease pro-
cess among ALS patients: decreased physical activity could
increase bone turnover, leading to increased release of lead
from bone. VALE addressed the latter possibility directly by
taking measured bone turnover into account and showed that
adjustment for or stratification by bone turnover biomarkers
did not substantively alter our results. Interestingly, we
found hints of a stronger lead-ALS association among in-
dividuals with lower bone resorption or higher bone forma-
tion, that is, individuals likely to have less release of lead
from bone to blood. These findings suggest that reverse
causality does not fully account for the association between
blood lead and ALS.
Despite the strong lead-ALS association observed, both
cases and controls had low levels of blood lead. However,
a small difference in blood lead levels may be biologically
significant given the low absolute lead level observed in the
controls. Furthermore, a small difference in current blood
lead may reflect large differences in past environmental lead
Associations Between Log2-Transformed Blood Lead Level and ALS Risk, Veterans With ALS and Lead Exposure Study, United
All Cases (n 5 184)ALS Cases (n 5 151)
95% CI ORb
Stratified by PINPc
194100 184100 2.61.9, 188.8.131.52, 2.7 151 822.51.8, 3.4 1.8 1.2, 2.5
PINP ?34.45 lg/L
PINP >34.45 lg/L
Stratified by CTXc
9750 90492.4 1.5, 3.71.5 0.9, 2.67641 2.21.4, 3.5 1.40.8, 2.3
97 509049 2.9 1.8, 4.82.3 1.4, 3.97239 2.7 1.6, 4.5 2.21.3, 3.8
CTX ?0.32 ng/mL
CTX >0.32 ng/mL
95 49 3117 3.01.5, 5.8 2.81.4, 5.5 2815 2.81.4, 184.108.40.206, 5.3
99 51149 81 2.01.3, 2.9 1.61.1, 2.4120651.91.3, 220.127.116.11, 2.3
Stratified by ALAD
ALAD115680 157 852.71.9, 3.82.01.3, 2.9 127 69 2.51.7, 18.104.22.168, 2.7
ALAD2 321625 14 1.90.8, 4.5 1.2 0.4, 3.12312 1.90.8, 4.6 1.1 0.4, 3.1
Abbreviations: ALAD, d-aminolevulinic acid dehydratase gene; ALS, amyotrophic lateral sclerosis; CI, confidence interval; CTX, C-terminal
telopeptides of type 1 collagen; OR, odds ratio; PINP, procollagen type 1 amino-terminal peptide.
aAdjusted for age (continuous; age at diagnosis for cases and age at interview for controls); in analyses stratified by PINP and CTX, 4 cases (3
with ALS) were excluded because of missing PINP/CTX level.
bAdjusted for age and log2-transformed CTX level; 4 cases (3 with ALS) were excluded because of missing CTX level.
cCategorizations were made at the medians of PINP/CTX among the controls; 4 cases (3 with ALS) were excluded because of missing PINP/
dTwo cases and 6 controls were excluded because of missing ALAD data.
Blood Lead and the Risk of ALS 1131
exposure or a long period of increased bone lead release after
cessation of environmental lead exposure (14). It is possible
that a long-term increase in release of lead from bone to
blood, slightly elevating blood lead level, might result in
greater exposure to neural target tissues. The mechanisms
relating lead neurotoxicity to ALS are still unclear. However,
several mechanisms proposed to play a role in ALS patho-
genesis, including oxidative stress, excitotoxicity, and mito-
chondrial dysfunction (24), are also involved in lead
Veterans may be exposed to lead from firing practice (28)
and other military-related sources, so the observed lead-
ALS association may partly explain the higher risk of
ALS noted for military service personnel compared with
the general population (29–32). However, lead exposure
can result from many different sources, including residential
history, nonmilitary occupational history, and hobbies, and
more research is necessary to determine which of these
sources (if any) best explains the observed increased lead
levels in the VALE cases.
The K59N polymorphic variant of the ALAD gene may
affect an individual’s blood and bone lead levels and thus
influence susceptibility to lead exposure (33). In VALE, lead
levels did not differ substantially by genotype in either cases
or controls; however, this result is not surprising given that an
effect of ALAD genotype on blood lead is primarily observed
at much higher blood lead levels (33). We found a significant
lead-ALS association among ALAD1-1 homozygotes but not
ALAD2 carriers, but we did not find a significant interaction
between ALAD genotype and blood lead, consistent with our
previous finding (11). In our previous study, we observed an
level—a finding we did not replicate in VALE (11). Reasons
for this difference between the 2 studies are not apparent,
although the present study had a larger sample size.
VALE has several strengths, including its sample size and
the available information on both bone turnover and ALAD
genotype. In addition, we utilized a highly sensitive assay
for lead measurement, which enabled us to detect small
differences in blood lead levels in a population with a low
absolute level of blood lead.
Limitations should also be appreciated. First, about 22%
of the cases were diagnosed more than 2 years before sam-
ple collection, and they may represent a selected group of
cases with better survival and with blood lead or bone turn-
over levels different from those of other ALS cases. How-
ever, the interval between diagnosis and sample collection
did not substantially affect lead levels, nor did excluding
individuals diagnosed more than 2 years before sample col-
lection alter the results. Second, the exposure data were
collected cross-sectionally, and thus we cannot entirely rule
out reverse causality. However, the fact that the lead-ALS
relation persisted after adjustment for or stratification by
a biomarker of bone resorption suggests that disease-related
lead mobilization from bone, the most likely reason to sus-
pect reverse causality, does not fully explain the association.
Third, we did not evaluate other genes that may modify the
relation between lead exposure and neurologic outcomes,
for example, the hemochromatosis gene, HFE (34). Finally,
residual confounding from other factors could not be com-
pletely ruled out. For example, socioeconomic status might
be associated with both blood lead level and the risk of ALS.
We did not detect a significant association between blood
lead levels and years of education, however, a commonly
used proxy of socioeconomic status, among either the cases
(data available for 100 cases, P ¼ 0.33) or controls (data
available for all controls, P ¼ 0.68).
In summary, we found that elevated blood lead level was
associated with higher odds of ALS among US veterans,
regardless of bone turnover or ALAD genotype. More stud-
ies are needed to support a causal relation between blood
lead and ALS risk. Although it would be difficult to collect
blood lead data prospectively, before ALS symptom onset,
some insight might be achieved by linking ALS to condi-
tions involving abnormal bone resorption.
Author affiliations: Epidemiology Branch, National Insti-
tute of Environmental Health Sciences, National Institutes
of Health, Department of Health and Human Services,
Research Triangle Park, North Carolina (Fang Fang, Dale
P. Sandler, Freya Kamel); Department of Medical Epidemi-
ology and Biostatistics, Karolinska Institutet, Stockholm,
Sweden (Fang Fang, Weimin Ye); Center for Human
Genetics, Department of Medicine, Duke University Medi-
cal Center, Durham, North Carolina (Lydia C. Kwee, Silke
Schmidt); Epidemiology Research and Information Center,
Durham VA Medical Center, Durham, North Carolina
(Lydia C. Kwee, Kelli D. Allen, Eugene Z. Oddone, Silke
Schmidt); Department of Medicine, Duke University Med-
ical Center, Durham, North Carolina (Lydia C. Kwee, Kelli
D. Allen, Eugene Z. Oddone, Silke Schmidt); Biostatistics
Sciences, National Institutes of Health, Department of
Health and Human Services, Research Triangle Park, North
Carolina (David M. Umbach); Social and Scientific Systems
Inc., Durham, North Carolina (Mary Watson); and Westat,
Durham, North Carolina (Jean Keller).
Drs. Silke Schmidt and Freya Kamel contributed equally
to this work.
This work was supported by the Intramural Research Pro-
gram of the National Institutes of Health, National Institute
of Environmental Health Sciences (Z01- ES49005-15) and
by grants from the National Institute of Environmental
Health Sciences/National Institutes of Health (R01 ES
013244) and the ALS Association (ALSA 1230). The
work of the National Registry of Veterans with ALS and
its DNA bank was supported by the Office of Research and
Development, Cooperative Studies Program, Department of
Veterans Affairs (CSP #500A, CSP #478).
The authors thank L. Darwin, the GENEVA study team
(V. Loiacono, C. Stanwyck, C. Williams), and the ALS
registry staff (B. Norman, L. DiMartino, K. Juntilla,
L. Marbrey, B. McCraw, H. Rowe, P. W. Williams) for their
assistance with the VALE study; K. Levine (Research
Triangle Institute) and B. J. Collins (National Toxicology
Program) for measuring blood lead levels; the University of
Connecticut General Clinical Research Center for bone
1132 Fang et al.
Am J Epidemiol 2010;171:1126–1133
turnover biomarker measurements; H. Munger for DNA Download full-text
sample management and genotyping; M. A. Hauser for
genotyping laboratory oversight; J. H. Lindquist and
C. J. Coffman for database management and statistical anal-
yses; and neurologists in the VA ALS Registry group for
reviewing medical records (E. J. Kasarskis, R. S. Bedlack,
J. C. Morgenlander, M. P. Rozear, A. Sabet, L. Sams).
The views expressed in this manuscript are those of the
authors and do not necessarily represent the views of the
Department of Veterans Affairs.
Conflict of interest: none declared.
1. Kamel F, Umbach DM, Hu H, et al. Lead exposure as a risk
factor for amyotrophic lateral sclerosis. Neurodegener Dis.
2. Deapen DM, Henderson BE. A case-control study of amyotro-
phic lateral sclerosis. Am J Epidemiol. 1986;123(5):790–799.
3. McGuire V, Longstreth WT Jr, Nelson LM, et al. Occupational
exposures and amyotrophic lateral sclerosis. A population-
based case-control study. Am J Epidemiol. 1997;145(12):
4. Gunnarsson LG, Bodin L, So ¨derfeldt B, et al. A case-control
study of motor neurone disease: its relation to heritability, and
occupational exposures, particularly to solvents. Br J Ind Med.
5. Campbell AM, Williams ER, Barltrop D. Motor neurone dis-
ease and exposure to lead. J Neurol Neurosurg Psychiatry.
6. Felmus MT, Patten BM, Swanke L. Antecedent events in
in motor neuron disease. Ann Clin Res. 1981;13(2):102–107.
8. Gresham LS, Molgaard CA, Golbeck AL, et al. Amyotrophic
lateral sclerosis and occupational heavy metal exposure:
a case-control study. Neuroepidemiology. 1986;5(1):29–38.
9. Armon C, Kurland LT, Daube JR, et al. Epidemiologic corre-
lates of sporadic amyotrophic lateral sclerosis. Neurology.
10. Chancellor AM, Slattery JM, Fraser H, et al. Risk factors for
motor neuron disease: a case-control study based on patients
from the Scottish Motor Neuron Disease Register. J Neurol
Neurosurg Psychiatry. 1993;56(11):1200–1206.
11. Kamel F, Umbach DM, Lehman TA, et al. Amyotrophic lateral
sclerosis, lead, and genetic susceptibility: polymorphisms in the
delta-aminolevulinic acid dehydratase and vitamin D receptor
genes. Environ Health Perspect. 2003;111(10):1335–1339.
12. Kamel F, Umbach DM, Munsat TL, et al. Lead exposure and
amyotrophic lateral sclerosis. Epidemiology. 2002;13(3):
13. Albers K, Nelson L, Tanner C, et al. Lead exposure and
amyotrophic lateral sclerosis in a Northern California popu-
lation [abstract]. Neuroepidemiology. 2009;33:74.
14. Hu H, Payton M, Korrick S, et al. Determinants of bone and
blood lead levels among community-exposed middle-aged to
elderly men. The Normative Aging Study. Am J Epidemiol.
acid dehydratase genotype and lead toxicity: a HuGE review.
Am J Epidemiol. 2001;154(1):1–13.
16. Allen KD, Kasarskis EJ, Bedlack RS, et al. The National
Registry of Veterans with Amyotrophic Lateral Sclerosis.
17. Brooks BR, Miller RG, Swash M, et al. El Escorial revisited:
revised criteria for the diagnosis of amyotrophic lateral scle-
rosis. Amyotroph Lateral Scler Other Motor Neuron Disord.
18. Schmidt S, Allen KD, Loiacono VT, et al. Genes and Envi-
ronmental Exposures in Veterans with Amyotrophic Lateral
Sclerosis: the GENEVA study. Rationale, study design and
demographic characteristics. Neuroepidemiology. 2008;30(3):
19. Hansen TV, Simonsen MK, Nielsen FC, et al. Collection of
blood, saliva, and buccal cell samples in a pilot study on the
Danish nurse cohort: comparison of the response rate and
quality of genomic DNA. Cancer Epidemiol Biomarkers Prev.
20. Rylander-Rudqvist T, Ha ˚kansson N, Tybring G, et al. Quality
oragene method—a pilot study on the cohort of Swedish men.
Cancer Epidemiol Biomarkers Prev. 2006;15(9):1742–1745.
21. Kamel F, Umbach DM, Munsat TL, et al. Association of
cigarette smoking with amyotrophic lateral sclerosis. Neuro-
22. Weisskopf MG, McCullough ML, Calle EE, et al. Prospective
study of cigarette smoking and amyotrophic lateral sclerosis.
Am J Epidemiol. 2004;160(1):26–33.
23. Nelson LM, McGuire V, Longstreth WT Jr, et al. Population-
based case-control study of amyotrophic lateral sclerosis in
western Washington State. I. Cigarette smoking and alcohol
consumption. Am J Epidemiol. 2000;151(2):156–163.
24. RothsteinJD.Current hypothesesforthe underlyingbiologyof
amyotrophic lateral sclerosis. Ann Neurol. 2009;65(suppl 1):
25. Bondy SC, Komulainen H. Intracellular calcium as an index of
neurotoxic damage. Toxicology. 1988;49(1):35–41.
26. Adonaylo VN, Oteiza PI. Lead intoxication: antioxidant de-
fenses and oxidative damage in rat brain. Toxicology. 1999;
27. Savolainen KM, Loikkanen J, Eerika ¨inen S, et al. Interac-
tions of excitatory neurotransmitters and xenobiotics in
excitotoxicity and oxidative stress: glutamate and lead.
Toxicol Lett. 1998;102-103:363–367.
28. Mancuso JD, McCoy J, Pelka B, et al. The challenge of con-
trolling lead and silica exposures from firing ranges in a spe-
cial operations force. Mil Med. 2008;173(2):182–186.
29. Horner RD, Kamins KG, Feussner JR, et al. Occurrence of
amyotrophic lateral sclerosis among Gulf War veterans.
30. Haley RW. Excess incidence of ALS in young Gulf War vet-
erans. Neurology. 2003;61(6):750–756.
31. Coffman CJ, Horner RD, Grambow SC, et al. Estimating the
occurrence of amyotrophic lateral sclerosis among Gulf War
(1990–1991) veterans using capture-recapture methods.
32. Weisskopf MG, O’Reilly EJ, McCullough ML, et al. Pro-
spective study of military service and mortality from ALS.
33. Hu H, Wu MT, Cheng Y, et al. The delta-aminolevulinic
acid dehydratase (ALAD) polymorphism and bone and
blood lead levels in community-exposed men: the Normative
Aging Study. Environ Health Perspect. 2001;109(8):
34. Onalaja AO, Claudio L. Genetic susceptibility to lead poi-
soning. Environ Health Perspect. 2000;108(suppl 1):23–28.
Blood Lead and the Risk of ALS1133
Am J Epidemiol 2010;171:1126–1133