Effect of plasma homocysteine level and urinary monomethylarsonic acid
on the risk of arsenic-associated carotid atherosclerosis
Meei-Maan Wu
a,b, Hung-Yi Chiou
a,⁎
, Yu-Mei Hsueh
a, Chi-Tzong Hong
c, Che-Long Su
d,
Shu-Feng Chang
d, Wen-Ling Huang
a, Hui-Ting Wang
a, Yuan-Hung Wang
a,
Yi-Chen Hsieh
a, Chien-Jen Chen
ea
School of Public Health, Taipei Medical University, 250 Wu-Hsing Street, Taipei 110, Taiwan, ROC
b
Graduate Institute of Medicine, College of Medicine, Fu-Jen Catholic University, Taipei, Taiwan, ROC
c
Department of Neurology, Wang-Fang Hospital, Taipei Medical University, Taiwan, ROC
dDivision of Neurology, Ming-Jong Hospital, Pingtung, Taiwan, ROC
eGraduate Institute of Epidemiology, College of Public Health, National Taiwan University, Taipei, Taiwan, ROC
Received 27 January 2006; revised 9 May 2006; accepted 9 May 2006 Available online 17 May 2006
Abstract
Arsenic-contaminated well water has been shown to increase the risk of atherosclerosis. Because of involving S-adenosylmethionine, homocysteine
may modify the risk by interfering with the biomethylation of ingested arsenic. In this study, we assessed the effect of plasma homocysteine level and
urinary monomethylarsonic acid (MMA
V) on the risk of atherosclerosis associated with arsenic. In total, 163 patients with carotid atherosclerosis and 163
controls were studied. Lifetime cumulative arsenic exposure from well water for study subjects was measured as index of arsenic exposure.
Homocysteine level was determined by high-performance liquid chromatography (HPLC). Proportion of MMA
V(MMA %) was calculated by dividing
with total arsenic species in urine, including arsenite, arsenate, MMA
V, and dimethylarsinic acid (DMA
V). Results of multiple linear regression analysis
show a positive correlation of plasma homocysteine levels to the cumulative arsenic exposure after controlling for atherosclerosis status and nutritional
factors (P
b 0.05). This correlation, however, did not change substantially the effect of arsenic exposure on the risk of atherosclerosis as analyzed in a
subsequent logistic regression model. Logistic regression analyses also show that elevated plasma homocysteine levels did not confer an independent
risk for developing atherosclerosis in the study population. However, the risk of having atherosclerosis was increased to 5.4-fold (95% CI, 2.0
–15.0) for
the study subjects with high MMA% (≥16.5%) and high homocysteine levels (≥12.7 μmol/l) as compared to those with low MMA% (b9.9%) and low
homocysteine levels (
b12.7 μmol/l). Elevated homocysteinemia may exacerbate the formation of atherosclerosis related to arsenic exposure in
individuals with high levels of MMA% in urine.
© 2006 Elsevier Inc. All rights reserved.
Keywords: Atherosclerosis; Arsenic; Homocysteine; Biomethylation; Risk factors
Introduction
Arsenic is a metalloid element and widely distributed on earth
because of its strong affinity with pyrite and high concentration in
hydrous iron oxides (
Nordstrom, 2002
). Humans are exposed to
arsenic in the environment mainly through groundwater supplies of
drinking water (
WHO, 1981; U.S.PHS, 1989
). Epidemiological
studies in Taiwan have shown that inorganic arsenic from
ground-water is associated with an increased risk of peripheral arterial
disease (
Tseng et al., 1996
), ischemic heart disease (
Chen et al.,
1996
), and cerebral infarction (
Chiou et al., 1997b
). A recent report
also indicated a close association of long-term arsenic exposure
with the progression of carotid atherosclerosis (
Wang et al., 2002
),
an indication of vessel narrowing in carotid artery. The
arsenic-associated vascular manifestation was also observed among the
residents in Chile, Mexico, Poland, and the United States, as well as
vineyard workers in Germany (
Engel and Smith, 1994; Lewis et al.,
1999
). Arsenic may act as an independent risk factor for
athe-rosclerotic vascular diseases in humans aside from the classic risk
factors of cigarette smoking, diabetes, hypertension, and
hyperlip-idemia. The mechanisms by which arsenic induces atherogenesis
⁎ Corresponding author. Fax: +886 2 23779188. E-mail address:[email protected](H.-Y. Chiou).
0041-008X/$ - see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.taap.2006.05.005
are not fully understood. Generation of reactive oxidants has been
related to arsenic toxicity in studies of cell culture and humans
(
Wang and Huang, 1994; Barchowsky et al., 1996; Wu et al., 2001,
2003
). However, the sources of formation of the reactive oxygen
species are not completely elucidated. Recent studies on the
metabolism of arsenic involving biomethylation process conclude
that it may result in the generation of reactive oxygen species and
free radicals during the process (
Yamanaka and Okada, 1994; Del
Razo et al., 2001; Kitchin and Ahmad, 2003
), suggesting a source
of oxidative stress.
Metabolism of inorganic arsenic in human bodies includes
sequential biomethylation processes by alternating reduction of
pentavalent arsenic to trivalent and an addition of a methyl group
to its trivalent form (
Cullen et al., 1984
). S-adenosylmethionine
(SAM) acts as the methyl donor in the arsenic biomethylation
and is subsequently demethylated to S-adenosylhomocysteine
(SAH). Historically, methylation of arsenic has been regarded as
a detoxification pathway because the arsenic metabolites,
mono-methylarsonic acid (MMA
V) and dimethylarsinic acid (DMA
V),
are less toxic than inorganic (arsenite and arsenate) (
Gebel,
2002
). However, recent experimental studies have shown that
the reactive intermediate metabolites, monomethylarsonous acid
(MMA
III) and dimethylarsinous acid (DMA
III), are more toxic
than their parent arsenite in a variety of mammalian cells (
Petrick
et al., 2000; Styblo et al., 2000; Mass et al., 2001; Ahmad et al.,
2002
). In contrast, studies on arsenic-exposed humans
conclud-ed that individuals with a lower capacity to biomethylate arsenic
have a higher risk of developing arsenic-associated diseases (
Del
Razo et al., 1997; Hsueh et al., 1997; Yu et al., 2000; Chen et al.,
2003; Tseng et al., 2005
). Interestingly, all their data also
indi-cated that study subjects with higher urinary MMA
Vpercentage
(MMA%) or lower DMA
Vpercentage (DMA%) suffered from a
higher risk of the reported diseases, including peripheral artery
disease related to atherosclerosis (
Tseng et al., 2005
). Whether
this risk is related to the presence of trivalent methylated
pro-ducts in tissues remains to be eluzcidated. Additionally, the two
sequential stages of methylation efficiency involving different
methylated products in individuals may likely have distinct
fea-tures of health effects.
Moderate elevation of homocysteine level in plasma has
re-cently been proposed as a significant predictor of atherosclerosis
and its related complications (
Hackam and Anand, 2003
).
Me-chanism studies have demonstrated that homocysteine may induce
vascular damage by promoting platelet activation, oxidative stress,
endothelial dysfunction, hypercoagulability, vascular smooth
mus-cle cell proliferation, and endoplasmic reticulum stress (
Lawrence
de Koning et al., 2003
). As homocysteine is produced from the
hydrolysis of SAH (
Finkelstein et al., 1971
), arsenic may contribute
to the increase of homocysteine levels by consuming the SAM pool
and therefore enhance the subsequent cardiovascular risk.
How-ever, this speculation requires careful examination. Plasma
Table 1
Multiple linear regression analyses on plasma homocysteine levelsain relation to homocysteine metabolism factors and cumulative arsenic exposure
Coefficient Standard error Variable (×100) (×100) P value Age (year) 1.12 0.26 b0.001 Gender (male vs. female) 34.4 4.32 b0.001 Folic acid (nmol/l) 0.91 0.46 0.052 Vitamin B12(pmol/l) −0.04 0.01 0.006
Atherosclerosis status (yes vs. no) 0.26 4.48 0.954 Cumulative arsenic exposure (μg/l-year) 0.46 0.19 0.016
a Values are log-transformed. Table 2
Traditional risk factors and carotid atherosclerosis
Characteristics Patients Controls Unadjusted Age –gender-adjusted n (%) n (%) OR (95% CI) OR (95% CI) Age (years) b60 29 (17.8) 71 (43.6) 1.0 1.0 60.0–69.9 79 (48.5) 59 (36.2) 3.3 (1.9–5.7)a 3.2 (1.8–5.5)a ≥70 55 (33.7) 33 (20.2) 4.1 (2.2–7.5)a 4.0 (2.2–7.3)a Gender Female 79 (48.5) 94 (57.7) 1.0 1.0 Male 84 (51.5) 69 (42.3) 1.4 (0.9–2.2) 1.3 (0.8–2.1) Body mass index (kg/m2)
b27 140 (85.9) 139 (86.3) 1.0 1.0 ≥27 23 (14.1) 22 (13.7) 1.0 (0.6–1.9) 1.2 (0.6–2.3) Current smoking No 95 (58.3) 119 (73.0) 1.0 1.0 Yes 68 (41.7) 44 (27.0) 1.9 (1.2–3.1)b 2.1 (1.0–4.4)c Total cholesterol (mg/dl) b200 73 (45.1) 88 (54.3) 1.0 1.0 ≥200 89 (54.9) 74 (45.7) 1.5 (0.9–2.2) 1.6 (1.0–2.5)c HDL cholesterol (mg/dl) b45 19 (14.0) 13 (8.8) 1.0 1.0 ≥45 117 (86.0) 134 (91.2) 0.6 (0.3–1.3) 0.8 (0.4–1.8) LDL cholesterol (mg/dl) b130 61 (44.8) 80 (54.4) 1.0 1.0 ≥130 75 (55.2) 67 (45.6) 1.5 (0.9–2.3) 1.6 (1.0–2.7)c Triglycerides (mg/dl) b130 104 (64.2) 113 (69.8) 1.0 1.0 ≥130 58 (35.8) 49 (30.3) 1.3 (0.8–2.0) 1.4 (0.8–2.2) Hypertension No 100 (61.4) 117 (72.2) 1.0 1.0 Yes 63 (38.7) 45 (27.8) 1.6 (1.0–2.6)c 1.5 (0.9–2.4) Diabetes mellitus No 144 (88.9) 145 (89.5) 1.0 1.0 Yes 18 (11.1) 17 (10.5) 1.0 (0.5–2.2) 1.0 (0.5–2.0) Homocysteine (μmol/l) b12.7 56 (34.4) 82 (50.3) 1.0 1.0 ≥12.7 107 (65.6) 81 (49.7) 1.9 (1.2–3.0)b 1.4 (0.9–2.4) OR, odds ratio; CI, confidence interval; LDL, low-density lipoprotein; HDL, high-density lipoprotein.
Differences from the total number of 163 cases and controls are due to missing values.
a Pb 0.001. b 0.001b P b 0.01. c 0.01 b P b 0.05.
homocysteine concentrations have been examined in Taiwanese
(
Chao et al., 1999; Lin et al., 2002
) and in else ethnic population for
the risk of cardiovascular diseases (
Hackam and Anand, 2003
); this
factor however has not been examined for a disease risk when
combined with arsenic exposure. We attempted to investigate the
additional effect of plasma homocysteine level on the risk of
atherosclerosis related to arsenic exposure. MMA% or DMA% in
urine involved in the two-stage methylation process is also taken
into account while evaluating the atherogenic effect from plasma
homocysteine level added on the risk estimates for arsenic
exposure.
Materials and methods
Study subjects. Study subjects were recruited from the Lanyang Basin of Ilan County in northeastern Taiwan. Research on arseniasis in this area was begun in the early 1990s (Chiou et al., 1997b). Characteristics of the study area, recruitment of the study cohort, baseline data obtained from questionnaire interviews, and determination of arsenic concentrations in well water were described in detail previously (Chiou et al., 1997b, 2001). In brief, the Lanyang Basin is one of two areas, in which well water with high arsenic level is clustered. However, the variation of arsenic concentration in well water from the Basin area is much more striking, ranging from undetectable (b0.15 μg/l) to 3.59 mg/l. During the years of 1991 to 1994, a total of 8088 residents aged≥40 years from 18 villages in four townships were interviewed and included as the study cohort (Chiou et al., 1997b, 2001; Chen et al., 2004). Well water samples were also collected for the determination of arsenic content at that time. In 1997–1998, an initial health examination was carried out for a subsample of 1318 residents (687 and 631 for year 1997 and 1998, respectively) from the cohort, including ultrasonographic assessment of the extracranial carotid artery (ECCA) being conducted. These examinee cohort are younger and more females than the original cohort as we expected for a study of community-based vascular disease. Before the health examination, a follow-up questionnaire was also given to update the information on lifestyle characteristics such as cigarette smoking and alcohol and tea consumption, as well as a detailed history of well water use since the previous interview in 1994. Urine sample and fasting blood were collected and stored at an appropriate temperature until use. A total of 605 examinees (88%) out of the 687 gave their consent to participate the current research project.
To assess the extent of carotid atherosclerosis for study subjects, a Hewlett-Packard SONO 1000 ultrasound system, equipped with a 7.5-MHz real-time B-mode scanner and a 5.6-MHz pulsed-Doppler B-mode scanner was used. The duplex scanning and operation on the participants were described in a previous study (Wang et al., 2002). For future and subsequent off-line analysis, all scans were recorded on super-VHS videotape. Indications of carotid atherosclerosis were evaluated mainly based on 2 indices: the maximal ECCA intimal–medial thickness (IMT) and the presence of ECCA plaque. The maximal IMT was measured in the far side of the common carotid artery (CCA) at the most stenotic location between 0 and 2 cm proximal to the carotid bifurcation. The ECCA plaque was assessed for 5 carotid artery segments, including the proximal CCA (0 to 1 cm proximal to the bifurcation), distal CCA (1 to 2 cm to the bifurcation), bulb, internal carotid artery, and external carotid artery. The presence of ECCA plaque was defined as irregular surface, lumen encroachment, wall thickening≥50% of the adjacent IMT, as well as structure heterogeneity such as acoustic shadow. All the measurements were bilateral, and mean of the measurements was presented for each artery segment for both indices. Patient subjects were diagnosed according to a maximal ECCA IMT of ≥ 1.0 mm or the presence of observable plaque in any of the 5 carotid artery segments. In the initial-stage screening, two hundred and seventy nine subjects with an indication of carotid atherosclerosis were identified (46.1%). This prevalence is slightly higher than that of a previous report in a Taiwanese population (39%) (Wang et al., 2002). This difference in frequency might be due to factors in lifestyle or the high prevalence of stroke in the study area (Chiou et al., 1997b) as yet to be investigated. For the present study, a random sample of 163 patient subjects was selected. These study subjects were not substantially different from the original 279 patients in the distributions of demographic characteristics. Age (±5 years)- and sex-matched controls (n = 163) with no indication of carotid atherosclerosis were chosen from the same cohort who had undergone the ultrasonographic assessment.
The diagnoses of atherosclerosis for all the 326 study subjects were reexamined and confirmed by two of our investigators (C.-L. Su and C.-T. Hong, neurologists).
Index for arsenic exposure. Well water samples were collected from each household, and the arsenic content in well water was determined during 1991– 1994, by a method of hydride-generation atomic absorption spectrometry (Chiou et al., 1997b). To reflect the overall exposure to ingested arsenic for each study subject, cumulative arsenic exposure from drinking well water was applied in addition to the arsenic concentration in well water of the household. The cumulative arsenic exposure was calculated as the sum of the products derived by multiplying the arsenic concentration in well water by the years of drinking well water during the periods of living in one's household throughout the subject's life. Information on the history of well water consumption as well as a detailed residential history were obtained from the baseline questionnaire data and updated from the follow-up questionnaire.
Biochemical variables and homocysteine metabolism assay. B i o c h e m i c a l variables, including total cholesterol, high-density lipoprotein (HDL) cholesterol, low-density lipoprotein (LDL) cholesterol, and triglycerides, were assessed in 1997. All laboratory analyses were performed using a standard automatic analyzer. Height, weight, systolic blood pressure, and diastolic blood pressure were measured according to standard protocols. Hypertension was defined as (1) an average systolic blood pressure of≥140 mm Hg, (2) an average diastolic blood pressure of ≥90 mm Hg, or (3) a history of being diagnosed as hypertensive or having taken antihypertensive medication. Subjects were considered to have diabetes, if they had ever been diagnosed by a physician or had a fasting blood sugar level of≥126 mg/ dl. For measures of total homocysteine level, plasma samples collected in 1997 were thawed and assayed by a method of high-performance liquid chromatography (HPLC) (Durand et al., 1998). Plasma folate and cobalamin levels were quantified using SimulTRAC-SNB Radioassay Kit according to commercial instructions (ICN Pharmaceuticals, Burlingame, CA).
Arsenic species in urine. 5-ml urine samples from each study subject were examined for arsenic speciation, including arsenite, arsenate, MMAVand DMAV,
by a method of HPLC combined with hydride generation AAS as described previously (Chiou et al., 1997a). We calculated the proportion of MMAVor DMAV
of total arsenic species and their metabolites and focused on the effect of MMAVor
DMAV percentage (MMA% or DMA%, respectively) in the risk estimates of
atherosclerosis. Urinary MMAVpercentages of 9.9% and 16.5% were taken as
cut-points, which approximately represent the lower, middle and upper tertiary value of the distribution of control subjects. The corresponding tertiary cut-points for the DMAVanalysis are 71% and 83%, respectively.
Table 3
Arsenic exposure and risk of carotid atherosclerosis
Characteristics Cases Controls Model I Model II n (%) n (%) OR (95% CI) OR (95% CI) Arsenic concentration in well water (μg/l)
≤50.00 25 (15.6) 39 (24.1) 1.0 (reference) 1.0 (reference) 50.01–100.00 46 (28.8) 49 (30.3) 1.6 (0.8–3.1) 1.9 (0.9–3.8) ≥100.01 89 (55.6) 74 (45.7) 2.1 (1.1–3.8)a 2.6 (1.3–5.0)b
Trend across tertiles 1.4 (1.1–1.9)a 1.6 (1.1–2.1)b
Cumulative arsenic exposure (μg/l-year)
≤1.70 34 (21.3) 57 (35.2) 1.0 (reference) 1.0 (reference) 1.71–4.20 43 (26.9) 53 (32.7) 1.5 (0.8–2.7) 1.7 (0.9–3.2) ≥4.21 83 (51.9) 52 (32.1) 2.4 (1.4–4.3)b 2.9 (1.6–5.3)c Trend across tertiles 1.6 (1.2–2.1)b 1.7 (1.3–2.3)c OR, odds ratio; CI, confidence interval. Model I, adjusted for age and gender; model II, model I with the addition of current smoking, total cholesterol, hypertension, and plasma homocysteine level. Differences from the total number of 163 cases and controls are due to missing data.
a 0.01b P b 0.05. b 0.001b P b 0.01. c
Statistical analysis. We first use linear regression method to analyze the relationship between plasma homocysteine level and arsenic exposure while holding constant the plasma levels of folic acid and vitamin B12. These two
nutrition factors are essentially involved in homocysteine metabolism (Lawrence de Koning et al., 2003). In the next atherosclerosis risk analysis, logistic regression model was used to analyze the dependence of disease risk on various risk factors in this study, including arsenic exposure, plasma homocysteine, and traditional risk factors of cardiovascular disease. The effect of a risk factor was expressed as an odds ratio (OR) and a 95% confidence interval (CI). All risk factors under study were defined as categorical variables in the regression model. To evaluate whether there was an interactive effect between plasma homocysteine level and urinary MMA% on the risk of developing carotid atherosclerosis, we estimated the risk associated with homocysteine level according to the lower, middle, or upper tertiary values of MMA% or of DMA%. The interaction of these two factors was assessed using the method, synergy index S, defined bySchlesselman and Stolley (1982). We further evaluated the combined effect of homocysteine level, MMAV percentage, and arsenic exposure on the atherosclerosis risk and therefore classified the study subjects into eight groups according to their respective median values. All analyses were performed using SAS (Win8e) statistical software, and the statistical significance level was defined as Pb 0.05.
Results
Relation of plasma homocysteine level with arsenic exposure
Linear regression coefficient estimates depending on plasma
homocysteine level for arsenic exposure and other predictors
are listed in
Table 1
. A significant positive association was
observed in the aged, male gender and cumulative arsenic
exposure, while the homocysteine level related negatively to
vitamin B
12. No association of homocysteine level was found
with plasma folate and the status of carotid atherosclerosis in the
study subjects.
Traditional risk factors and carotid atherosclerosis
Table 2
shows the frequency distribution and the ORs with the
95% CIs for the classic risk factors in the 163 patients and 163
controls. Aging and current smoking were risk factors with the
strongest effects on carotid atherosclerosis in this study population.
Total cholesterol and LDL cholesterol were significantly higher in
case subjects as compared with controls. In contrast, the effects of
hypertension and plasma homocysteine level lost significance after
adjusting for age and gender differences in the distribution between
cases and controls. Other factors, including BMI, HDL cholesterol
or triglycerides, and diabetes, revealed no evidence of an
associa-tion with the development of carotid atherosclerosis in these study
subjects.
Association of arsenic exposure with carotid atherosclerosis
To assess the risk of carotid atherosclerosis associated with
levels of arsenic exposure, we first divided both indices of arsenic
exposure into tertiles according to the distribution of the controls
and then examined the trend of the ORs across the tertiles
(
Table 3
). As shown in the table, the age–gender-adjusted analysis
demonstrated a significantly higher risk of carotid atherosclerosis
Table 4
Interaction between plasma homocysteine levels and monomethylarsonic acid percentage (MMA%) for the risk of carotid atherosclerosis Homocysteine
level
MMA%b 9.9 9.9≤ MMA% b16.5 MMA%≥16.5
Patient Control Adjusted OR Patient Control Adjusted OR Patient Control Adjusted OR (μmol/l) n (%) n (%) (95% CI) n (%) n (%) (95% CI) n (%) n (%) (95% CI) b12.7 20 (38.5) 29 (52.7) 1.0 23 (43.4) 21 (40.4) 1.0 12 (21.8) 32 (58.2) 1.0
≥12.7 32 (61.5) 26 (47.3) 0.8 (0.3–2.1) 30 (56.6) 31 (59.6) 0.9 (0.4–2.3) 43 (78.2) 23 (41.8) 5.4 (2.0–15.0)a OR, odds ratio; CI, confidence interval.
Model was adjusted for age, gender, current smoking, total cholesterol, hypertension, and cumulative arsenic exposure.
a Pb 0.05 for the comparison between the strata of high or low homocysteine level.
Fig. 1. Adjusted odds ratios (aOR) of atherosclerosis risk by cumulative arsenic exposure, plasma homocysteinemia level, and urinary monomethylarsonic acid percentage (MMA%). The reference group was the study subjects who were exposed to low cumulative arsenic exposure (≤1.7 μg/l-year), low plasma homocysteine level (b12.7 μmol/l), and had low MMA% (b13.4%). Data have been adjusted for age, gender, current smoking, total cholesterol, and history of hypertension. P for a trend test among Groups I to II: 0.006.
in the upper tertile of arsenic concentration in well water compared
with the first tertile (OR, 2.1; 95% CI, 1.1–3.8). Adjusting for
current smoking, total cholesterol, hypertension, and plasma
homocysteine level did not attenuate the relationship (OR, 2.6;
95% CI, 1.3
–5.0). There was also a significant association
bet-ween arsenic and carotid atherosclerosis using cumulative arsenic
exposure as an index of the exposure level in this population (OR,
2.9; 95% CI, 1.6–5.3 after multivariate adjustment). The linear
trends across the tertiles were significant for all models (P
b 0.05).
Interaction between plasma homocysteine and urinary MMA%
or DMA%
As indicated above, the distribution of high or low
homo-cysteine levels was not statistically different between control and
patient groups in the age–sex-adjusted analysis in
Table 2
.
How-ever, when we further perform a stratified analysis, according to
urinary MMA% of study subjects, the association between
homo-cysteine levels and atherosclerosis risk was different in strata of
lower, middle and upper urinary MMA%, indicating a possible
interaction in risk modification. As shown in
Table 4
, in subjects
with urinary MMA% above the upper tertiary value of 16.5%,
elevated plasma homocysteine level was significantly associated
with a 5.4-fold increased risk (95% CI, 2.0–15.0) for carotid
atherosclerosis. In contrast, the risk from the high plasma
homo-cysteine level was not increased in the subjects with urinary MMA
% less than 16.5% (OR, 0.8; 95% CI, 0.3–2.1, and OR, 0.9; 95%
CI, 0.4–2.3, for the lower and middle tertiary group, respectively).
On the other hand, no biological gradient among the strata of low,
middle or high DMA
Vpercentages in urine samples (data not
shown) is observed. The synergistic index (S = 0.95) did not reach
statistical significance (χ
2test, P = 0.162) in interaction estimates.
Combined effect of plasma homocysteine, urinary MMA%, and
arsenic exposure
In a multivariate logistic regression analysis (
Fig. 1
), the risk
of carotid atherosclerosis was estimated for each combination of
arsenic exposure, plasma homocysteine, and urinary MMA%,
using exposure to low arsenic, low homocysteine level, and low
MMA
Vpercentage as the reference group. As is expected and
shown in the
Fig. 1
, arsenic alone is a major risk factor in this
study population (OR, 1.7; 95% CI, 0.6–5.2). Addition of high
homocysteine level and high MMA
Vpercentage further
in-creased the risk ratio to the arsenic-exposed individuals by 60%
(OR, 2.7; 95% CI, 1.0–7.8). A trend test indicates that the
atherosclerosis risk increases along with the accumulating
number of the three risk factors (P for trend: 0.006).
Discussion
Our observation that carotid atherosclerosis is associated
with ingested arsenic from well water is consistent with the
results of our previous study carried out on a different
arsenic-exposed population in southwestern Taiwan (
Wang et al., 2002
).
In the current study, we further tested the hypothesis that arsenic
exposure increases plasma homocysteine level and the
sub-sequent risk for carotid atherosclerosis. We examined changes
in plasma homocysteine levels of 326 arsenic-exposed study
subjects and found that the homocysteine levels were positively
correlated to the cumulative arsenic exposure through drinking
well water. However, this correlation did not change
substan-tially the independent effect of arsenic exposure on the risk of
atherosclerosis in the study population. The adjusted OR
(1.7-fold) for the effect of cumulative arsenic exposure was
nonetheless statistically significant after controlling for plasma
homocysteine level. There was only a slight change of OR from
1.8-fold, the corresponding value by dropping the
homocys-teine variable from the full fitted model II in
Table 3
(data not
shown). We also found that the levels of plasma homocysteine
were not statistically related to the risk of carotid atherosclerosis
in these study subjects. In other words, arsenic exposure might
have an effect on plasma homocysteine levels in the study
subjects, yet the biological significance of this correlation
remains to be elucidated. Arsenic, acting as an independent risk
factor after adjustment for other potential confounding factors,
including plasma homocysteine level, should, at least, have
exerted on a distinct causal pathway.
Several possible mechanisms of arsenic-induced
atheroscle-rosis have been recently proposed based on experimental data and
epidemiological evidence (
Kitchin, 2001; Simeonova and Luster,
2004
). Accumulating evidence demonstrated that arsenic could
cause cellular redox alteration, impaired nitric oxide (NO)
ho-meostasis, and enhanced coagulation activity, which are relevant
to the dysfunction of endothelial cells (
Simeonova and Luster,
2004
). Endothelial dysfunction is thought to be an early event in
atherosclerosis progression (
Libby et al., 2002
), resulting in
in-flammatory cell infiltration and platelet-thrombus formation (
Si-meonova and Luster, 2004
). Exposure of endothelial cells to
arsenite has been shown to induce NF-κB activation through
re-active oxygen species (
Barchowsky et al., 1996, 1999
). It has
been demonstrated that arsenite induces expression of genes
en-coding for inflammatory mediators including MCP-1, IL-6 and
IL-8 (
Simeonova et al., 2003; Lee et al., 2005
). Promoter regions
of these genes contain multiple binding sites for the NF-κB
tran-scription factor. It has also been reported that arsenic increases
cyclooxygenase-2 protein expression through peroxynitrite
ge-neration (
Bunderson et al., 2002
), suggesting a link between
re-active nitrogen species and arsenic-induced inflammatory states.
More recently, Bunderson et al. have also reported an association
of arsenic-induced atherosclerosis with the increased expression
of prostacyclin in experimental animals (
Bunderson et al., 2004
).
Our and other reports based on arsenic-exposed human study
subjects also support these laboratory findings (
Wu et al., 2001,
2003; Pi et al., 2002
). Taken together, arsenic-associated vascular
disorders found in humans may likely arise from changes in
ex-pression levels of a variety of genes that participate in
athe-rosclerosis through a mechanism of oxidative interference.
Most, though not all, observational studies have shown that
moderately elevated plasma homocysteine levels are associated
with an increased risk for premature atherosclerosis and
throm-botic disease (
Hackam and Anand, 2003; Fruchart et al., 2004
). In
experimental animals, homocysteine has been reported to
acce-lerate atherosclerosis and amplify proatherogenic processes when
combined with other risk factors of cardiovascular disease, such as
hyperlipidemia and hypertension (
Matthias et al., 1996; Wang
et al., 2003
). Consistent with this finding, some studies have
contended that the observed association between
homocysteine-mia and atherosclerotic events is not independent of conventional
cardiovascular risk factors (
Collaboration, 2002; Kolling et al.,
2004
). In the present study, we found no independent predictive
role of elevated homocysteine levels in the risk of carotid
athe-rosclerosis in this arsenic-exposed population. However, the risk
estimates OR for elevated plasma homocysteine level were
dose-dependent on the MMA% in urine samples, which indicated
hete-rogeneity among the study subjects. The increased efficiency of
metabolic methylation from arsenite to MMA
Vor reduced
ef-ficiency from MMA
Vto DMA
Vin individuals might somehow be
interrelated to plasma homocysteine levels in the development of
atherosclerosis. Biomethylation is the major pathway for
metab-olism of arsenic, during which ROS and free radicals are
concur-rently produced (
Cullen et al., 1984
). Study subjects with higher
levels of MMA% are supposedly at greater increased risk from
oxidative injuries to the vascular system. Homocysteine
concentra-tions at pathological or physiological levels have been shown to
decrease the activity of glutathione peroxidase-1 (GPx-1) (
Handy
et al., 2005
), an antioxidant enzyme. Accumulation of the ROS free
radicals and subsequent oxidative damages contributing to
athe-rosclerosis risk might therefore be enhanced. Whether GPx-1 enzyme
activity is involved in the atherosclerosis in the study subjects
with high levels of plasma homocysteine and high levels of MMA
% after arsenic exposure, however, requires further examination.
On the other hand, there is no biological gradient in risk
among the strata of low, middle or high DMA% in urine (data not
shown). DMA
III-induced carcinogenesis has been described in
animal models (
Yamanaka and Okada, 1994
). Although the
trivalent methylated metabolites have been detected in urine of
humans chronically exposed to arsenic, their associations with
disease risk remains to be elucidated (
Vahter, 2002
). Several
indices of methylation efficiency have been used in previous
reports on exposed humans. Higher MMA% or lower DMA% is
the most consistent predictors among the indices for the risk of
arsenic-associated diseases (
Del Razo et al., 1997; Yu et al., 2000;
Chen et al., 2003; Tseng et al., 2005
). It is unclear how an increased
concentration of MMA
Vrelative to DMA
Vwould contribute
significantly to an increased risk of arsenic-induced health effects.
Although the mechanisms remain not fully elucidated, formation
of MMA
IIIand the by-product SAH during the first methylation
step may provide possible explanations (
Buchet and Lauwerys,
1988; Thompson, 1993; Yi et al., 2000; Drobna et al., 2005
).
MMA
IIIis a reactive product harmful to tissues, and the SAH may
inhibit the second step of methylation process; the latter of which
could also account for the lower percentage of DMA
Vobserved in
humans at higher risk in the same studies. In this study, no
correlation of increased risk in parallel with lower DMA% was
beyond our expectation. Alternatively, data variation because of
small sample size may also explain. More population-based
studies are needed to examine the contribution of each methylated
metabolite to the observed risk following exposure to arsenic.
To assess the additional risk of atherosclerosis from the joint
effect of plasma homocysteine and MMA%, on the top of the
arsenic-exposed individuals, we calculated combined risk of the
three risk factors and compare it to the group of low arsenic
exposure, low plasma homocysteine, and low MMA%. Although
arsenic alone could cause atherosclerosis in the carotid arteries of
study subjects, a combination of high plasma homocysteine and
high MMA% may further add a risk of
∼60% (from 1.7- to
2.7-fold) to the arsenic-exposed individuals. Elevated plasma
homo-cysteine level may result from low consumption of folic acid or
vitamin B
12or of both in Western populations (
Selhub et al.,
1993
). In our study on an oriental population, we also found a
significantly inverse association between plasma levels of
homo-cysteine and vitamin B
12. However, effect of the folate on
homo-cysteine levels was not found in this population. Perhaps inherent
or acquired heterogeneity of study subjects resulting in different
risk profiles in the population studied. Although a relatively small
risk for vascular disease may be difficult to detect, the combined
effect from elevated homocysteine level and high MMA% may
still increase a significant risk for atherosclerosis. Like many
population-based studies, the observed correlation of the three
factors in this population may have occurred as a random event as
well. Whether there is causal relatedness should be further
iden-tified by experimental animals or confirmed by human data from
different populations.
There are some potential limitations of this study. First, genetic
variants of homocysteine metabolism enzymes factors were not
determined for subjects, which might have contributed to some of
the unexplained variation in this study. The association between
homocysteine levels and atherosclerosis risk may be thus
under-estimated. Second, as plasma collection for the assay of
homocys-teine level was conducted at the almost same time as the assessment
of ECCA in each study subject, the induction period for the
acce-leration of an atherosclerotic event due to homocysteine imposition
might not have been long enough. A follow-up health examination
on the study subjects in the future may overcome this limitation.
Third, a larger sample size is needed to adjust for the genetic and
nongenetic influences of the disease while assessing the effect of
plasma homocysteine levels on atherosclerosis risk. In particular,
justifying a small to moderate effect of homocysteine in the
pre-sence of a strong environmental risk factor such as arsenic requires
data from large-scale studies.
In conclusion, this study demonstrated that long-term
exposure to arsenic from well water is significantly associated
with an increased risk of developing carotid atherosclerosis, and
that the coexistence of high homocysteinemia level and high
urinary MMA% may exacerbate atherosclerosis formation
caused by arsenic in the carotid artery in humans. Factors
involved in arsenic methylation, particularly the genetic
make-up of the methyltransferase or reducing enzymes in the
formation of MMA
III, as well as the genetic or nongenetic
factors in the homocysteine metabolism likely act as risk
modifiers in the development of atherosclerosis associated with
arsenic. The proposition that GPx-1 enzyme may be interfered
in association with homocysteine level needs to be further tested
in human subjects. More studies on exposed humans or
experimental animals are warranted to confirm the observed
correlation of the combination of arsenic exposure
homocys-teinemia and high MMA% levels in this study.
Acknowledgments
This work was supported by grants
NSC91-3112-B-B10-006,NSC92-2811-B-038-002, NSC93-2321-B-038-014, and
NSC94-2321-B-038-004 from the National Science Council
of Taiwan, ROC. Additional supports were received from the
Topnotch Stroke Research Center, Ministry of Education, and
from the center of Excellence for Clinical Trial and Research in
Neurology Specialty, Department of Health, Executive Yuan of
Taiwan, ROC.
References
Ahmad, S., Kitchin, K.T., Cullen, W.R., 2002. Plasmid DNA damage caused by methylated arsenicals, ascorbic acid and human liver ferritin. Toxicol. Lett. 133, 47–57.
Barchowsky, A., Dudek, E.J., Treadwell, M.D., Wetterhahn, K.E., 1996. Arsenic induces oxidant stress and NF-kappa B activation in cultured aortic endothelial cells. Free Radic. Biol. Med. 21, 783–790.
Barchowsky, A., Klei, L.R., Dudek, E.J., Swartz, H.M., James, P.E., 1999. Stimulation of reactive oxygen, but not reactive nitrogen species, in vascular endothelial cells exposed to low levels of arsenite. Free Radic. Biol. Med. 27, 1405–1412.
Buchet, J.P., Lauwerys, R., 1988. Role of thiols in the in-vitro methylation of inorganic arsenic by rat liver cytosol. Biochem. Pharmacol. 37, 3149–3153. Bunderson, M., Coffin, J.D., Beall, H.D., 2002. Arsenic induces peroxynitrite generation and cyclooxygenase-2 protein expression in aortic endothelial cells: possible role in atherosclerosis. Toxicol. Appl. Pharmacol. 184, 11–18. Bunderson, M., Brooks, D.M., Walker, D.L., Rosenfeld, M.E., Coffin, J.D.,
Beall, H.D., 2004. Arsenic exposure exacerbates atherosclerotic plaque formation and increases nitrotyrosine and leukotriene biosynthesis. Toxicol. Appl. Pharmacol. 201, 32–39.
Chao, C.L., Tsai, H.H., Lee, C.M., Hsu, S.M., Kao, J.T., Chien, K.L., Sung, F.C., Lee, Y.T., 1999. The graded effect of hyperhomocysteinemia on the severity and extent of coronary atherosclerosis. Atherosclerosis 147, 379–386. Chen, C.J., Chiou, H.Y., Chiang, M.H., Lin, L.J., Tai, T.Y., 1996. Dose-response
relationship between ischemic heart disease mortality and long-term arsenic exposure. Arterioscler. Thromb. Vasc. Biol. 16, 504–510.
Chen, Y.C., Su, H.J., Guo, Y.L., Hsueh, Y.M., Smith, T.J., Ryan, L.M., Lee, M.S., Christiani, D.C., 2003. Arsenic methylation and bladder cancer risk in Taiwan. Cancer Causes Control 14, 303–310.
Chen, C.L., Hsu, L.I., Chiou, H.Y., Hsueh, Y.M., Chen, S.Y., Wu, M.M., Chen, C.J., 2004. Ingested arsenic, cigarette smoking, and lung cancer risk: a follow-up study in arseniasis-endemic areas in Taiwan. JAMA 292, 2984–2990. Chiou, H.Y., Hsueh, Y.M., Hsieh, L.L., Hsu, L.I., Hsu, Y.H., Hsieh, F.I., Wei, M.L.,
Chen, H.C., Yang, H.T., Leu, L.C., Chu, T.H., Chen-Wu, C., Yang, M.H., Chen, C.J., 1997a. Arsenic methylation capacity, body retention, and null genotypes of glutathione S-transferase M1 and T1 among current arsenic-exposed residents in Taiwan. Mutat. Res. 386, 197–207.
Chiou, H.Y., Huang, W.I., Su, C.L., Chang, S.F., Hsu, Y.H., Chen, C.J., 1997b. Dose-response relationship between prevalence of cerebrovascular disease and ingested inorganic arsenic. Stroke 28, 1717–1723.
Chiou, H.Y., Chiou, S.T., Hsu, Y.H., Chou, Y.L., Tseng, C.H., Wei, M.L., Chen, C.J., 2001. Incidence of transitional cell carcinoma and arsenic in drinking water: a follow-up study of 8,102 residents in an arseniasis-endemic area in northeastern Taiwan. Am. J. Epidemiol. 153, 411–418.
Collaboration, H.S., 2002. Homocysteine and risk of ischemic heart disease and stroke: a meta-analysis. JAMA 288, 2015–2022.
Cullen, W.R., McBride, B.C., Pickett, A.W., Reglinski, J., 1984. The wood preservative chromated copper arsenate is a substrate for trimethylarsine biosynthesis. Appl. Environ. Microbiol. 47, 443–444.
Del Razo, L.M., Garcia-Vargas, G.G., Vargas, H., Albores, A., Gonsebatt, M.E., Montero, R., Ostrosky-Wegman, P., Kelsh, M., Cebrian, M.E., 1997. Altered profile of urinary arsenic metabolites in adults with chronic arsenicism. A pilot study. Arch. Toxicol. 71, 211–217.
Del Razo, L.M., Quintanilla-Vega, B., Brambila-Colombres, E., Calderon-Aranda, E.S., Manno, M., Albores, A., 2001. Stress proteins induced by arsenic. Toxicol. Appl. Pharmacol. 177, 132–148.
Drobna, Z., Waters, S.B., Devesa, V., Harmon, A.W., Thomas, D.J., Styblo, M., 2005. Metabolism and toxicity of arsenic in human urothelial cells expressing rat arsenic ( + 3 oxidation state)-methyltransferase. Toxicol. Appl. Pharmacol. 207, 147–159.
Durand, P., Prost, M., Blache, D., 1998. Folate deficiencies and cardiovascular pathologies. Clin. Chem. Lab. Med. 36, 419–429.
Engel, R.R., Smith, A.H., 1994. Arsenic in drinking water and mortality from vascular disease: an ecologic analysis in 30 counties in the United States. Arch. Environ. Health 49, 418–427.
Finkelstein, J.D., Kyle, W., Harris, B.J., 1971. Methionine metabolism in mammals. Regulation of homocysteine methyltransferases in rat tissue. Arch. Biochem. Biophys. 146, 84–92.
Fruchart, J.C., Nierman, M.C., Stroes, E.S., Kastelein, J.J., Duriez, P., 2004. New risk factors for atherosclerosis and patient risk assessment. Circulation 109, III15–III19.
Gebel, T.W., 2002. Arsenic methylation is a process of detoxification through accelerated excretion. Int. J. Hyg. Environ. Health 205, 505–508. Hackam, D.G., Anand, S.S., 2003. Emerging risk factors for atherosclerotic
vascular disease: a critical review of the evidence. JAMA 290, 932–940. Handy, D.E., Zhang, Y., Loscalzo, J., 2005. Homocysteine down-regulates
cellular glutathione peroxidase (GPx1) by decreasing translation. J. Biol. Chem. 280, 15518–15525.
Hsueh, Y.M., Chiou, H.Y., Huang, Y.L., Wu, W.L., Huang, C.C., Yang, M.H., Lue, L.C., Chen, G.S., Chen, C.J., 1997. Serum beta-carotene level, arsenic methyla-tion capability, and incidence of skin cancer. Cancer Epidemiol. Biomark. Prev. 6, 589–596.
Kitchin, K.T., 2001. Recent advances in arsenic carcinogenesis: modes of action, animal model systems, and methylated arsenic metabolites. Toxicol. Appl. Pharmacol. 172, 249–261.
Kitchin, K.T., Ahmad, S., 2003. Oxidative stress as a possible mode of action for arsenic carcinogenesis. Toxicol. Lett. 137, 3–13.
Kolling, K., Ndrepepa, G., Koch, W., Braun, S., Mehilli, J., Schomig, A., Kastrati, A., 2004. Methylenetetrahydrofolate reductase gene C677T and A1298C polymorphisms, plasma homocysteine, folate, and vitamin B12 levels and the extent of coronary artery disease. Am. J. Cardiol. 93, 1201–1206.
Lawrence de Koning, A.B., Werstuck, G.H., Zhou, J., Austin, R.C., 2003. Hyperhomocysteinemia and its role in the development of atherosclerosis. Clin. Biochem. 36, 431–441.
Lee, P.C., Ho, I.C., Lee, T.C., 2005. Oxidative stress mediates sodium arsenite-induced expression of heme oxygenase-1, monocyte chemoattractant protein-1, and interleukin-6 in vascular smooth muscle cells. Toxicol. Sci. 85, 541–550.
Lewis, D.R., Southwick, J.W., Ouellet-Hellstrom, R., Rench, J., Calderon, R.L., 1999. Drinking water arsenic in Utah: a cohort mortality study. Environ. Health Perspect.. 107, 359–365.
Libby, P., Ridker, P.M., Maseri, A., 2002. Inflammation and atherosclerosis. Circulation 105, 1135–1143.
Lin, J.S., Shen, M.C., Cheng, W.C., Tsay, W., Wang, Y.C., Lin, B.B., Hung, M.H., 2002. Age, sex and vitamin status affect plasma level of homocysteine, but hyperhomocysteinaemia is possibly not an important risk factor for venous thrombophilia in Taiwanese Chinese. Br. J. Haematol. 117, 159–163. Mass, M.J., Tennant, A., Roop, B.C., Cullen, W.R., Styblo, M., Thomas, D.J.,
Kligerman, A.D., 2001. Methylated trivalent arsenic species are genotoxic. Chem. Res. Toxicol. 14, 355–361.
Matthias, D., Becker, C.H., Riezler, R., Kindling, P.H., 1996. Homocysteine induced arteriosclerosis-like alterations of the aorta in normotensive and hypertensive rats following application of high doses of methionine. Atherosclerosis 122, 201–216. Nordstrom, K., 2002. Worldwide occurrences of arsenic in ground water. Science
296, 2143–2144.
Petrick, J.S., Ayala-Fierro, F., Cullen, W.R., Carter, D.E., Vasken Aposhian, H., 2000. Monomethylarsonous acid (MMA(III)) is more toxic than arsenite in Chang human hepatocytes. Toxicol. Appl. Pharmacol. 163, 203–207. Pi, J., Yamauchi, H., Kumagai, Y., Sun, G., Yoshida, T., Aikawa, H.,
by chronic exposure of Chinese residents to arsenic contained in drinking water. Environ. Health Perspect. 110, 331–336.
Schlesselman, J.J., Stolley, P.D., 1982. Case-Control Studies: Design, Conduct, Analysis. Oxford University Press, New York.
Selhub, J., Jacques, P.F., Wilson, P.W., Rush, D., Rosenberg, I.H., 1993. Vitamin status and intake as primary determinants of homocysteinemia in an elderly population. JAMA 270, 2693–2698.
Simeonova, P.P., Luster, M.I., 2004. Arsenic and atherosclerosis. Toxicol. Appl. Pharmacol. 198, 444–449.
Simeonova, P.P., Hulderman, T., Harki, D., Luster, M.I., 2003. Arsenic exposure accelerates atherogenesis in apolipoprotein E(−/−) mice. Environ. Health Perspect. 111, 1744–1748.
Styblo, M., Del Razo, L.M., Vega, L., Germolec, D.R., LeCluyse, E.L., Hamilton, G.A., Reed, W., Wang, C., Cullen, W.R., Thomas, D.J., 2000. Comparative toxicity of trivalent and pentavalent inorganic and methylated arsenicals in rat and human cells. Arch. Toxicol. 74, 289–299.
Thompson, D.J., 1993. A chemical hypothesis for arsenic methylation in mammals. Chem. Biol. Interact. 88, 14–89.
Tseng, C.H., Chong, C.K., Chen, C.J., Tai, T.Y., 1996. Dose-response relationship between peripheral vascular disease and ingested inorganic arsenic among residents in blackfoot disease endemic villages in Taiwan. Atherosclerosis 120, 125–133.
Tseng, C.H., Huang, Y.K., Huang, Y.L., Chung, C.J., Yang, M.H., Chen, C.J., Hsueh, Y.M., 2005. Arsenic exposure, urinary arsenic speciation, and peripheral vascular disease in blackfoot disease-hyperendemic villages in Taiwan. Toxicol. Appl. Pharmacol. 206, 299–308.
U.S.PHS, 1989. Toxicological Profile for Arsenic. Public Health Service, Wa-shington, DC.
Vahter, M., 2002. Mechanisms of arsenic biotransformation. Toxicology 181–182, 211–217.
Wang, T.S., Huang, H., 1994. Active oxygen species are involved in the induction of micronuclei by arsenite in XRS-5 cells. Mutagenesis 9, 253–257.
Wang, C.H., Jeng, J.S., Yip, P.K., Chen, C.L., Hsu, L.I., Hsueh, Y.M., Chiou, H.Y., Wu, M.M., Chen, C.J., 2002. Biological gradient between long-term arsenic exposure and carotid atherosclerosis. Circulation 105, 1804–1809.
Wang, H., Jiang, X., Yang, F., Gaubatz, J.W., Ma, L., Magera, M.J., Yang, X., Berger, P.B., Durante, W., Pownall, H.J., Schafer, A.I., 2003. Hyperhomocys-teinemia accelerates atherosclerosis in cystathionine beta-synthase and apolipoprotein E double knock-out mice with and without dietary perturbation. Blood 101, 3901–3907.
WHO, 1981. Environmental Health Criteria 18: Arsenic. World Health Organi-zation, Geneva.
Wu, M.M., Chiou, H.Y., Wang, T.W., Hsueh, Y.M., Wang, I.H., Chen, C.J., Lee, T.C., 2001. Association of blood arsenic levels with increased reactive oxi-dants and decreased antioxidant capacity in a human population of north-eastern Taiwan. Environ. Health Perspect. 109, 1011–1017.
Wu, M.M., Chiou, H.Y., Ho, I.C., Chen, C.J., Lee, T.C., 2003. Gene expression of inflammatory molecules in circulating lymphocytes from arsenic-exposed human subjects. Environ. Health Perspect. 111, 1429–1438.
Yamanaka, K., Okada, S., 1994. Induction of lung-specific DNA damage by metabolically methylated arsenics via the production of free radicals. Environ. Health Perspect. 102 (Suppl 3), 37–40.
Yi, P., Melnyk, S., Pogribna, M., Pogribny, I.P., Hine, R.J., James, S.J., 2000. Increase in plasma homocysteine associated with parallel increases in plasma S-adenosylhomocysteine and lymphocyte DNA hypomethylation. J. Biol. Chem. 275, 29318–29323.
Yu, R.C., Hsu, K.H., Chen, C.J., Froines, J.R., 2000. Arsenic methylation capacity and skin cancer. Cancer Epidemiol. Biomark. Prev. 9, 1259–1262.
