Changes in Exposure to Arsenic Following the Installation of an Arsenic Removal Treatment in a Small Community Water System

Abstract

Arsenic in drinking water poses a threat to public health world-wide. In March 2001, the EPA revised the maximum contaminant level (MCL) for arsenic in drinking water downward from 50 µg/L to 10 µg/L and required all U.S. small community water systems (CWSs) and non-community water systems (NCWSs) to comply by 23 January 2006. Much of the financial burden associated with complying with and maintaining this new drinking water MCL was shouldered by local community governments. For example, the Walker River Paiute Tribe operated a CWS on the Walker River Paiute Indian Reservation that needed upgrading to meet the new arsenic MCL. In collaboration with the Walker River Paiute Tribe, we conducted a study to assess whether reducing the arsenic concentration in drinking water to meet the new MCL reduced the arsenic body burden in local community members who drank the water. Installing a drinking water treatment to remove arsenic dramatically reduced both the drinking water concentrations (to below the current EPA MCL of 10 µg/L) and the community members’ urinary concentrations of total As, AsIII, and AsV within a week of its full implementation. Additional assistance to small water systems to sustain new drinking water treatments may be warranted.

1. Introduction

Arsenic is a toxic element that occurs naturally in the Earth’s crust, and inorganic arsenic compounds occur in water, soil, and food. Studies have noted differences in relative potency, with trivalent arsenites [As (III)] tending to be more toxic than pentavalent arsenates [As (V)] [1]. However, in most cases, the differences in the relative potency are reasonably small (about 2- to 3-fold), and the different forms of arsenic may be interconverted, both in the environment and in the human body [1]. One organic form of arsenic, called arsenobetaine (AsB), is found in fish and shellfish and is much less harmful than the inorganic forms [1]. In the United States, it is estimated that 90% of arsenic exposure comes from seafood consumption [2].

In regions where arsenic concentrations are particularly high in the Earth’s crust, arsenic can be found in groundwater, seafood, plants, and soil. Of the various sources of exposure, arsenic in drinking water poses the greatest threat to public health world-wide [3]. Chronic exposure to arsenic can lead to arsenic poisoning, and has been linked to a variety of cancers, including lung and bladder cancer. Other health effects that can take many years to develop from chronic arsenic exposure include skin keratosis, gastrointestinal symptoms, and cardiovascular disease [4,5,6,7]. Acute exposure to arsenic in drinking water can be fatal [8].

To protect public health in the United States, the Environmental Protection Agency (EPA) promulgated a maximum contaminant level (MCL) of 50 µg/L for arsenic in drinking water provided by community water systems (CWS) and non-community water systems (NCWS). A CWS is a public water system that supplies water to the same population year-round, including most cities and towns, apartment buildings, and mobile home parks with their own water supplies; a NCWS is a public water system that serves at least 25 of the same people more than six months of the year, such as schools, churches, nursing homes, and factories [9].

Since the original arsenic rule was adopted, reports from Taiwan and other areas showed that arsenic exposure at concentrations lower than 50 µg/L was associated with increased risk of lung and bladder cancer, although the relationships are potentiated by smoking [10,11]. In response, EPA revised the MCL for arsenic downward from 50 µg/L to 10 µg/L in March 2001. By 23 January 2006, all CWSs and NCWSs in the United States had to comply with the new standard. EPA estimated that there were 3000 CWSs (serving approximately 11 million people) and 1100 NCWSs (serving approximately 2 million people) that needed to take action to meet the new MCL for arsenic. EPA also estimated that over 177 million dollars would be spent nationally for treatment costs, primarily for installing and operating treatment technologies to remove arsenic [12].

EPA [13] published a handbook to help CWSs determine how to address high levels of arsenic in finished drinking water. For example, finished water with lower levels of arsenic can be produced by abandoning the contaminated source and replacing it with an uncontaminated source or blending the finished water with uncontaminated water before moving it to the distribution system. The handbook [13] also provides recommendations for enhancing existing treatment processes such as enhancing coagulation/filtration; adding treatment processes such as ion exchange or activated alumina; or using point-of-use treatment systems that use reverse osmosis, iron-based sorbent, or activated alumina to remove arsenic. The choice of treatment options depends on the concentration of arsenic in raw water, resources available to install additional treatments, funding to sustain the additional treatment, and availability of an acceptable method to dispose of treatment process residuals [13].

While the EPA handbook [13] provided CWSs with options to remove arsenic from drinking water, it did not provide methods to test whether the arsenic removal effectively reduced exposure in the community members drinking the water. The National Research Council (NRC) (US) Subcommittee on Arsenic in Drinking Water published Biomarkers of Arsenic Exposure [14], which includes methods to assess exposure to arsenic, including analysis of hair and nails for more long-term exposures and urine for short-term exposures. The NRC publication includes studies verifying human exposures to arsenic; however, it does not include any studies that used urinary arsenic to examine changes in arsenic concentrations associated with changes in drinking water treatment. In our study, arsenic concentrations in urine were used to assess arsenic exposure before and after drinking water treatment to remove arsenic was implemented.

A considerable amount of the financial burden to install and operate arsenic removal technologies was shouldered by the communities with small CWSs [12]. Further, there were no published studies that examined whether compliance with the new rule by small CWSs reduced the community’s exposure to arsenic in drinking water. To examine this, small CWSs were contacted to assess their interest in investigating whether implementing the new MCL decreased urinary arsenic (a biomarker of arsenic exposure) in community members drinking the water. The American Indian Walker River Paiute Tribe, which operated a CWS on the Walker River Paiute Indian Reservation, agreed to participate. The Walker River Paiute Indian Reservation is located within three counties in rural Midwestern Nevada about 100 miles southeast of Reno, Nevada. The community has a population of 1200 residing among a high desert land base surrounded by mountains, desert lakes, and marshland [15]. The community water system serving this community sits on the Eastern boundary of the town of Schurz and provides tap water for about 900 people on the reservation. The arsenic levels in the Walker River CWS were well above the previous 50 µg/L limit, with mean concentrations of 79 and 105 µg/L measured, respectively, when the wells were installed in 1985 and 1986 [16]. These concentrations represented the third highest of all Native American Indian Tribal community water systems that were regulated by the EPA [16]. The CWS would not be able to meet the new arsenic regulation of 10 µg/L without investing in arsenic removal treatment technologies [16]. In the early 2000s, the Tribe received a USD 1.2-million-dollar grant from the EPA to build a new drinking water treatment facility that included arsenic removal technology. The Indian Health Service installed an iron coagulation microfiltration process to reduce arsenic concentrations in the public water system [16].

In collaboration with the Walker River Paiute Tribe, we conducted a study to assess whether reducing the arsenic concentration in drinking water reduced the arsenic concentration in urine specimens from community members who drank the water.

2. Materials and Methods

2.1. Study Design

This study was approved by the Institutional Review Board (IRB) of the Centers for Disease Control and Prevention (IRB #4751).

We designed a quasi-experiment pre-post study. We measured arsenic in people’s urine specimens before and after they were exposed to the “treatment”, which was a modification of the drinking water treatment train to remove arsenic. This study is “quasi” experimental because we did not control the exposure (i.e., the amount of arsenic in drinking water), rather, the exposure was determined by the elements of the drinking water treatment train employed by the drinking water utility.

The timeframe included (1). a pre-evaluation of urine and drinking water arsenic levels prior to the implementation of the treatment technology and (2). an evaluation of urine and drinking water arsenic levels 1 week and 3 months after the new technology was implemented (post 1 week and post 3-month sampling periods). We chose the dates for the post-implementation data collections based on the literature indicating that arsenic has a short half-life (about 10 h for inorganic arsenic [17]) in people.

2.2. Community Selection

We selected the community to work with using the following selection criteria: (1). The community was served by a CWS; (2). Prior to installation of treatment technology, water from the CWS was near or exceeded the previous EPA drinking water standards for arsenic (~40 to >50 µg/L); (3). The community planned to add technology to their CWS to reduce arsenic levels; (4). The town has a population ≥1000 people; (5). A large proportion (e.g., >50%) of the population utilizes the drinking water from the CWS; and (6). The community expressed interest in participating in this study.

2.3. Participant Selection

To participate in our study, we required Walker River residents to (1). receive water from the CWS, (2). not have installed point-of-use filtration or other household devices to remove arsenic from their drinking water, (3). reside in their current home ≥1 year, (4). not have traveled for more than two weeks during the background assessment, and (5). not have traveled for three days during each of the follow-up assessments. We called every 5th name in the local phone book until we identified 54 volunteers who met our criteria and were willing to participate.

2.4. Questionnaires

Participants completed a baseline questionnaire upon recruitment. We collected information on demographics, occupation, diet, and other potential sources of arsenic (e.g., smoking status, occupational exposures, exposures from showering and bathing, additional exposures in and around the home).

We administered a short questionnaire to each participant prior to each sample collection period (three times before installation of the arsenic removal system and one week and three months after installation). This questionnaire included information on any potential changes in water sources since the last sample collection period (e.g., went on vacation), as well as potential exposures to sources of arsenic at work or in the home that were not from drinking water or food during the three days prior to sample collection.

2.5. Water Intake Diary

For the three days prior to each sample collection period, participants completed a self-administered daily water intake diary in which they recorded the frequency and approximate volume of tap water, drinks made with tap water, and foods made with tap water they consumed. We used this information to improve our assessment of the arsenic body burden from drinking water.

2.6. Drinking Water Sample Collection and Analysis

We collected water samples from the community water system and from the point-of-use in each participating household during each sampling period using 1 L acid-cleaned polyethylene bottles. We collected the water samples concurrently with urine samples. We shipped water samples overnight on ice to the Laboratory Sciences Division, Colorado Department of Public Health and Environment and analyzed them for arsenic using EPA method 200.8 [18].

2.7. Urine Sample Collection and Analysis

Study participants provided a first-void urine specimen in the pre-screened containers provided during each sample collection period. We verified the chain of custody for the urine specimens and shipped them overnight on dry ice to the Division of Laboratory Sciences, National Center for Environmental Health, for analysis. We measured total arsenic by inductively coupled plasma dynamic reaction cell mass spectrometry (ICP-DRC-MS) [19].

We determined the concentrations of arsenite (AsIII), and arsenate (AsV) using high-performance liquid chromatography (HPLC) to separate the species coupled to an inductively coupled plasma dynamic reaction cell mass spectrometer (ICP-DRC-MS) to detect the arsenic species [19].

If any laboratory samples were below the limit of detection (LOD), we divided the LOD by the square root of 2 for imputation.

2.8. Dates of Study Activities

• May–June 2006: Recruited study participants.
• June 2006: Collected 3 baseline water samples, urine specimens, and 3-day water consumption diaries for each of 54 study participants.
• 14 April–28 April 2008: Arsenic removal treatment initiated and stabilized.
• July 2008–September 2008: Collected 2 water samples, urine specimens, and 3-day water consumption diaries for each of 29 of the original study participants 1 week after and again 3 months after the arsenic removal treatment was implemented and stabilized.

2.9. Statistical Analysis

Using demographic questionnaire data and laboratory values of the study population, descriptive statistics were used to assess the metabolite concentrations between the sampling periods. SAS Version 9.4 (Cary, NC, USA) software was used for all statistical analyses. We calculated the geometric means and 95% confidence intervals for the drinking water and urinary arsenic species. Geometric means were used due to the skewed data and small sample size. For the baseline period and the 1-week post treatment implementation period, the results from the analysis of urine specimens and water samples were averaged together by sampling period for each ID. There was only one urine specimen and water sample collected for the 3-month post treatment implementation period. Geometric means for urine and water arsenic concentrations were calculated stratified by the demographic factors obtained from the questionnaire.

The results were adjusted for creatinine when able and results not adjusted for creatinine were provided for comparison as appropriate. Creatinine adjustment was not conducted on metabolite concentrations that had over 60% <LOD values post the baseline sampling period. Median percent reduction was calculated for the geometric means from the baseline sampling period. If <LOD occurred post-baseline, then median percent reduction in the metabolite was approximately 100%. To determine if there was a significant statistical difference in the metabolite concentrations between the different sampling periods, the data was log-transformed, and a Spearman correlation was conducted.

Values used in statistical analyses were the geometric mean of results from multiple specimens or samples or the value from a single specimen or sample. That is, the geometric mean of the results from three water samples and three urine specimens collected per participant at baseline and the geometric mean of the two specimens or samples collected per study participant one week post arsenic treatment installation and results of the one specimen or sample collected per/study participant three months post arsenic treatment installation.

3. Results

Demographics

A summary of the demographic characteristics of our study population is presented in

Table 1. Demographic characteristics of the study population at baseline and in the two follow-up time periods.

Characteristics	Baseline (N = 54) N (%)	Follow-Up (N = 29) * N (%)
Sex
Female	33 (61.1)	17 (58.6)
Male	21 (38.3)	12 (41.3)
Age (Years)
Mean (SD)	48 (16.4)	49 (15.0)
Median (range)	47 (20–84)	47 (20–76)
Race
White	5 (9.3)	2 (6.9)
Alaska Native	41 (75.9)	23 (79.3)
Black	0	0
Native Hawaiian/API	5 (9.3)	3 (10.3)
Other	3 (5.6)	1 (3.5)
Ethnicity
Not Hispanic or Latino	31 (77.5) †	16 (84.2) §
Hispanic/Latino	7 (17.5)	1 (5.3)
Don’t know/no response	2 (5.0)	0
Smoking status
Current smoker	13 (24.1)	6 (20.7)
Ever smoked	12 (22.2)	9 (31.0)
Never smoked	27 (50.0)	14 (42.3)
Do not know/no response	2 (5.0)	0
Live with a smoker
Yes	18 (33.3)	4 (15.4) ¶
No	36 (66.7)	22 (84.6)
Occupation (by category) **‡
Education	2 (3.9)	1 (3.6)
Health	5 (9.8)	0
Office (Admin, legal, HR, management, etc.)	14 (27.5)	8 (28.6)
Worked from home (e.g., homemaker, babysitter, etc.)	15 (29.4)	7 (25.0)
Disabled/Retired/Unemployed	6 (11.8)	6 (21.4)
Other (laborer, operator, maintenance)	9 (17.6)	6 (21.4)

Notes: * 29 people consistently participated in all sample collection periods and were therefore included in “Follow-up” cohort (post 1 week and post 3 months). ^† Missing = 14. ^§ Missing = 10. ^¶ Missing = 3. ** Missing = 3 for baseline. ^‡ Missing = 1 for follow-up.

; it includes results from 54 people who completed the baseline data collection activities during June 2006 and 29 of the original 54 people who completed the follow-up activities during July–September 2008. Most participants were female and had a mean age of 48 years (SD: 16.4). Seventy-six percent of participants reported their race as Alaska Native and the same percentage also reported not being Hispanic or Latino. Fewer than 25% of participants reported being current smokers and two-thirds reported living in a home with no smokers. When asked about location of occupation, over half reported working in an office or at home. During the follow-up period, the demographic characteristics of the 29 participants remained similar to the baseline population and no significant differences were observed.

For each of the visits, we asked participants to report any food consumption that may contribute to the urinary arsenic concentrations (i.e., shellfish and fish) as well as the amount of tap water used for cooking and drinking.

Table 2. Reported consumption of arsenic-containing food sources and drinking water at baseline, one week and three months after the treatment to remove arsenic was implemented for study participants who completed data collection for baseline and at least one follow-up period.

Characteristic	Baseline (N = 29) * N (%)	Post 1 Week (N = 29) N (%)	Post 3 Months (N = 25) N (%)
Any shellfish consumption during sample collection period? †
Yes	4 (13.8)	3 (10.3)	1 (4.0)
No	25 (86.2)	26 (89.7)	24 (96.0)
Any fish consumption during sample collection period? †
Yes	9 (31.0)	7 (24.1)	4 (16.0)
No	20 (69.0)	22 (75.9)	21 (84.0)
Any rice consumption during sample collection period? †
Yes	16 (44.8)	16 (44.8)	10 (34.5)
No	13 (55.2)	13 (55.2)	15 (51.7)
Any travel out of town during sample collection period? †
Yes	26 (89.7)	22 (75.9)	14 (56.0)
No	3 (10.3)	7 (24.1)	11 (44.0)
Frequency of tap water usage for cooking
Always use tap	27 (93.1)	27 (93.1)	27 (93.1)
Never use tap	2 (6.1)	2 (6.1)	2 (6.1)
Frequency of tap water usage for drinking
Always use tap	27 (93.1)	27 (93.1)	27 (93.1)
Never use tap	2 (6.9)	2 (6.9)	2 (6.9)
Average daily tap water intake for drinking (cups)
Mean (SD)	20.3 (15.7)	21.6 (12.4)	20.4 (12.5)
Median	16.5	21.0	16.0
Range	0–89	0–46	6–60

Note(s): * Analysis was conducted only on the 29 people who participated in all three sample collection periods (baseline, post 1 week, post 3 months). † Missing = 4 in post 3 months.

includes data for the 29 study participants who completed all study activities during the three sampling periods and summarizes their responses at baseline, one week, and three months (July 2008–September 2008) after the arsenic removal technology was installed in the Walker River CWS. For all three periods, most participants reported eating no shellfish (range: 86.2–96.0%) or fish (range: 69.0–84.0%). Nearly half of the participants reported eating rice that was cooked using tap water (range: 34.5–44.8%). During the three study periods, participants were also asked about travel outside the reservation to further assess exposure; during the baseline period, 26 (86.9%) participants did not travel; however, during the three-month follow-up period, 14 (56.0%) participants reported traveling during the sampling period.

Among all respondents, 27 (93.1%) also indicated that they always use tap water for cooking and indicated that they always use tap water for drinking. Participants also recorded the number of cups (8 ounces) of water that they drank during each sampling period. During the baseline period, the average number of cups reported was 20.3 (SD 15.7), with a range of 0–89 cups of water. The mean number of cups remained similar during the post-one-week and post-three-month follow-up periods.

Where appropriate, to account for differences in the hydration of study participants, we calculated creatinine-adjusted concentrations by dividing the As concentration in micrograms per liter urine by the creatinine concentration in grams per liter urine. As concentrations were reported as weight of As per gram of creatinine [20].

The creatinine-adjusted and crude baseline total urinary arsenic concentrations (µg/L), stratified by demographic characteristics, are summarized in

Table 3. Creatinine-adjusted and unadjusted total urinary arsenic concentrations at baseline by demographic characteristics. N = 85 specimens (3 per study participant for the 29 participants who completed data collection at baseline and at least one follow-up period).

Demographics		N (%)	Creatinine-Adjusted Geometric Mean (µg/L)	95% CI	Unadjusted Geometric Mean (µg/L)	95% CI
Sex	Female	51 (59.3)	62.6	52.2, 75.1	69.2	59.3, 80.8
	Male	35 (40.7)	60.0	45.3, 79.6	55.5	37.7, 81.8
Age	20–35	15 (27.4)	56.1	42.7, 73.5	83.9	64.1, 110.0
	36–50	30 (34.8)	46.0	35.5, 59.6	52.1	34.2, 79.3
	51–65	24 (27.9)	64.5	47.1, 88.5	61.7	48.1, 79.1
	65+	17 (19.8)	104.6	75.9, 144.2	72.0	49.1, 105.6
Race	White	6 (7.0)	38.9	20.4, 74.0	43.4	21.8, 86.4
	Alaska Native	69 (80.2)	66.1	55.4, 78.9	67.3	55.7, 81.3
	Black	0	-	-	-	-
	Native Hawaiian/Asian/Pacific Islander	8 (9.3)	50.6	29.7, 86.4	46.8	15.3, 143.1
	Other	3 (3.5)	50.9	26.4, 98.2	72.9	17.6
Smoking Status	Current smoker	18 (20.9)	53.6	34.3, 83.6	57.6	36.4, 91.1
	Ever smoked	27 (31.4)	79.0	61.2, 102.1	62.5	43.0, 90.1
	Never smoked	41 (47.7)	55.5	45.2, 68.1	66.4	52.7, 83.7
Lives with smoker *	Yes	68 (80.0)	87.7	60.3, 127.5	76.8	53.4, 110.4
	No	17 (20.0)	56.8	48.0, 67.2	57.9	47.1, 71.1
Overall		86 (100)	61.6	52.8, 71.8	63.3

Note: * Missing = 1.

for the 29 participants who completed all study activities. The baseline arsenic concentrations represent the geometric mean values of three urine samples for all but one individual (for that individual, the arsenic concentration is a mean of two samples) for a total of 86 samples. The geometric mean for all baseline total urinary arsenic concentrations in our study population was 61.6 µg/L (95% CI: 52.8–71.8 µg/L). When stratified by sex, the baseline urinary total arsenic concentrations for females and males were similar, at 62.6 (95% CI: 52.2–75.1) and 60.0 µg/L (95% CI: 45.3–79.6), respectively. Differences in concentrations, however, were observed when stratifying by age group category, race, and smoking status. Participants >65 years in age had the highest geometric mean concentration of total urinary arsenic, at 104.6 µg/L. The Alaska Native category also had the highest total urinary arsenic concentration, but this may have been influenced by the sample size. When examining concentrations by smoking status, participants who reported ever or current smoking had higher mean urinary total arsenic concentrations than those who reported never smoking (53.6 µg/L and 79.0 µg/L vs. 55.5 µg/L, respectively). In addition, participants who reported living with a smoker had higher mean urinary total arsenic concentrations than those who reported not living with a smoker (87.7 µg/L vs. 56.8 µg/L, respectively).

To better understand the relationship between the observed creatinine adjusted urinary total arsenic concentrations and the participant’s water intake, Spearman correlations were used to compare creatinine adjusted urinary total As, AsV, and total As in drinking water by sampling period (

Table 4. Spearman correlations (ρ) between creatinine adjusted urinary arsenic concentrations and reported tap water intake.

Sampling Period	Creatinine Adjusted Urinary as Total			Creatinine Adjusted Urinary AsV
	N	ρ	p-Value	N	ρ	p-Value
All periods	168	0.326	<0.0001	168	0.098	0.2059
Baseline	82	0.654	<0.0001	82	0.314	0.0041
Post 1 week	57	0.566	<0.0001	57	0.023	0.8250
Post 3 months	29	0.506	0.0051	29	0.124	0.5200

).

A significant correlation between total creatinine adjusted urinary arsenic concentration and reported tap water intake was observed for the baseline period (p < 0.0001); after the system was installed, the relationship remained strong but was lower than the baseline (p = 0.0051). This suggests that a large proportion of the observed total creatinine adjusted urinary arsenic can be explained by drinking water intake. A similar trend is observed with AsV, where the strongest and significant correlation is only found in the baseline period, prior to the installation of the water system. Subsequent measurements of creatinine adjusted urinary arsenic are lower, likely owing to the decreased intake in AsV from the drinking water.

Table 5. Summary of arsenic concentrations in urine and water showing concentrations of total urinary arsenic, urinary As III, and urinary AsV at baseline, and 1 and 3 months after arsenic removal treatment implementation and the percent reduction from baseline. N = 29 for baseline and post 1-week, N = 25 for post 3 weeks.

Biomarker	Unadjusted Geometric Mean (95% CI) (µg/L)	Median Percent Reduction from Baseline Period (%)	Creatinine Adjusted Geometric Mean (95% CI) (µg/L)	Median Percent Reduction from Baseline Period (%)
Urinary As Total
Baseline	70.7 (55.6, 90.0)		64.2 (50.1, 82.3)
Post 1 week	14.5 (11.7, 18.0)	75.4 *	14.7 (12.1, 18.0)	75.5
Post 3 months	14.2 (10.6, 19.0)	82.1 *	12.7 (9.5, 17.0)	76.7
Urinary AsIII (µg/L)
Baseline	3.9 (2.9, 5.4)
Post 1 week	<LOD	-
Post 3 months	<LOD	-
Urinary AsV (µg/L)
Baseline	2.9 (2.3, 3.7)
Post 1 week	<LOD	-
Post 3 months	<LOD	-
Total As in Drinking Water (µg/L)
Baseline	73.9 (71.9, 75.9)
Post 1 week	7.5 (6.3, 8.8)	89.6 *
Post 3 months	2.9 (2.7, 3.2)	96.0 *
AsV in Drinking Water (µg/L)
Baseline	70.1 (67.9, 72.2)
Post 1 week	6.4 (5.4, 7.5)	90.3 *
Post 3 months	3.0 (2.9, 3.0)	95.8 *

Note: * p < 0.0001.

shows the changes in urine and water arsenic concentrations over time. The concentration of all forms of arsenic decreased in both water samples and urine specimens over time. Among the 29 participants, the creatinine-adjusted geometric mean concentrations of total urinary arsenic decreased from 64.2 µg/L (95% CI: 50.1, 82.3) to 14.7 µg/L (95% CI: 12.1, 18.0) one week after the installation of the new water system, representing a median percent reduction of 75.5%. Similarly, a small reduction in concentration was also observed between the baseline and post 1 week sampling period for AsIII and AsV, with changes of 3.9 µg/L (95% CI: 2.9, 5.4) and 2.9 µg/L (95% CI: 2.3, 3.7), respectively.

Water samples also showed decreasing arsenic concentrations over time. After the installation of the water system, the geometric mean total arsenic concentration in the drinking water fell from 73.9 µg/L (95% CI: 71.9, 75.9) to 7.5 µg/L (95% CI: 6.3, 8.8), representing a median percent reduction of 89.6%. The arsenic concentrations were further lowered to 2.9 µg/L after three months, representing a median percent reduction of 96.0% from baseline.

In addition to examining the overall trends by study period, we examined the changes in total urinary arsenic for each study participant. Figure 1a,b represent the total urinary arsenic concentration at baseline and one week and three months after arsenic removal technology was installed, for each participant. Apart from two participants (as illustrated in Figure 1b), all participants showed a reduction in total arsenic in their urine. A larger number of participants were observed to have an increased total urinary arsenic concentration from the post 1 week to post 3-month sampling periods. These increases could be associated with many factors, including drinking untreated water.

Similarly to the total urinary arsenic concentrations, AsV decreased across each of the sampling periods. Notably, from the post 1 week to post 3-month sampling period, the concentrations of AsV in most participants were below the method detection limit. All participants had a decrease in urinary AsIII between the baseline and post 1 week sampling periods. A few participants had increased AsIII concentrations between the baseline period and post 3 months and between post one week and post three months. These increases could stem from multiple factors including consuming food with AsIII (e.g., shellfish) and drinking water that had not been completely treated (i.e., there was a water treatment system malfunction).

4. Discussion

The U.S. EPA estimated in 2000 that 4.4% of CWSs would exceed the MCL of 10 ppb [21] and in 2006, researchers estimated that only 2% of the U.S. population would have drinking water arsenic concentrations above the new MCL [22]. However, the proportion of tribally owned CWSs providing drinking water with arsenic above the MCL was more than three times higher, and the proportion of tribal population at risk was over seven times higher than these national estimates [16]. Following the construction by the Indian Health Service of two new wells in the mid-1980s, the arsenic concentrations in the Walker River Paiute Tribe community water system were among the highest of all Native American community water systems regulated at that time by the U.S. EPA. All CWS were required to meet the new MCL of 10 µg/L by January 2006 [23] and in 2006, the Indian Health Service constructed a water treatment facility that included arsenic removal treatment to meet the updated Safe Drinking Water Standard for arsenic. Our study was designed to determine if the new drinking water treatment plant reduced the concentration of arsenic in drinking water, thereby reducing the risks of exposure and, presumably, reducing risks for illnesses associated with arsenic exposure from drinking water.

Overall, implementation of the iron coagulation microfiltration process succeeded in reducing the concentrations of total arsenic concentrations in the finished drinking water to well below the current MCL of 10 µg/L. The updated MCL is a health-based standard developed specifically in response to data indicating that health risks are associated with exposure to even low levels of arsenic [24]. Thus, the arsenic removal technology is a preventive measure that protects community health from myriad chronic diseases, including diabetes and hypertension as well as central nervous system damage from acute exposure [24]. Concentrations of AsV in water also decreased (AsIII was not measured in water), limiting exposure to inorganic arsenic. More importantly, dramatic decreases in the urinary biomarkers for arsenic exposure were observed for total arsenic and the inorganic congeners AsV and AsIII. This is especially important for AsIII, which is more toxic than AsV [24]. Thus, the successful removal of most of the arsenic by the Walker River Paiute Tribe water treatment system has both short- and long-term public health benefits. Below, specific findings from our study are discussed.

Among study participants, the urinary total arsenic concentrations were similar among males and females (see Table 3). Older participants (ages 51+) had higher urinary total arsenic concentrations that younger participants (ages 20–50) did. Participants who responded as white had the lowest urinary total arsenic concentrations compared with the other racial groups, which is consistent with trends in the National Health and Nutrition Examination Survey (NHANES) [25]. Study participants who reported living with a smoker had the highest urinary total arsenic concentrations (see Table 3). These concentrations were higher than those of current smokers, suggesting that another source of arsenic (i.e., drinking water) was important in determining an individual’s total exposure.

In our study, the baseline total urine arsenic concentration observed in our study population was (geometric mean, 95% confidence interval) was 64.2 µg/L (50.1–82.3). This fell within the 95th percentile for survey years 2005–2006 and 2007–2008, ages 20+ (geometric mean, 90% confidence interval: 71.4 µg/L, 57.7–98.3; 59.0 µg/L, 44.2–75.6, respectively) of the U.S. population reported in NHANES (National Center for Environmental Health, 2024). One week post treatment, participants’ total arsenic levels were reduced to those comparable to the 75th percentile for NHANES study years 2005–2006 and 2007–2008, ages 20+ (geometric mean, 90% confidence interval: 18.9 µg/L, 15.8–22.9; 16.2 µg/L, 14.5–18.6, respectively). The total arsenic concentrations remained low post 3 months (geometric mean, 90% confidence interval 14.2 µg/L, 10⁻⁶–19.0).

Our initial results suggested that the local drinking water treatment system needed to be updated to improve arsenic removal to meet the new standard promulgated by EPA and protect the local community from exposure. Following the drinking water treatment upgrades, concentrations of total arsenic the drinking water supply decreased dramatically from a geometric mean of 70.1 µg/L (95% CI: 67.9–72.2) to well below the new MCL of 10 µg/L in finished water. As noted above, we saw a concomitant decrease in urinary arsenic one week after the new system went online. Further, concentrations of total arsenic in both urine specimens and drinking water continued to decrease between the post one week and post three months sampling periods. These results were highly correlated, indicating that drinking water was a significant source of As exposure (see Table 4 and Figure 1a–c).

Urinary concentrations of AsIII and AsV remained low and water concentrations of AsV in water continued to decrease through the post three-month sampling period. AsIII is somewhat more toxic than AsV and the oxidative states can be interconverted in both the human body and the environment [1]. Thus, the reduction in AsV in this community’s drinking water was important in reducing health risks from exposure to both As oxidative states.

Our study demonstrated the importance of not only initial, but also ongoing support for small drinking water systems. During parts of the summer of 2008, the drinking water treatment facility was shut down: 25 July from 8:15 a.m. to 2 p.m. (about 6 h); 5 August–11 August because there was no caustic available (about 6 days); and August 18–22 to repair a broken pipe that delivered caustic to the treatment train (about 4 days). We found that, between the one-week and three-month post treatment improvement samples, several study participants had increased urinary concentrations of AsIII compared with baseline and/or the post one-week samples. These increases could stem from multiple factors, including consuming food containing AsIII (e.g., shellfish). However, most of this community did not report eating shellfish during the study period (see Table 2). It is more likely that the increases in urinary AsIII were the result of temporary drinking water treatment interruptions such as those mentioned above.

The primary limitation for this study was small sample size and reliance on self-reporting. However, in our study the decreases in arsenic concentrations in drinking water and biologic specimens collected from the community demonstrated the effectiveness of the new drinking water treatment to reduce exposure to arsenic. Our study also demonstrates the value of supporting small drinking water systems that may be otherwise unable to meet changing and challenging drinking water quality regulations. Finally, our study suggests that sustainable support for logistical issues (e.g., obtaining treatment chemicals) and system maintenance may ensure ongoing successful drinking water treatment.

5. Conclusions

The installation and implementation of drinking water treatments to remove arsenic dramatically reduced both the drinking water concentrations (to below the current EPA MCL of 10 µg/L) and the community members’ urinary concentrations of total As, AsIII, and AsV within a week of its full implementation. Spikes in total arsenic urinary concentrations between one week and 3 months post-treatment may have been associated with a temporarily missing chemical needed to remove arsenic. Additional assistance to small water systems to sustain new drinking water treatments may be warranted. It would be useful to revisit this community to assess the sustainability of the arsenic removal treatment to ensure a continued reduction in exposure to arsenic from drinking water.