Arsenic in drinking water: An analysis of global drinking water regulations and recommendations for updates to protect public health.

Evidence-based public health policy often comes years or decades after the underlying scientific breakthrough. The World Health Organization's (WHO's) provisional 10 μg/L arsenic (As) drinking water guideline was set in 1993 based on "analytical achievability." In 2011, an additional proviso of "treatment performance" was added; a health-based risk assessment would lead to a lower and more protective guideline. Since the WHO does not require United Nations member states to submit copies of national drinking water regulations, there is no complete database of national drinking water standards or guidelines. In this study, we collated and analyzed all drinking water regulations for As from national governments worldwide. We found regulations for 176 countries. Of these countries, 136 have drinking water regulations that specify 10 μg/L As or less, while 40 have regulations that allow more than 10 μg/L of As; we could not find any evidence of regulations for 19 countries. The number of people living in countries that do not meet the WHO's guideline constitutes 32% of the global population. Global As regulations are also strongly tied to national income, with high income countries more likely to meet the WHO's guideline. In this study, we examined the health risk assessments that show a clear need for reducing As exposure to levels far below the current WHO provisional guideline. We also show that advances in analytical chemistry, drinking water treatment, and the possibility of accessing alternative drinking water supplies without As suggest that both low-income countries with limited resources and high-income countries with adequate resources can adopt a lower and more protective national drinking water standards or guidelines for As. Thus, we recommend that regulators and stake holders of all nations reassess the possibilities for improving public health and reducing health care expenses by adopting more stringent regulations for As in drinking water.

1. Introduction

Arsenic (As) is a common drinking water contaminant that is often found in groundwater wells [1-6]. Even at very low concentrations, chronic consumption of As in drinking water has been strongly associated with a variety of cancers and other adverse health effects in humans [7-13]. At least 226 million people in 56 countries are exposed to unsafe concentrations of As in drinking water and food [14].

It often takes years or decades for an advance in science to cause a change in public health policy [15]. Notably, we will demonstrate that all national standards and guidelines for As in drinking water are based on outdated assumptions. For example, many low-and middle-income countries still use the World Health Organization’s (WHO’s) 1963 drinking water standard for As of 50 micrograms per liter (μg/L), even though the WHO lowered its recommended maximum concentration for As to 10 μg/L in 1993 [16, 17]. Most high-income countries use a 10 μg/L As drinking water standard, consistent with the current WHO recommended maximum concentration. However, many regulators and stakeholders may not be aware that the WHO 10 μg/L guideline is deemed “provisional” [17, 18]. Although risk assessment data indicate a lower guideline would be more appropriate, the WHO retains the 10 μg/L for As based on an assumed practical quantification limit of routine laboratories, as well as an assumed practical drinking water treatment limit [17, 18]. Much new technology for the testing and treatment of As has been developed since the provisional 10 μg/L WHO guideline was first set based on practical concerns rather than health data, but the provisional guideline value has not been updated since 1993 [17, 18].

This causes a significant threat to global public health. Current research shows that more protective national standards and guidelines for As in drinking water are technologically feasible and urgently necessary to protect public health. In this study we collate all current national standards for As in drinking water into a database that can be used to examine patterns of As regulations based on geographic regions and income. We review the basis of the WHO’s drinking water guideline for As and why it is deemed “provisional”. We examine risk assessment data that indicate that an As guideline of 10 μg/L does not provide sufficient protection against cancer and other adverse health effects. Finally, we review current quantification and treatment technologies, demonstrating the technological feasibility of reducing the maximum allowable concentration of As in drinking water in both high and low income countries.

2. Materials and methods

2.1 International drinking water standards for arsenic

In order to better understand the state of current regulations for As in drinking water worldwide, we collated all available national standards. We began with a list of the 193 United Nations member states and supplemented this list with 2 other states for which the World Bank provides income group data, Taiwan and Kosovo [19, 20]. We then searched for official laws or decrees promulgated by these national governments regulating As in drinking water. For each country we began our search for national laws and decrees on As in drinking water at the Law Library of Congress website (http://www.loc.gov/law/help/guide/nations.php). When possible, the official online government gazette of a country was also used for this search. If necessary, the FAOLEX database of national laws and regulations on food, agriculture, and renewable natural resources [21], Google Scholar [22], and Google were also used for this search. We also searched catalogs of national standards agencies. When necessary, Google Translate [23] was used to translate between English and the official language of a country. Common search terms included the name of the country, the official gazette of the country, “drinking water quality standards”, “drinking water standards”, “drinking water”, “water”, “arsenic”, “μg/L”, and “mg/L” in both English and the official language of the country. We also compared our search results with those of previous partial surveys of international drinking water regulations [24-27].

For countries in which our search methods could not locate a national law or decree, we continued our search, seeking secondary evidence for regulations such as peer-reviewed articles, dissertations, theses, or similar documents that state a drinking water standard for As in a country. Secondary evidence of regulations was only used when primary evidence was not found. We searched for secondary evidence of regulations using the name of the country, the official gazette of the country, “drinking water quality standards”, “drinking water standards”, “drinking water”, “water”, “arsenic”, “μg/L”, and “mg/L” in both English and the official language of the country. We also used internet searches to determine the national agency or organization responsible for setting drinking water quality standards in these countries and contacted representatives of these agencies via email and Facebook in a national language requesting help obtaining the standards.

If we were unable to find either primary or secondary evidence for a drinking water regulation for As in a country, we assumed that there was most likely no national standard or guideline. There were 19 countries for which we could not find any evidence of drinking water regulations for As.

To understand the relationships between population, income, and As standards, we collated population, gross domestic product (GDP), and GDP per capita data for each country [20, 28, 29]. We selected 2019 population and GDP data since it is the most current data before the demographic and economic upheavals caused by the COVID-19 pandemic. Although COVID-19 was first detected during 2019, as of January, 2020, international economists were not yet expecting it to have a major impact on the world economy [30].

2.2 Data analysis and statistics

We used R version 4.1.1, “Kick Things”, released August 10, 2021 to calculate descriptive statistics and perform hypothesis testing. For hypothesis testing, we assumed a 95% confidence level for significance. Multiple statistical comparisons of the data were not made, so no corrections for multiple comparisons were applied. Figures and maps were also created with R using the R packages ggplot2 and ggmaps.

3. Results and discussion

The results of our search for national As regulations are listed in Table 1. In addition to the name of each country and concentration of As specified in the regulation, we have also included the 2019 Gross Domestic Product (GDP) per capita from the World Bank [20], the 2019 WB income category [31], whether the stated standard is determined independently by the country or tied to international regulations or guidelines such as the WHO drinking water guideline, and when the regulation was most recently updated.

Table 1

International drinking water standards for arsenic and the 2019 gross domestic product (GDP) per capita in United States dollars for 195 countries.

Country	As standard (μg/L)	Year of publication	GDP / Capita (2019) [20]	WB Income Class (2019) [31]	Regulatory Link	Regulation typea
Africa
Algeria [32]	10	2011	$3,974	LM		Law/Decree
Angola [33]	50	2011	$2,791	LM		Law/Decree
Benin [34]	50	2001	$1,219	LM		Law/Decree
Botswana [35–37]b	10	2009	$7,961	UM		Standards Org.
Burkina Faso [38]	10	2005	$787	Low	WHO: 1996	Law/Decree
Burundi [39, 40]	10	2000	$261	Low	EAS: 2000	Standards Org.
Cameroon [18, 41]	10	2007	$1,507	LM	WHO	Gov. Org.
Cape Verde [42]	10	2004	$3,604	LM		Law/Decree
Central African Republic [18, 43]	10	2017	$468	Low	WHO	Law/Decree
Chad [44]	10	2010	$710	Low		Law/Decree
Comoros [45]	50	1994	$1,370	LM	WHO	Law/Decree
Congo	NA		$2,280	LM
Democratic Republic of the Congo	NA		$581	Low
Djibouti [46]	50	2001	$3,415	LM		Law/Decree
Egypt [47, 48]	10	2007	$3,019	LM		Gov. Org.
Equatorial Guinea	NA		$8,132	UM
Eritrea	NA		NA	Low
Ethiopia [49]	10	2013	$856	Low		Standards Org.
Gabon [50]	50	2011	$7,767	UM		Gov. Org.
Ghana [51–53]	10	2017	$2,202	LM		Standards Org.
Guinea [54]	10	1997	$963	Low		Law/Decree
Guinea-Bissau	NA		$697	Low
Ivory Coast [18, 55]	10	2017	$2,276	LM	WHO	Law/Decree
Kenya [56–58]	10	2018	$1,817	LM		Standards Org.
Lesotho [59]	NA		$1,118	LM
Liberia [18, 60]	10	2017	$622	Low	WHO	Gov. Org.
Libya [61–63]	10	2015	$7,686	UM		Standards Org.
Madagascar [64]	50	2004	$523	Low		Law/Decree
Malawi [65–67]	50	2013	$412	Low		Standards Org.
Mali [68]	10	2007	$879	Low		Gov. Org.
Mauritania [69]	10	2015	$1,679	LM	WHO	Gov. Org.
Mauritius [70]	10	1996	$11,099	High		Law/Decree
Morocco [71–73]	10	2006	$3,282	LM		Standards Org.
Mozambique [74]	10	2004	$504	Low		Law/Decree
Namibia [75, 76]	300	1988	$4,957	UM		Law/Decree
Niger [18,77]	10	2017	$554	Low	WHO	Law/Decree
Nigeria [78]	10	2015	$2,230	LM		Standards Org.
Rwanda [79, 80]	10	2014	$820	Low		Standards Org.
São Tomé and Príncipe [81]	NA		$1,947	LM
Senegal [18, 82]	10	1996	$1,447	LM		Gov. Org.
Seychelles [83]	10	2012	$17,448	High		Law/Decree
Sierra Leone [84]	NA		$528	Low
Somalia	NA		NA	Low
South Africa [85, 86]	10	2015	$6,001	UM		Law/Decree
South Sudan [87]	50	2011		Low		Gov. Org.
Sudan [88, 89]	7	2009	$713	Low		Standards Org.
Swaziland [85, 86, 90, 91]	10	2015	$4,090	LM		Standards Org.
Tanzania [92–94]	10	2018	$1,089	LM	EAS	Standards Org.
The Gambia [95]	50	2008	$778	Low		Gov. Org.
Togo [96]	10	2015	$679	Low		Gov. Org.
Tunisia [97]	10	2013	$3,317	LM		Standards Org.
Uganda [98]	10	2014	$794	Low	EAS	Standards Org.
Zambia [99]	10	2010	$1,305	LM		Standards Org.
Zimbabwe [100–104]	10	2014	$1,464	LM		Standards Org.
The Americas
Antigua and Barbuda [105]c	10	2003	$17,113	High	CARICOM	Law/Decree
Argentina [107]	10	2019	$9,912	UM		Law/Decree
Bahamas [108]c	50	2010	$34,864	High	CARICOM	Standards Org.
Barbados [109]c	10	2017	$18,148	High	CARICOM	Gov. Org.
Belize [110]c	NA		$4,815	UM	CARICOM
Bolivia [111]	10	2018	$3,552	LM		Gov. Org.
Brazil [112]	10	2021	$8,655	UM		Law/Decree
Canada [113]	10	2020	$46,195	High		Gov. Org.
Chile [114]	10	2007	$14,896	High		Law/Decree
Colombia [115]	10	2007	$6,429	UM		Gov. Org.
Costa Rica [116]	10	2015	$12,244	UM		Law/Decree
Cuba [117]	50	2017	NA	UM		Standards Org.
Dominica [18, 118]c	10	2017	$8,111	UM	CARICOM	Gov. Org.
Dominican Republic [119]	50	2001	$8,282	UM		Gov. Org.
Ecuador [120]	100	2015	$6,184	UM		Law/Decree
El Salvador [121]	10	2009	$4,187	LM		Standards Org.
Grenada [122]c,d	1	2005	$10,809	UM	CARICOM	Law/Decree
Guatemala [124]	10	2013	$4,620	UM		Standards Org.
Guyanac	NA		$6,610	UM	CARICOM
Haiti [125]c	10	2017	$1,272	Low	WHO	Gov. Org.
Honduras [126]	10	2007	$2,575	LM		Law/Decree
Jamaica [127]c	NA		$5,582	UM	CARICOM
Mexico [128]e	10	2019	$10,069	UM		Law/Decree
Nicaragua [129]	50	2000	$1,913	LM		Law/Decree
Panama [130]f	10	2007	$15,731	High		Law/Decree
Paraguay [131]g	500	2000	$5,415	UM		Law/Decree
Peru [133]	10	2017	$6,978	UM		Law/Decree
Saint Kitts and Nevisc	NA		$19,939	High	CARICOM
Saint Luciac	NA		$11,611	UM	CARICOM
Saint Vincent and the Grenadinesc	NA		$7,458	UM	CARICOM
Suriname [134]c	NA		$6,360	UM	CARICOM
Trinidad and Tobagoc	NA		$17,398	High	CARICOM
United States [135]	10	2018	$65,298	High		Gov. Org.
Uruguay [136]	20	2010	$16,190	High		Standards Org.
Venezuela [137]	10	1998	NA	UM		Law/Decree
Asia
Afghanistan [138]	50	2013	$507	Low		Standards Org.
Armenia [139–141]	50	2005	$4,623	UM		Law/Decree
Azerbaijan [142, 143]	50	1985	$4,794	UM	CIS	Standards Org.
Bahrain [144, 145]	10	2012	$23,504	High	GCC	Standards Org.
Bangladesh [146]	50	2019	$1,856	LM		Gov. Org.
Bhutan [147]	10	2018	$3,316	LM		Gov. Org.
Brunei [17, 148]	10	1993	$31,087	High	WHO: 1993	Gov. Org.
Cambodia [149]	50	2004	$1,643	LM		Gov. Org.
China [150]	50	2006	$10,217	UM		Standards Org.
Georgia [151]	10	2014	$4,698	UM		Law/Decree
India [152]h	10	2012	$2,100	LM		Standards Org.
Indonesia [153]	10	2010	$4,142	UM		Gov. Org.
Iran [154]	10	2010	NA	UM		Standards Org.
Iraq [155, 156]	10	2009	$5,955	UM		Standards Org.
Israel [157]	10	2016	$43,592	High		Law/Decree
Japan [158]	10	2015	$40,247	High		Gov. Org.
Jordan [159]	10	2015	$4,405	UM		Standards Org.
Kazakhstan [160]	50	2015	$9,812	UM		Law/Decree
Kuwait [145, 161]	10	2011	$32,000	High	GCC	Gov. Org.
Kyrgyzstan [162]	50	2004	$1,309	LM		Law/Decree
Laos [163]	50	2009	$2,535	LM		Gov. Org.
Lebanon [164]	50	1999	$7,584	UM		Standards Org.
Malaysia [165]	10	2004	$11,414	UM		Gov. Org.
Maldives [166]	10	2017	$10,627	UM		Gov. Org.
Mongolia [167–169]	10	2018	$4,340	LM		Standards Org.
Myanmar [170–172]	50	2014	$1,408	LM		Gov. Org.
Nepal [173]	50	2005	$1,071	LM		Gov. Org.
North Korea	NA		NA	Low
Oman [145,174]	10	2012	$15,343	High	GCC	Standards Org.
Pakistan [175]	50	2010	$1,285	LM		Gov. Org.
Philippines [176]	10	2016	$3,485	LM		Gov. Org.
Qatar [145, 177]	10	2014	$62,088	High	GCC	Gov. Org.
Saudi Arabia [145, 178]	10	2015	$23,140	High	GCC	Gov. Org.
Singapore [179]	10	2019	$65,233	High		Law/Decree
South Korea [180, 181]	10	2015	$31,846	High		Gov. Org.
Sri Lanka [182]	50	2019	$3,853	LM		Law/Decree
Syria [183–185]	50	2007	NA	Low		Standards Org.
Taiwan [186]	10	2017	NA	High		Gov. Org.
Tajikistan [143, 187]	50	1985	$871	LM	CIS	Standards Org.
Thailand [188, 189]	50	2008	$7,807	UM		Gov. Org.
Timor-Leste [18, 190]i	10	2017	$577	LM	WHO	Gov. Org.
Turkey [191, 192]	10	2019	$9,127	UM		Law/Decree
Turkmenistan [193–195]	NA		NA	UM
United Arab Emirates [145, 196]j	10	2014	$43,103	High	GCC	Gov. Org.
Uzbekistan [197]	50	2006	$1,725	LM		Law/Decree
Vietnam [198]	10	2009	$2,715	LM		Gov. Org.
Yemen [199]	10	1999	$774	Low		Gov. Org.
Australia and Oceania
Australia [200]	10	2017	$55,060	High		Gov. Org.
Federated States of Micronesia [135, 201]	10	2018	NA	LM	US EPA	Law/Decree
Fiji [202, 203]k	10	2011	$6,176	UM		Standards Org.
Kiribati [18, 24]	10	2017	$1,655	LM	WHO	Gov. Org.
Marshall Islands [205]	50	1994	NA	UM		Law/Decree
Nauru [18, 24]l	10	2017	$11,724	High	WHO	Gov. Org.
New Zealand [207]	10	2018	$42,745	High		Gov. Org.
Palau [208]	50	1996	$14,902	High		Gov. Org.
Papua New Guinea [209]	50	2006	$2,829	LM		Law/Decree
Samoa [210–212]	10	2016	$4,209	UM		Gov. Org.
Solomon Islands [18, 213]	10	2017	$2,374	LM	WHO	Law/Decree
Tonga [18, 24, 214]m	10	2017	$4,903	UM	WHO	Gov. Org.
Tuvalu [18, 215]	10	2017	$4,059	UM	WHO	Gov. Org.
Vanuatu [216]	10	2019	$3,115	LM		Law/Decree
Europe
Albania [217]	50	2016	$5,353	UM		Law/Decree
Andorra [218]n	10	2007	$40,886	High		Law/Decree
Austria [219]	10	2001	$50,138	High	EU	Law/Decree
Belarus [220]	10	2015	$6,663	UM		Law/Decree
Belgium [221]	10	2003	$46,421	High	EU	Law/Decree
Bosnia and Herzegovina [222]	10	2010	$6,109	UM		Law/Decree
Bulgaria [223]	10	2001	$9,828	UM	EU	Law/Decree
Croatia [224]	10	2019	$14,936	UM	EU	Law/Decree
Cyprus [225]	10	2001	$20,815	High	EU	Law/Decree
Czech Republic [226]	10	2014	$23,495	High	EU	Law/Decree
Denmark [227]	5	2015	$60,170	High	EU	Law/Decree
Estonia [228]	10	2002	$23,723	High	EU	Gov. Org.
Finland [229]	10	2014	$48,783	High	EU	Gov. Org.
France [230, 231]	10	2017	$40,494	High	EU	Law/Decree
Germany [232]	10	2017	$46,445	High	EU	Law/Decree
Greece [233]	10	2017	$19,583	High	EU	Law/Decree
Hungary [234]	10	2002	$16,732	High	EU	Law/Decree
Iceland [235]	10	2001	$66,945	High		Law/Decree
Ireland [236]	10	2014	$78,661	High	EU	Law/Decree
Italy [237, 238]	10	2016	$33,228	High	EU	Gov. Org.
Kosovo [239]	10	2012	$4,345	UM	EU	Law/Decree
Latvia [240]	10	2017	$17,829	High	EU	Law/Decree
Liechtenstein [241]	10	2018	NA	High		Law/Decree
Lithuania [242]	10	2017	$19,602	High	EU	Law/Decree
Luxembourg [243]	10	2017	$114,705	High	EU	Law/Decree
Malta [244]	10	2009	$29,821	High	EU	Law/Decree
Moldova [245]	10	2019	$4,504	LM		Law/Decree
Monaco [246]	10	2017		High		Law/Decree
Montenegro [247]	10	2012	$8,909	UM		Law/Decree
Netherlands [248]	10	2011	$52,331	High	EU	Law/Decree
Norway [249]	10	2016	$75,420	High		Law/Decree
Poland [250]	10	2017	$15,693	High	EU	Law/Decree
Portugal [251]	10	2017	$23,252	High	EU	Law/Decree
Republic of North Macedonia [252]	10	2018	$5,954	UM		Law/Decree
Romania [253]	10	2019	$12,920	High	EU	Law/Decree
Russia [254]	50	2001	$11,774	UM		Gov. Org.
San Marino [255]	10	2012	NA	High		Law/Decree
Serbia [256]	10	2019	$7,412	UM		Law/Decree
Slovakia [257]	10	2006	$19,266	High	EU	Law/Decree
Slovenia [258]	10	2015	$25,946	High	EU	Law/Decree
Spain [259]	10	2003	$29,367	High	EU	Law/Decree
Sweden [260]	10	2017	$51,615	High	EU	Law/Decree
Switzerland [261]	10	2020	$81,994	High		Gov. Org.
Ukraine [262]	10	2010	$3,465	LM		Gov. Org.
United Kingdom [263]o	10	2016	$42,330	High		Law/Decree

Abbreviations: CARICOM = Caribbean Community; CIS = Commonwealth of Independent States; EAS = East Africa States; EU = European Union; GCC = Gulf Cooperation Council; Gov. Org. = governmental organization (ministry, agency, department); GDP = gross domestic product; LM = Lower Middle; NA = NA; WHO = World Health Organization; UM = Upper Middle; US EPA (United States Environmental Protection Agency)

a Law/Decree includes legislation and presidential or royal decrees; Gov. Org. includes ministries and governmental agencies; Standards Org. includes standards bureaus and standards agencies.

b Standard withdrawn October 14, 2016 [36].

c Caribbean Community and Common Market (CARICOM) standard for packaged purified drinking water is 50 μg/L [106].

d The official government document lists this value as “1,0 ug/l”, which is a magnitude lower than the contemporaneous WHO value [122, 123]. Since the other water contaminants follow the WHO values, this value is likely a typographical error for “10 ug/l”.

e As standard only currently applies to municipalities of > 500K inhabitants; to municipalities of > 50K inhabitants by 2022 and to all by 2025 [128].

f Applies specifically to bottled water [130].

g The official government document lists this value as “0.5 mg/l”, which is a magnitude higher than a previous WHO value [131, 132]. Since the document references the WHO guidelines, and a former WHO guideline for As was 50 μg/L, this value is likely a typographical error for “0.05 mg/l”.

h Permissible limit in the absence of alternative source: 50 μg/L [152].

i Author suggests that East Timor uses WHO guidelines as local standards [190].

j Standards are specifically for the Emirate of Abu Dhabi [196].

k The maximum contaminant level for As in bottled water is 50 μg/L [204].

l A Nauru government document states that there are no standards [206], but the 2018 WHO survey noted that Nauru uses WHO guidelines as standards [18, 24].

m A secondary source states that there are no national standards [214], but the 2018 WHO survey noted that Tonga uses WHO guidelines as standards [18, 24].

n The decree states that in some cases 50 μgl/L is the maximum allowable threshold taking into account the detection limit of the analytical equipment [218].

o Applies specifically to private water supplies [263].

3.1 Arsenic regulations

We found evidence of regulations for a maximum allowable concentration of As in drinking water for 176 countries. There were 19 countries for which we could not find any evidence of drinking water regulations. By comparison, in a 2015 survey of global drinking water regulations, the WHO found regulations for As in 102 out of 104 countries [24] while the International Water Resources Association’s 2018 comparison of water quality guidelines included drinking water standards for 10 countries [25].

Of the As regulations that we found, the lowest maximum allowable concentration of As was 1 μg/L, the highest was 500 μg/L, and the mode was 10 μg/L. The lowest maximum allowable concentration for As of 1 μg/L may be a typographical error in the government document establishing this allowable concentration since the other contaminants listed in the document followed the values listed in the WHO’s drinking water guidelines at the time [122, 123]. The highest maximum allowable concentration of 500 μg/L was also likely a typographical error in the official government record [131]. If these 2 likely errors are set aside, the lowest maximum concentration for As was 5 μg/L and the highest maximum concentration was 300 μg/L. When the likely typographical errors are adjusted to their probable intended values, 2 countries have a maximum allowable concentrations less than 10 μg/L, 134 countries have maximum allowable concentrations of 10 μg/L, and 40 countries have maximum allowable concentrations of greater than 10 μg/L. For all statistical analyses that follow, we retained the published regulatory values without adjustments for possible typographical errors.

The oldest regulation for As in drinking water currently in force is dated 1985 [142, 143, 187], while the most recent regulation dates to 2021 [112]. The average date of the regulations is 2011, while the mode is 2017 (Fig 1). Arsenic standards are established by national legislation or decree in 85 countries, ministries or agencies in 58 countries, and national standards boards in 33 countries. Many standards published by national standards boards are copyrighted documents and must be purchased or licensed for a fee in order to be accessed.

Fig 1

Comparison of publication dates of arsenic drinking water standards.

3.2 Arsenic regulations, population, and per capita income

Seventy percent (136) of the 195 countries in our survey have regulations for As in drinking water that are equal to or more protective than the WHO’s drinking water guideline of 10 μg/L, while 21% (40) of the 195 countries have regulations that are less protective than the WHO’s 10 μg/L guideline, and we could not find regulations for 10% (19) of the 195 countries. Sixty-six percent of the world’s population lives in countries with As drinking water regulations equal to or more protective than the WHO’s 10 μg/L guideline, 32% live in countries with regulations that are less protective than the WHO’s 10 μg/L guideline, and 2% live in countries where we could not find evidence of a drinking water guideline for As (Fig 2).

Fig 2

Comparison of populations covered by different levels of arsenic drinking water standards.

Arsenic regulations are also strongly tied to national income as represented by GDP per capita. The sum of all GDPs of the countries with 2019 GDP data and an As regulation equal to or more protective than the WHO guideline of 10 μg/L divided by the sum of the population of these countries was $13,587 per capita. In contrast, the sum of all GDPs of countries with 2019 GDP data and an As regulation less protective than the WHO guideline divided by the sum of the population of these countries was $7,601 per capita. The sum of all GDPs of countries for which we had 2019 GDP data but could not find an As regulation divided by the sum of the population of these countries was $1,227 per capita. The GDPs per capita of countries with As regulations equal to or more protective than the WHO guideline of 10 μg/L were significantly higher (n = 129, M = $17,678) than those of countries with As regulations less protective than the current 10 μg/L WHO guideline (n = 36, M = $5,384) (F(2,177) = 7.55, p < .001).

A graph of GDP per capita versus national drinking water standard or guideline for As is shown in Fig 3. Maps of GDP per capita and national drinking water standards or guidelines for As are shown in Figs 4 and 5, respectively.

Fig 3

Gross domestic product (GDP) per capita in current (2019) United States dollars (US $) versus national drinking water standard or guideline for arsenic in micrograms per liter (μg/L) for 180 countries [20].

Fig 4

A map of 2019 gross domestic product (GDP) per capita in United States dollars (US $) for 180 countries (map base: [264]).

Fig 5

A map of national drinking water standards or guidelines for arsenic in micrograms per liter (μg/L) for 180 countries (map base: [264]).

The World Bank classifies economies into one of 4 groups: low income, lower middle income, upper middle income, and high income [31]. We found no difference in recency of As regulations by income group (F(3,175) = 0.63, p = 0.60). For the countries that have As regulations, the mean As regulation for low income countries was 21 μg/L, for lower middle income countries the mean was 24 μg/L, for upper middle income countries the mean was 38 μg/L, and for high income countries the mean was 11 μg/L (Fig 6). An analysis of variance (ANOVA) on these means yielded significant variation between income classes (F(3, 172) = 3.15, p = .03). A post hoc Tukey test showed that the mean As standard of the high income class differed significantly from that of the upper middle income class (p = .01) while the remaining differences were not significant (p>0.05).

Fig 6

Comparison of mean national arsenic drinking water standards by income group.

Sixteen countries tie their maximum allowable As concentrations directly to the WHO drinking water guidelines. A large number of other countries have maximum allowable As concentrations that equal the current or previous WHO drinking water guidelines. This underscores the importance of the WHO drinking water guidelines for protecting the health of world citizens.

3.3 The drinking water guideline set by the World Health Organization

The WHO first established a drinking water standard for As, 200 μg/L, in 1958 based on health concerns [123, 265]. As adverse health effects from As exposure received more study, this As drinking water standard was lowered to 50 μg/L in 1963 [16, 123]. In 1971, the World Health Organization (WHO) deemed its 50 μg/L As standard a “tentative limit” [132] and noted, “Some epidemiological studies have suggested that arsenic is carcinogenic but no real proof of the carcinogenicity to man of arsenic in drinking-water has been forthcoming. It would seem wise to keep the level of arsenic in drinking-water as low as possible” [132].

In 1971, 50 μg/L was “as low as possible” for a limit because the available methods for the routine analysis of As in drinking water were not accurate, precise, sensitive, or affordable enough to reliably measure lower concentrations [132]. For example, the recommended method of polarographic estimation was often not accurate due to interferences, and the incomplete reduction of analyte [132, 266]. Similarly, the commonly used spectrophotometric determination of As by the silver diethyldithiocarbamate (AgSCSN(CH2CH3)2) method was not precise due to the sometimes incomplete generation of arsine gas (AsH3(g)) from the sample matrix [132, 267]. In contrast, atomic absorption spectroscopy (AAS) was accurate, precise, and sensitive, but not affordable in many low-income countries [132, 267].

In 1984, the WHO began publishing drinking-water “guidelines” instead of “standards” [268]. Guidelines were intended to provide guidance to national regulators and stakeholders, but they are specifically not to be taken as international standards. National regulators are encouraged to take local conditions, resources, and hazards into account when setting national standards [268, 269]. The 1984 WHO drinking-water guideline for As was maintained at 50 μg/L, the former WHO standard [268].

In 1993, the WHO replaced its earlier “tentative limit” of 50 μg/L with a “provisional guideline” of 10 μg/L for As in drinking water [17]. By this time, a skin cancer risk in humans was known, and other cancer risks were suspected. “Inorganic arsenic is a documented human carcinogen and has been classified by IARC [International Agency for Research on Cancer] in Group 1 [‘This category is used when there is sufficient evidence of carcinogenicity in humans.’]. A relatively high incidence of skin and possibly other cancers that increase with dose and age has been observed in populations ingesting water containing high concentrations of arsenic” [17]. The WHO calculated a health-based value of 0.17 μg/L, but noted that this value was below the practical quantification limit of 10 μg/L [17]. Therefore, the 10 μg/L drinking water guideline for As was initially “provisional” because of what is now called “analytical achievability” [17, 269]. Thus, the 1993 WHO 10 μg/L drinking water guideline for As was based on practical analytical concerns rather than health data, since health data would have led to a lower guideline [17].

The WHO has maintained this 10 μg/L provisional drinking water guideline for As in all subsequent editions and addendums of Guidelines for Drinking-water Quality since 1993 [18, 123, 270–272]. In 2006, the WHO stated that As in drinking water not only causes skin cancer, but also causes bladder and lung cancers [271]. “There is overwhelming evidence from epidemiological studies that consumption of elevated levels of arsenic through drinking-water is causally related to the development of cancer at several sites, particularly skin, bladder and lung” [271]. In 2011, the WHO added kidney cancer to this list [272]. “The International Programme on Chemical Safety (IPCS) concluded that long-term exposure to arsenic in drinking-water is causally related to increased risks of cancer in the skin, lungs, bladder and kidney, as well as other skin changes, such as hyperkeratosis and pigmentation changes” [272]. However, the WHO did not lower their drinking water guideline for As “because [the] calculated guideline value is below the achievable quantification level” [272]. Again, “guideline values are not set at concentrations of substances that cannot reasonably be measured. In such circumstances, provisional guideline values are set at the reasonable analytical limits” [269, 272].

In 2011, the WHO added a second provision to its provisional 10 μg/L drinking water guideline for As [272]. This second provision is based on “treatment performance” [272]. More specifically, “It is technically feasible to achieve arsenic concentrations of 5 μg/l or lower using any of several possible treatment methods. However, this requires careful process optimization and control, and a more reasonable expectation is that 10 μg/l should be achievable by conventional treatment (e.g. coagulation)” [272]. In 2017, the WHO maintained its 10 μg/L provisional drinking water guideline for As, the “analytical achievability” provision for this guideline, and the “treatment performance” provision for this guideline [18, 269].

In summary, the current WHO 10 μg/L provisional drinking water guideline for As is not set according to a health-based risk assessment, since this would require a lower maximum concentration, but rather, it is based on the detection limits for the routine analysis of As in drinking water, which have not been updated since 1993, and treatment technologies for As, which have not been updated since 2011 [18, 272].

3.4 Health-based drinking water guidelines

In 2000, the United States Environmental Protection Agency (U.S. EPA) proposed a non-enforceable health-based goal, or Maximum Contaminant Level Goal (MCLG) for As of 0 μg/L [273]. This 0 μg/L MCLG remains in force [135]. In 1999, the U.S. EPA noted that there was a Practical Quantitation Limit (PCL) for As of 3 μg/L [274]. However, instead of using this PQL as the enforceable standard Maximum Contaminant Level (MCL), 5 μg/L was proposed as a national standard based on a cost/benefit analysis [273]. After this proposed MCL of 5 μg/L was set out for public comment, the MCL was raised to 10 μg/L before being adopted as the national standard, over the objections of the U.S. EPA’s own Scientific Advisory Board, who argued for a lower MCL [275, 276].

In contrast, the California Environmental Protection Agency (CalEPA) has set a public health goal (PHG) of 0.004 μg/L for As in drinking water [7, 8]. This PHG is based on human health effects; it is not influenced by detection limits or the performance of drinking water treatment systems. Under California law, a PHG must be based on current scientific evidence and give a negligible risk of adverse health effects over a lifetime of exposure [277]. In addition, a PHG must not consider economic, technical, or other societal factors [277]. As a result, the CalEPA 0.004 μg/L PHG is 2,500 times lower than the WHO provisional drinking water guideline and the U.S. EPA drinking water standard of 10 μg/L.

More specifically, the CalEPA PHG is based on the risk of death from lung and bladder cancers after a lifetime of exposure to As in drinking water [8]. These risks were calculated from epidemiological studies in Taiwan, Chile, and Argentina (Eqs 1, 2, 3, 4, 5 and 6) [8]. Since the risk of death from lung and bladder cancers is greater than that from skin cancer, kidney cancer, and noncancer health effects, the risks from these 3 less significant effects were not factored into the CalEPA PHG (Eqs 1 and 2) [8].

The nonsignificant digits, “75”, are shown as a subscript (Eqs 1, 2, 3, 4, 5 and 6). These nonsignificant digits are included in all steps of a calculation to prevent rounding error. The CalEPA rounded their PHG to 1 significant figure as follows (Eq 2) [8].

This 0.004 μg/L PHG is estimated to result in 1 excess cancer death in 1,000,000 people (Eq 2) [8]. That is, if 1,000,000 people drank water with 0.004 μg of As/L over their lifetimes, it is estimated that 1 of these 1,000,000 people would die from cancer because of this exposure to As. The excess death could be prevented if the dose of the carcinogen is lowered or eliminated [8].

In contrast, the 10 μg/L value used as the WHO provisional drinking water guideline and the U.S. EPA MCL is estimated to result in 2,500 excess cancer deaths in 1,000,000 people (Eq 3), or in 1 excess cancer death in 400 people (Eq 4).

By comparison, the National Research Council estimated that drinking 10 μg/L of As over a lifetime results in 3,700 excess cancer deaths in 1,000,000 males and 3,000 excess cancer deaths in 1,000,000 females (Eq 3) [8].

The 50 μg/L value still used by many countries as national drinking water standards is estimated to result in 12,000 excess cancer deaths in 1,000,000 people (Eq 5), or 1 excess cancer death in 81 people (Eq 6).

3.5 Low-cost methods for improving public health by reducing arsenic exposure

3.5.1 Advances in inexpensive analytical chemistry methods

One reason for using a 50 μg/L drinking water standard for As instead of the more protective WHO 10 μg/L guideline is the high cost of atomic absorption spectrometers or other sophisticated instruments for measuring total As to 10 μg/L or lower. However, recent developments in analytical methods now make it possible to quantify As to 10 μg/L or lower without expensive equipment.

3.5.2 Spectrophotometry

One example of a low-cost method for quantifying As to 10 μg/L or lower without expensive equipment is the arsenomolybdate method, validated in 2005 [267]. By design, the arsenomolybdate method uses the same equipment as the commonly used silver diethyldithiocarbamate (AgSCSN(CH2CH3)2) method for measuring As [267, 278]. In the arsenomolybdate method, As is removed from the sample by reduction to arsine gas (AsH3(g)), collected in an absorber by oxidation to arsenic acid (H3AsO4), colorized by a sequential reaction to arsenomolybdate, and quantified by spectrophotometry at 835 nm (nanometers) [267]. The method detection limit is 7 μg/L (Table 2) [267]. This detection limit is intended to equal the “concentration of a substance that can be measured and reported with 99% confidence that the analyte concentration is greater than 0” [279, 280]. In summary, the arsenomolybdate method is more accurate, precise, and environmentally safe than the AgSCSN(CH2CH3)2 method; and it is more accurate and affordable than the graphite furnace atomic absorption spectroscopy (GFAAS) method [267].

Table 2

Common drinking water standards, guidelines, and public health goals for total arsenic (As) in micrograms per liter (μg/L), the detection limits for total As in μg/L by spectrophotometry, and the estimated cancer risks at these concentrations (Eqs 1, 2, 3, 4, 5 and 6).

These cancer risks are in bold font and rounded to 2 figures.

As concentration	NumberofExcessCancerDeaths1,000,000People	1ExcessCancerDeathNumberofPeople
50 μg/L (drinking water standard common in lower income countries)	12,000ExcessCancerDeaths1,000,000People	1ExcessCancerDeath81People
10 μg/L (WHO provisional drinking water guideline; drinking water standard common in higher income countries)	2,500ExcessCancerDeaths1,000,000People	1ExcessCancerDeath400People
7.5 μg/L (detection limit by spectrophotometry using suspended nanoparticles [286])	1,900ExcessCancerDeaths1,000,000People	1ExcessCancerDeath540People
7 μg/L (detection limit by spectrophotometry using arsenomolybdate [267])	1,700ExcessCancerDeaths1,000,000People	1ExcessCancerDeath580People
4 μg/L (detection limit by spectrophotometry using an arsenoantimonomolybdenum blue-malachite green complex [283])	990ExcessCancerDeaths1,000,000People	1ExcessCancerDeath1,000People
4 μg/L (detection limit by spectrophotometry using suspended microparticles [284])	990ExcessCancerDeaths1,000,000People	1ExcessCancerDeath1,000People
0.5 μg/L (detection limit by spectrophotometry using suspended nanoparticles [285])	120ExcessCancerDeaths1,000,000People	1ExcessCancerDeath8,100People
0.3 μg/L (detection limit by spectrophotometry using a molybdoarsenate-malachite green complex [282])	74ExcessCancerDeaths1,000,000People	1ExcessCancerDeath13,000People
0.004 μg/L (Public Health Goal Set by the California Environmental Protection Agency [8])	1.0ExcessCancerDeath1,000,000People	1ExcessCancerDeath1,000,000People

Other advances in inexpensive analytical chemistry methods include the use of cationic dyes, such as malachite green (C6H5C(C6H4N(CH3)2)2+), that react with oxyanions, such as derivatized As and derivatized phosphorus (P), to form an ionic solid. These solids are either suspended with a surfactant or dissolved with an organic solvent and measured by spectrophotometry [281-283]. More specifically, one method for the determination of total As uses an arsenoantimonomolybdenum blue-malachite green complex [283]. This complex is suspended with Triton™ X-350 and analyzed at 640 nm [283]. The detection limit is 4 μg/L and is defined as the concentration of standard solution that has a 0.01 absorbance (Table 2) [283]. This method is subject to interferences unless the As is removed from the sample by reduction to AsH3 gas before color development [283]. Another method for the determination of total As uses a molybdoarsenate-malachite green complex [282]. This complex is concentrated by filtration onto a nitrocellulose membrane filter [282]. This complex and filter are dissolved with 2-methoxyethanol (methyl cellosolve; CH3OCH2CH2OH), and the filtrate is analyzed at 627 nm [282]. The detection limit is 0.3 μg/L; however, the criteria used to calculate this detection limit was not given (Table 2) [282]. Interference from phosphate is corrected by a selective reduction procedure, interference from ferric iron (Fe(III)) is corrected by a cation exchange procedure, and interference from silicate is corrected by an acidic digestion procedure [282]. Alternatively, all of these interferences are corrected if As is removed from the sample by reduction to AsH3 gas before color development.

More recently, another cationic dye, ethyl violet (C(C6H4N(CH2CH3)2)3+), was reacted with derivatized oxyanions of As to form suspended particles of ionic complexes [284, 285]. More specifically, one method for the determination of total As uses a molybdoarsenate-ethyl violet complex [284]. This complex forms suspended microparticles and an apparently homogenous blue solution [284]. Prior to color development, interference from phosphate is corrected by an anion exchange procedure, and interference from silica is corrected by reaction with sodium fluoride (NaF) [284]. After color development, the excess dye is converted to a colorless carbinol species in strong acid and the molybdoarsenate-ethyl violet complex is analyzed at 612 nm [284]. This decolorization of excess dye significantly reduces the absorbance of reagent blanks and permits a 4 μg/L detection limit (Table 2) [284]. This detection limit is defined as 3σ/m, “where σ is the standard deviation of 5 measurements of the reagent blank, and m is the slope of the calibration graph” [284]. This method was modified and uses ethyl violet, an isopolymolybdate-iodine tetrachloride complex, and molybdoarsenate to form suspended nanoparticles that are analyzed at 550 nm as a determination of total As [285]. Prior to color development, interference from ferric iron is corrected by reaction with ethylenediaminetetraacetic acid (EDTA), interference from phosphate is corrected by an anion exchange procedure, and interference from silica is corrected by reaction with NaF [285]. The detection limit is 0.5 μg/L (Table 2) [285]. This detection limit is defined as 3σ/m [285].

Another method for the determination of total As uses a reduction and selective extraction of arsenite, As(III), into an ionic liquid functionalized with gold (Au) nanoparticles that are analyzed visually or potentially with a spectrophotometer [286]. The reducing agent is ascorbic acid (C6H8O6) [286]. The ionic liquid is prepared by mixing ultrapure water (in this case, resistivity = 18.3 MΩ⋅cm), tetradecyl (trihexyl) phosphonium chloride (Cyphos® IL-101; [C14(C6)3P]Cl), and Triton™ X-114 [286]. This ionic liquid is functionalized by adding chloroauric acid (HAuCl4) and potassium tetrahydroborate (KBH4) [286]. The functionalized ionic liquid is red in the absence of As(III) and blue in the presence of As(III) [286]. The estimated detection limit by naked eye is 7.5 μg/L (Table 2) [286]. This method is highly selective; 1.0 micromolar (μM) concentrations of K+, Na+, Ca2+, Mg2+, Al3+, Ni2+, Fe3+, Cr3+, Zn2+, Mn2+, Pb2+, Cd2+, Hg2+, SO42−, PO43−, CO32−, NO2−, and SiO32− do not significantly interfere [286]. Higher, but unspecified concentrations, of Cl−, NO3−, and SCN− do not significantly interfere [286].

3.5.3 Summary of advances in inexpensive analytical chemistry methods

In summary, no country needs to use the less protective 50 μg/L standard or guideline due to the expense of analytical chemistry methods. There are many affordable methods for measuring total As to the more protective WHO 10 μg/L provisional drinking water guideline, or to concentrations as low as 0.3 μg/L (Table 2).

3.5.4 Advances in inexpensive drinking water treatment technologies

Another reason for using a 50 μg/L drinking water standard instead of the more protective WHO 10 μg/L guideline is the high expense of drinking water treatment systems. However, advances in inexpensive drinking water treatment technologies have produced technologies that can now remove As to concentrations that are lower than 10 μg/L [287]. Selected examples of these advances follow.

Batch rectors used optimized pH adjustment, oxidation, coagulation, and filtration to economically remove As from drinking water to “about 5 μg/L” during field trials of 10 households and 6 schools in Assam, India [288]. Sodium bicarbonate (NaHCO3) was used to adjust pH, potassium permanganate (KMnO4) was used to oxidize ambient As(III) and Fe(II) to relatively insoluble As(V) and Fe(III), and iron (III) chloride (FeCl3) was used to coagulate the oxidized As and Fe [288]. The treated water was allowed to settle for 1 to 2 hours [288]. The supernatant was filtered through sand-gravel filters [288]. The households each used a 10 L batch rector, 5 schools each used a 25 L batch reactor, and 1 school used a 200 L batch reactor [288]. The concentration of As in the influent ranged from 100 μg/L to 240 μg/L [288]. The concentration of As in the effluent was less than 5 μg/L if the initial concentration of Fe in the influent was less than 1,000 μg/L, and the concentration of As in the effluent was less than 8 μg/L if the initial concentration of Fe in the influent was greater than 2,500 μg/L [288]. The estimated recurring cost per cubic meter (m3) of treated water was approximately US $0.16 per/m3 [288].

An electrocoagulation batch reactor economically removed As from drinking water to concentrations that were always less than 5 μg/L during a 3.5 month field trial in West Bengal, India [289]. More specifically, electrolytic oxidation of a sacrificial iron (Fe) anode produced Fe(III) precipitates. These Fe(III) precipitates reacted with As in the influent and produced As-Fe(III) precipitates. These As-Fe(III) precipitates aggregated or flocculated together. These flocs were mixed with a small amount of alum and removed by gravitational settling [289]. The concentration of As in the influent was 266 ± 42 μg/L and the average concentration of As in the effluent was 2.1 ± 1.0 μg/L [289]. This batch reactor treated 31,000 L (50 batches) of drinking water during the 3.5 month field trial [289]. The cost of treated water was US $0.83/m3 to US $1.04/m3 [289].

Above ground adsorbent filters economically removed As from drinking water to concentrations that were consistently less than 10 μg/L during field trials of 20 households and 3 schools in West Bengal, India [290]. These filters were made of activated carbon, charcoal, fine granular sand, activated laterite, and raw laterite [290]. Laterites are highly weathered soils that are dominated by clay sized particles of iron hydrous oxides and aluminum hydrous oxides [291]. The laterite was activated by sequential treatment with hydrochloric acid (HCl) and sodium hydroxide (NaOH) [290]. The concentration of As in the influent ranged from 50 μg/L to 500 μg/L [290]. The concentration of As in the effluent was always less than 10 μg/L [290]. These filters have a relatively large As removal capacity, 32.5 milligrams (mg)/g, and as a result have a relatively long service life, at least 5 years [290]. The cost of the treated water was less than US $0.35/m3 [290].

Below ground adsorbent filters economically removed As from drinking water to concentrations that were consistently less than 10 μg/L during field trials of 4 households in Hangjinhouqi County, Inner Mongolia, China [292]. These filters were made by mixing 1 mass unit of a locally abundant limestone with 2 mass units of a locally abundant Fe-mineral (hematite and goethite) [292]. This mixture was placed below ground, around the well screen of a conventional tube well [292]. The limestone likely increased the pH of the groundwater [293]. This increase in pH likely enhanced the oxidation of soluble As(III) with dissolved oxygen (O2(g)) to make insoluble As(V) [294]. This As(V) was removed from the groundwater by precipitation with the dissolved calcium (Ca2+(aq)) from the limestone and by adsorption to the surface of the Fe-mineral [292]. The concentration of As in the unfiltered groundwater ranged from 318 μg/L to 635 μg/L [292]. The concentration of As in the filtered groundwater was always less than 10 μg/L [292]. “The filtration system was continuously operated for a total volume of 365,000 L, which is sufficient for drinking water supplying a rural household of 5 persons for 5 years at a rate of 40 L per person per day” [292]. The cost of the treated water was less than US $0.10/m3 [292].

A groundwater extraction, aeration, and reinjection system economically removed As from drinking water to concentrations that were consistently less than 10 μg/L during a village scale field trial in West Bengal, India [295]. More specifically, groundwater was extracted from the aquifer with a submersible electric pump [295]. This groundwater was aerated by spraying it into an above ground plastic tank with “ordinary plastic shower heads” and letting it sit in the tank for at least 30 minutes [295]. The final concentration of O2(aq) in this aerated water ranged from 4 mg/L to 6 mg/L [295]. Approximately, 15% to 20% of this aerated water as reinjected into the aquifer by gravity, and the remaining 80% to 85% of this aerated water was used for drinking [295]. The concentration of As in the influent was not clearly specified [295]. The concentration of As in the effluent was always less than 10 μg/L [295]. The cost of the treated water was US $0.50/m3 [295].

In summary, no country needs to use the less protective 50 μg/L standard or guideline due to the expense of treatment technologies. There are many affordable methods for treating water to reduce As concentrations to lower than 10 μg/L (Table 3).

Table 3

Common drinking water standards, guidelines, and public health goals for total arsenic (As) in micrograms per liter (μg/L), the effluent concentrations of total As in μg/L from water treatment systems used in the low-income world, and the estimated cancer risks at these concentrations (Eqs 1, 2, 3, 4, 5 and 6).

These cancer risks are in bold font and rounded to 2 figures.

As concentration	NumberofExcessCancerDeaths1,000,000People	1ExcessCancerDeathNumberofPeople
50 μg/L (drinking water standard common in lower income countries)	12,000ExcessCancerDeaths1,000,000People	1ExcessCancerDeath81People
10 μg/L (WHO provisional drinking water guideline; drinking water standard common in higher income countries)	2,500ExcessCancerDeaths1,000,000People	1ExcessCancerDeath400People
<10 μg/L (treatment performance using above ground adsorbent filters [290]; below ground adsorbent filters [292]; groundwater extraction, aeration, and reinjection system [295])	<2,500ExcessCancerDeaths1,000,000People	<1ExcessCancerDeath400People
5 μg/L (treatment performance using an optimized pH adjustment, oxidation, coagulation, and filtration in a batch reactor [288])	1,200ExcessCancerDeaths1,000,000People	1ExcessCancerDeath810People
<5 μg/L (treatment performance using an electrocoagulation batch reactor [289])	<1,200ExcessCancerDeaths1,000,000People	<1ExcessCancerDeath810People
0.004 μg/L (Public Health Goal Set by the California Environmental Protection Agency [8])	1.0ExcessCancerDeath1,000,000People	1ExcessCancerDeath1,000,000People

3.6 Methods for improving public health by reducing arsenic exposure that require higher expenditures

3.6.1 Other advances in analytical chemistry methods

Some countries use a 10 μg/L drinking water standard or guideline because they assume that this continues to be the limit of quantification for routine analytical chemistry laboratories. However, recent advances in analytical chemistry have developed methods that can be used in routine laboratories with detection limits that are at 0.1 μg/L or less [18, 296]. With these advances 10 μg/L As should no longer be considered the practical limit of quantification, so revised As standards or guidelines could be more protective of public health.

3.6.2 Inductively coupled plasma-mass spectrometry

Advances in inductively coupled plasma-mass spectrometry (ICP-MS) give detection limits for total As that are at 0.1 μg/L or less [18, 296]. Moreover, ICP-MS is commonly used in routine drinking water testing laboratories because it can simultaneously measure the concentrations of almost all of the elements on the periodic table. This reduces cost by allowing each sample to be analyzed for multiple elements in just a few seconds.

An ICP-MS ionizes the As or other analytes in an inductively coupled plasma (ICP) and then separates and quantifies these ions in a mass spectrometer (MS) [296]. The drinking water sample is turned into an aerosol and delivered to an argon (Ar) plasma [296]. This plasma is a flame like object at the top of an ICP torch, and has a sufficient concentration of Ar cations (Ar+(g)) and free electrons (e−(g)) to make the gas electrically conductive (Eq 7) [296].

This torch is surrounded by an induction coil [296]. This coil transmits power from a radio frequency (RF) generator to the plasma [296]. This energy maintains the ionization of the Ar gas, and the temperature of the plasma from about 5,500 kelvin (K) to 8,000 K [296]. These temperatures are approximately 2 or 3 times hotter than all flame spectroscopic methods [296]. As a result, these higher temperatures help make ICP-MS more sensitive than flame atomic absorption spectrometry (FAAS), flame atomic emission spectrometry (FAES), and flame atomic fluorescence spectrometry (FAFS) [296]. The reaction for the ionization of As in a plasma follows (Eq 8).

These Ar+(g; Eq 7) ions, As+(g; Eq 8) ions, and any other ions exit the ICP and enter the MS [296]. An MS is a mass filter and a mass detector [296, 297]. The most common type of MS used in atomic mass spectroscopy is the quadrupole mass analyzer [296]. This analyzer uses direct current (DC) and radiofrequency (RF) fields to filter ions [296, 297]. These ions are directed in between 4 parallel rods, the quadrupole [296, 296]. These rods are connected to a source of variable DC voltages; 2 rods are positively charged and 2 rods are negatively charged [296]. In addition, variable RF alternating current (AC) voltages are applied to each pair of rods [296]. The DC and AC voltages on the rods are simultaneously changed; however, the ratio of these voltages is held constant [296]. This makes the ions oscillate between the rods [296]. At any instant, only the ions with the desired mass and charge exit the quadrupole and are detected at an ion transducer; all other ions hit the rods, are converted to 0 charge, and are not detected [296]. In this way, an entire spectrum of atomic masses can be scanned in less than 0.1 seconds [296]. Ultimately, the MS detector measures the ratio of mass to charge (m/z) for each cation [296-297]. Since ions with multiple charges are rarely produced, the charge (z) is normally assumed to equal 1 and m/z is normally the mass of the cation [296-297].

The only naturally occurring isotope for As is 75As; that is, all of the As in nature has 75 protons and neutrons [298]. Therefore, all of the As in nature has m/z = 75. Unfortunately, if a drinking water sample has chlorine (Cl), calcium (Ca), or sulfur (S), these common atoms can react with the plasma to form 40Ar35Cl+(g), 38Ar37Cl+(g), 37Cl21H+(g), 40Ca35Cl+(g), or 40Ar34S1H+(g) [299, 300]. These are polyatomic interferences; they all have m/z = 75 and cannot be distinguished from 75As+(g) [299, 300]. No other isotope of As can be used to eliminate this systematic error [298]. Mathematical corrections can be used to estimate the contribution of these polyatomic interferences to the signal at m/z = 75; however, the combined uncertainties in these corrections tend to inflate the detection limit [301].

Fortunately, advances in collision/reaction cell (C/RC) technology are being used to eliminate these polyatomic interferences and give lower detection limits [301, 302]. If a polyatomic ion exits an ICP, enters a C/RC, and collides with an inert gas, such as helium (He(g)), the interference can be eliminated by breaking the polyatomic ion apart (collision-induced dissociation, CID) or by slowing the polyatomic ion down (kinetic energy discrimination, KED) [300]. Or if a polyatomic ion exits an ICP, enters a C/RC, and collides with a reactive gas, such as hydrogen (H2(g)), the interference can be eliminated by changing the mass of the polyatomic ion [300]. For example, an Agilent Technologies 7500c ICP-MS using a C/RC with 0.5 mL/minute of He(g) and 3.8 mL/minute of H2(g) eliminated interferences from 1 g/L of sodium chloride (NaCl) and gave a 0.025 μg/L detection limit for total As (Table 4) [302]. In this case, this detection limit is the lowest concentration that can be quantified at the 99.86% confidence level [302, 303]. This 0.025 μg/L detection limit is 400 times less than a 10 μg/L drinking water standard (Table 4).

Table 4

Common drinking water standards, guidelines, and public health goals for total arsenic (As) in micrograms per liter (μg/L), the detection limits for total As in μg/L by inductively coupled plasma-mass spectrometry (ICP-MS), inductively coupled plasma-tandem mass spectrometry (ICP-MS/MS), and hydride generation-gas chromatography-photoionization detection (HG-GC-PID), and the estimated cancer risks at these concentrations (Eqs 1, 2, 3, 4, 5 and 6).

These cancer risks are in bold font and rounded to 2 figures.

As concentration	NumberofExcessCancerDeaths1,000,000People	1ExcessCancerDeathNumberofPeople
50 μg/L (drinking water standard common in lower income countries)	12,000ExcessCancerDeaths1,000,000People	1ExcessCancerDeath81People
10 μg/L (WHO provisional drinking water guideline; drinking water standard common in higher income countries)	2,500ExcessCancerDeaths1,000,000People	1ExcessCancerDeath400People
0.029 μg/L (detection limit by ICP-MS using a Knotted Reactor [304])	7.2ExcessCancerDeaths1,000,000People	1ExcessCancerDeath140,000People
0.025 μg/L (detection limit by ICP-MS using a Collision/Reaction Cell [302])	6.2ExcessCancerDeaths1,000,000People	1ExcessCancerDeath160,000People
0.004 μg/L (public health goal set by the California Environmental Protection Agency [8])	1.0ExcessCancerDeath1,000,000People	1ExcessCancerDeath1,000,000People
0.0016 μg/L (detection limit by ICP-MS/MS using a collision/reaction cell with O2(g) [305])	0.40ExcessCancerDeaths1,000,000People	1ExcessCancerDeath2,500,000People
0.00082 μg/L (detection limit by HG-GC-PID [307])	0.20ExcessCancerDeaths1,000,000People	1ExcessCancerDeath4,900,000People
0.0002 μg/L (detection limit by ICP-MS/MS using a collision/reaction cell with 10% CH3F(g) and 90% He(g) [306])	0.050ExcessCancerDeaths1,000,000People	1ExcessCancerDeath20,000,000People

Another advance uses flow injection analysis (FIA) and a knotted reactor to remove interference precursors from the sample matrix and concentrate total inorganic As before analysis by ICP-MS [304]. Arsenate (As(V)) was reduced to arsenite (As(III)) in a solution of 1% (mass/volume) L-cysteine (HSCH2CHNH2COOH) and 0.03 molar (M) nitric acid (HNO3) [304]. This As(III) was complexed with a solution of 0.1% (mass/volume) ammonium pyrrolidine dithiocarbamate ((CH2)4NCS2NH4) [304]. This complex was absorbed on the inner wall of a knotted reactor, in this case a 150-centimeter (cm) long by 0.5-millimeter (mm) inside diameter (ID) piece of polytetrafluoroethylene (PTFE) tubing [304]. The interference precursors from the sample matrix were washed away and the total inorganic As was concentrated when this complex was absorbed in the reactor [304]. This complex was desorbed from the reactor with 1 molar (M) HNO3 and eluted into a Perkin-Elmer-Sciex ELAN 5000 ICP-MS [304]. This gave a 0.029 μg/L detection limit for total inorganic As (Table 4) [304]. This detection limit is defined as 3 times the sample standard deviation (s), presumably from the measurement of reagent blanks [304].

3.6.3 Inductively coupled plasma-tandem mass spectrometry

Advances in inductively coupled plasma-tandem mass spectrometry (ICP-MS/MS) give detection limits for total As that are 0.01 μg/L or less [301]. The first MS is typically used as a mass filter and lets through only ions at m/z = 75 (75As+(g), 40Ar35Cl+(g), 38Ar37Cl+(g), 37Cl21H+(g), 40Ca35Cl+(g), and 40Ar34S1H+(g)) [299-301]. If present, these ions enter a C/RC and the 75As+(g) selectively reacts with either oxygen (O2(g)) or fluoromethane (CH3F(g)) to produce 75As16O+(g) at m/z = 91 or 75As12C1H2+(g) at m/z = 89, respectively [301, 305, 306]. The second MS is used to filter the interferences at m/z = 75, and quantify the 75As16O+(g) at m/z = 91 or 75As12C1H2+(g) at m/z = 89 [301, 305, 306]. This use of 2 mass spectrometers in tandem increases sensitivity and lowers the detection limit [301].

For example, if O2(g) is used in the C/RC, a 0.0016 μg/L detection limit for total As was observed (Table 4) [306]. This detection limit is defined as the lowest concentration that can be quantified at the 99.7% confidence level [303, 305]. If 10% CH3F(g) and 90% He(g) are used in the C/RC, a 0.0002 μg/L detection limit for total As was observed (Table 4) [306]. This detection limit is defined as 3s/m, where s is the sample standard deviation from 10 measurements of a reagent blank, and m is the average slope from 10 calibration graphs [306].

ICP-MS/MS is a very new technology; the first commercial instruments were sold in 2012 [301]. As a result, ICP-MS/MS is mostly used for research and is not commonly used in drinking water testing laboratories. However, ICP-MS/MS will likely become more commonly used for drinking water testing in the future.

3.6.4 Hydride generation-gas chromatography-photoionization detection

Advances in hydride generation-gas chromatography-photoionization detection (HG-GC-PID) give a detection limit for total inorganic As at 0.00082 μg/L [307]. This detection limit is defined as 3s, where s is the sample standard deviation from 5 measurements of a reagent blank [307]. The hydride generation step uses 50 mL of sample, 2.0 mL of concentrated hydrochloric acid (HCl), 3.0 mL of 1.0 M potassium iodide (KI), and 4.0 mL of 4.0% (weight/volume) sodium borohydride (NaBH4) to reduce arsenate (As(V)) and arsenite (As(III)) to arsine gas (AsH3(g)) [307]. Helium (He(g)) is used as a carrier gas to move the AsH3(g) from the reaction vessel, to a trap at −50°C, a trap at −196°C, a gas chromatograph (GC), and a photoionization detector (PID) [307]. If present, water vapor (H2O(g)) is an interference and is removed from the AsH3(g) in a trap that is cooled to −50°C with dry ice (CO2(s)) and 2-propanol (CH3CHOHCH3) [307]. After this step, the AsH3(g) is concentrated in a trap that is cooled to −196°C with liquid nitrogen (N2(l)) [307]. If present, stibine gas (SbH3(g)) is an interference and is separated from the AsH3(g) in a GC with a Carbopack™ B HT column [307]. Finally, AsH3(g) is quantified using a 10.2 electron volt (eV) PID [307].

3.6.5 Summary of other advances in analytical chemistry methods

In summary, no country needs to use the less protective 10 μg/L standard or guideline due to the expense of analytical chemistry methods. There are many methods for measuring total As to lower and the more protective concentrations (Table 4).

3.6.6 Other advances in drinking water treatment technologies

Another reason for continuing to use a 10 μg/L drinking water standard is “treatment performance” [18, 272]. However, recent advances in treatment technologies allow removal of As to concentrations that are significantly lower than 10 μg/L and would allow lower standards that would be more protective of public health. Selected examples of these advances follow.

By law, the Netherlands has a 10 μg/L drinking water standard [248]; however, the Netherlands voluntarily uses a less than 1 μg/L drinking water guideline to better protect public health. In 2015, “the Association of Dutch Drinking water Companies (Vewin) voluntarily agreed on a guideline of <1 μg/L for As in drinking water” [308]. “This policy is based on a two-step assessment of As in drinking water, including i) an assessment of excess lung cancer risk for Dutch population and ii) a cost-comparison between the health care provision for lung cancer and As removal from water to avoid lung cancer” [308]. The average concentrations of As in raw water from 241 public supply well fields in the Netherlands ranges from <0.5 μg/L to 69 μg/L; the treatment of this water to <1 μg/L saves the Netherlands from 7.2 million Euros (M€)/year to 14 M€/year [310]. This 7.2 M€/year to 14 M€/year includes the savings in health care costs from not having to treat excess lung cancer cases and the engineering costs from treating drinking water [308].

In the Netherlands, As is typically removed from raw well water by an optimized aeration and rapid sand filtration process [308]. This aeration oxidizes the soluble Fe(II) and As(III) that is naturally found in raw well water to insoluble Fe(III) and As(V) [308]. A soluble Fe(III) coagulant, such as FeCl3, is sometimes added to the raw well water [308]. This produces As-Fe(III) precipitates that are removed in a rapid sand filter [308]. This sand filter sometimes uses a coarse granular top layer and finer bottom layer [308]. This optimized process routinely removes As to <1 μg/L [308].

An optimized reverse osmosis process can also economically remove As from drinking water to concentrations that are significantly lower than 10 μg/L [309]. This process uses a 2-stage membrane cascade and can supply drinking water at 0.5 μg of As/L to a population of 20,000 people for US $1,041/day or US $0.52/m3 [309]. A 2-stage membrane cascade has 2 reverse osmosis units connected in series. The feed water enters the first stage, the retentate or reject water from the first stage is discarded, and the permeate from the first stage enters the second stage. The retentate from the second stage is added to the feed water of the first stage. The permeate from the second stage is disinfected and distributed as drinking water. This process uses polyamide membranes [309]. The most significant cost is energy consumption; it is 35% of the total cost [309].

In summary, no country needs to use the less protective 10 μg/L standard or guideline due to limitations of treatment technologies. There are many treatment methods for reducing total As concentrations to lower more protective concentrations (Table 5).

Table 5

Common drinking water standards, guidelines, and public health goals for total arsenic (As) in micrograms per liter (μg/L), the effluent concentrations of total As in μg/L from water treatment systems used in the high-income world, and the estimated cancer risks at these concentrations (Eqs 1, 2, 3, 4, 5 and 6).

These cancer risks are in bold font and rounded to 2 figures.

Drinking Water Standard, Drinking Water Guideline, Public Health Goal, or Effluent Concentration	NumberofExcessCancerDeaths1,000,000People	1ExcessCancerDeathNumberofPeople
50 μg/L (drinking water standard common in lower income countries)	12,000ExcessCancerDeaths1,000,000People	1ExcessCancerDeath81People
10 μg/L (WHO provisional drinking water guideline; drinking water standard common in higher income countries)	2,500ExcessCancerDeaths1,000,000People	1ExcessCancerDeath400People
<1 μg/L (treatment performance using an optimized aeration and rapid sand filtration process [308])	<250ExcessCancerDeaths1,000,000People	<1ExcessCancerDeath4,000People
0.5 μg/L (treatment performance using an optimized reverse osmosis 2-stage membrane cascade [309])	120ExcessCancerDeaths1,000,000People	1ExcessCancerDeath8,100People
0.004 μg/L (public health goal set by the California Environmental Protection Agency [8])	1.0ExcessCancerDeath1,000,000People	1ExcessCancerDeath1,000,000People

3.7 Implementation of regulations

Unfortunately, having legal standards for drinking water quality does not guarantee that these standards will be implemented, or that efforts will be made by suppliers to meet those standards, especially in regions where government and resources are limited. Implementation of standards can also be complicated when multiple jurisdictions are involved [310]. Transparency is vital for the effectiveness of guidelines and standards to protect public health [310]. In countries where standards bureaus control drinking water standards and release them only for a fee, transparency is severely curtailed. Compliance requires additional efforts and resources.

In contrast, in regions where government is effective and resources are sufficient, such as in Denmark and the Netherlands, reducing the national drinking water standard for As to below 5 μg/L in was found to be both technically feasible and affordable [248, 311]. Even in regions in which there are limited resources for implementation and compliance, establishing and updating drinking water standards and guidelines can be of use to stakeholders who need to determine whether a water source is safe, and to put pressure on water suppliers to improve the quality of the water [311].

4. Conclusions

As demonstrated in this study, national and international guidelines and standards for As in drinking water need to be updated in accordance with current research and technologies. Updating regulations has the potential to improve public health by reducing As exposures worldwide. Technologies are now available that could enable both resource-limited and high income countries to adopt more protective drinking water regulations for As.

The WHO 10 μg/L drinking water guideline for As is provisional because it does not sufficiently protect public health. The value of 10 μg/L was specified on the basis of “analytical achievability” or “treatment performance” [17, 18, 132, 272]. However, since the value for this guideline of 10 μg/L for was selected by the WHO in 1993, numerous new technologies and methods have been developed so that 10 μg/L no longer represents a practical limit for either analytical achievability or As treatment performance. This global drinking water guideline should be updated to better protect public health.

Many countries, especially lower income countries continue to use the long-outdated (1963) former WHO standard of 50 μg/L as a maximum allowable concentration for As [16]. Because many of these countries also have high populations, nearly one third of the world’s population lives in jurisdictions with a 50 μg/L standard for As. This concentration of 50 μg/L is associated with high levels of morbidity and mortality and can no longer be justified by the high cost of As quantification or treatment since new low-cost analytical and treatment methods are now available. Lowering the maximum allowable concentration from 50 μg/L to 10 μg/L or lower is urgently needed to avoid countless preventable cancer deaths and to better protect public health.

A variety of solutions are required to update and set more protective national drinking water standards and guidelines for As. In fact, each country might set a hierarchy of drinking water standards and guidelines for As. For example, low-income countries with limited resources might need to rely on relatively inexpensive spectrophotometers for testing and relatively simple systems for treating drinking water. If such a country or parts of a country had As-free water (Figs 2 and 3), then a 50 μg/L standard based on analytical achievability could be lowered to a more protective 0.3 μg/L standard through simple testing and water sharing (Table 2). This would lower the lifetime cancer risk from 12,000 excess cancer deaths in 1,000,000 people (1 excess cancer death in 81 people) to 74 excess cancer deaths in 1,000,000 people (1 excess cancer death in 13,000 people; see Table 2). If such a country or parts of the country did not have As-free water (Figs 2 and 3), then a 50 μg/L standard based on “treatment performance” could be lowered to a more protective <5 μg/L standard through treatment (Table 3). This would lower the lifetime cancer risk from 12,000 excess cancer deaths in 1,000,000 people (1 excess cancer death in 81 people) to <1,200 excess cancer deaths in 1,000,000 people (<1 excess cancer death in 810 people; see Table 3).

Countries with greater economic resources have access to more expensive instruments for testing and relatively sophisticated systems for treating drinking water. If such a country or parts of the country had As-free water (Figs 2 and 3), then a 10 μg/L standard based on analytical achievability could be lowered to a more protective ≤0.004 μg/L standard through testing and selective water use (Table 4). This would lower the lifetime cancer risk from 2,500 excess cancer deaths in 1,000,000 people (1 excess cancer death in 400 people) to ≤1 excess cancer deaths in 1,000,000 people (Table 4). If such country or parts of the country did not have As-free water (Figs 2 and 3), then a 10 μg/L standard based on treatment performance to a more protective 0.5 μg/L standard through more complete As treatment (Table 5). This would lower the lifetime cancer risk from 2,500 excess cancer deaths in 1,000,000 people (1 excess cancer death in 400 people) to 120 excess cancer deaths in 1,000,000 people (1 excess cancer death in 8,100 people; see Table 5).

Arsenic regulations, populations, GDPs, and reference links for 195 countries.

(XLSX)

Click here for additional data file.

1 Dec 2021

6 Dec 2021

Journal Requirements:

1. Thank you for stating the following in the Competing Interests/Financial Disclosure:

(TPH has provided expert witness testimony on the subject of prevention of falls in hospitals for K&L Gates Law Firm and Minter Ellison Law Firm.)

We note that you received funding from a commercial source: (K&L Gates Law Firm, Minter Ellison Law Firm)

Please provide an amended Competing Interests Statement that explicitly states this commercial funder, along with any other relevant declarations relating to employment, consultancy, patents, products in development, marketed products, etc.

Within this Competing Interests Statement, please confirm that this does not alter your adherence to all PLOS ONE policies on sharing data and materials by including the following statement: "This does not alter our adherence to PLOS ONE policies on sharing data and materials.” (as detailed online in our guide for authors http://journals.plos.org/plosone/s/competing-interests). If there are restrictions on sharing of data and/or materials, please state these. Please note that we cannot proceed with consideration of your article until this information has been declared.

Please respond by return email with your amended Competing Interests Statement and we will change the online submission form on your behalf.

-------------------

Response

The references to K&L Gates Law Firm, Minter Ellison Law Firm are likely for another manuscript. We have never heard of these firms and we do not have an author with the initials TPH. We do not have any commercial funders so we believe that we do not need to amend our Competing Interests Statement.

No changes made.

---------------------------

2. In your Data Availability statement, you have not specified where the minimal data set underlying the results described in your manuscript can be found. PLOS defines a study's minimal data set as the underlying data used to reach the conclusions drawn in the manuscript and any additional data required to replicate the reported study findings in their entirety. All PLOS journals require that the minimal data set be made fully available. For more information about our data policy, please see http://journals.plos.org/plosone/s/data-availability.

Upon re-submitting your revised manuscript, please upload your study’s minimal underlying data set as either Supporting Information files or to a stable, public repository and include the relevant URLs, DOIs, or accession numbers within your revised cover letter. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories. Any potentially identifying patient information must be fully anonymized.

Important: If there are ethical or legal restrictions to sharing your data publicly, please explain these restrictions in detail. Please see our guidelines for more information on what we consider unacceptable restrictions to publicly sharing data: http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions. Note that it is not acceptable for the authors to be the sole named individuals responsible for ensuring data access.

We will update your Data Availability statement to reflect the information you provide in your cover letter.

----------------

Response

All data are included in the supplementary file “supplementary file s1.xlsx”. This file is referenced with its caption on the very last line of our document, immediately following the References section.

No changes made.

-------------------

3. We note that Figures 4 and 5 in your submission contain map images which may be copyrighted. All PLOS content is published under the Creative Commons Attribution License (CC BY 4.0), which means that the manuscript, images, and Supporting Information files will be freely available online, and any third party is permitted to access, download, copy, distribute, and use these materials in any way, even commercially, with proper attribution. For these reasons, we cannot publish previously copyrighted maps or satellite images created using proprietary data, such as Google software (Google Maps, Street View, and Earth). For more information, see our copyright guidelines: http://journals.plos.org/plosone/s/licenses-and-copyright.

We require you to either (1) present written permission from the copyright holder to publish these figures specifically under the CC BY 4.0 license, or (2) remove the figures from your submission:

A. You may seek permission from the original copyright holder of Figure 4 and 5 to publish the content specifically under the CC BY 4.0 license.

We recommend that you contact the original copyright holder with the Content Permission Form (http://journals.plos.org/plosone/s/file?id=7c09/content-permission-form.pdf) and the following text:

“I request permission for the open-access journal PLOS ONE to publish XXX under the Creative Commons Attribution License (CCAL) CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). Please be aware that this license allows unrestricted use and distribution, even commercially, by third parties. Please reply and provide explicit written permission to publish XXX under a CC BY license and complete the attached form.”

Please upload the completed Content Permission Form or other proof of granted permissions as an "Other" file with your submission.

In the figure caption of the copyrighted figure, please include the following text: “Reprinted from [ref] under a CC BY license, with permission from [name of publisher], original copyright [original copyright year].”

B. If you are unable to obtain permission from the original copyright holder to publish these figures under the CC BY 4.0 license or if the copyright holder’s requirements are incompatible with the CC BY 4.0 license, please either i) remove the figure or ii) supply a replacement figure that complies with the CC BY 4.0 license. Please check copyright information on all replacement figures and update the figure caption with source information. If applicable, please specify in the figure caption text when a figure is similar but not identical to the original image and is therefore for illustrative purposes only.

The following resources for replacing copyrighted map figures may be helpful:

USGS National Map Viewer (public domain): http://viewer.nationalmap.gov/viewer/

The Gateway to Astronaut Photography of Earth (public domain): http://eol.jsc.nasa.gov/sseop/clickmap/

Maps at the CIA (public domain): https://www.cia.gov/library/publications/the-world-factbook/index.html and https://www.cia.gov/library/publications/cia-maps-publications/index.html

NASA Earth Observatory (public domain): http://earthobservatory.nasa.gov/

Landsat: http://landsat.visibleearth.nasa.gov/

USGS EROS (Earth Resources Observatory and Science (EROS) Center) (public domain): http://eros.usgs.gov/#

Natural Earth (public domain): http://www.naturalearthdata.com/

-------------------------

Response:

We created the maps in Figures 4 and 5 ourselves using the ggmap package in R. As requested in the reference document for the ggmap package (https://cran.r-project.org/web/packages/ggmap/citation.html), we have cited the authors of the ggmap package (Kahle and Wickham, 2013) since they provided the source code for the base maps on a CC by 4.0 basis. The ggmap source code was based on the public domain Natural Earth project. The figure captions of our maps cite Kahle and Wickham, 2013 as reference [264] as the source for these maps. These maps were not made with any copyrighted materials and they are suitable for CC by 4.0 licensing as is. We revised Section 2.2 (Data analysis and statistics) to explicitly state that we created the maps in R with the ggmaps package.

----------------------------

4. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information

--------------------------

Response:

The caption for our Supporting Information file appears on the very last line of our document, immediately following the References section.

No changes made.

-------------------

5. Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

-------------------

Response:

We have reviewed the reference list and it is complete and correct. We have not cited any retracted papers. We have fixed several typos and added missing doi links where needed. We deleted two references in response to reviewer comments and renumbered subsequent references.

-----------------------

Additional Editor Comments:

The manuscript submitted has merit in the field and serves as an excellent reference of current situations globally with respect to Arsenic exposures and the story leading us to where we are today. I do think there should be some substantive revisions of sections brought up by reviews, but the overall content of the manuscript remains sounds. Thank you for your submission!

----------------------

Response:

Thank you for reading the manuscript carefully and soliciting helpful reviewer comments. This was a monumental effort, and we are grateful that our research will be made available with the publication of this paper for others to use.

-----------------------

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: No

________________________________________

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: No

________________________________________

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

________________________________________

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: No

________________________________________

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: the title of a project are compatible with their work results, and the idea of the project is very crucial, in particular, the effects of Arsenic element on public health, and the WHO always promote the researchers to focus that kind of study

Reviewer #2: Arsenic in drinking water: an analysis of global drinking water regulations and recommendations for updates to protect public health

Comments for the authors

In this study, the authors collated and analyzed all drinking water regulations for As from national governments worldwide and related them with the health risks. The data was not interpreted scientifically. The presentation of the data is poor.

Abstract: The authors have only mentioned the data collection and analysis. what is the purpose of the study? what are the recommendations from this study.? Please revise the abstract.

--------------------

Response:

We have revised the abstract to state the purpose and recommendations of the study more explicitly.

-----------------

Please revise the very first statement of the introduction ‘…that is often found in groundwater wells’

-------------------

Response:

Done.

-------------------

226,000,000 people… 226 million people

---------------------

Response:

Done.

-------------------

many low-and middle-income countries still use the World Health Organization’s (WHO’s) 1963 drinking water standard... I don’t agree with this assumption.

------------------------

Response:

Please note that this is not an assumption. It is a fact that is directly supported by the data that we collated for this study. See our data in Table 1 that show that as of 2021, 36 countries still use the WHO’s 1963 drinking water standard. All but 2 of these countries are classified as low or middle income by the World Bank. As mentioned in the Abstract, the percentage of population living in countries that do not meet the current WHO drinking water guideline for As constitutes 32% of the total global population.

No changes made.

------------------

From the previous few decades, groundwater As has gained considerable attention and most of the literature data present the comparison with WHO 10 µg/L.

--------------------------

Response:

It is fortunate that most studies in the literature use the current WHO 10 µg/L guideline for comparison. However, many countries still legally use the former 1963 WHO guideline, as shown in Table 1 and explained in our previous response.

No changes made.

--------------------------

I think the authors should pay more attention to the regulation limits. The Department of Environmental Protection for New Jersey has proposed the As limit of 5 µg/L in drinking water.

There are several other threshold limits from different areas that are not mentioned in this review.

---------------------

Response:

The title of this paper is “Arsenic in drinking water: an analysis of global drinking water regulations and recommendations for updates to protect public health.”. This paper focuses on national regulations, not the regulations of individual states, provinces, districts, etc. within any one nation such as the United States. Thus, we do not cover regulations for the individual states of the United States. We do cover As regulations for every national government in the world that we could locate in our extensive search for national regulations.

No changes made.

-------------------

Moreover, these regulation limits are made depending on the hydrogeochemistry of the specific region. Depending on the exposure scenario, the health risk indices and threshold guidelines are formed. The authors completely missed this type of useful information.

-------------------

Response:

As discussed in our Abstract, Introduction, and Section 3.3 (The drinking water guideline set by the World Health Organization), the World Health Organization’s (WHO’s) drinking water guidelines take into account health risk assessments, analytical capabilities, and treatment performance. The WHO guidelines are not based on hydrogeochemistry. More specifically, the 1963 WHO drinking water standard for As of 50 micrograms per liter (µg/L) was based entirely on “analytical achievability”. Similarly, the WHO’s 10 µg/L guideline established in 1993 was also based entirely on “analytical achievability”; however, in 2011 the additional proviso of “treatment performance” was added to the current WHO WHO’s 10 µg/L guideline. The WHO also states that these values cause an unacceptable risk of death from “skin and possibly other cancers” and cannot be set at a lower, more protective health-based level because it cannot be measured in a routine testing laboratory or removed by conventional treatment plants.

Furthermore, in Section 3.3, we wrote “National regulators are encouraged [by the WHO] to take local conditions, resources, and hazards into account when setting national standards [268,269].” We cited the WHO’s 2018 document “Developing drinking-water quality regulations” [our reference 269] in support of this statement.

No changes made.

------------------

Section 2: It seems like the authors have collected data only from US states. Please clarify.

-----------------------

Response:

The title of this paper is “Arsenic in drinking water: an analysis of global drinking water regulations and recommendations for updates to protect public health”. There is no suggestion in the title that any regulations from individual US states will be discussed, nor is there any survey of data from US states.

The title of our Section 2.1 is “International drinking water standards for arsenic”. As detailed in the first 2 sentences of that section, the entries in the database that we created and present in Table 1 include the 193 United Nations member states plus Taiwan and Kosovo. Individual US state regulations were not part of this database.

No changes made.

-------------------

Section 2.1: I suggest avoiding the use of authorial (we or I). The use of passive voice can be suitable.

------------------------

Response:

Although PLOS does not have explicit advice regarding passive or active voice, many other leading publishers such as Sage Journals (https://journals.sagepub.com/author-instructions/phr), Cambridge University Press (https://www.cambridge.org/core/journals/enterprise-and-society/information/instructions-contributors), NatureProtocols (https://www.nature.com/nprot/for-authors/preparing-your-submission), and the Canadian Medical Association Journal (https://www.cmaj.ca/submission-guidelines) all explicitly recommend or require active voice rather than passive voice. We have not encountered any scientific journals that require passive voice to avoid agency. Thus, we believe it is acceptable to use active voice in scientific writing.

No changes made.

--------------------

This study is more like a meta-analysis review. The authors have collected abundant data and categorized it according to the threshold value, income, health, etc. I question the robustness of statistics.

Which software was used to extract values of the data plotted in a graph?

---------------------

Response:

As stated in Section 2.2 (Data analysis and statistics), we used R version 4.1.1 for all statistics and data analysis. All of the figures and maps were also made with R. We added a sentence to the end of Section 2.2 to explicitly state that we used R to create the figures and maps as well as to perform the statistical analyses.

---------------------

The authors did not follow guidelines of the analysis used in societies in academic fields.

The authors actually used very simple calculations based on simple empirical equations (just simple calculations with Excel). Though they provided those values, I do not think it has much meaning and much scientific advancement.

I think they just collected data but did not interpret it in a mechanistic way.

-------------------

Response:

As noted in Section 2.2 (Data analysis and statistics), we used R for the statistical calculations, not Excel. We believe our use of statistics is in accordance with standard practices for scientific writing. After comparing our reporting of statistics to the PLOS guidelines (https://journals.plos.org/plosone/s/submission-guidelines.#loc-statistical-reporting), to be fully in compliance, we added a statement to Section 2.2 about why we did not use corrections for multiple comparisons. That is, we did not use corrections for multiple comparisons in this paper because we did not make multiple statistical comparisons of the data.

As noted in Sections 3.1 and 3.2, this study is the most comprehensive study of national drinking water regulations for arsenic that has ever been published to date. It provides and analyses data for almost twice as many countries (195) as a 2018 WHO survey (104 countries). As a result, our Table 1 and the references cited in the table constitute the most complete database of national drinking water regulations published to date. By virtue of its completeness of scope, this study provides comparative information about the drinking water regulations for arsenic worldwide that has never been available before.

The statistical analyses of this data set provide the first firm data concerning the relationship between the protectiveness of national regulations for arsenic in drinking water and national per capita income. More specifically, “the GDPs per capita of countries with As regulations equal to or more protective than the WHO of guideline of 10 µg/L were significantly higher (n = 129, M = $17,678) than those of countries with As regulations less protective than the current 10 µg/L WHO guideline (n = 36, M = $5,384) (F(2,177) = 7.55, p<.001)”. It also shows that the recency of regulations is not related to national per capita income. An extremely important finding of this study is that nearly 1/3 of the world’s population live in jurisdictions that still use the 1963 WHO standard of 50 μg As/L as their national standard. We believe these findings are crucial for understanding the relationship between regulations and risks to public health, and that they are critical for directing future efforts to protect health.

No changes made.

-------------------

Section 3.4: All of the data processing and calculation should be mentioned under the methodology section.

----------------

Response:

This Section, (3.4 Health-based drinking water guidelines) of the Results and Discussion section, provides discussion of several regulatory documents and their implications for public health. The equations in this section were not part of our research on current international drinking water regulations, so it is not appropriate to discuss the calculations in the Materials and Methods section, which describes the materials and methods for our collation of international drinking water regulations and their analysis.

The equations are provided here for completeness and transparency to demonstrate the implications of regulations that are based on “analytical achievability”, “treatment performance”, the risk of death from various cancers, and public health. This is essential for the following discussion on various tangible ways to reduce the risk of death from cancer and protect public health.

No changes made.

-----------------

Section 3.5: the section seems totally irrelevant to the current study. how can you relate this section with the data collected from various studies? Did the authors categorize the analysis techniques also…??? Even the developing countries have advanced methods of As analysis.

------------------

Response:

As stressed in Section 4 (Conclusions), our overarching recommendation is that international guidelines and national standards for As in drinking water are out of date and need to be updated in accordance with the advances of science.

In Section 3.2 (Arsenic regulations, population, and per capita income), we showed that many countries, especially low and medium income countries, still use the 1963 WHO standard of 50 μg As/L as their national standards. This standard is based on analytical achievability from 58 years ago. In Section 3.5 (Low-cost methods for improving public health by reducing arsenic exposure), we enumerate examples of currently available low-cost methods for analysis and treatment of As contaminated water. This enumeration is an essential step for our argument that it is feasible for low and middle income countries to adopt more stringent standards regarding As in drinking water. Again, approximately one-third of the global population live in countries that do not meet the current 10 µg/L WHO drinking water guideline for As. We believe that reducing the income inequities in international drinking water regulations would be a positive step towards reducing the global health gap for lower income people.

No changes made.

-----------------

Section 3.5.2: We can not start any new section with ‘for example’…example of what?

----------------

Response:

We have reworded this sentence for clarity.

----------------

Table 2: how the data was verified.

---------------------

Response:

The comparison figures in the 2 right-hand columns of Table 2 are drawn from equations 1-6, as stated in the table caption. As described in detail in Section 3.4 (Health-based drinking water guidelines), these equations and estimates of cancer deaths are drawn directly from published regulatory documents, for which we provide citations in Section 3.4.

No changes made.

------------------------

The consumption of As-contaminated water could not be the only reason for developing cancer.

-------------------

Response:

It is quite correct that “The consumption of As-contaminated water could not be the only reason for developing cancer”. This is why regulators rely on the concept of “Excess cancer deaths”, a common metric in risk assessment used for calculating health-based drinking water quality guidelines and standards. We revised our description of excess cancer deaths in Section 3.4 (Health-based drinking water guidelines) to clarify this point.

-----------------

It is ambiguous to relate the cancer data with instrument detection limits.

-----------------------

Response:

The purpose of Table 2 is to facilitate direct comparisons between regulatory limits, detection limits, and regulatory risk assessments. We believe these comparisons will be useful for prioritizing efforts to protect public health.

No changes made.

----------------

The conclusion is simply a summary of the results. What is the take-home message of this study?

-------------------

Response:

We have added a new paragraph to Section 4 (Conclusions) to state our take-home message more explicitly.

--------------------

All the figure captions and indications are wrongly mentioned in the manuscript.

--------------

Response:

We have checked the figure captions and in-text citations and verified that they are in fully in compliance with PLOS guidelines (https://journals.plos.org/plosone/s/figures). Two figures were numbered incorrectly; they have been deleted in accordance with the following comment.

-------------------

The figures related to the structure of the dyes or molecules seem irrelevant to the study.

-----------------------

Response:

We have deleted these figures and their associated citations.

----------------------------

There are several grammatical and punctuation mistakes in the manuscript.

------------------

Response:

Thank you for this observation. We have proofread the document again and fixed several typographical, grammar, and punctuation errors.

________________________________________

21 Jan 2022

Arsenic in drinking water: an analysis of global drinking water regulations and recommendations for updates to protect public health

PONE-D-21-35059R1

Dear Dr. Frisbie,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Aaron Specht

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Comments were addressed and manuscript is ready for publication

11 Mar 2022

PONE-D-21-35059R1

Arsenic in drinking water: an analysis of global drinking water regulations and recommendations for updates to protect public health

Dear Dr. Frisbie:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Aaron Specht

Academic Editor

PLOS ONE