Articles

Burden of disease attributable to unsafe drinking water,

sanitation, and hygiene in domestic settings: a global
analysis for selected adverse health outcomes
Jennyfer Wolf, Richard B Johnston, Argaw Ambelu, Benjamin F Arnold, Robert Bain, Michael Brauer, Joe Brown, Bethany A Caruso, Thomas Clasen,
John M Colford Jr, Joanna Esteves Mills, Barbara Evans, Matthew C Freeman, Bruce Gordon, Gagandeep Kang, Claudio F Lanata, Kate O Medlicott,
Annette Prüss-Ustün, Christopher Troeger, Sophie Boisson, Oliver Cumming

Summary
Lancet 2023; 401: 2060–71 Background Assessments of disease burden are important to inform national, regional, and global strategies and to guide
Published Online investment. We aimed to estimate the drinking water, sanitation, and hygiene (WASH)-attributable burden of disease for
June 5, 2023 diarrhoea, acute respiratory infections, undernutrition, and soil-transmitted helminthiasis, using the WASH service
[Link]
levels used to monitor the UN Sustainable Development Goals (SDGs) as counterfactual minimum risk-exposure levels.
S0140-6736(23)00458-0
See Comment page 2017
Methods We assessed the WASH-attributable disease burden of the four health outcomes overall and disaggregated
Department of Environment,
Climate Change and Health,
by region, age, and sex for the year 2019. We calculated WASH-attributable fractions of diarrhoea and acute respiratory
World Health Organization, infections by country using modelled WASH exposures and exposure–response relationships from two updated meta-
Geneva, Switzerland analyses. We used the WHO and UNICEF Joint Monitoring Programme for Water Supply, Sanitation and Hygiene
(J Wolf PhD, R B Johnston PhD, public database to estimate population exposure to different WASH service levels. WASH-attributable undernutrition
J E Mills MSc, B Gordon MSc,
K O Medlicott BE, S Boisson PhD,
was estimated by combining the population attributable fractions (PAF) of diarrhoea caused by unsafe WASH and
A Prüss-Ustün PhD); Division of the PAF of undernutrition caused by diarrhoea. Soil-transmitted helminthiasis was fully attributed to unsafe WASH.
Water and Health, Ethiopian
Institution of Water Resources,
Findings We estimate that 1·4 (95% CI 1·3–1·5) million deaths and 74 (68–80) million disability-adjusted life-years
Addis Ababa University, Addis
Ababa, Ethiopia (DALYs) could have been prevented by safe WASH in 2019 across the four designated outcomes, representing
(Prof A Ambelu PhD); FI Proctor 2·5% of global deaths and 2·9% of global DALYs from all causes. The proportion of diarrhoea that is attributable to
Foundation, University of unsafe WASH is 0·69 (0·65–0·72), 0·14 (0·13–0·17) for acute respiratory infections, and 0·10 (0·09–0·10) for
California, San Francisco, CA,
undernutrition, and we assume that the entire disease burden from soil-transmitted helminthiasis was attributable to
USA (B F Arnold PhD); UNICEF
Middle East and North Africa, unsafe WASH.
Amman, Jordan (R Bain MEng);
Institute for Health Metrics Interpretation WASH-attributable burden of disease estimates based on the levels of service established under the
and Evaluation (M Brauer ScD)
SDG framework show that progress towards the internationally agreed goal of safely managed WASH services for all
and Department of Health
Metrics Sciences would yield major public-health returns.
(C Troeger MPH), University of
Washington, Seattle, WA, USA; Funding WHO and Foreign, Commonwealth & Development Office.
School of Population and
Public Health, University of
British Columbia, Vancouver, Copyright © 2023 World Health Organization. Published by Elsevier Ltd. All rights reserved. This is an Open Access
BC, Canada (M Brauer); article published under the CC BY 3·0 IGO license which permits unrestricted use, distribution, and reproduction in
Department of Environmental any medium, provided the original work is properly cited. In any use of this article, there should be no suggestion that
Sciences and Engineering,
WHO endorses any specific organisation, products or services. The use of the WHO logo is not permitted. This notice
Gillings School of Global Public
Health, University of North should be preserved along with the article’s original URL.
Carolina at Chapel Hill, Chapel
Hill, NC, USA (J Brown PhD); Introduction hepatitis;11 conditions related to naturally occurring and
The Hubert Department of
Global Health (B A Caruso PhD)
Despite substantial progress, unsafe drinking water, synthetic chemical exposures such as arsenicosis,
and Gangarose Department of sanitation, and hygiene (WASH) services continue to fluorosis, and lead poisoning;11 and longer term
Environmental Health pose important risks to health.1,2 Billions of people do not consequences such as childhood stunting.12 Additional
(Prof T Clasen PhD, use safely managed drinking water and sanitation pathways through which unsafe WASH might negatively
Prof M C Freeman PhD), Rollins
School of Public Health, Emory
services3 and hundreds of millions do not use even basic affect health include inflammation and changes in
University, Atlanta, GA, USA; WASH services (panel). Unsafe WASH increases the the gut microbiome.13,14 There is also evidence of
Division of Epidemiology and risk of diarrhoea,4 chronic undernutrition because of compromised WASH conditions affecting educational
Biostatistics, School of Public repeated bouts of diarrhoea,5,6 acute respiratory outcomes, cognitive development, and wellbeing
Health, University of California,
Berkeley, CA, USA
infections,7 and soil-transmitted helminthiasis.8 Although including mental health, contributing to bodily injury,
(Prof J M Colford Jr MD PhD); the evidence is scarce with regard to quantifying the and resulting in physical and sexual violence, particularly
School of Civil Engineering, exposure–response relationships, unsafe WASH is among women and girls.15,16
University of Leeds, Leeds, UK associated with various other adverse outcomes. Previous WASH-attributable burden of disease
(Prof B Evans MSc); Wellcome
Examples include: trachoma;9 schistosomiasis;10 assessments have used comparative risk assessment

2060 [Link] Vol 401 June 17, 2023

 Articles

Trust Research Laboratory,

Research in context Division of Gastrointestinal
Sciences, Christian Medical
Evidence before this study infections is taken from two current meta-analyses that College, Vellore, Tami Nadu,
The disease burden attributable to unsafe drinking water, include intervention studies about WASH improvements and India (Prof G Kang PhD);
sanitation, and hygiene (WASH) has been estimated multiple these health outcomes. Instituto de Investigación
Nutricional, Lima, Peru
times by different institutions, such as WHO and the Institute
Added value of this study (Prof C F Lanata MD); School of
for Health Metrics and Evaluation (IHME). These analyses have Medicine, Vanderbilt
This study estimates the WASH-attributable burden of diarrhoea,
used different counterfactuals ranging from basic WASH University, Nashville, TN, USA
acute respiratory infections, undernutrition, and soil-transmitted (C F Lanata); Department of
services to high-quality piped water with additional
helminthiasis that can be prevented by meeting SDG targets 6.1 Epidemiology (C F Lanata) and
treatment, sewered sanitation systems or basic sanitation Department of Disease Control,
and 6.2. In contrast to previous estimates that were based on
reaching high community coverage, and handwashing with Faculty of Infectious Tropical
other minimum risk-exposure levels, this WASH-attributable
soap with assigned exposure levels having low or no assumed Disease (O Cumming MSc),
disease assessment reflects levels of service established under the London School of Hygiene &
disease risk. Previous WASH-attributable burden of disease
SDG framework. Although comprehensive data on SDG Tropical Medicine, London, UK
assessments done by WHO and IHME have estimated the
indicators are still scarce, our estimates show the additional value Correspondence to:
WASH-attributable burden for diarrhoea and acute respiratory
of collecting information on these service levels to reflect the full Dr Jennyfer Wolf, Department of
infections. WHO’s assessment also included WASH- Environment, Climate Change
burden of disease associated with unsafe WASH.
attributable undernutrition, soil-transmitted helminthiasis, and Health, World Health
trachoma, malaria, and schistosomiasis, with a total disease Implications of all the available evidence Organization, Geneva 1211,
Switzerland
burden of about 1·6 million deaths and over 100 million Although continued improvements in WASH are reducing the wolfj@[Link]
disability-adjusted life-years from unsafe WASH. When global burden of disease from diarrhoea, acute respiratory
counterfactuals are set at the higher service levels as reflected infections, undernutrition, and soil-transmitted helminthiasis,
by the Sustainable Development Goal (SDG) targets 6.1 there are important health benefits that can be obtained in
and 6.2, the attributable burden of disease is larger. However, reaching SDG targets 6.1 and 6.2. Accurate estimates will require
reliable exposure data for these higher levels of service, and continued efforts to establish the effect of WASH on other
epidemiological evidence of their effects on health outcomes, diseases and wellbeing outcomes and to estimate exposure and
remain more difficult to obtain. Evidence for the exposure– exposure–response data for the SDG WASH indicators, especially
response relationship between WASH exposures and safely managed drinking water and sanitation services.
diarrhoea, and WASH exposures and acute respiratory

methods for diarrhoea, acute respiratory infections,1,2 and pollution and contamination by 2030. Our estimates can
schistosomiasis1 and have also included other outcomes support efforts to improve use of safe WASH services and
such as malaria, undernutrition, soil-transmitted inform the recommendations of the ongoing Lancet
helminthiasis, and trachoma using other methods that Commission on water, sanitation and hygiene, and health.19
allowed more limited underlying exposure and exposure–
response data.1 Including these outcomes, WHO Methods
estimated that 1·6 million deaths in 2016 could be Study design
attributed to unsafe WASH,1 whereas the Institute of For this burden of disease assessment, unsafe WASH
Health Metrics and Evaluation (IHME) estimated that spans a range of use of drinking water, sanitation, and
1·7 million deaths in 2019 resulted from WASH- hygiene services and technologies and behaviours, which
attributable diarrhoea and acute respiratory infections.2 influence the risk for disease transmission. For grouping
In this study, we present burden of disease estimates the population into exposure categories, we used
attributable to unsafe WASH for the year 2019 for different levels of WASH services using the terms
diarrhoea, acute respiratory infections, undernutrition, unimproved, limited, basic, and safely managed as
and soil-transmitted helminthiasis. Disease burden defined by the WHO and UNICEF Joint Monitoring
assessments raise awareness about the importance of Programme for Water Supply, Sanitation and Hygiene
different risk factors, translate scientific results into (JMP) for monitoring progress against SDG For more on JMP see https://
population-level estimates of health effects, and assist in targets 6.1 and 6.2. We did not include disease burden [Link]/
setting priorities and choosing interventions with the assessment attributable to other risks such as unsafe
largest expected public-health effect.17 This analysis water-resource management or unsafe water bodies.
estimates the WASH-attributable burden of disease based For the assessment of the WASH-attributable burden of
on the UN Sustainable Development Goal (SDG) targets diarrhoea and acute respiratory infections we used the
for WASH,18 which are recognised as attainable policy results of the most up-to-date evidence4–7 on interventions
goals (panel).17 These estimates are used to track progress improving WASH access, use, and related disease outcome.
towards SDG target 3.9, which calls upon countries to We followed guidelines for accurate and transparent
substantially reduce the number of deaths and illnesses reporting (appendix 1 pp 15–16). Analyses were done with See Online for appendix 1
from hazardous chemicals and air, water, and soil Stata (version 14).

[Link] Vol 401 June 17, 2023 2061

 Articles

risk-assessment approach was not possible because of a

Panel: Drinking water, sanitation, and hygiene (WASH) service levels3 and Sustainable paucity of data on both exposure and the exposure–
Development Goal (SDG) WASH indicators response relationship; we therefore used other methods
Drinking water service level (tables 1, 2). We produced disease burden estimates for
• Safely managed: Drinking water from an improved source that is accessible on the males and females of all ages living in 183 WHO
premises, available when needed and free from faecal and priority chemical member states,20 representing 99·5% of the global
contamination population. Disease burden attributable to WASH was
• Basic: Drinking water from an improved source, provided collection time is not more estimated for 27 low-income countries (LICs), 54 lower-
than 30 min for a round trip, including queuing middle-income countries (LMICs), and 51 upper-middle-
• Limited: Drinking water from an improved source, for which collection time exceeds income countries (UMICs; appendix 1 pp 1–2).21 As most
30 min for a round trip, including queuing of the 51 high-income countries (HICs) have near
• Unimproved: Drinking water from an unprotected dug well or unprotected spring universal access to safely managed drinking water and
• Surface water: Drinking water directly from a river, dam, lake, pond, stream, canal, or sanitation services and as the available epidemiological
irrigation canal evidence from intervention studies linking drinking
Improved drinking water sources include piped water, boreholes or tubewells, protected water, sanitation, and disease outcomes originates
dug wells, protected springs, rainwater, and packaged or delivered water. mainly from LMICs,4 we do not estimate the burden of
disease attributable to unsafe water and sanitation in
Sanitation service level HICs. However, handwashing with soap and water after
• Safely managed: Use of improved facilities that are not shared with other households toilet use is not universally practiced in HICs, and we do
and where excreta are safely disposed of in situ or removed and treated offsite estimate disease burden from diarrhoea and acute
• Basic: Use of improved facilities that are not shared with other households respiratory infections due to inadequate hygiene in
• Limited: Use of improved facilities that are shared with other households HICs.
• Unimproved: Use of pit latrines without a slab or platform, hanging latrines, or bucket Contrary to the previous WASH-attributable burden of
latrines disease assessment done by WHO,1 we did not include
• Open defecation: Disposal of human faeces in fields, forests, bushes, open bodies of malaria as a health outcome, because the attribution was
water, beaches or other open places, or with solid waste made to unsafe water resource management, rather than
Improved sanitation facilities include flush or pour flush toilets connected to piped sewer unsafe WASH. We also did not include schistosomiasis
systems, septic tanks, or pit latrines; pit latrines with slabs (including ventilated pit and trachoma, because the available exposure–response
latrines); and composting toilets. data relate to modest WASH improvements such as
Hygiene service level basic WASH services.10 Previous burden of disease
• Basic: Availability of a handwashing facility with soap and water at home assessments22 have estimated population attributable
• Limited: Availability of a handwashing facility lacking soap or water at home fractions (PAFs) for other vector-borne diseases such as
• No facility: No handwashing facility at home dengue and lymphatic filariasis through structured
Handwashing facilities might be located within the dwelling, yard, or plot. They can be fixed expert surveys, a method which has been shown to have
or mobile and include a sink with tap water, buckets with taps, tippy-taps, and jugs or basins numerous limitations.23
designated for handwashing. Soap includes bar soap, liquid soap, powder detergent, Since these estimates of the WASH-attributable burden
and soapy water but does not include ash, soil, sand, or other handwashing agents. of disease are used to track SDG 3.9, a consultation with
national authorities was run by WHO from April 5, 2022,
SDG WASH targets and indicators to June 30, 2022, to seek feedback on estimates before
6.1: By 2030, achieve universal and equitable access to safe and affordable drinking water finalisation.
for all
6.1.1: Proportion of population using safely managed drinking water services Comparative risk assessment
6.2: By 2030, achieve access to adequate and equitable sanitation and hygiene for all and The standard approach for estimating the burden of
end open defecation, paying special attention to the needs of women and girls and disease attributable to a given risk factor is comparative
those in vulnerable situations risk assessment (figure 1).24,25 Comparative risk assess­
6.2.1: proportion of population using (A) safely managed sanitation services and (B) a ments systematically evaluate changes in population
hand-washing facility with soap and water health as a consequence of changing the distribution of
a risk factor’s exposure in the population.26 The approach
requires the distribution of population’s exposure (p) to
Disease outcomes with a strong epidemiological link the relevant risk factor levels (j) and the exposure–
to unsafe WASH were included if they had sufficient response relationships (relative risk [RR]) between
data available to allow the quantification of the WASH- different exposure levels (pj) and the health outcome,
attributable burden preventable by improving WASH usually based on a pooled analysis of high quality
services. The WASH-attributable burden of diarrhoea interventions (appendix 1 p 10).27 A counterfactual
and acute respiratory infections was estimated using minimum exposure level—corresponding to the
comparative risk assessment. For undernutrition and removal or reduction of exposure—is specified and
soil-transmitted helminthiasis a standard comparative- compared with the current distribution of the risk

2062 [Link] Vol 401 June 17, 2023

 Articles

factor’s exposure in the population of interest. For this

Prevalence of Association
analysis, we used counterfactual minimum risk- WASH between WASH
exposure levels that aligned as closely as possible with minimum risk counterfactual and
the SDG indicators (panel, figures 2–4), to the extent that exposure outcome (against
counterfactual lowest level of
available exposure and exposure–response data in 2019* exposure)
permitted.
Diarrhoea
Safely managed drinking water 37·9% 0·48 (0·26–0·87),
Population exposed (pj) (27·1–49·9) p=0·0174
To estimate the population exposure to different WASH
Basic sanitation connected to 29·7% 0·53 (0·30–0·93),
service levels (figure 1), we used the public database of sewer (23·9–36·1) p=0·0304
the JMP, which draws upon nationally representative Handwashing with soap after 26·4% 0·7 (0·64–0·76),
surveys, censuses, and administrative data to produce potential faecal contact (23·4–29·6) p<0·00014
national, regional, and global estimates of WASH service Acute respiratory infections
use. The JMP extracts data from national data sources Handwashing with soap after 26·4% 0·83 (0·76–0·90),
and matches them to global indicators and definitions, to potential faecal contact (23·4–29·6) p<0·00017
the extent possible, to ensure comparability and Data are prevalence (95% CI) or relative risk (95% CI), p value. WASH=drinking
consistency. These harmonised extracts are used as water, sanitation, and hygiene. *Aggregated across included countries.
inputs to a regression model to produce draft estimates, Table 1: Counterfactual and outcome association for diarrhoea and
which undergo a country consultation as an additional acute respiratory infections
quality-control measure before finalisation.
As some countries did not have sufficient national
data, we applied multilevel modelling to estimate the WASH counterfactual Methods Limitations
use of different WASH services and household water exposure
treatment for 2019 using a two-level random slope Undernutrition As for diarrhoea in table 1 Combining the PAF of Considers only one of
model for use of different drinking water and sanitation (because based on WASH malnutrition attributable to multiple potential pathways
services and a two-level random intercept model for use diarrhoea PAFs) diarrhoea with the PAF of linking unsafe WASH and
diarrhoea attributable to undernutrition and therefore
of household water treatment and access to basic unsafe WASH might represent only a
handwashing facilities. For countries without fraction of WASH-
datapoints, we used the regional mean estimate of the attributable undernutrition
model. Further details on input data and the modelling Soil-transmitted Safely managed drinking Complete attribution of Assumes that all soil-
helminthiasis* water, safely managed overall disease burden transmitted helminthiasis
approach have previously been published (appendix 1 sanitation, and estimates could be prevented through
pp 2–3).28 handwashing with soap safe WASH
We adjusted estimates of safely managed drinking WASH=drinking water, sanitation, and hygiene. PAF=population attributable fraction. *Ascaris lumbricoides, Trichuris
water services to incorporate nationally representative trichiura, and hookworms.
survey data on the proportion of households meeting
Table 2: Counterfactual and methods for undernutrition and soil-transmitted helminthiasis
all three criteria for safely managed drinking water
(quality, availability, and accessibility) at the household
level.29 This adjustment was made because JMP Estimates for access to basic handwashing facilities at
estimates of safely managed drinking water assess home do not necessarily reflect the actual practice of
microbiological water quality at the point of collection, handwashing with soap, since many people do not wash
not at the point of consumption, and therefore do not hands after toilet use even if soap and water are available.32
account for the possibility of contamination after We therefore adjusted these estimates based on the
collection;30 and, due to the paucity of data, these results of a meta-analysis of the association between
estimates are based on the minimum of quality, presence of a handwashing station with soap and water
availability, and accessibility at urban and rural levels and observed handwashing practices.33
for each country, rather than the proportion of the
population meeting all three criteria at the household Relative risks linking exposure and health outcome (RRj)
level (appendix 1 pp 4–5). The exposure–response relationships (figure 1, appendix 1
For sanitation, we would ideally have chosen a p 6) linking the different WASH exposure levels
counterfactual of safely managed sanitation; however, (figures 2–4) and diarrhoea or acute respiratory infections
this was not possible due to the current paucity of data have been estimated in two meta-analyses.4,7 For drinking
on both exposure and exposure–response. Instead, water, the highest diarrhoea-risk reduction relates to
using the latest epidemiological evidence,4 we used a improved drinking water on premises with higher water
counterfactual of basic sanitation connected to sewer quality, for sanitation to basic sanitation connected to
networks (figure 3), which is not necessarily equivalent sewer, and for hygiene to handwashing promotion usually
to safely managed sanitation as much of the wastewater with the provision of hygiene infrastructure such as
collected in sewers is not safely treated.31 handwashing stations with water and soap (tables 1, 2).4

[Link] Vol 401 June 17, 2023 2063

 Articles

Required data inputs

Population stratified in exposure levels (pj) Estimated in this work
n
Σ j=1 pj (RRjc – 1)
PAF= n
Σ j=1 pj (RRj – 1)+1

Relative risk at exposure level j (RRj) linking exposure Attributable disease burden estimates (deaths and DALYs)
level (pj) and outcome by outcome(country, age, sex)=PAF × total burden(country, age, sex)

WHO global health estimates (deaths and

DALYs) by outcome and country, age, or sex
(total burden(country, age, sex))

Figure 1: General approach for calculating the PAF and the attributable disease burden in comparative risk assessment
Sources for data inputs are listed in appendix 1 (p 6). DALY=disability-adjusted life-year. PAF=population attributable fraction. n=total number of exposure levels.

Safe drinking water* consistently available

at all times

Safely managed drinking water (SDG 6.1)

Free of contamination + On premises + Available when needed

Point-of-use water treatment

(as a proxy for higher water
quality at point of use) 52% estimated diarrhoea risk
reduction between the
highest risk exposure group
and counterfactual
Basic drinking water

Surface water, unimproved or

limited drinking water

Figure 2: Conceptual model for disease burden assessment attributable to unsafe drinking water and associated reduction in risk of diarrhoea4
Blue box with solid line: counterfactual minimum risk exposure used in the analysis with associated percentage reduction in risk of diarrhoea compared with the
highest risk-exposure group from a meta-analysis.4 Blue box with dashed line: additional plausible risk reduction, which cannot currently be estimated due to a
paucity of exposure or exposure–response data. SDG=Sustainable Development Goal. *Safe drinking water does not represent any substantial risk to health over a
lifetime of consumption. The position of the boxes is not directly proportional to the expected health effect.

Safe sanitation (safe collection, storage,

treatment, disposal, or use of human
and animal excreta with full coverage
at the community level)

Safely managed sanitation (SDG 6.2)

Sewers Onsite systems
Wastewater is treated or Excreta emptied and or Excreta treated and
offsite treated offsite disposed in situ

Basic sanitation connected to sewer

47% estimated
Basic sanitation not diarrhoea risk
connected to sewer reduction between
the highest risk
exposure group and
Open defecation or counterfactual
unimproved or limited
sanitation

Figure 3: Conceptual model for disease burden assessment attributable to unsafe sanitation and associated reduction in risk of diarrhoea4
Blue box with solid line: counterfactual minimum risk exposure used in the analysis with associated percentage reduction in risk of diarrhoea compared with the
highest risk-exposure group from Wolf and colleagues’4 meta-analysis. Blue boxes with dashed lines: additional plausible risk reduction, which cannot currently be
estimated due to a paucity of exposure or exposure–response data. The position of the boxes is not directly proportional to the expected health effect.
SDG=Sustainable Development Goal.

2064 [Link] Vol 401 June 17, 2023

 Articles

via surfaces, and handwashing with soap can remove

or destroy pathogens on hands, thereby reducing
Essential hygiene conditions and
practices including food hygiene transmission.7 The burden of diarrhoea attributable to
unsafe WASH and the burden of acute respiratory
infections attributable to unsafe hygiene are estimated
Handwashing with soap after
potential faecal contact
using comparative risk assessment (figure 1). We used
the following standard formula
30% estimated diarrhoea risk
reduction and 17% acute R
respiratory infection risk reduction PAFWASH=1 – ∏ (1 – PAFr)
between the highest risk exposure r=1
group and counterfactual

No handwashing with soap after to estimate the burden of diarrhoea attributable to the
potential faecal contact risk factor cluster of unsafe WASH combined:40 where r
is the individual risk factor (unsafe drinking water,
Figure 4: Conceptual model for disease burden assessment attributable to sanitation, or hygiene), and R is the total number of
unsafe hygiene and associated reductions in risk of diarrhoea and acute risk factors (three) accounted for in the cluster.
respiratory infections4,7
Blue box with solid line: counterfactual minimum risk exposure used in the
analysis with associated percentage reductions in risk of diarrhoea4 and risk of Undernutrition
acute respiratory infections7 compared with no handwashing. Blue box with For calculating the WASH-attributable burden of
dashed line: additional plausible risk reductions that cannot currently be undernutrition, we did not use a comparative risk
estimated due to a paucity of exposure or exposure–response data. The position
of the boxes is not directly proportional to the expected health effect.
assessment as the available exposure–response relation­
ships from systematic reviews and meta-analyses relate
to modest WASH improvements such as household
PAF and attributable burden estimates water treatment, basic drinking water, sanitation, or
The PAF is the proportion of total morbidity or mortality hygiene provision, and hygiene education alone or in
due to a condition or disease that could have been combination12,41–43 and do not reach full safety. We
prevented by reducing the risk factor to a counterfactual followed a previously published approach (appendix 1
defined by the minimum risk-exposure level.27 The PAF is p 5).1 Undernutrition can be a consequence of repeated
estimated through combining population exposure and bouts of diarrhoea.5,13 We multiplied country-level PAFs
corresponding relative risks (figure 1). To estimate the of diarrhoea attributable to unsafe WASH as estimated
risk factor-attributable burden, the PAF is multiplied with in this study with previously published PAFs of protein-
the total disease burden. We use WHO global health energy malnutrition attributable to diarrhoea (appendix 2
estimates (disability-adjusted life-years [DALYs] and p 1).6 In the Global Burden of Diseases, Injuries, and
deaths) for the included health outcomes by country, sex, Risk Factors Study (GBD),6 protein–energy malnutrition
and age groups in 2019 (appendix 2 pp 3–6).34,35 Tables is exclusively considered an outcome of childhood See Online for appendix 2
with included International Classification of Diseases-10 underweight and wasting. Troeger and col­ leagues6
codes are listed in appendix 1 (pp 6–8). Further estimated the attributable fractions for protein–energy
information on the preparation of WHO Global Health malnutrition attributable to under­weight and wasting
Estimates is provided in two technical reports.36,37 independently from a counterfactual model that
Disease burden can be caused by different risks. quantified the expected shift in the distribution of
Although the PAF for each risk factor, such as unsafe weight-for-age and weight-for-height in the absence of
WASH, is a proportion and is bounded by 0 and 1, the diarrhoea compared with the observed distributions.44
sum of the PAFs from the individual relevant exposures We multiplied the resulting PAF of undernutrition
can exceed 1.38 For example, attributing 100% of soil- attributable to unsafe WASH with WHO total disease
transmitted helminthiasis to unsafe WASH does not burden figures for protein energy malnutrition for
preclude some proportion of this WASH-attributable children younger than 5 years (appendix 1 p 5).
disease burden being eliminated through other
interventions, such as deworming medication. Soil-transmitted helminthiasis
Estimation of uncertainty intervals at the country, This assessment includes infections with the major
regional, and global levels was done with Monte Carlo soil-transmitted helminths that infect humans:
simulation (appendix 1 p 5). Ascaris lumbricoides, Trichuris trichiura, and hookworms
(Necator americanus and Ancylostoma duodenale).
Diarrhoea and acute respiratory infections Soil-transmitted helminths are transmitted by eggs
Diarrhoeagenic pathogens are transmitted via previously present in human faeces of an infected individual, which
described environmental routes,39 which can be then enter soil in the absence of safe sanitation. Infection
interrupted with safe WASH. Respiratory pathogens are occurs through ingestion of eggs attached to vegetables,
transmitted through the air, person-to-person contact, or in contaminated water sources or soil, or—by

[Link] Vol 401 June 17, 2023 2065

 Articles

in 2019. Additionally, 112 000 (92 000–134 000) children

PAF (95% CI) Deaths (95% CI) DALYs (95% CI)
younger than 5 years died from acute respiratory infections
Diarrhoea 0·69 1 035 000 54 590 000 attributable to unsafe hygiene in 2019 (appendix 2 pp 2–4).
(0·65–0·72) (929 000–1 160 000) (50 033 000–59 562 000)
We estimate that 69% of diarrhoea, 14% of acute
Acute respiratory infections 0·14 356 000 16 578 000
(0·13–0·17) (320 000–405 000) (14 257 000–19 481 000) respiratory infections, and 10% of undernutrition, and
Undernutrition 0·10 8000 825 000 assume 100% of the burden of soil-transmitted
(0·09–0·10) (7000–9000) (755 000–905 000) helminthiasis, could be have been prevented with safe
Soil-transmitted helminthiasis* 1·0† 2000 1 942 000 WASH in 2019 (table 3). There are considerable differences
(2000–3000) (1 862 000–2 028 000) in the WASH-attributable disease burden between income
WASH=drinking water, sanitation, and hygiene. PAF=population attributable fraction. DALY=disability-adjusted groups: there were 270 000 deaths in LICs, 975 000 deaths
life-year. *Ascaris lumbricoides, Trichuris trichiura, and hookworms. †Assumed value. in LMICs, and 112 000 deaths in UMICs, compared with
Table 3: WASH-attributable disease burden by health outcome, 2019
44 000 deaths in HICs (though deaths from unsafe water
and sanitation were not estimated for HICs). There
is also substantial regional variation ranging from
hookworms—active penetration of skin by larvae. There 510 000 WASH-attributable deaths in WHO’s Africa region
is no direct person-to-person transmission or infection and 593 000 deaths in WHO’s South-East Asia region to
from fresh faeces as the excreted eggs require 5–10 days 33 000 deaths in WHO’s European region. Furthermore,
to mature in the soil before becoming infective. we found substantial variation in the WASH-attributable
A lumbricoides, T trichiura, and hookworms do not fraction of the different diseases. For example,
multiply in the human host and re-infection only occurs 18% of the diarrhoea-related disease burden in HICs
as a result of contact with infective stages in the could be prevented through safe WASH compared with
environment.45 76% in LMICs in WHO’s Africa region and
Based on these requirements for transmission, we 66% in LMICs in WHO’s South-East Asia region
assumed that transmission would be interrupted if (appendix 1 p 12).
everyone used safely managed WASH services and
practiced handwashing with soap after potential faecal Discussion
contact. We estimate that 1·4 million deaths and 74 million DALYs
Sources of exposure, exposure–response, and overall could have been prevented through the universal
disease burden data are listed in appendix 1 (p 6). provision of safe WASH in 2019, accounting for
2·5% of all deaths and 2·9% of all DALYs in the global
Role of the funding source population and 7·6% of all deaths and 7·5% of all DALYs
The funders had no role in study design, data collection, in children younger than 5 years. Diarrhoea accounts for
data analysis, data interpretation, writing of the the majority of the WASH-attributable burden with over
manuscript, or the decision to submit the manuscript for 1 million deaths, about 55 million DALYs, and a
publication. preventable fraction of 69%, followed by acute respiratory
infections attributable to unsafe hand hygiene with about
Results 356 000 deaths, 17 million DALYs, and a preventable
After adjusting the estimates for safely managed drinking fraction of 14%.
water, 37·9% (95% CI 29·1–49·9) of the population in Our estimates are lower than those from GBD 2019,2
LMICs used safely managed drinking water (calculated which estimated 1·7 million deaths and 88 million DALYs.
at the household level), 29·7% (23·9–36·1) used basic Differences arise from the expanded attributable health
sanitation connected to sewer networks, and 26·4% outcomes from more up-to-date epidemiological evidence
(23·4–29·6) of the global population washed hands with and, especially, the adoption of different counterfactual
soap after potential faecal contact. Estimates of WASH minimum risk-exposure levels. GBD 20192 used high-
exposure levels and matching exposure–response quality piped water that is boiled or filtered at point of use
relationships are shown in tables 1 and 2 and regional as the minimum risk-exposure level for drinking water. To
aggregates are in appendix 1 (pp 8–10). Country-level estimate this exposure level, piped water was divided into
exposure estimates derived through multilevel modelling basic-quality and high-quality piped water based on the
for this analysis and official exposure estimates from the results of a systematic review done in 2013, which included
global SDG database maintained by the JMP are available a mixture of private and community piped drinking-water
in appendix 2 (pp 7–9). services and predominantly measured drinking water
The WASH-attributable disease burden combined quality at one point in time.46 On the exposure–response
across the four outcomes amounts to 1 401 000 deaths side, GBD 2019 used a pooled RR of 0·09 (corresponding
(95% CI 1 283 000–1 542 000) and 73 935 000 DALYs to a 91% reduction in the risk of diarrhoea for drinking
(68 248 000–80 186 000) in 2019. An estimated 273 000 water improvements alone) by multiplying the effect of
(252 000–296 000) deaths from diarrhoea among children filtering or boiling at the household level with the effect of
younger than 5 years were attributable to unsafe WASH providing high-quality piped water.47 The epidemiological

2066 [Link] Vol 401 June 17, 2023

 Articles

studies providing exposure–response relationships for onchocerciasis, and lymphatic filariasis for which the
point-of-use interventions are usually done in settings risk factor was unsafe water resource management,
with a high level of unimproved drinking water sources, which concerns very different interventions and
which calls into question the approach of multiplying the strategies as compared with safe WASH service
effect of point-of-use interventions with the effect of provision. Schistosomiasis and trachoma, the
providing high-quality piped water. Additionally, point-of- two outcomes related to unsafe WASH, add only about
use interventions often require intensive supervision and 5000 additional deaths and 900 000 DALYs—a very small
follow-up to sustain water treatment, which could lead to proportion of the WASH-attributable disease burden
incorrectly estimated health effects from reporting bias. A estimated in this work. As for both trachoma and
comparison of WHO and IHME WASH-attributable schistosomiasis we were not able to establish a
disease burden and minimum risk-exposure levels over counterfactual of higher level WASH service provision
time is included in appendix 1 (p 13). (ie, extending beyond basic WASH services) so we did
This analysis expands upon previous minimum risk- not include their WASH-attributable burden of disease
exposure levels by using the SDG ambition of universal assessment in this analysis.
access to safely managed WASH services as minimum However, unsafe WASH affects health in many more
risk-exposure levels. We consider these levels more ways: a high burden of paediatric asymptomatic carriage
attainable than a theoretical zero-risk counterfactual of certain bacteria and parasites due to unsafe WASH is
because they reflect the ambition set out in the strongly associated with adverse consequences, including
internationally agreed SDGs and have already been growth faltering,48 and new estimates for the burden of
achieved by numerous countries across different regions disease attributable to antimicrobial resistance highlight
of the world. In addition, nationally representative the importance of community-based WASH to prevent
exposure data on safely managed services is available for these infections.49 WASH can affect other non-infectious-
an increasing number of countries aggregated and disease health outcomes—including physical and sexual
published by the JMP and is combined with matching trauma from violence, bodily injury such as from water
exposure–response relationships available through the carriage, mental health, and general wellbeing—
up-to-date meta-analyses.4,7 Our results, therefore, offer a particularly among women and girls.15,50,51 More evidence
policy-relevant complement to previous analyses and is needed to understand these additional outcomes and
might influence the choice of future counterfactual to generate sufficient data to enable their inclusion in
minimum risk exposures in evaluations of health effects future analyses.
of other risk factors included under the SDGs. In addition, our analysis is confined to domestic
Our analysis has important limitations. Due to settings, so does not include the disease burden
insufficient data, we excluded multiple potentially attributable to WASH in other settings such as schools,
important health outcomes, which makes it probable health-care facilities, workplaces, detention centres,
that we underestimated the true burden. We restricted refugee camps, markets, and other public spaces.
our estimation of disease burden attributable to WASH There are further limitations relating to the methods
to the four health outcomes for which there was sufficient and data for the outcomes that we did include. For
information to estimate the exposure–response diarrhoea and acute respiratory infections we were able
relationship (ie, intervention studies that estimated the to use a standard comparative risk assessment approach
health effect for higher levels of WASH services), PAFs based on systematic reviews that include a large number
that could be taken from the published literature, or of rigorous studies. However, most of the underlying
certain requirements in the disease-transmission studies are open-label trials with subjective health
pathway that allowed the assumption of attributing the outcomes (eg, self-reported diarrhoea or respiratory
total disease burden of a certain disease—here, soil- symptoms), so there is a risk of overestimating their
transmitted helminthiasis—to unsafe WASH. A list of effectiveness.52 In addition, many interventions were
other adverse health outcomes linked to unsafe WASH tightly controlled studies and the transportability of their
or related exposures, including chemical exposures, is results to population-level changes in WASH conditions
included in appendix 1 (pp 14–15). We estimated on an remains uncertain.
exploratory basis the additional attributable disease For undernutrition and soil-transmitted helminthiasis,
burden had we considered other outcomes for which a we judged there to be insufficient data for a comparative
PAF has been estimated in previous analyses, including risk assessment approach and adopted alternative
trachoma, schistosomiasis, malaria, lymphatic filariasis, methods. For soil-transmitted helminthiasis, we attributed
onchocerciasis, and dengue (appendix 1 p 14). We the complete disease burden to unsafe WASH based on
multiplied these PAFs with overall disease burden knowledge about the disease transmission pathway. For
estimates by outcome and for the year 2019. From this undernutrition, we used published PAFs of protein–
result, we estimated an additional 360 000 attributable energy malnutrition from diarrhoea and combined them
deaths and 31 million DALYs. Of these, 99% of deaths with the PAFs of diarrhoea from unsafe WASH. This
and 97% of DALYs were from malaria, dengue, two-step approach quantified only the WASH-attributable

[Link] Vol 401 June 17, 2023 2067

 Articles

burden of protein–energy malnutrition that is a handwashing facilities.33 These adjustments introduce

consequence of diarrhoea. There might be multiple uncertainty and do not adequately account for other
additional mechanisms by which WASH could contribute important aspects of safely managed services, for
to undernutrition, such as via parasitic infections,53 the example that drinking water is free from priority
asymptomatic carriage of certain enteric pathogens,48 and chemicals, such as arsenic and fluoride. The minimum
environmental enteric dysfunction.12,54 risk-exposure level for sanitation is basic sanitation
Our results reflect the selected counterfactuals and connected to sewer networks. This selection is based on a
assume no residual attributable disease burden where 2022 systematic review and meta-analysis4 that showed a
the complete population has met the SDG targets. lower RR compared with basic sanitation without
However, there remains a considerable WASH sewered connections. Basic santitation connected to
preventable fraction of disease in countries with universal sewer networks do not assess all aspects of safely
or near universal access to safely managed water,55 managed sanitation, such as the extent of wastewater
meaning that further improvements would yield further treatment in centralised plants,31 and excludes other
gains. Additionally, there remain many millions of forms of safely managed sanitation (eg, onsite
people across HICs who do not have safe WASH but technologies where waste is emptied and treated off-site
who are not adequately captured in official data.56 Almost or safely disposed in situ). For both sanitation and water,
a million people living in cities in the USA have been there remains a paucity of rigorous epidemiological
estimated to be without access to basic sanitation57 and research to quantify the health effects of different types
nearly half a million households have been estimated to of safely managed WASH services. Due to the scarcity of
not have complete piped water services.58 Drinking water the available data, there is also a risk of misclassification
and sanitation-attributable diarrhoea burden in HICs or absence of data for safely managed water and
might add several million additional DALYs to our sanitation services that will hopefully reduce over time as
estimates, but due to the paucity of data—especially on countries adapt national monitoring systems for the
the exposure–response side—we do not include them in higher service levels called for in the SDG framework.
current estimates. In populations that have not achieved We considered the regional distribution of the WASH-
SDG targets for safely managed WASH services, our attributable disease burden, by the income level of
results further motivate complementary interventions countries, and by sex and age. Our findings of substantial
that reduce the WASH-attributable disease burden, such variation by income status are supported by previous
as anthelmintics to treat soil-transmitted helminthiasis,59 analyses1,64 that found that the WASH attributable disease
and small-quantity nutritional supplementation to burden is highest in LMICs, and that the proportion of
complement breast­feeding.60 high-burden diseases, such as diarrhoea and acute
Our results also reflect the preventable fraction of respiratory infections, in these countries is much higher.
disease attributable to WASH systems functioning under Due to the paucity of data, we assume the same exposure
current climate conditions. Global warming might alter and exposure–response relationship and therefore the
the incidence or severity of several WASH preventable same PAFs between WASH and disease outcomes for
infectious diseases.61 Drought, flooding, and changes in males and females, despite extensive research
water quality can interfere with water supply systems, documenting that women and men have vastly different
resulting in intermittency and systems breakdown.62 Both WASH experiences. Though we present WASH-
higher intensity of rainfall and drought can reduce the attributable burden of disease estimates for males and
operational effectiveness of sanitation systems leading to females at the country level, these vary only due to
localised failure.63 Both effects could reduce the total differences in the overall disease burden figures. Women
prevented disease burden for populations with safe and girls might have increased exposures based on
WASH. gendered expectations that they collect and treat water,
Our classification of exposure to different WASH manage dependents’ faeces, and care for ill children.65–67
service levels relies on nationally representative survey Furthermore, all genders simultaneously have multiple
data and administrative data, which might not accurately intersecting identities (eg, race, socioeconomic status, and
capture all aspects of safely managed services and have ability) that can influence exposure and exposure–
considerable gaps for higher levels of service, including response relationships related to WASH.51 The same
for some populous countries. To estimate exposure to issues relate to differential disease burden in different age
safely managed drinking water, we adjusted estimates groups.
using nationally representative survey data to account for For estimates that are closer to the true attributable
the probability of water not contaminated at source being burden of disease, better data are needed to more
contaminated at the point of consumption and to assess accurately characterise the population exposure to safely
the proportion of households meeting all three criteria managed service levels and to quantify the exposure–risk
(ie, accessibility, availability, and quality). For hygiene, we relationship between WASH services, particularly high
account for the probability of handwashing with soap service levels including all aspects of safely managed
occurring after faecal contact conditional on presence of WASH and different service types, and a broader range

2068 [Link] Vol 401 June 17, 2023

 Articles

of health outcomes. Future research should be done in purpose and is either published as supplementary material or can be
all, including high-income, settings. accessed through the corresponding author.
Further research is also needed to better understand Acknowledgments
differences in exposure and exposure–response We acknowledge the fundamental work of the initial expert group
(appendix 1 p 17). This work was funded by WHO and the UK Foreign,
relationships between different sex and age groups to Commonwealth & Development Office. RBJ, JW, KOM, JEM, BG, AP-U,
identify disparities in access to safe WASH services that and SB are staff members or consultants of WHO. RB is a staff member
are responsible for persistent inequalities. of UNICEF. MB employed by the Institute for Health Metrics and
As new data become available, future estimates might Evaluation. The authors alone are responsible for the views expressed in
this publication, which do not necessarily represent the views, decisions,
consider lower minimum risk-exposure levels, even or policies of the institutions with which they are affiliated.
extending safely managed WASH services by including,
References
for example, continuously available drinking water free 1 Prüss-Ustün A, Wolf J, Bartram J, et al. Burden of disease from
from microbial and chemical contamination, safe inadequate water, sanitation and hygiene for selected adverse
sanitation systems that extend to animal waste, and health outcomes: an updated analysis with a focus on low- and
middle-income countries. Int J Hyg Environ Health 2019;
hygiene practices that extend beyond handwashing after 222: 765–77.
faecal contact to including food hygiene, ideally in a 2 GBD 2019 Risk Factors Collaborators. Global burden of 87 risk
combined safe WASH counterfactual exposure scenario.68 factors in 204 countries and territories, 1990–2019: a systematic
analysis for the Global Burden of Disease Study 2019. Lancet 2020;
Combining up-to-date epidemiological evidence with 396: 1223–49.
exposures based as closely as possible on the levels of 3 WHO, UNICEF. Progress on household drinking water, sanitation
service established under the SDG framework, we and hygiene 2000–2020: five years into the SDGs. 2021. https://
[Link]/report/jmp-2021-wash-households (accessed
estimate that meeting the WASH targets of SDG 6 could July 5, 2021).
prevent at least 1·4 million deaths and 74 million DALYs 4 Wolf J, Hubbard S, Brauer M, et al. Effectiveness of interventions to
annually. Although comprehensive data on SDG improve drinking water, sanitation, and handwashing with soap on
risk of diarrhoeal disease in children in low-income and middle-
indicators are still scarce, our estimates show the income settings: a systematic review and meta-analysis. Lancet
additional value of collecting information on these service 2022; 400: 48–59.
levels, particularly in low-resource settings, to accurately 5 Checkley W, Buckley G, Gilman RH, et al. Multi-country analysis of
the effects of diarrhoea on childhood stunting. Int J Epidemiol 2008;
reflect the full burden of disease associated with WASH. 37: 816–30.
As the UN report the need to accelerate progress towards 6 Troeger C, Colombara DV, Rao PC, et al. Global disability-adjusted
the goal of universal access to safe WASH, our study life-year estimates of long-term health burden and undernutrition
attributable to diarrhoeal diseases in children younger than 5 years.
shows the substantial health gains that can be obtained by Lancet Glob Health 2018; 6: e255–69.
redoubling efforts to achieve the SDG WASH targets. 7 Ross I, Bick S, Ayieko P, et al. Effectiveness of handwashing with
Contributors soap for preventing acute respiratory infections in low- and middle-
JW, RBJ, RB, MB, TC, JMC, JEM, MCF, BG, AP-U, SB, and OC income countries—a systematic review and meta-analysis. Lancet
2023; 401: 1681–90.
conceptualised the paper. JW, RB, CT, and SB accessed and verified the
8 Garn JV, Wilkers JL, Meehan AA, et al. Interventions to improve
data. JW and RB analysed the data. JW, RB, MB, CT, SB, and OC
water, sanitation, and hygiene for preventing soil-transmitted
interpreted the data. JW wrote the original draft. All authors reviewed
helminth infection. Cochrane Database Syst Rev 2022;
and edited the draft. OC supervised the study. All authors had full access 6: CD012199.
to all the data in the study and had the final responsibility for the
9 Freeman MC, Garn JV, Sclar GD, et al. The impact of sanitation on
decision to submit for publication. infectious disease and nutritional status: a systematic review and
Declaration of interests meta-analysis. Int J Hyg Environ Health 2017; 220: 928–49.
JW reports grants from UK Foreign, Commonwealth & Development 10 Grimes JE, Croll D, Harrison WE, Utzinger J, Freeman MC,
Office (FCDO), during this study. MB reports grants from Bill & Templeton MR. The relationship between water, sanitation and
Melinda Gates Foundation during this study. CFL reports non-financial schistosomiasis: a systematic review and meta-analysis.
PLoS Negl Trop Dis 2014; 8: e3296.
support from WHO COVID-19 Vaccine Effectiveness Working Group,
grants from CureVac AG, grants from PATH, grants from HilleVax, 11 WHO. Safer water, better health. Geneva: World Health
Organization, 2019.
and personal fees from Valneva outside the submitted work. JB reports
grants from Bill & Melinda Gates Foundation, grants from US Centers 12 Dangour AD, Watson L, Cumming O, et al. Interventions to
improve water quality and supply, sanitation and hygiene practices,
for Disease Control & Prevention, and grants from Columbia World
and their effects on the nutritional status of children.
Projects during this study. MCF reports Consulting for Rickett’s Global Cochrane Database of Systematic Reviews 2013; 8: CD009382.
Scientific Advisory Committee. RBJ reports grants from UK FCDO
13 Guerrant RL, DeBoer MD, Moore SR, Scharf RJ, Lima AAM.
during this study, grants from Agence française de développement, The impoverished gut—a triple burden of diarrhoea, stunting
grants from Australian Government Department of Foreign Affairs and and chronic disease. Nat Rev Gastroenterol Hepatol 2013;
Trade, grants from Austrian Development Agency, grants from Bill & 10: 220–29.
Melinda Gates Foundation, grants from German Federal Ministry for 14 Ginn O, Rocha-Melogno L, Bivins A, et al. Detection and
Economic Development and Cooperation, grants from The Netherlands quantification of enteric pathogens in aerosols near open
Directorate-General for International Cooperation, grants from Swiss wastewater canals in cities with poor sanitation. Environ Sci Technol
Agency for Development and Cooperation, and grants from United 2021; 55: 14758–71.
States Agency for International Development outside the submitted 15 Caruso BA, Conrad A, Patrick M, et al. Water, sanitation, and
work. All other authors declare no competing interests. women’s empowerment: a systematic review and qualitative
metasynthesis. PLoS Water 2022; 1: e0000026.
Data sharing
16 Sclar GD, Garn JV, Penakalapati G, et al. Effects of sanitation on
All data collected for this analysis including exposure, exposure–
cognitive development and school absence: a systematic review.
response data, overall disease statistics, and analytic code are available Int J Hyg Environ Health 2017; 220: 917–27.
immediately following publication without end date to anyone for any

[Link] Vol 401 June 17, 2023 2069

 Articles

17 Murray CJ, Ezzati M, Lopez AD, Rodgers A, Vander Hoorn S. 39 Wagner EG, Lanoix JN. Excreta disposal for rural areas and small
Comparative quantification of health risks conceptual framework communities. Geneva: World Health Organization, 1958.
and methodological issues. Popul Health Metr 2003; 1: 1. 40 Lim SS, Vos T, Flaxman AD, et al. A comparative risk assessment of
18 Greenland S. Causality theory for policy uses of epidemiological burden of disease and injury attributable to 67 risk factors and risk
measures. In: Murray CJL, Salomon JA, Mathers CD, et al, eds. factor clusters in 21 regions, 1990–2010: a systematic analysis for
Summary measures of population health: concepts, ethics, the Global Burden of Disease Study 2010. Lancet 2012; 380: 2224–60.
measurement, and application. Geneva: World Health 41 Gizaw Z, Worku A. Effects of single and combined water, sanitation
Organization, 2002: 291–302. and hygiene (WASH) interventions on nutritional status of children:
19 Commissioners of the Lancet Commission on Water, Sanitation and a systematic review and meta-analysis. Ital J Pediatr 2019; 45: 77.
Hygiene, and Health. The Lancet Commission on water, sanitation 42 Bekele T, Rawstorne P, Rahman B. Effect of water, sanitation and
and hygiene, and health. Lancet 2021; 398: 1469–70. hygiene interventions alone and combined with nutrition on child
20 WHO. Countries. [Link] (accessed growth in low and middle income countries: a systematic review
May 24, 2022). and meta-analysis. BMJ Open 2020; 10: e034812.
21 World Bank. World bank country and lending groups. 2021. https:// 43 Gera T, Shah D, Sachdev HS. Impact of water, sanitation and
[Link]/knowledgebase/articles/906519-world- hygiene interventions on growth, non-diarrheal morbidity and
bank-country-and-lending-groups (accessed May 5, 2022). mortality in children residing in low- and middle-income countries:
22 Prüss-Ustün A, Wolf J, Corvalán C, Bos R, Neira M. Preventing a systematic review. Indian Pediatr 2018; 55: 381–93.
disease through healthy environments: a global assessment of the 44 Troeger C, Colombara DV, Rao PC, et al. Supplementary appendix
environmental burden of disease from environmental risks. to: global disability-adjusted life-year estimates of long-term
Geneva: World Health Organization, 2016. health burden and undernutrition attributable to diarrhoeal
23 Guyatt GH, Oxman AD, Vist GE, et al. GRADE: an emerging diseases in children younger than 5 years. Lancet Glob Health
consensus on rating quality of evidence and strength of 2018; 6: e255–69.
recommendations. BMJ 2008; 336: 924–26. 45 WHO. Soil-transmitted helminth infections. 2020. [Link]
24 Murray CJ, Lopez AD. On the comparable quantification of health [Link]/news-room/fact-sheets/detail/soil-transmitted-helminth-
risks: lessons from the Global Burden of Disease Study. infections (accessed Sept 22, 2021).
Epidemiology 1999; 10: 594–605. 46 Bain R, Cronk R, Wright J, Yang H, Slaymaker T, Bartram J.
25 Ezzati M, Lopez AD, Rodgers A, Vander Hoorn S, Murray CJ, Fecal contamination of drinking-water in low- and middle-income
Comparative Risk Assessment Collaborating Group. Selected major countries: a systematic review and meta-analysis. PLoS Med 2014;
risk factors and global and regional burden of disease. Lancet 2002; 11: e1001644.
360: 1347–60. 47 GBD 2019 Risk Factors Collaborators. Global burden of 87 risk
26 Ezzati M. Annex 4.1: comparative risk assessment in the Global factors in 204 countries and territories, 1990–2019: a systematic
Burden of Disease Study and the environmental health risks. In: analysis for the Global Burden of Disease Study 2019. Lancet 2020;
Methodology for assessment of environmental burden of disease. 396 (suppl 1): 1223–49.
Geneva: World Health Organization, 2000: 31–33. 48 Rogawski ET, Liu J, Platts-Mills JA, et al. Use of quantitative
27 Prüss-Ustün A, Mathers C, Corvalán C, Woodward A. Introduction molecular diagnostic methods to investigate the effect of
and methods: assessing the environmental burden of disease at enteropathogen infections on linear growth in children in low-
national and local levels. Geneva: World Health Organization, 2003. resource settings: longitudinal analysis of results from the MAL-ED
28 Wolf J, Bonjour S, Prüss-Ustün A. An exploration of multilevel cohort study. Lancet Glob Health 2018; 6: e1319–28.
modeling for estimating access to drinking-water and sanitation. 49 Antimicrobial Resistance Collaborators. Global burden of bacterial
J Water Health 2013; 11: 64–77. antimicrobial resistance in 2019: a systematic analysis. Lancet 2022;
29 Bain R, Johnston R, Khan S, Hancioglu A, Slaymaker T. Monitoring 399: 629–55.
drinking water quality in nationally representative household 50 Sclar GD, Penakalapati G, Caruso BA, et al. Exploring the relationship
surveys in low- and middle-income countries: cross-sectional between sanitation and mental and social well-being: a systematic
analysis of 27 multiple indicator cluster surveys 2014–2020. review and qualitative synthesis. Soc Sci Med 2018; 217: 121–34.
Environ Health Perspect 2021; 129: 97010. 51 Crider YS, Ray I. Water and development: a gender perspective.
30 WHO, UN Children’s Fund. Safely managed drinking water: 2022. [Link]
thematic report on drinking water 2017. Geneva: World Health (accessed July 13, 2022).
Organization and United Nations Children’s Fund, 2017. 52 Savović J, Jones HE, Altman DG, et al. Influence of reported study
31 Jones ER, van Vliet MTH, Qadir M, Bierkens MFP. Country-level design characteristics on intervention effect estimates from
and gridded estimates of wastewater production, collection, randomized, controlled trials. Ann Intern Med 2012; 157: 429–38.
treatment and reuse. Earth Syst Sci Data 2021; 13: 237–54. 53 Garzón M, Pereira-da-Silva L, Seixas J, Papoila AL, Alves M.
32 Ram PK, Sahli M, Arnold B, et al. Global scaling up handwashing: Subclinical enteric parasitic infections and growth faltering in
validity of rapid measures of handwashing behavior: an analysis of infants in São Tomé, Africa: a birth cohort study.
data from multiple impact evaluations in the global scaling up Int J Environ Res Public Health 2018; 15: E688.
handwashing project. Washington: Water and Sanitation Program, 54 Cumming O, Cairncross S. Can water, sanitation and hygiene help
World Bank Group, 2014. eliminate stunting? Current evidence and policy implications.
33 Wolf J, Johnston R, Freeman MC, et al. Handwashing with soap Matern Child Nutr 2016; 12 (suppl 1): 91–105.
after potential faecal contact: global, regional and country estimates. 55 Lee D, Gibson JM, Brown J, Habtewold J, Murphy HM. Burden of
Int J Epidemiol 2019; 48: 1204–18. disease from contaminated drinking water in countries with high
34 WHO. Global health estimates: leading causes of death. 2019. access to safely managed water: a systematic review. medRxiv 2022;
[Link] published online Nov 8. [Link]
health-estimates/ghe-leading-causes-of-death (accessed (preprint).
Jan 19, 2023). 56 Brown J, Acey CS, Anthonj C, et al. The effects of racism, social
35 WHO. Global health estimates: leading causes of DALYs. 2019. exclusion, and discrimination on achieving universal safe water and
[Link] sanitation in high-income countries. Lancet Glob Health 2023;
health-estimates/global-health-estimates-leading-causes-of-dalys 11: e606–14.
(accessed Jan 19, 2023). 57 Capone D, Cumming O, Nichols D, Brown J. Water and sanitation in
36 WHO. WHO methods and data sources for country-level causes of urban America, 2017–2019. Am J Public Health 2020; 110: 1567–72.
deaths 2000–2019. Geneva: World Health Organization, 2020. 58 Mueller JT, Gasteyer S. The widespread and unjust drinking water and
37 WHO. WHO methods and data sources for global burden of disease clean water crisis in the United States. Nat Commun 2021; 12: 3544.
estimates 2000–2019. Geneva: World Health Organization, 2020. 59 Taylor-Robinson DC, Maayan N, Donegan S, Chaplin M, Garner P.
38 Rowe AK, Powell KE, Flanders WD. Why population attributable Public health deworming programmes for soil-transmitted helminths
fractions can sum to more than one. Am J Prev Med 2004; in children living in endemic areas. Cochrane Database Syst Rev 2019;
26: 243–49. 9: CD000371.

2070 [Link] Vol 401 June 17, 2023

 Articles

60 Dewey KG, Wessells KR, Arnold CD, et al. Characteristics that 65 Graham JP, Hirai M, Kim S-S. An analysis of water collection labor
modify the effect of small-quantity lipid-based nutrient among women and children in 24 sub-Saharan African countries.
supplementation on child growth: an individual participant data PLoS One 2016; 11: e0155981.
meta-analysis of randomized controlled trials. Am J Clin Nutr 2021; 66 Sorenson SB, Morssink C, Campos PA. Safe access to safe water in
114 (suppl 1): S15–42. low income countries: water fetching in current times. Soc Sci Med
61 Mora C, McKenzie T, Gaw IM, et al. Over half of known human 2011; 72: 1522–26.
pathogenic diseases can be aggravated by climate change. 67 Williams RN, Sclar GD, Routray P, Majorin F, Blais L, Caruso BA.
Nat Clim Chang 2022; 12: 869–75. A qualitative assessment of mothers’ perceptions and behaviors in
62 Delpla I, Jung A-V, Baures E, Clement M, Thomas O. Impacts of response to an intervention designed to encourage safe child feces
climate change on surface water quality in relation to drinking management practices in rural Odisha, India. J Water Sanit Hyg Dev
water production. Environ Int 2009; 35: 1225–33. 2022; 12: 375–86.
63 Hyde-Smith L, Zhan Z, Roelich K, Mdee A, Evans B. Climate 68 Roche R, Bain R, Cumming O. A long way to go—estimates of
change impacts on urban sanitation: a systematic review and failure combined water, sanitation and hygiene coverage for
mode analysis. Environ Sci Technol 2022; 56: 5306–21. 25 sub-Saharan African countries. PLoS One 2017; 12: e0171783.
64 Prüss-Ustün A, Bartram J, Clasen T, et al. Burden of disease from
inadequate water, sanitation and hygiene in low- and middle-
income settings: a retrospective analysis of data from 145 countries.
Trop Med Int Health 2014; 19: 894–905.

[Link] Vol 401 June 17, 2023 2071