s

Exposure levels were selected according to theavailability of exposure data and corresponding expo-sure-risk information (Wolf et al. 2013, 2014) andincluded the following: (i) using an unimproved watersource; (ii) using an improved water source other thanpiped to premises; (iii) using basic piped water on pre-mises (improved source); and (iv) using a water filter orboiling water in the household (on water from an unim-proved or improved source).

o estimate uncertainty intervals, we developed a MonteCarlo simulation of the results with 5000 draws of theexposure distribution, and of the relative risks.

epidemiological

diarrhoeal diseases, have been undertaken in recent dec-ades (Murray & Lopez 1996; WHO 2002, 2004, 2009;Pr€uss-Ust€un et al. 2008; Lim et al. 2012). These reportimportant changes in the roles of various risk factors(Clasen et al. 2014).Inadequate drinking water, sanitation and hygiene(WASH) are important risk factors, particularly in low-income settings. In 2011, an estimated 768 million peo-ple relied on ‘unimproved’ water supplies (as defined bythe WHO/UNICEF Joint Monitoring Program for Waterand Sanitation – JMP), which are thought to have highlevels of pathogen contamination (WHO & UNICEF2013a). Many more use sources that are classified as‘improved’ but are still unsafe for consumption (Bainet al. 2014). More than 2.5 billion people lack access toan improved sanitation facility (WHO & UNICEF2013a). Inadequate hand hygiene practices have beenestimated to affect 80% of the population globally (Free-man et al. 2014b).The health risks from inadequate WASH have beendocumented previously (Esrey et al. 1991; Fewtrell et al.2005; Waddington et al. 2009). However, the unpub-lished review on which the 2010 Global Burden of Dis-ease (GBD) study is based (Lim et al. 2012) departedfrom earlier reviews by finding no additional benefit fromfurther improvements such as higher water quality orcontinuous piped supply over the exposure defined asusing ‘other improved water supplies’ (Engell & Lim2013). A more recent systematic review, however, is lar-gely consistent with previous evidence (Wolf et al. 2014).Estimating the impact of WASH on diarrhoeal diseaseshas commonly been assessed with comparative riskassessment methods (Ezzati et al. 2002; WHO 2004; Limet al. 2012), although other methods such as populationintervention models could also be considered (Clasenet al. 2014). Other diseases cannot currently be estimatedwith such methods due to insufficient evidence andrequire alternative approaches. As these would requireconsiderable additional assessments and analyses, theyare not analysed in detail in this article.Accrual of substantive recent evidence, as well astrends in the total diarrhoea burden, justifies the revisionof methods and estimates of the burden of diarrhoealdisease associated with inadequate WASH. While theestimate presented focuses mainly on low- and middle-income settings, the approach used can accommodate awider range of settings. An overview of previous findingson the impacts of WASH on other diseases thandiarrhoea is also provided.

l

Though very highnutrient inputs and intense eutrophication are required todrive a weakly stratified ecosystem to hypoxia or anoxia,even moderate increases in nutrient inputs and eutrophica-tion can lead to these conditions in a strongly stratifiedwater body (NRC 2000). Stratification in coastal waters isdriven both by temperature and salinity gradients; bothgradients are responsive to global change as freshwaterinputs vary with rising temperatures and as surface windand ocean circulation patterns are altered. These condi-tions are found in the northern Gulf of Mexico hypoxic-zone region: currently, bottom-water hypoxia is much moreprobable when the difference in surface- and bottom-watersalinities is greater than 4.1 ppt.

ustained bottom-water anoxia can result in the complete loss of animals inbottom sediments.

eutrophic

uman activity has greatly accelerated the flows ofnutrients to estuaries and other coastal marineecosystems over the past half century, increasing primaryproduction and causing widespread eutrophication(Nixon 1995; NRC 2000; Rabalais 2002). As of the late1990s, global fluxes in rivers to coastal oceans were sometwofold greater for nitrogen (N; Galloway et al. 2004;Boyer and Howarth 2008) and two- to threefold greaterfor phosphorus (P; Howarth et al. 1995) than before theIndustrial and Agricultural Revolutions. During the past10 years, in many industrialized countries, P inputs havesharply decreased as a result of improvements in waste-water treatment plants, but N pollution has remainedhigh. The global increase in nutrient fluxes has not beenevenly distributed, and some regions have seen muchgreater changes than others (Figure 1). Two-thirds of theestuaries in the US are degraded as a result of nutrientpollution, which can cause dissolved oxygen depletion(hypoxia and anoxia), loss of critical habitat such as sea-grass beds and coral reefs, changes in ecological structureincluding loss of biodiversity, and increased incidencesand duration of harmful algal blooms

s

Such actions include, for example,enhancing microbial activity by adding appropriateelectron donors or acceptors to the system (29), orintroducing abiotic reactants into contaminatedgroundwaters such as zero-valent metals in per-meable reactive barriers (30)

geogenic contamination

synthetic membranes.

For example, a recent study has shownthat when five estrogenic compounds are mixedin concentrations all below levels at which theirindividual effects can be detected, their cumula-tive impact on fish was detrimental

A bout one-fifth of the world_s popula-tion does not have access to safe water,and two-fifths suffer the consequencesof unacceptable sanitary conditions (1). Patho-gens in water cause more than 2 million deathsannually; most are children under the age of 5.The increasing chemical pollution of surfaceand groundwaters, with largely unknown long-term effects on aquatic life and on humanhealth, could easily lead to a problem of similaror even greater magnitude. More than one-thirdof the Earth_s accessible renewable freshwateris used for agricultural, industrial, and domes-tic purposes, and most of these activities leadto water contamination with numerous syn-thetic and geogenic compounds (Table 1). Ittherefore comes as no surprise that chemicalpollution of natural waters has already becomea major public concern in almost all parts of theworld.