Children’s Environmental Health Indicators in Context of the Sustainable Development Goals for Small Island Developing States

The unique environmental vulnerability of small island developing states (SIDS) is likely to impact negatively on children’s health. Children’s environmental health indicators (CEHI) are standardized measures that can be used to assess the environmental exposures and their resulting health outcomes in children. This study sought to utilize the United Nations (UN) global Sustainable Development Goals (SDGs), with their associated targets and indicators, as a framework for a CEHI proposal for SIDS. Exposure-side indicators were taken from key themes from the 2012 Rio+20 UN Conference on Sustainable Development, and health-side indicators were selected based on the most significant contributors to the burden of disease in children. The multiple-exposures–multiple-effect (MEME) framework was then used to show the relationships between environmental exposures and children’s health outcomes. The framework was populated with available data from the World Bank’s DataBank. Whilst there was some data available at a population level, major gaps in both exposure-side and health-side indicators were revealed. In order to progress children’s environmental health in SIDS, a further piece of work is required to propose a fully prioritized set of exposure-side and health-side CEHIs; based on, but not exclusively linked to, the SDGs.


Introduction
The key to a peaceful, prosperous, and sustainable world is healthy, safe, and empowered children [1]. Because children are particularly vulnerable to environmental risks to their sustained and healthy development [2], creating a safe and healthy environment is one of the essential elements for improving child survival, helping children to thrive, and achieving health equity [3]. In fact, studies have identified a significant link between the burden of disease and environmental risks [4,5]. When selecting exposure-side indicators, we first set the strategic themes in relation to the leading topics of the 2012 Rio+20 UN Conference on Sustainable Development, which were food; water, sanitation, and hygiene (WASH); energy; cities; and climate change [14]. We modified the Rio+20 theme of 'cities' to 'cities and human settlements' to correspond with SDG 11, as well as to ensure that the many settlements outside cities in SIDS were considered. Because geographical characteristics and the development status of SIDS can make their cities and human settlements particularly vulnerable to climate change [15], we linked climate change to cities and human settlements and concentrated on showing the multiple effects of climate change and unsustainable settlements on children's health. In addition, water contamination was identified as one of the major problems in waste management in SIDS [16], waste management was incorporated as part of WASH.
For health-side indicators, we particularly focused on indicators related to the most significant contributors to the burden of disease in children of LMICs [9]. Therefore, we selected respiratory diseases, diarrheal diseases, malaria, and malnutrition as our main health outcomes. We also selected a health-side indicator addressing disasters as one of our main health-side indicators, because SIDS are prone to natural disasters, often associated with extreme weather events and climate change [17].
To construct children's environmental health indicators (CEHI) for SIDS, the multipleexposures-multiple-effect (MEME) framework, which was specifically developed for improving, monitoring, and evaluation of children's environmental health [2], was populated with the selected SDG indicators and targets. The MEME framework was chosen because it allows one to describe the complex relationship between exposures and health outcomes with integrating socio-economic contexts and policy actions into the framework. Within each strategic theme, exposure-side indicators and context indicators were selected and health-side indicators were placed alongside the related exposure or context. Health outcomes were set by using SDG targets. Related SDGs were stated separately to show the general objectives of the whole CEHI process. Because some measures that are required to describe the context of the relationship between environmental exposures and health outcomes were not readily available in the indicators list, we also reviewed indicators from the MEME framework and included its indicator in the list.
With robust data provided, the SDG indicators integrated into the MEME framework would enable monitoring and evaluation of children's environmental health. We searched the World Bank's DataBank [18] for the data from 2010 to 2017 for data availability. We used the DataBank as a sole source of data because other international organizations' methods of country classifications were different from the World Bank's. The World Bank classifies small states into three groups: Caribbean When selecting exposure-side indicators, we first set the strategic themes in relation to the leading topics of the 2012 Rio+20 UN Conference on Sustainable Development, which were food; water, sanitation, and hygiene (WASH); energy; cities; and climate change [14]. We modified the Rio+20 theme of 'cities' to 'cities and human settlements' to correspond with SDG 11, as well as to ensure that the many settlements outside cities in SIDS were considered. Because geographical characteristics and the development status of SIDS can make their cities and human settlements particularly vulnerable to climate change [15], we linked climate change to cities and human settlements and concentrated on showing the multiple effects of climate change and unsustainable settlements on children's health. In addition, water contamination was identified as one of the major problems in waste management in SIDS [16], waste management was incorporated as part of WASH.
For health-side indicators, we particularly focused on indicators related to the most significant contributors to the burden of disease in children of LMICs [9]. Therefore, we selected respiratory diseases, diarrheal diseases, malaria, and malnutrition as our main health outcomes. We also selected a health-side indicator addressing disasters as one of our main health-side indicators, because SIDS are prone to natural disasters, often associated with extreme weather events and climate change [17].
To construct children's environmental health indicators (CEHI) for SIDS, the multipleexposures-multiple-effect (MEME) framework, which was specifically developed for improving, monitoring, and evaluation of children's environmental health [2], was populated with the selected SDG indicators and targets. The MEME framework was chosen because it allows one to describe the complex relationship between exposures and health outcomes with integrating socio-economic contexts and policy actions into the framework. Within each strategic theme, exposure-side indicators and context indicators were selected and health-side indicators were placed alongside the related exposure or context. Health outcomes were set by using SDG targets. Related SDGs were stated separately to show the general objectives of the whole CEHI process. Because some measures that are required to describe the context of the relationship between environmental exposures and health outcomes were not readily available in the indicators list, we also reviewed indicators from the MEME framework and included its indicator in the list.
With robust data provided, the SDG indicators integrated into the MEME framework would enable monitoring and evaluation of children's environmental health. We searched the World Bank's DataBank [18] for the data from 2010 to 2017 for data availability. We used the DataBank as a sole source of data because other international organizations' methods of country classifications were different from the World Bank's. The World Bank classifies small states into three groups: Caribbean small states (CSS), Pacific Island small states (PSS), and other small states [19]. Data for PSS were used as the closest equivalent for SIDS because the PSS group contains the largest number of LMICs and exclusively covers island states: nine states included in the PSS were Fiji, Kiribati, Marshall Islands, Federated States of Micronesia, Samoa, Solomon Islands, Tonga, Tuvalu, and Vanuatu [20]. CSS contains seven LMICs, including Belize, Dominica, Grenada, Guyana, Jamaica, Saint Lucia, and Saint Vincent and the Grenadines [20]. Table 1 shows indicators from both the SDGs and the MEME frameworks. The exposure-side and health-side indicators, SDG targets, and goals were classified according to the strategic themes and placed in the MEME framework. Eleven exposure-side or context indicators related to food, WASH, energy, and cities and human settlements associated with climate change were selected from the global indicator framework and the MEME framework. For instance, poverty was viewed as a context indicator for food as poverty can create a vicious cycle that involves malnutrition [21]. Indicators on water, sanitation, hygiene, and waste management were matched to WASH as stated above.

Results
Considering that energy is related to access to clean fuel and use of renewable energy sources [22], energy was associated with indoor and outdoor air pollution. CO 2 emissions were placed as a context indicator for climate change because CO 2 emissions are generally regarded as one of the main causes of climate change [23]. In addition, a MEME indicator on assessing the number of children with an increased risk of malaria transmission was used as a exposure-side indicator for climate change because climate change has been suspected to increase malaria-transmissible areas [24].
Health outcomes were composed of seven SDG health-side indicators on malnutrition, diarrheal diseases, respiratory diseases, malaria, and natural disasters. Thus, these indicators were mapped to food, WASH, energy, and climate change, respectively. Food was related to malnutrition because of the link between inadequate food intake and malnutrition. WASH was linked to the WASH mortality rate instead of mortality as a result of diarrheal disease, because there is no health-side indicator addressing diarrheal diseases in the global indicator framework for the SDGs.
For energy, because the use of biomass fuel is a major source of household air pollution [25], access to clean fuels and technologies for cooking and heating can reduce the mortality rate attributed to household air pollution. Fossil fuel combustion can generate a number of air pollutants including sulfur oxides, nitrogen oxides, and particulate matter [26], all of which can cause outdoor air pollution. It has been inferred that these air pollutants could result in pulmonary diseases [27]. Therefore, indicator 3.4.1 was used as a health-side indicator for energy because it includes chronic respiratory disease as health outcomes of interest.
Although there is a wide variety of effects of climate change on children's health [28], climate change was specifically linked with malaria and natural disasters because of their relevance to SIDS. Furthermore, as climate change has been suspected to be linked to increased risks of mosquito-borne diseases [29], malaria was selected as a health outcome for climate change. As the frequencies of extreme events, including droughts, floods, and extreme temperatures, can be affected by climate change [30], climate change is thought to be a potential contributor to natural disasters.
Actions consisted of nine SDG targets corresponding to outcomes. Targets tackling health outcomes tend to involve multiple health outcomes in one target. Thus, health outcome targets were included if those targets state the health outcome of interest. For instance, WASH has four targets, two of which are health outcome targets. Because target 3.3 states water-borne diseases and target 3.4 states mortality due to water contamination, presumably related to diarrheal diseases, targets 3.3 and 3.4 were used as action indicators for WASH. Similarly, target 3.4 could be incorporated as an action indicator for energy because non-communicable diseases encompass chronic respiratory diseases. Because target 3.3 has malaria as one of the health outcomes of interests, it also links to climate change. The target on natural disasters was also regarded as an action indicator for climate change. Targets on poverty reduction and improving food security were employed for food.
Seven SDGs that indicate general objectives of each CEHI process are shown. SDG 3, "Ensure healthy lives and promote well-being for all at all ages [12]", was included as a common goal and each strategic theme has corresponding goals.  [31]. * Context/exposure/health outcome/action numbering taken from Sustainable Development Goals (SDG) target and indicator numbering as described in Figure 1. Abbreviations: AIDS, Acquired Immune Deficiency Syndrome; PM 10 , particulate matter less than or equal to 10 µm in diameter; PM 2.5 , particulate matter less than or equal to 2.5 µm in diameter. Table 2 provides data variables and data availability for the selected exposures. One indicator could have a number of data variables in the DataBank. Labels of data variables presented here are identical to those in the DataBank. Discrepancies between, as well as within, strategic domains were observed. Accessibility to data on climate change was relatively high. Data on WASH and energy were only available at a basic level, such as percentage of people using basic drinking water services and mean annual concentration of PM 2.5 . An advanced level of data, including people having access to safely managed drinking water or people having exposure to PM 2.5 concentration exceeding WHO interim target-3 level, was not readily accessible. For WASH, indicators on waste management and water quality did not even have data variables specified in the DataBank. Data for poverty were not found in the Databank. Even though some of the exposure-side indicators were disaggregated by location of residence, none of the data were child specific. As for population 7.1.2 Access to electricity (% of population; total, rural, and urban) 11.6.2 PM 2.5 air pollution, mean annual exposure (micrograms per cubic meter) 11.6.2 PM 2.5 air pollution, population exposed to levels exceeding WHO guideline value (% of total) 11.6.2 PM 2.5 air pollution, population exposed to levels exceeding WHO Interim Target-1 value (% of total) Data not available 11.6.2 PM 2.5 air pollution, population exposed to levels exceeding WHO Interim Target-2 value (% of total) 11.6.2 PM 2.5 air pollution, population exposed to levels exceeding WHO Interim Target-3 value (% of total)  Table 3 shows data variables and data availaiblty for the selected health outcomes. Health data were rarely available for PSS. Data have been collected for only two health outcomes; namely prevalence of undernourishment and mortality from cardiovascular disease, cancer, diabetes, or chronic respiratory disease between the exact ages 30 and 70. However, data were not child-specific. For other indicators, data were either not available or data variables did not exist in the Databank. Health-side indicators on food were the only indicators that are child-specific, but no data have been collected.

Discussion
The SDGs and MEME indicators were formulated around the MEME framework to illustrate the relationship between the selected environmental risks and health consequences in children. The four strategic themes, namely, food, WASH, energy, and climate change, associated with cities and human settlements were selected because of their significance in the SDGs and relevance to situations in SIDS. For instance, food insecurity has been a major concern in SIDS [32], and SIDS heavily rely on combustion of fossil fuels [33] to generate energy. Depletion of fresh water, lack of proper sanitation facilities, and inappropriate treatment of waste water have been environmental threats to SIDS [34], and SIDS are vulnerable to the imminent impacts of climate change [35].
The strategic themes were linked to the leading causes of childhood mortality [6,9], preventable through improving environmental conditions. The limitation of this approach was that we linked a single environmental exposure to a specific health outcome, but there are many cases where multiple exposures can cause multiple health outcomes in a complex relationship. However, use of the MEME framework can reduce the bias because it allows a more systematic approach to the complex relationship between exposures and health outcomes.
The indicators in the framework can be used to generate scientific data for SIDS. In addition, because of their relevance to the SDGs, the indicators can inform policy makers of the glimpse of the current environmental health status. To determine the data availability and simulate future usage, data from the World Bank's DataBank were presented.
PSS were used as a probe for SIDS because PSS encompassed the largest number of LMICs and exclusively contained island states. CSS include Belize and Suriname, which are not island countries [36]. Because the common characteristics can stem from the geographical peculiarity of islands [37], using PSS as an example was plausible. The shortcoming that was found in the scrutiny of data was that because the SDG indicators were developed to be used at country level to feed into a global level perspective, they are not ideally fitted for assessing children's environmental health status for SIDS. Therefore, SIDS should develop complementary/supplementary indicators that meet their national needs.
In addition, none of the exposure data were disaggregated by age. Similarly, the health-side indicators in the SDGs are mostly not child-specific. Even though the guidelines for SDG indicator measurement include data disaggregation into age bands where possible [38], it is unclear whether source data from SIDS contain the detail required to achieve this. We know that children are disproportionately affected by environmental exposures because of their unique exposure pathways and their growing bodies and systems [39], making it especially important to collect disaggregated data by age group to track progress in children's health.
Furthermore, data robustness could also be enhanced through collection and summarization of age metadata. Analysis of age-disaggregated data would then enable identification of susceptible age groups and analysis of health status gaps between age groups. In addition, this information would help prioritize the allocation of limited resources in SIDS. Thus, it is inferred that disaggregating these indicators by age will be required before adopting the indicators for assessing children's environmental health status in SIDS.
Collection of data on both environmental risks and the related health outcomes provides a basis for identifying the effects of environmental interventions on health outcomes. Data availability was appraised for each strategic theme and was found to be mostly unavailable-especially in health outcomes. Only 2 out of 18 data variables have data collected in health outcomes.
In addition to the lack of data availability, there was a tendency for a number of health outcomes to be included in one health-side target and separate data variables were not provided in the data list. For instance, health-side target 3.3; "By 2030, end the epidemics of Acquired Immune Deficiency Syndrome (AIDS), tuberculosis, malaria, and neglected tropical diseases and combat hepatitis, water-borne diseases, and other communicable diseases [12]", did not have a disaggregated data variable for water-borne diseases in the Databank. If a data variable is missing, the corresponding data would not be accessible either.
Similarly, exposure data were not appropriately elaborated in the data. For instance, energy only focused on access to clean fuel, while improved stove could be more feasible in LMICs settings. Hence, it is crucial to specify a complete set of indicators in the data variables to collect robust data and monitor children's environmental health status for SIDS.

Conclusions
This study proposes a coherent set of the most important CEHIs for SIDS, using the MEME framework to show the linkages between exposure-side indicators and health-side indicators. The SDG goals, targets, and indicators, supplemented by indicators from the MEME children's environmental health framework, provide a common reference point. There are many advantages to developing a set of CEHIs that are based on SDG targets and indicators. SIDS are required to regularly report progress on SDGs so a proposal for collection of data , which contributes to this reporting, will align with existing focus areas.
Whilst most of the SDG indicators are not child-specific, the guidelines for data collection request disaggregation into age and gender groups. This disaggregated data is unavailable in the World Bank DataBank, but some may be available from other sources. Not all of the SDG indicators are ideally suited for assessing children's environmental health, and SIDS should develop complementary/supplementary indicators for more specific exposure-side and health-side indicators.
Setting up a data collection process and system based on a suitable set of indicators will provide evidence that will help SIDS to identify and target their most important children's environmental health issues.