Feed the Future
This project is part of the U.S. Government's global hunger and food security initiative.

More Trees, Less Diarrheal Disease: Analyses of Earth Observation and Survey Data Show the Importance of Ecosystem Services for Human Health and Nutrition

This post was written by Kiersten Johnson, US Agency for International Development, Bureau for Food Security.

This month saw the release of a dire UN report from the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services (IPBES) warning that nature is declining globally at rates unprecedented in human history. 

“Land degradation, biodiversity loss and climate change are three different faces of the same central challenge: the increasingly dangerous impact of our choices on the health of our natural environment.”

-- Robert Watson, Chair, Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services

While decades of research have found that anthropogenic impacts on natural systems affect the health of our natural environment, it is only more recently that researchers have begun to empirically demonstrate at scale the extent to which degradation of natural systems impact the health, food security and nutrition of human populations. 

Many of these analyses have been critically facilitated by the availability of data sensed remotely by Earth-observing satellites.  As these Earth observations data have become increasingly accessible, possibilities have expanded for integrating these data with population-based datasets like the Demographic and Health Surveys -- the gold-standard survey program that collects cross-nationally comparable data on critical aspects of population, health and nutrition throughout the developing world.  Integrating these two types of data has allowed researchers to begin to explore associations between the natural environment (e.g., forest cover) and important health- and nutrition-related outcomes like diarrheal disease -- a leading cause of malnutrition and the second-leading cause of death globally among children under the age of 5 (World Health Organization 2017).

The first effort to explore these relationships was a 2013 study using survey data on recent experience of diarrhea among children under 5 from the Malawi DHS (Johnson et al. 2013) and a forest cover variable obtained via Earth observations imagery (MODIS Vegetation Continuous Fields).  It was hypothesized that places with greater tree cover obtained more water filtration ecosystem services from those tree-covered areas. The results supported the hypothesis, demonstrating that children living in places with greater forest cover were less likely to experience diarrheal disease (see Figure 1).

Figure 1. This figure shows that Malawian children living in places with greater forest cover had statistically significantly lower odds of experiencing diarrheal disease in the two weeks prior to the survey interview. The figure also shows that children living in places with greater forest cover also had statistically significantly higher odds of consuming vitamin A-rich foods.

A subsequent study (Herrera et al. 2017) made use of a unique dataset on watersheds to look at the relationship between upstream watershed disturbance, which incorporates satellite remote sensing data on tree cover, and downstream experience of childhood diarrheal disease, introducing a causal element into an otherwise cross-sectional analysis.  This analysis was conducted using pooled datasets from 35 Demographic and Health Surveys, and found that higher upstream tree cover is statistically significantly associated with lower probability of childhood diarrheal disease downstream for rural households (see Figure 2). The authors concluded that maintaining natural capital within watersheds can be an important public health investment, especially for populations with low levels of built capital.

Figure 2. This figure shows the factors associated with experience of childhood diarrheal disease in this pooled multi-country study. Variables reducing the probability of diarrheal disease have odds ratios lower than 1; they are found to the left of the red vertical line. The bottom two variables indicate that the associations between upstream watershed variables (human activity and tree cover) and the probability of diarrheal disease are significant (confidence intervals do not cross the red vertical line) for rural, but not urban households.

A new study in preparation (Das, Horn, Barnum, Mulligan and Johnson 2019) looks at the same question for the country of India, using data from the National Family Health Survey (the name for the Demographic and Health Survey in India) and the same unique dataset on watersheds to look at the relationship between upstream watershed disturbance and downstream experience of childhood diarrheal disease. Preliminary findings of the India analysis are consistent with the prior studies described here: children’s experience of diarrheal disease in India is lower in places where there in greater forest canopy cover.

Figure 3. The map on the left shows the modeled prevalence of diarrhea disease in the youngest children under 5 years of age for each surveyed household, with the lighter color indicating a lower prevalence in diarrhea disease. The map on the right shows the percent forest cover in India with the darker color indicating greater forest cover. Bar graphs show that the prevalence of diarrheal disease was significantly lower in areas with at least some forest cover.

The findings of these studies demonstrate a consistent inverse association between forest cover and diarrheal disease: more trees, less disease.  They contribute to the growing evidence base supporting the argument that a healthy, biodiverse environment is essential for healthy, well-nourished people.

This figure shows where the research in today's post contributes to the Feed the Future Results Framework