Climate-Smart Health

Climate change and other environmental challenges have real and immediate impacts on human health, especially in vulnerable nations like Madagascar. For example:

• Harvard Chan researcher Chris Golden has found that many Malagasy people rely on wild foods as a source of crucial nutrients. Climate-induced changes in fisheries, over-hunting of wild mammals, and habitat destruction could all worsen Madagascar’s severe malnutrition epidemic.
• Climate change and ecological destruction affect the transmission of infectious diseases: increased temperature and rainfall helps malaria-carrying mosquitos thrive, while habitat destruction can damage the predator-prey dynamics that keep parasitic diseases like schistosomiasis under control.
• Drought, fire, and floods may induce crop failure and malnutrition.
• Increasing cyclonic activity can lead to the destruction of existing healthcare facilities and to long-standing repercussions to mental health.

In collaboration with the Harvard T.H. Chan School of Public Health, the climate-smart health project aims to develop AI tools to 1) detect and predict climate change-induced health outcomes with diverse sources of data, including health surveys and satellite images, and 2) optimize intervention planning for limited health resources allocations to reduce the impact of climate on public health in vulnerable communities in Madagascar.

Micronutrient deficiency (MND), which is a form of malnutrition that can have serious health consequences, is difficult to diagnose in early stages without blood draws, which are expensive and time-consuming to collect and process. It is even more difficult at a public health scale seeking to identify regions at higher risk of MND. To provide data more widely and frequently, we propose an accurate, scalable, lowcost, and interpretable regional-level MND prediction system. Specifically, our work is the first to use satellite data, such as forest cover, weather, and presence of water, to predict deficiency of micronutrients such as iron, Vitamin B12, and Vitamin A, directly from their biomarkers. We use realworld, ground truth biomarker data collected from four different regions across Madagascar for training, and demonstrate that satellite data is a viable data source for predicting regional-level MND, which surprisingly exceeds the performance of baseline predictions based only on survey responses. Our method could be broadly applied to other countries where satellite data is available, and potentially create high societal impact if these predictions are used by policy makers, public health officials, or healthcare providers.