Photo credit: Neil Palmer/CIAT

Malnutrition in children and adults results from a number of factors related to the availability and accessibility of quality food as well as the presence of infectious diseases. As such, the spatial relationship between agricultural systems (farming and livestock) and nutritional and dietary outcomes in human population can provide valuable information to aid in the design of appropriate policy interventions.

HarvestChoice leads an extensive effort to compile harmonized cross-country data from nationally representative household production surveys and censuses in Africa South of the Sahara (SSA). We extract information on farm household enterprise mixes, input and technology use, market participation, and other relevant factors to characterize agricultural production systems.

In the past, our team created data tools using household surveys, including a mapping tool of various dietary quantity indicators at sub-national level for selected SSA countries. Thanks to the support of the CGIAR Research Program on Agriculture for Nutrition and Heath (A4NH), HarvestChoice has recently added many spatial layers on nutrition and dietary outcomes based on elaborations from the Demographic and Health Surveys (DHS).

Our newly added indicators help visualize how diet quality and nutritional outcomes are distributed across sub-national regions in SSA. Such data can support analysis of how diet and nutrition are related to the characteristics of markets, environment, and agricultural systems, and can provide the context for understanding the scalability of research outcomes.

DHS data provide population-level representative data to estimate key nutrition indicators and to help inform national strategies and action plans. They offer some unique advantages for the proposed update of mapping tools due to geo-coded survey clusters [1] (in recent waves), comparable data across countries, good coverage, self-reported as well as biomarker measures of nutritional outcomes, and multiple waves (for some countries) to enable the analysis of trends in nutritional outcomes [2]. DHS have been conducted in 41 SSA countries (30 GIS-coded datasets since 2003), most of which have been identified as having the greatest burden of undernutrition [3].

DHS surveys are usually representative at national, regional and area (urban/rural) level, highly reliable, and homogenous across countries, especially relative to national food consumption and Living Standards Measurement Study (LSMS) surveys. They are particularly strong on measures such as nutrition, health, gender-related variables, and have useful information on wealth (consumer durables, housing characteristics), education, and access to services (water, sanitation, health facilities, schools).

DHS data are, however, weak on agricultural activities and income, as well as infrastructure. Combining HarvestChoice spatial data holdings on agriculture, biophysical characteristics, and infrastructure with DHS data can produce a well-rounded set of variables for a large number of countries and can facilitate cross-country as well as country-level [4] work on nutrition analysis and agriculture, including research on determinants of food security and linkages between agriculture, nutrition, and health.

In our pipeline for the future is the creation of a mapping tool to enable spatially-specific investment decisions on crop choices. The tool will support data visualization on crops and areas where intervention could result in the greatest potential impact.  It will then provide a data and mapping platform for a variety of uses, from outreach to policy, including investment options, and can support investment decisions and prioritization of activities that seek to enhance nutrition and health through appropriate agricultural interventions, policies, and practices.

In the meantime, thanks to Mappr and Tableau [5], you can already start exploring spatial linkages between agricultural production, food security, and nutritional outcomes with HarvestChoice data holdings.


[1] Note that DHS cluster locations are randomly displaced up to 2 kilometers in urban areas and up to 5 kilometers in rural areas, with additional 1% of the rural clusters displaced up to 10 kilometers to protect the confidentiality of survey respondents, In addition, although DHS data are nationally and regionally representative and geo-coded cluster data allows for regrouping in to a different (lower) levels of representativeness, caution should be exercised when reclassifying DHS clusters to levels that do not align with DHS regions.

[2] Nutritional outcomes in the DHS include anthropometric data for children under five years of age, women anthropometric data, biomarker data on micronutrient deficiencies such Iron and Vitamin A deficiencies, infant and young child feeding practices, dietary diversity of women (who gave birth in ‘the last three years’), and Iron and Vitamin supplementation of women (who gave birth in ‘the past five years’) and children age 6-59 months.

[3] Monica Kothari and Noureddine Abderrahim  (2010), Nutrition Update 2010, ICF Macro International.

[4] Nonetheless existing approaches using DHS data have not taken advantage of spatial modeling to expand the use of DHS data below the survey region level. Andrew Tatem, Susana Adamo, Nita Bharti, et al. (2012), ”Mapping Populations at Risk: Improving Spatial Demographic Data for Infectious Disease Modeling and Metric Derivation,” Population Health Metrics, 10(8).

[5] Links to Tableau viz: A4NH Viz 1A4NH Viz 2A4NH Viz 3


HarvestChoice, 2015. "Nutrition and Food Security Variables Now Available ." International Food Policy Research Institute, Washington, DC., and University of Minnesota, St. Paul, MN. Available online at

Apr 1, 2015