Syndicate content

Sub-national Malnutrition Data

Vanessa Moreira da Silva's picture
Also available in: Español | Français | 中文 | العربية

There is a growing perception that spatial disparities in development indicators within countries are becoming more pronounced. Sub-national data are needed to inform policy makers on such matters. However, data on the sub-national level is less frequent (curated in a global setting) because sub-national administrative areas change frequently.

Now, for the first time, a global database makes sub-national data on child-malnutrition indicators available easily and openly accessible. They are based on the WHO Global Database on Child Growth and Malnutrition which is a carefully maintained database (good survey source and metadata) covering almost all developing countries. The database contains key anthropometric indicators that are internationally comparable and consistent between sub-national and country-level data. Also, it is the source for the current WDI data at the country level. Along with various child-malnutrition indicators and metadata, this database contains the “names” of sub-national areas in the various countries, but not sub-national coding schema.

Kenya is a country facing severe drought (in some regions), poverty, diseases (HIV, malaria) and lack of access to food among other problems causing malnutrition rates to reach the highest levels in more than a decade.

International organizations and the Kenyan government have invested greatly into reducing malnutrition in Kenya over the years. In response to their efforts, rates have dropped on average but they are still very high in some regions. In the North Eastern region, a drought prone region near the Somalia border where insecurity is high (Somali refugees fleeing drought and conflict back home), malnutrition rates are among the highest in country.

Building upon the standards and procedures for identifying sub-national administrative areas and for managing codification schema already in place, for easier updates, we merged the WHO Global Database on Child Growth and Malnutrition with the most updated administrative codes from the following sources:

  • Global Administrative Unit Layers (GAUL)
    The GAUL initiative is implemented by the Food and Agricultural Organization (FAO) and based on collaborative work by international agencies and national authorities that generate and collect spatial information on administrative units. The GAUL dataset is updated annually and disseminated free of charge via FAOs open source GeoNetwork platform. Each new version includes all historical updates.
  • Global Administrative Areas (GADM)
    GADM is a collaborative effort by researchers and organizations to freely provide a spatial database of the location of countries and sub-national administrative areas for use in GIS and similar software. The primary focus is on providing sub-national area boundary files, not on a schema for coding these. However, for each sub-national area, the database provides some attribute information, foremost being sub-national area name and variant names.
  • Second Administrative Level Boundaries Dataset (UNSALB)
    This interagency initiative was launched in the context of the activities of United Nations Geographic Information Working Group (UNGIWG) and provides open access to a global standardized GIS layer containing the delimitation of the administrative boundaries down to the 2nd sub-national level. The SALB coding scheme can be applied to any country and allows for tractable updating to reflect changes in the naming and delimitation of the sub-national administrative areas.


The WHO Global Database on Child Growth and Malnutrition does not contain the survey sample expansion weighted population data (only lists the sample size) for each sub-national area that would enable users to compute the population-weighted national aggregates. Also, standard deviations are not available which would be useful for a more careful comparison.

Even though every effort has been made to maximize the comparability of statistics across countries and over time users are advised that data may differ in terms of data collection methods, population coverage and estimation methods used. Differences between the estimates presented in the WHO Global Database on Child Growth and Malnutrition may arise because of differences in reporting periods or in the availability of data. For detailed information on the methodology used for estimating regional and global trends of child malnutrition see

UNICEF, WHO and the World Bank, recommend not using the data for wasting and severe wasting indicators to make inter-temporal comparisons. Wasting and severe wasting indicators are very responsive to infection and changes in food availability. A child’s weight relative to its height can drop quickly but also bounce back up with appropriate interventions or a stabilization of a crisis. Malnutrition prevalence estimates are generated from household surveys that only allow for a snapshot view at one short point in time. Stunting, underweight, and overweight are more stable and less reactive to rapid changes in the conditions children live in.

Data currently disseminated publically via

Related links:


Good to see the greater detail and disaggregation possible - but this is still at levels of aggregation that can be misleading. Averages for Nairobi are misleading because the concentration of middle and high income groups pulls these up. Obscures the very poor conditions evident in the informal settlements in which half of Nairobi's population lives. See the work of the African Population and Health Research Centre on this and on how to get a level of spatial disaggregation that is actually useful in identifying where needs are concentrated

Submitted by iMitwe on

Especially, this open initiative is not known in some African country ( my country Burundi for example). Great achievement.

Add new comment