Kingsley-amg/ghana-malnutrition

Child malnutrition in Africa with a focus on Ghana: a three-level R analysis (continent → country → 16 regions) with a regional choropleth and equity breakdowns. DHS + World Bank + geoBoundaries data. # Child Malnutrition in Africa, with a Focus on Ghana A geospatial and equity analysis of child malnutrition built in **R**, zooming from the **African continent**, into **Ghana's national trends**, down to its **16 regions** and social groups. Uses open data from the **DHS Program**, the **World Bank**, and **geoBoundaries**. > Ghana is a relative success on chronic undernutrition (stunting ~17%, well below > the African median), but the national average hides a sharp **north-south > divide** (Northern 30% vs Eastern 10%) and steep **wealth and education > gradients** (poorest children ~3× as likely to be stunted as the richest). --- ## 🔗 Read the report **▶ Full report (HTML)**, knitted from R Markdown, with the regional map, charts and interpretation. A **PDF version** is in `report/`. ## 🔍 What the analysis covers (three-level zoom) 1. **Africa context**, where Ghana ranks among African countries on stunting. 2. **Ghana over time**, national trends and the emerging "double burden" (chronic undernutrition falling while overweight/wasting persist). 3. **Within Ghana**, a **choropleth map of stunting across the 16 regions**, a ranked regional comparison, and an **equity analysis** by wealth quintile, mother's education, residence and sex. ## 🔑 Key findings - Ghana's stunting (~17%) is **well below the African median** (~30%). - A **2.8× north-south gap**: ~30% stunting in the Northern region vs ~10% in Eastern. - Steep **social gradient**: ~3× more stunting in the poorest vs richest wealth quintile; ~4× by mother's education (no education vs higher). - National progress now depends on **reaching the poorest regions and households**, not averages. ## 🗂️ Structure ``` ghana-malnutrition/ ├── 01_extract.R # pull DHS + World Bank data and Ghana region boundaries ├── ghana_malnutrition.Rmd # the full analysis (source) ├── data/ # extracted CSVs + ghana_adm1.geojson ├── docs/index.html # knitted HTML report (GitHub Pages) └── …