Dataset for: Dietary staple protein quality and meningitis mortality across 54 African countries: an ecological cross-sectional study
Description
Country-level dataset for 54 African countries comprising the primary outcome (meningitis cause fraction, GBD 2023), disease covariates (diarrhoea, malaria, HIV cause fractions, GBD 2023), care variables (household air pollution, unsafe sanitation, SDI, HiB3, PCV3, GBD 2023), and dietary variables (animal protein, staple protein supply for wheat, rice, maize, millet, sorghum and cassava, FAOSTAT food balance sheets 2019–2023). Used in beta regression analysis of dietary staple protein quality and meningitis mortality. Includes a data dictionary sheet defining all variables.
Files
Steps to reproduce
Country-level data were compiled for 54 African countries. The primary outcome — all-age meningitis cause fraction for 2023 — was obtained from the Global Burden of Disease (GBD) 2023 Results Tool via the Global Health Data Exchange (GHDx) at https://vizhub.healthdata.org/gbd-results/. Disease covariates (diarrhoea, malaria, and HIV cause fractions) and care variables (household air pollution, unsafe sanitation SEV, socio-demographic index, HiB3 and PCV3 vaccination coverage) were also obtained from GBD 2023 via GHDx. Dietary variables (staple protein supply for wheat, rice, maize, millet, sorghum and cassava; total animal protein; non-cereal plant protein; total kilocalories per capita per day) were obtained from FAOSTAT Food Balance Sheets (2010–) at https://www.fao.org/faostat/en/#data/FBS, averaged over 2019–2023. For nine countries where current FAOSTAT data were under revision, data from the 2018–2022 series were used. For two countries (Eritrea, Equatorial Guinea) where data were never compiled, values were imputed from their respective GBD regions. Statistical analysis was performed in R version 4.5.3 using beta regression (betareg package v3.2.4) with a logit link function. Four model specifications were tested for each of six staple crops, comprising a primary model and three pre-specified sensitivity analyses. Full analytical code is provided in the accompanying R script (Sanjeev_Meningitis_Analysis_Final.R).