Copula geoadditive modelling of co-occurring anaemia and malaria among children aged 6–59 months in Ghana, The Gambia, Rwanda, and Uganda: a joint spatial analysis

Résumé

Anaemia and malaria frequently co-occur in sub-Saharan African children due to shared biological mechanisms and overlapping socio-economic risk factors, yet most studies model each condition in isolation. This study applied a bivariate Frank copula geoadditive regression framework, implemented using the Generalised Joint Regression Modelling (GJRM) package in R, to jointly model anaemia and malaria in 17,208 children aged 6–59 months across Ghana, The Gambia, Rwanda, and Uganda using nationally representative Demographic and Health Survey data. The bivariate model incorporated logit link functions for both marginal outcomes, second-order P-spline smooth functions for child age, and structured and unstructured spatial random effects at the administrative region level across 38 units. Overall anaemia prevalence was 49.3% and malaria prevalence was 12.0%, with 8.9% of children experiencing both conditions simultaneously. The Frank copula confirmed moderate positive dependence between outcomes (Kendall’s τ = 0.185, p < 0.001), demonstrating that children with one condition face substantially elevated risk of the other beyond shared risk factors. Key determinants included female sex (protective for anaemia, OR = 0.86), rural residence (80% higher malaria odds, OR = 1.80, 95% CI 1.55–2.08), household poverty, large household size, and altitude. Wealth demonstrated a strong dose-response gradient, with the richest quintile experiencing up to 65% lower malaria odds relative to the poorest. Joint probability maps identified northern and eastern Uganda as bearing the greatest co-occurrence burden (20%–40% probability), while The Gambia and Rwanda demonstrated persistent anaemia despite near-elimination of malaria. These findings confirm that malaria control alone is insufficient to resolve anaemia in low-transmission settings and support geographically differentiated, outcome-specific intervention strategies for integrated epidemiological analysis of co-occurring childhood conditions.

Citer ce document

Accès au document

Texte intégral en lecture en ligne, réservé aux abonnés SPHAERO et aux membres de l'institution. Se connecter

Auteur(s)

Statistiques

Consultations : 1

Téléchargements : 0