Spatial and temporal variation of malaria incidence in children under 10 years in a pyrethroid-resistant vector area in southern Benin.

BACKGROUND

Spatial and temporal identification of malaria-endemic areas is a key component of vector-borne disease control. Strategies to target the most vulnerable populations, the periods of high transmission and the most affected geographical areas, should make vector-borne disease control and prevention programmes more cost-effective. The present study focuses on the spatial and temporal dynamics of malaria cases and the exogenous factors influencing the transmission in an area with pyrethroid-resistant mosquito vector populations.

METHODS

A prospective cohort study of 1806 children under 10 years of age was conducted over 20 months to assess the risk of malaria incidence in the Cove-Zagnanado-Ouinhi (CoZO) health zone located in southern Benin. Childhood malaria data were used to identify malaria hotspots according to months of follow-up using spatial scanning methods based on the Kulldoff algorithm. Stability scores were calculated by season to assess incidence heterogeneity. Incidence values by month were aggregated with meteorological data; and demographic data were merged to detect cross-correlation between incidence and meteorological variables. Generalized equation estimators were chosen for their ability to handle intra-group correlation, ensuring robust and interpretable results despite the complexity of the data to identify factors explaining the spatio-temporal heterogeneity of malaria incidence in the CoZO health zone.

RESULTS

Malaria incidence ranged from 1.41 (95% IC 0.96-2.08) to 13.91 (95% IC 12.22-15.84) cases per 100 child-months. Spatial heterogeneity in malaria transmission hotspots was observed over the study period, with relative risks ranging from 1.59 (p-value = 0.032) to 16.24 (p-value = 0.002). There was a significant negative association (correlation coefficient = - 0.56) between malaria incidence and temperature; and a slightly positive association (correlation coefficient = 0.58) between malaria incidence and rainfall. A significant association between malaria incidence with average house altitude (adjusted incidence rate ratio [aIRR] 1 (95% IC 0.99-1) P < 0.001), soil type aIRR 0.54 (0.39-0.75) p < 0.001 and temperature (incidence rate ratio [IRR] 0.69 (0.66-0.73) p < 0.001).

CONCLUSION

This study uses innovative technologies such as remote sensing and geographic information systems (GIS) to analyse the environmental, meteorological and geographical factors influencing malaria transmission, thereby identifying high-risk areas and associated factors. It demonstrates that these tools improve the accuracy of control strategies, while highlighting the crucial role of the environment and human behaviour, paving the way for more targeted interventions against malaria and other vector-borne diseases.