Predicting Malaria Transmission Dynamics in Dangassa, Mali: A Novel Approach Using Functional Generalized Additive...

Area of Science:

• Epidemiology
• Environmental Health
• Biostatistics

Background:

• Mali is committed to malaria pre-elimination by 2030.
• Predictive modeling can enhance malaria prevention strategies.

Purpose of the Study:

• To develop and compare functional regression models for predicting malaria incidence based on historical meteorological data.
• To identify specific meteorological patterns associated with increased malaria risk.

Main Methods:

• Utilized a five-year dataset (2012-2017) from 1400 individuals in Dangassa, Mali.
• Employed Functional Generalized Spectral Additive Model (FGSAM), Functional Generalized Linear Model (FGLM), and Functional Generalized Kernel Additive Model (FGKAM).
• Analyzed rainfall, temperature, humidity, and wind speed over the 18 weeks preceding malaria incidence.

Main Results:

• Malaria incidence peaked 10-12 weeks after high humidity (>65%), consecutive rainfall, and low wind speed (<1.8 m/s).
• The FGLM demonstrated superior predictive performance compared to FGSAM and FGKAM.
• FGSAM offered a balance between flexibility and simplicity.

Conclusions:

• Specific meteorological conditions can serve as early indicators for potential malaria outbreaks.
• The developed models provide a valuable tool for proactive malaria prevention and control strategies in Mali.