Abstract
Common childhood diseases vary seasonally in Mali, much of the Sahel, and other parts of the world, yet patterns for multiple diseases have rarely been simultaneously described for extended periods at single locations. In this retrospective longitudinal (1996–2004) investigation, we studied the seasonality of malaria, acute respiratory infection and diarrhoea time-series in the district of Niono, Sahelian Mali. We extracted and analysed seasonal patterns from each time-series with the Multiplicative Holt–Winters and Wavelet Transform methods. Subsequently, we considered hypothetical scenarios where successful prevention and intervention measures reduced disease seasonality by 25 or 50% to assess the impact of health programmes on annual childhood morbidity. The results showed that all three disease time-series displayed remarkable seasonal stability. Malaria, acute respiratory infection and diarrhoea peaked in December, March (and September) and August, respectively. Finally, the annual childhood morbidity stemming from each disease diminished 7–26% in the considered hypothetical scenarios. We concluded that seasonality may assist with guiding the development of integrated seasonal disease calendars for programmatic child health promotion activities.
Acknowledgements
We are indebted to Moussa Dembele, who died prior to the writing of this manuscript and without whom this investigation would not have been possible. We also would like to thank R. Sakai, S. Karambe, C. Traoré, the Malaria Research & Training Center (FMPOS, Université de Bamako, Bamako, Mali) and Columbia University (New York, NY, USA) for their support. The Climate and Society Programme (Earth Institute, Columbia University, New York, NY, USA) partially funded data collection and travel expenses; Columbia University funded DCM; finally, Malaria Research & Training Center, FMPOS, Université de Bamako (Mali), supported data collection and analysis in Mali.