Emerging trends of malaria-dengue geographical coupling in the Southeast Asia region

Sarita Azad, Píetro LióJournal of Vector Borne Diseases 2014 · DOI

TL;DR

This study investigates the complex interplay between malaria, dengue fever, and the El Niño-Southern Oscillation (ENSO) phenomenon in the Southeast Asia region (SEAR). Using advanced wavelet analysis, the researchers found that while dengue incidence exhibits clear periodic patterns, malaria's link to ENSO is less evident. However, a significant coupling between dengue and ENSO was identified, particularly in the last decade, suggesting potential for early warning systems.

What Problem Does This Paper Address?

Malaria and dengue fever are two of the most significant mosquito-borne diseases impacting public health and economic development in the Southeast Asia region (SEAR). Despite their prevalence, the intricate relationships between these diseases and their potential drivers, such as climate patterns like ENSO, remain incompletely understood. Specifically, the paper aims to:

  • Quantify the temporal coupling between malaria and dengue cases across SEAR.
  • Investigate the statistical association between malaria and dengue with ENSO fluctuations.
  • Identify geographical patterns and clustering of these disease dynamics within SEAR.

Understanding these relationships is crucial for developing effective, integrated disease control strategies in a region highly vulnerable to these epidemics.

Methods

The study employs two primary analytical techniques:

  • Wavelet Analysis: This is the core methodology used to analyze time-series data. Wavelet analysis is particularly adept at identifying patterns and periodicities in non-stationary data, meaning data whose statistical properties change over time. It allows researchers to examine how relationships between different time series (like disease incidence and climate indices) evolve across various time scales and frequencies. The paper specifically uses the Morlet wavelet and focuses on band-averaged wavelet power to assess cross-coherence between time series. This technique is well-suited for detecting transient cycles and their timing within epidemiological data.
  • Hierarchical Clustering: This statistical technique is used to group countries based on the similarity of their malaria and dengue case patterns over the study period. By clustering countries, the researchers aim to identify geographical groupings that share similar disease dynamics, providing insights into cross-country disease links and potential shared influencing factors.

Datasets

The study utilizes the following datasets:

  • Malaria and Dengue Case Data: Annual reported cases for malaria and dengue fever for SEAR countries from 1985 to 2009. Data were obtained from the World Health Organization (WHO) website. Notably, malaria cases from Maldives were excluded due to low reported incidence, and dengue cases from Korea were excluded as there were no reported cases.
  • El Niño-Southern Oscillation (ENSO) Data: Time series of Sea Surface Temperature (SST) anomalies, indicative of ENSO fluctuations, were obtained from the NOAA Climate Prediction Centre.

Key Insights and Ideas

  • Dengue's Periodic Nature: The study strongly suggests that dengue incidence in SEAR follows distinct periodic patterns, making it more predictable in its temporal fluctuations.
  • Malaria's Weaker ENSO Link: Unlike dengue, the direct periodic association between malaria outbreaks and ENSO in SEAR appears less pronounced.
  • Dengue-ENSO Coupling: A significant and statistically robust cross-coherence was found between dengue fever and ENSO, particularly within the 2-4 year wavelet band and notably in the last decade of the study period (1999-2009). This highlights ENSO as a potential predictor for dengue outbreaks.
  • Geographical Clustering of Disease Dynamics: Hierarchical clustering revealed distinct groupings of SEAR countries based on their malaria and dengue patterns. This demonstrates that similar disease dynamics can occur across different countries, suggesting shared underlying environmental or socioeconomic drivers.
  • Cross-Disease, Cross-Country Links: The study identified interesting correlations between malaria in one country and dengue in another, suggesting complex interdependencies that warrant further investigation.

Key Results

  • Overall Correlation: The overall Pearson correlation coefficient between malaria and dengue cases in SEAR was found to be -0.63, indicating an inverse relationship, with both time series appearing to be in anti-phase over the last decade.
  • Wavelet Cross-Coherence: A statistically significant cross-coherence (at 95% confidence level) was observed between dengue and ENSO in the 2-4 year wavelet band, particularly prominent from 1999-2002.
  • Country-Specific Correlations: While SEAR-wide correlations varied, individual countries showed significant positive correlations between malaria and dengue (e.g., Indonesia, Myanmar, Timor-Leste) or negative correlations in others.
  • Clustering Outcomes: Figure 5 illustrates four distinct clusters of SEAR countries based on their malaria and dengue case patterns, with Cluster 1 showing a homogeneous group for malaria cases in Nepal, Thailand, Sri Lanka, Bangladesh, Bhutan, and India.

Limitations and Future Work

  • Data Granularity: The study relies on annual data, which may mask finer temporal dynamics or short-term fluctuations.
  • Causality vs. Correlation: While wavelet analysis can suggest potential causality through cross-coherence, it does not definitively prove it. Further research is needed to establish causal links.
  • Socioeconomic and Environmental Factors: The paper acknowledges that while climate is a factor, a deeper investigation into the biology and ecology of vectors, as well as social and socioeconomic factors, is required to fully understand inter-country relationships and epidemic emergence.
  • Future Work: The authors suggest that the identified ENSO-dengue coupling could be leveraged for early warning systems for dengue in SEAR. Further detailed investigations into the specific factors driving the observed geographical clustering and cross-country disease links are also recommended.

Summary updated

Abstract

The present analysis suggests that the dengue incidence patterns in SEAR are periodic. There is not much evidence of malaria and ENSO having periodic association in the region; however, dengue fever and ENSO shows statistical significant cross-coherence in the 2-4 yr wavelet band and the results are statistically significant in the last decade. This study also provides statistical evidence of geographical clustering which quantitatively demonstrate the cross-country and cross-epidemic situations that exist across SEAR.
BibTeX
@article{azad2014emerging,
  author = {Sarita Azad and Píetro Lió},
  title = {{Emerging trends of malaria-dengue geographical coupling in the Southeast Asia region}},
  journal = {Journal of Vector Borne Diseases},
  year = {2014},
  doi = {10.4103/0972-9062.141707},
}
Figure