Multi-pronged surveillance to understand the spatiotemporal correlations among macaques, vectors and humans in Plasmodium knowlesi malaria transmission

TL;DR

This study explores how P. knowlesi malaria spreads among macaques, mosquitoes, and humans in Malaysia, showing that infected macaques can predict human infection risks.

Contribution

The study integrates multi-pronged surveillance data from three hosts to reveal spatiotemporal correlations in P. knowlesi transmission.

Findings

P. knowlesi-positive macaques correlate with historical human cases within 6 km and beyond.

Macaque infection rates in peridomestic-agriculture areas are higher than in urban areas for P. knowlesi, P. inui, and P. coatneyi.

Abstract

Plasmodium knowlesi, a non-human primate (NHP) malaria parasite, has become a major public health concern in Malaysia and is now the leading cause of human malaria infections in the country. The transmission of P. knowlesi involves a complex cycle among humans, non-human primates and vectors. Numerous studies have focused on these hosts individually, but comprehensive research that integrates field data from all three hosts is lacking. This study aims to integrate multi-pronged surveillance data from macaques, vectors and human blood samples to better understand the epidemiology of P. knowlesi malaria in Peninsular Malaysia.

Field sampling data (both previously published and unpublished) collected from humans, macaques and mosquito vectors by this research group in Peninsular Malaysia between 2019 and 2022 were integrated. The data collected for each host type within the same site and month were aggregated as a single sampling event. Partial correlations of outcomes between different host sampling sites were analysed by controlling for inter-host sampling site proximity and temporal difference. Spatiotemporal correlations were analysed between the sampling outcomes and historical human P. knowlesi malaria cases reported within defined distances (up to 20 km) from the sampling sites across different time lead windows (range from −12 to 12 months).

Partial correlation analysis, controlled for inter-host sampling-site spatial proximity and temporal difference, showed a statistically significant positive partial correlation between the proportion of field-sampled human P. knowlesi-positive cases and the average number of Anopheles Leucosphyrus-group mosquitoes sampled per night within a 10-km proximity constraint (rs = 0.228, P = 0.042). A consistently statistically significant positive correlation was found between the proportion of P. knowlesi-positive macaques and the number of historical human P. knowlesi cases reported in defined spatial proximity to macaque sampling sites, particularly within spatial radii of 6 km and beyond, across both backward and forward time leads. Other NHP malaria parasites, P. cynomolgi, P. inui, P. coatneyi and P. fieldi, exhibited heterogeneous patterns in macaques and vectors, particularly in terms of geographical distribution and mixed-species infection. The proportions of macaque samples positive for P. knowlesi, P. inui and P. coatneyi were statistically higher in the peridomestic–agriculture area as compared with the urban area.

A key finding from this study is that the proportion of P. knowlesi infection in macaques may serve as a useful proxy for persistent transmission in an area, potentially indicating increased risk of human infection in nearby communities. This highlights the value of wildlife surveillance in predicting and managing zoonotic malaria risk. Integrating insights from epidemiology, ecology, veterinary science and public health is essential for effectively controlling zoonotic diseases such as P. knowlesi malaria and reducing their impact on both human and animal populations.

The online version contains supplementary material available at 10.1186/s13071-025-07082-6.

Figures