Filariasis Cases Cluster Near Mosquito Breeding Grounds on Indonesian Island, Study Finds

On the Indonesian island of Belitung, a parasitic disease that public health campaigns were supposed to have pushed toward extinction is quietly persisting, and new research suggests the reason may be written on the landscape itself. A cross-sectional survey conducted in 2019 across four sub-villages of two lymphatic filariasis-endemic villages in Belitung Regency has revealed that human cases of the disease are not scattered randomly through the population but instead cluster tightly around bodies of water capable of hosting Mansonia mosquitoes, the vectors responsible for transmitting the parasitic worms that cause brugian filariasis. The findings, published in Acta Parasitologica, offer programme planners a quantitative, map-based rationale for targeting mosquito habitats alongside the mass drug administration campaigns that have long formed the backbone of elimination efforts.

Lymphatic filariasis is a disfiguring and disabling neglected tropical disease caused by thread-like nematode worms, in this region primarily Brugia malayi, whose larval stages circulate in human blood and are picked up by mosquitoes during a blood meal. Although Indonesia has invested heavily in mass drug administration, delivering annual rounds of antiparasitic drugs to entire communities to drive down the parasite reservoir, the disease remains endemic in several island communities, and Belitung District has previously been flagged as an area at risk of recrudescence after the cessation of treatment campaigns. Understanding why transmission lingers in some places despite years of drug pressure is one of the central puzzles of the global elimination programme, and spatial epidemiology has emerged as one of the most promising tools for solving it.

The research team, led by Santoso Santoso of the Baturaja Public Health Laboratory Center with collaborators from Indonesia’s National Research and Innovation Agency and partner institutions, focused on two villages: Suak Gual, in Selat Nasik Subdistrict, and Lasar, in Membalong Subdistrict, the latter encompassing the sub-villages of Batumana, Dudat, and Ulin. In total, 671 residents participated in the survey. Because the parasites that cause brugian filariasis exhibit nocturnal periodicity, meaning their microfilariae surge into the peripheral bloodstream at night when their mosquito vectors are active, blood collection had to take place after dark. Thick blood smears were prepared from each participant and examined microscopically for the presence of parasites, providing a direct measure of who was infected and who was not.

Crucially, the team did not stop at counting cases. Using handheld GPS receivers, they recorded the precise coordinates of every participant’s residence, regardless of infection status, and mapped every water body in the study area that displayed characteristics consistent with a potential Mansonia breeding habitat. This ecological detail matters because Mansonia mosquitoes have an unusual and demanding larval biology: unlike many culicine mosquitoes that lay eggs on open water surfaces, Mansonia larvae attach their breathing siphons to the roots and stems of floating aquatic plants, drawing oxygen directly from the plant’s aerenchyma tissue. This means their breeding sites are defined not simply by standing water but by the presence of specific macrophyte vegetation, typically in swamps, ponds, and slow-moving water bodies, a constraint that makes their habitats discrete, mappable features of the landscape rather than diffuse environmental conditions.

With cases and habitats both geolocated, the researchers deployed a battery of spatial statistics to test whether the distribution of infections deviated from randomness. The Clark-Evans test, a classic nearest-neighbour analysis, compares the average distance between each case and its nearest neighbouring case with the distance expected if the points were distributed at random across the same area. An R statistic below 1 indicates clustering, values near 1 indicate randomness, and values above 1 indicate a regular, evenly spaced pattern. The team also applied the cross-K function, which extends this logic to two point patterns at once, asking whether cases and habitats are more closely associated with each other than chance would predict, and buffer zone analysis, which measures how many cases fall within set distances of the nearest habitat.

The results were striking. In Suak Gual, the Clark-Evans test returned an R value of 0.380 with a p-value below 0.001, indicating extremely strong clustering of cases, and in Batumana the R value of 0.432 was similarly significant. These findings held up after Bonferroni correction, a conservative statistical adjustment that raises the bar for significance when multiple tests are performed, and they remained robust across all boundary-window widths the researchers tested, a sensitivity check that guards against the possibility that the apparent clustering is an artefact of how the study area edges were defined. Ulin showed a nominally significant R value of 0.654 with p equal to 0.026, but this did not survive the Bonferroni correction, which raised the threshold to p equal to 0.106, nor a narrower boundary window, and the authors accordingly report it as a weaker and less consistent finding rather than a confirmed cluster.

Dudat told a different story altogether. There, the R value of 0.856 with a p-value of 0.434 indicated no significant deviation from spatial randomness, and the researchers attribute this to the sub-village’s comparatively dense and widespread distribution of potential habitats. The logic is intuitive: where breeding sites are everywhere, no single location confers an elevated risk, and cases accumulate broadly rather than concentrating in hotspots. This contrast between sub-villages is itself informative, because it suggests that the spatial signature of filariasis transmission is not a fixed property of the parasite or the human population but a function of the local ecology, varying with the abundance and arrangement of vector habitats across the terrain.

Perhaps the most operationally consequential number in the study is this: 76.5 percent of cases, 26 of the 34 infected individuals identified, lived within 500 metres of the nearest potential Mansonia habitat. That figure provides a concrete, defensible radius for intervention planning. Rather than treating an entire endemic village as a uniform risk zone, programmes could concentrate larval source management, environmental modification of aquatic plant habitats, and intensified surveillance on households within a half-kilometre buffer of mapped breeding sites. The approach aligns with a growing body of literature on residual transmission, the stubborn persistence of infection after mass drug administration, which modelling studies suggest is strongly shaped by the spatial aggregation of vectors and hosts. Where transmission is clustered, targeted vector control can achieve disproportionate reductions in risk for a given budget.

The study’s authors are careful about the limits of their evidence. The water bodies they mapped were classified as potential Mansonia habitats on the basis of their physical and ecological characteristics, not on the basis of mosquito larvae actually collected from them, and no entomological confirmation of vector presence or parasite infection in mosquitoes was performed. The survey was also cross-sectional, capturing a single moment in time, so it cannot establish the direction of causality or track how transmission dynamics unfold across seasons. Night-time blood smears, while the standard field method for detecting nocturnally periodic microfilariae, are less sensitive than modern antibody and antigen assays, meaning some low-level infections may have been missed. The researchers explicitly state that direct entomological surveys are still needed to confirm vector presence and to test the effectiveness of spatially targeted control in practice.

Even with those caveats, the study adds Belitung to a growing list of settings, from American Samoa to Ghana to Sri Lanka, where spatial analysis has exposed the heterogeneous geography of filariasis persistence and demonstrated that post-treatment surveillance can be made smarter by mapping. For a disease on the cusp of global elimination, the last kilometres are the hardest, and the tools that matter most are those that tell programmes precisely where to look. On this Indonesian island, the map now points toward the swamps and ponds where aquatic plants anchor the mosquito larvae that keep the parasite’s life cycle turning, and toward the households within walking distance of them, where the next phase of the elimination campaign may find its most valuable targets.

Subject of Research: Spatial clustering of lymphatic filariasis cases relative to potential Mansonia mosquito breeding habitats in endemic villages of Belitung Regency, Indonesia

Article Title: Spatial Clustering of Lymphatic Filariasis in Relation to Potential Mansonia Mosquito Habitats in Endemic Communities of Belitung Regency, Indonesia

Article References: Santoso, S., Suryaningtyas, N. H., Wigati, R. A., Sukendra, D. M., Yunarko, R., Widiastuti, D., Oktarina, R., Hidayat, S., Nurcahyo, W., & Ernawan, B. (2026). Spatial Clustering of Lymphatic Filariasis in Relation to Potential Mansonia Mosquito Habitats in Endemic Communities of Belitung Regency, Indonesia. Acta Parasitologica, 71(5), Article 230. https://doi.org/10.1007/s11686-026-01414-x

Image Credits: AI Generated

DOI: 10.1007/s11686-026-01414-x

Keywords: lymphatic filariasis, Brugia malayi, Mansonia, spatial analysis, Clark-Evans test, cross-K function, vector breeding habitats, Belitung, Indonesia, mass drug administration, neglected tropical diseases, spatial epidemiology