Human behavior determinants of exposure to Anopheles vectors of malaria in Sumba, Indonesia

Abstract

Malaria vector control interventions in Sumba, Indonesia, have not been able to eliminate malaria. Human drivers of exposure to Anopheles bites were investigated as part of a larger clinical trial evaluating the impact of a spatial repellent product on malaria incidence. Human behavioral observations (HBOs) evaluating temporal and spatial presence, sleeping behaviors, and insecticide treated net (ITN) use, were collected parallel to entomological collections-indoor and outdoor human landing catches (HLCs), and house hold surveys. Data demonstrates that mosquito access to humans, enabled by structurally open houses, is evident by the similar entomological landing rates both inside and outside households. The presence of animals inside houses was associated with increased mosquito entry-however, the number of humans present inside houses was not related to increased mosquito landing. Analyzing mosquito landing rates with human behavior data enables the spatial and temporal estimation of exposure to Anopheles bites, accounting for intervention (ITN) presence and usage. Human behavior adjusted exposure to Anopheles bites was found to be highest in the early in the evening, but continued at lower levels throughout the night. Over the night, most exposure (53%) occurred when people were indoors and not under the protection of nets (asleep or awake) followed by exposure outside (44%). Characterized gaps in protection are outdoor exposure as well as exposure indoors-when awake, and when asleep and not using ITNs. Interestingly, in the primary trial, even though there was not a significant impact of the spatial repellent on vector biting rates by themselves (16%), when factoring in human behavior, there was approximately 28% less exposure in the intervention arm than in the placebo arm. The treated arm had less human behavior adjusted bites in all spaces evaluated though there was proportionally higher exposure indoors. This analysis points to the importance of using HBOs both towards understanding gaps in protection as well as how interventions are evaluated. To mitigate ongoing transmission, understanding context specific spatial and temporal exposure based on the interactions of vectors, humans and interventions would be vital for a directed evidence-based control or elimination strategy.

Fig 1. Primary study area.

Primary study area with treated clusters in orange and untreated clusters in blue. Diamonds indicate HLC/HBO sentinel houses. The inset indicates the location of the study site in Sumba Island, Nusa Tenggara Timur Province. Map source: Natural Earth (https://www.naturalearthdata.com/).

Fig 2. Indoor (A) and outdoor (B) HLRs by cluster over the course of the night.

Though there were mosquitoes present throughout the night, there was high heterogeneity in mosquito landing rates between clusters. The solid green and red line indicate average HLRs indoors and outdoors.

Fig 3

a. Human behavior proportions from all clusters. ●: Indoor-ITN, ♦: Indoor, ▲: Outdoor-awake and ■: Outdoor-sleep. b. Average human behaviors. Average human behaviors spent in different categories over the course of a night. c. Average human behavior per treated and untreated clusters. d. Stacked bar graph of the proportions of human behavior overlaid with entomological indoor and outdoor landing rates.

Fig 4. Human behavior adjusted.

4a. Human behavior adjusted temporal exposure and ITN protection over the night with proportional spatial exposure and ITN protection in the inset. ITNs reduce exposure by approximately 27%. 4b. Quantified temporal and spatial (inset) gaps in protection i.e. human behavior adjusted exposure alone, demonstrate that the majority of exposure (53%) occurs when people are indoor and not protected by ITNs.

Fig 5. Overall behavior adjusted exposure.

Overall behavior adjusted exposure over the course of a night for treated, untreated and all clusters together. Treated clusters demonstrate less exposure overall with key differences seen during times when people are indoors and not under ITN protection, that consistent with expectations from a spatial repellent product.