Eco-epidemiological analysis of malaria transmission in Goden village, Burkina Faso

Introduction. Malaria remains the leading parasitic disease globally and a major public health challenge in Africa. Burkina Faso is among the most affected countries, with intense and perennial transmission. Despite the large-scale implementation of vector control strategies, including the distribution of insecticide-treated nets (ITNs) since 2010, transmission remains high. The prolonged use of pyrethroid insecticides has led to physiological and behavioral resistance in Anopheles vectors, reducing the effectiveness of control efforts. Moreover, transmission— which is intrinsically heterogeneous—is further shaped by ecological and social factors that drive local variations in human–vector contact, weakening the protective effect of ITNs. Study site. The rural village of Goden, in Burkina Faso, represents an ideal study site where malaria, caused almost exclusively by Plasmodium falciparum, remains highly prevalent despite several large-scale ITNs distribution campaigns. Over more than a decade of ITNs implementation, extensive data have been collected on the main malaria vectors, Anopheles coluzzii and Anopheles arabiensis, including their abundance, insecticide resistance, infection rates, and behavioral traits. Research objectives. This Ph.D. research investigated the factors driving malaria transmission within the village of Goden by focusing on three complementary but interconnected aspects: (i) human–vector interactions and how behavioral patterns influence exposure; (ii) the dynamics of insecticide resistance in Anopheles populations; and (iii) the spatial distribution of vector abundance and transmission risk. Methods. Entomological collections were conducted in the village using standard methods, including human landing catches (HLC, in 2020) for 16 hours (16:00–08:00) both indoors and outdoors, and pyrethrum spray catches (PSC) indoors (in 2019 and 2020). In parallel, household surveys were carried out to collect data on daily activities and ITNs use. Species identification was performed using PCR to distinguish members of the Anopheles gambiae complex. Molecular detection of Plasmodium in thoraxes and abdomens, together with blood meal analysis, allowed for the estimation of the main entomological parameters of transmission (human blood index, HBI; infected human blood meals, IHBM; sporozoite rate, SR; oocyst rate, OR; entomological inoculation rate, EIR). Molecular analyses of mosquito samples collected during the longitudinal study in 2011, 2015, and 2020 enabled the identification of target-site mutations (knockdown resistance, kdr) associated with pyrethroid resistance. Statistical analyses were used to identify the main patterns of human–vector contact (generalized additive models, GAM), explore trends in mosquito abundance (generalized linear and mixed models, GLM and GLMM), and describe the spatial distribution of transmission parameters (spatial interpolation, IDW, Kernel density, and GAM and GAMLSS models). Results and discussion. By combining host-seeking mosquito collections (HLC) with surveys on population habits and ITNs use, critical gaps in protection were identified. Despite nearly universal ITNs coverage in the village (95%) and a decrease in SR compared to previous surveys (SR: 2% in 2020; SR: 6% in 2015), a high biting 1 pressure was observed (up to 20 Anopheles/person/hour). This suggests that while ITNs still provide partial individual protection, their lethal effect might be reduced, consistent with the high frequency of pyrethroid resistance detected in the village. The analysis of biting rhythms revealed an extension of biting activity for both vectors and reduced protection during the early evening (16:00–18:00) and morning (07:00–08:00) hours when people are awake and unprotected. The peak of transmission risk (00:00–04:00) coincided with the peak activity of A. coluzzii, the most abundant and ubiquitous vector (indoors and outdoors) and with a higher SR (2.7% vs 0.05% for A. arabiensis; χ² = 2.96; p < 0.05). The estimated EIR (6.7 infectious bites/person in 16 hours) was substantially reduced after adjusting for ITNs use and sleeping habits (1.2–1.9 infectious bites/person), equivalent to about 0.7 infectious bites per unprotected person. This highlighting that, even with high ITNs coverage and use, transmission persists due to temporal and spatial overlap between human and vector activities. Genetic analysis revealed a temporal increase in kdr resistance mutations (1014F, 1014S, 402L) and the presence of multi-resistant vectors. In A. coluzzii, the 1014F allele decreased while 402L increased (both around 50% in 2020), whereas in A. arabiensis, L1014S was replaced by L1014F. These contrasting trends indicate differences in selective pressures related to the ecological behavior of the two vectors. The PSC analysis (59 houses, in 2019) showed A. coluzzii to be strongly anthropophilic (HBI 53% vs. 13% in A. arabiensis, p < 0.05), more infectious (SR 12% vs. 4%, p < 0.05), and more abundant indoors, consistent with previous studies (PSC in 2011). The study revealed high parasite circulation within the human population (43% of human blood meals infected) and a heterogeneous distribution of entomological parameters (HBI, IHBM, SR). The highest HBI, IHBM, and SR values were found in the northwestern area of the village, where (i) the population was more exposed to mosquito bites, (ii) bites were more infectious, and (iii) mosquitoes were likely to obtain a higher number of infected human blood meals. Infected mosquitoes were more anthropophilic (31% vs. 4%, p < 0.05), suggesting a direct link between the human reservoir and increased biting rates on humans. Statistical models highlighted a strong effect of spatial coordinates on HBI and mosquito abundance, while environmental variables showed limited impact. The spatial component was further confirmed in preliminary analyses of PSC data from 2020, which revealed a potential link between mosquito abundance and proximity to water bodies, as well as possible effects of host availability on the number of blood-fed mosquitoes collected. Conclusions. Overall, the results demonstrate that malaria transmission in the village of Goden is driven by the interplay between human behavior, insecticide resistance, and spatial heterogeneity. Together, these factors sustain residual transmission despite the near-universal coverage of ITNs. Effective malaria control therefore requires strategies that go beyond mass ITNs distribution, simultaneously addressing human exposure and behavior, resistance management, and the spatial targeting of interventions. Potential measures include community education and outdoor protection, insecticide rotation, distribution of next-generation 2 nets, and targeted interventions in high-risk areas, such as larval habitat management. Continuous monitoring of vector behavior and ecology, resistance patterns and human exposure is essential to inform adaptive and context-specific control programs.