Integrating approaches to study land use change and hotspots of malaria transmission in rural Madagascar: an observational study

Changes in how land is used can, by altering the ecology of microhabitats suitable for the Anopheles mosquito vector, modify the potential for malaria infection in a region. This can in part explain why malaria transmission is spatiotemporally variable and why localised foci of a substantially higher burden than regional or national averages are often observed. However, numerous other factors also contribute to the non-random distribution of malaria infections, such as human demography and behaviour, immune history and current health status, and access to adequate health services. Therefore, we aimed to integrate ecological, demographic, and socioeconomic data to better understand the forces driving malaria transmission hotspots in five sample regions in Madagascar. This research can be applied to develop land use practices or interventions that better mitigate malaria transmission. Large, interdisciplinary datasets from two prospective cohort studies of 701 and 856 people and one cross-sectional study of 5598 people (both sexes and all ages) from a total of 31 localities in Madagascar were analysed. Clinical (including malaria infection outcome) and health and socioeconomic survey data for each individual participant were paired with mosquito vector habitat transects and remote-sensing geospatial data in hierarchical models. To better examine transmission patterns in the spatiotemporal distribution of malaria infections, genetic analyses using a panel of polymorphic genetic markers were done to examine genetic diversity of malaria parasites. We observed a non-random distribution of malaria infections at multiple spatial scales, providing evidence of hotspots in malaria transmission at the individual, household, and community levels. For example, malaria prevalence varied from 5% to 40% between sites within 15 km. Malaria prevalence in many of these rural communities in Madagascar was substantially higher than expectations based on the published national or regional estimates of about 5–9%. Genetic data, especially in the northeastern localities, were consistent with the presence of stable foci of high transmission in remote areas where the landscape is dominated by recent and ongoing conversion of forest to agricultural land, primarily for rice cultivation. Remote communities in Madagascar, communities at the frontier of the rapid ecological transformation that is happening globally, have a disproportionately higher malaria burden than regional and nation