Data and code from: Investigating the Yanomami malaria outbreak: Gold mining and malaria

The Yanomami, an Indigenous group from the Amazon, confront multifaceted challenges endangering their health and cultural integrity. Of immediate concern is the humanitarian crisis caused by surges in malaria amid increasing illegal gold mining in their territory. Leveraging satellite imagery and panel regression analyses, we quantified the effect of land use changes on malaria incidence on their land (2016-2023). We observed a ~300 % increase in malaria cases during this period, associated with increases in illegal gold mining. An increase of one standard deviation in gold mining is associated with a 20-46 % rise in malaria incidence one to two years later. We found that changes in forest areas significantly affect malaria rates: for every one standard deviation increase in the perimeter of forest edges, malaria cases rise by 55 %. Our findings highlight the major impact of illegal gold mining and the resulting fragmentation of forests on the high malaria burden experienced by the Yanomami.

Methods
Data

In December 2023, we obtained annual data on malaria cases in Brazil between January 2003 and September 2023 from the Brazilian Ministry of Health database provided by the Indigenous Special Secretary of Health (SIVEP-SESAI Malaria, Malaria em áreas indígenas table) and filtered the data to include only cases reported at the Yanomami Indigenous Sanitary Special District. Malaria data in the Yanomami territory is reported at the polo base level (i.e., health district subdivisions) and typically associated with a coordinate point indicating the likely site of infection (hereafter, ‘infection sites’). We selected the variables: year, infection site (i.e., coordinate point location associated with diagnosed malaria cases), parasite species, population, and polo base. In total, 28 polo bases reported malaria in 64 distinct infection sites. We filtered the malaria data (~480,000 malaria records) to only include cases among the Yanomami people (~120,000 malaria records). The population size at each polo base (i.e., Indigenous health division), dated from 2010 to 2023, was used to calculate malaria incidence. Although this approach might deflate incidence in polo bases containing multiple infection sites, it allows assessment of changes in incidence over time while maintaining the highest resolution on focal land use change, which is the primary aim of this research. For the years before 2010, estimates from 2010 were used to calculate incidence. Forest and mining cover data were obtained through the Google Earth Engine platform using the MapBiomas Brazil 9.0 database (https://brasil.mapbiomas.org). Climate data (i.e., annual mean temperature and total precipitation) was obtained from the Climate Research Unit (CRU) database.