STATISTICAL COMPARISON OF THE GEOCHEMISTRY OF VOLCANIC BEDROCKS AND SOILS FROM PODOCONIOSIS-ASSOCIATED REGIONS ON THE AFRICAN CONTINENT

Podoconiosis is a disease causing swelling in the lower limbs attributed to minerals moving from the soil, through the skin, and into the lymphatic system. The disease is largely limited to rift zones around the world with the highest incidence in Africa. Although particulate silicon and aluminum have been suggested as causative agents of the disease, an association between the minerals and podoconiosis has been identified in only a few cases. The underlying bedrocks and soils from rift zones associated with podoconiosis have not been fully investigated with specific focus toward the disease. Statistical analyses have proven valuable for classifying the geological composition of both rocks and soils, with multivariate approaches revealing underlying components or combinations of components in complex data sets that might otherwise remain unidentified. We applied multivariate analyses to a compilation of bedrock samples from rift zones from the GEOROC database to look for statistical patterns of relative element abundance associated with podoconiosis-endemic regions.

Using both descriptive and multivariate statistics, we analyzed ten major oxides of the bedrock of four rift locations (Cape Verde Islands, East African Rift, Mid-African Rift, Red Sea Rift) on the African continent. The Hawaiian Islands were included as a control group since they are not known to have cases of podoconiosis. Descriptive analyses showed separation of each of the African locations into two geochemically distinct populations, one of which had no counterpart in the Hawaiian Islands. Principle component analysis also indicated a geochemically unique population within the African locations. Separation of the populations is most influenced by K₂O, MgO, Na₂O and SiO₂. We propose that the unique African population, which is high in alkali, magnesium and silicon, is responsible for the irritant particles associated with podoconiosis. We also propose that a comparison of soil samples from Ethiopia (a known podoconiosis hotspot) will show the same statistical separation of geochemical populations between podoconiosis-endemic and non-endemic locations.