Geophagy as a risk factor for geohelminth infections: a longitudinal study of Kenyan primary schoolchildren.

Geophagy among primary schoolchildren and its impact on geohelminth infection were studied in western Kenya. In a cross-sectional survey, 204 children aged 10-18 years were interviewed about geophagy and examined for helminth infections (Ascaris lumbricoides, Trichuris trichiura, Schistosoma mansoni, hookworm). Children found infected were treated with albendazole (600 mg in a single dose) and/or praziquantel (40 mg/kg in a single dose). The children were re-examined after 4, 8 and 11 months. In the initial cross-sectional survey, 77% of the children ate soil daily and 48% of all soil samples were contaminated with eggs of A. lumbricoides (median egg count in contaminated samples was 3/2 g of soil; range 1-15). Twenty-nine children (14.2%) were infected with A. lumbricoides and 87 (42.6%) with T. trichiura. Significant associations between geophagy and infection intensity with A. lumbricoides and T. trichiura, but not S. mansoni or hookworm, were found. Re-infection with A. lumbricoides was twice as common among geophagous children as among non-geophagous children (27.4% vs. 12.0%; P = 0.030). The intensity of reinfections was higher in geophagous children (median 773 eggs/g vs. 95 eggs/g; P = 0.027). The relative risk for A. lumbricoides reinfection was 2.28 for geophagous children (95% confidence interval [95% CI] 1.02-5.11), and the fraction of reinfection attributable to geophagy was 56.0% (95% CI 1.7%-80.4%). There was a significant difference in T. trichiura infection intensity between geophagous and non-geophagous children (median no. of eggs/g 68 vs. 20; P = 0.049), but not in reinfection rates. No difference between the groups was seen in S. mansoni or hookworm reinfection rate or intensity, or in the families' socioeconomic or educational status. Geophagy was associated with an increased risk of reinfection with A. lumbricoides and possibly with T. trichiura. Neither family background nor infection with non-orally transmitted helminths was associated with geophagy, suggesting that this association was not due to confounding, but causal. Geophagy is therefore likely to be a source of ascariasis and possibly trichiuriasis among primary schoolchildren.