Determinants of infection with Schistosomiasis haematobia using logistic regression

A population-based stratified random sample of 10,039 inhabitants of rural communities in Minya Governorate, Egypt, were evaluated for risk factors for Schistosoma haematobium infection using multivariate analysis. Data were obtained by personal interview recording demographics, information on exposure to canal water, history of infection, and other risk factors for infection and examining urine samples for S. haematobium ova. Logistic regression analysis was used to adjust for confounders while assessing the role of each risk factor for infection. Using logistic regression allowed detection of several confounders and interactions which influenced other independent variables. Differences in exposure patterns to canal water among age and gender subgroups explained only a small portion of the variation in infection rates, thus favoring the alternative explanation: development of age-acquired immunity. The association of age with reduced prevalence of S. haematobium was the only relationship increasing (odds ratio [OR] = 2.95- 4.30) with logistic regression. Male gender was a risk factor for infection but did not increase with logistic regression (OR = 2.33-2.03). The protective effects of education, only noted in schoolage children (OR = 0.59- 0.51), were believed to be due to a school-based screening and treatment program.