Association between Drinking-Water Source and Detected Microbiological Contamination in Selected Urban Communities of Akure Metropolis, Southwestern Nigeria: A Pilot Study

Abstract

Access to safe drinking water remains a public-health priority in urban Nigerian communities where households depend on groundwater and packaged water. Few studies have directly compared well, borehole and sachet-water sources within the same urban setting while accounting for physicochemical covariates. This quantitative laboratory-based cross-sectional pilot study assessed the association between drinking-water source and detected microbiological contamination in 45 purposively selected samples (15 well, 15 borehole and 15 sachet) from Oba-Ile, Oke-Ijebu and Aule, Akure Metropolis, Ondo State. Physicochemical parameters were measured with calibrated meters, while microbiological assessment used culture on Nutrient and MacConkey agar followed by Gram staining and microbial-load assessment. The contamination outcome was defined operationally from culture findings and indicator-organism isolation; the primary statistical analysis used the culture-based detection classification rather than the numerical microbial-load values. Under this operational definition, detected microbiological contamination occurred in 35/45 samples (77.8%), including 14/15 wells, 14/15 boreholes and 7/15 sachet samples. Among combined groundwater samples, 28/30 (93.3%) were classified as contaminated compared with 7/15 (46.7%) sachet samples. Groundwater had higher odds of detected contamination than sachet water (crude OR = 16.00, 95% CI: 2.76–92.72; Fisher’s exact p = 0.0009). Turbidity was higher among contaminated samples (4.40 vs 3.62 NTU; p = 0.043) and showed a weak positive correlation with the contamination outcome (ρ = 0.307, p = 0.040). In exploratory logistic regression adjusted for turbidity, groundwater remained associated with detected contamination (AOR = 11.42, 95% CI: 1.91–68.47; p = 0.008). Because the sample was purposive and pilot-sized, the estimates are intended to describe the sampled communities rather than city-wide prevalence or disease risk.