Skip to main content
. 2022 Jan 31;106(3):915–922. doi: 10.4269/ajtmh.21-0780

Table 2.

Sociodemographic factors associated enterotoxigenic Bacteroides fragilis infection in each of the eight study sites

Factors Incidence rate ratio of B. fragilis detection per 100 child-months (95% CI) P value
Female gender 0.96 (0.94–0.97) < 0.001
Duration of exclusive breastfeeding, mo 0.98 (0.96–0.99) < 0.001
LAZ at enrollment 0.98 (0.97–0.99) < 0.001
Maternal height, cm 1.0 (0.99–1.0) 0.352
Maternal age, y 1.0 (0.99–1.0) 0.004
Maternal education > 6 y 0.89 (0.88–0.91) < 0.001
Improved drinking water source 0.92 (0.90–0.95) < 0.001
Routine treatment of drinking water 1.0 (0.98–1.02) 0.732
Lack of improved sanitation 1.21 (1.18–1.25) < 0.001
Improved floor material in households 0.93 (0.90–0.95) < 0.001
More than 2 people living in a single room 1.04 (1.02–1.05) < 0.001
Mother has more than 3 living children 1.10 (1.07–1.11) < 0.001
Household ownership of chicken or cattle 1.15 (1.13–1.17) < 0.001

LAZ = length-for-age z-score. A Poisson regression model was used. The dependent variable was the number of infections during follow-up (1–24 months) and the offset variable was the log of the total number of follow-ups. All analyses were adjusted for different study sites and all variables included in the multivariable model.