In 1992, the authors studied Helicobacter pylori infection and exposures relevant to person-to-person, waterborne, foodborne, and zoonotic transmission in a census sample of 684 2-9-year-old children in Aldana, Nariño, a rural community in the Colombian Andes. H. pylori prevalence, as determined by the 13C-urea breath test, was 69%, and prevalence increased from 53% in 2 year-olds to 87% in 9 year-olds. Beginning at 3 years of age, a higher percentage of males compared with females were infected. Odds ratios were estimated by multivariate logistic regression to control for mutual confounding by transmission-pathway proxy variables and socioeconomic indicators. Among transmission-pathway proxies, the strongest predictor of H. pylori status was the number of persons who lived in the home, with the number of children apparently being of greater importance than the number of adults. Swimming in rivers, streams, or pools increased the odds of infection, as did using streams as a drinking water source. Children who frequently consumed raw vegetables were more likely to have the infection, and children who had contact with sheep also had increased prevalence odds. Because the results did not implicate a single mode of transmission, the possibility of multiple pathways is indicated.

The results of this population-based study suggest that nutritional factors may play a role in determining susceptibility to H. pylori infection.