Search

Helicobacter pylori infection has been linked to a spectrum of gastroduodenal diseases of broad public health impact, yet the natural history of this frequently asymptomatic infection remains poorly understood. Evidence suggests that initial acquisition occurs primarily during childhood and may persist throughout life. The seroprevalence of H. pylori antibodies was examined in 365 primary schoolchildren aged 4-7 years in a low-income United States/Mexico border community from January to May 1996. Overall, 21% of the 365 children tested positive, with a significant monotonic decrease in seroprevalence by 1-year age intervals (36% in children aged 4 years, 24% in those aged 5 years, 20% in those aged 6 years, and 14% in those aged 7 years). The odds ratio for each 1-year age increase was 0.76 (95% confidence interval: 0.6, 1.0) after adjustment for relevant covariates. Given that H. pylori antibodies diminish after infection clears, this trend suggests that transient infection may be common in young children. In contrast, hepatitis A virus seroprevalence increased with age. There was a moderate association (odds ratio = 1.47, 95% confidence interval: 0.8, 2.9) of H. pylori with hepatitis A virus seroprevalence that weakened after adjustment for age and socioeconomic status (odds ratio = 1.26, 95% confidence interval: 0.6, 2.5). Follow-up studies are needed to clarify the natural history of Helicobacter pylori infection and identify predictors of initial acquisition, persistence, and recurrence.

Rates of Helicobacter pylori infection are traditionally higher in developing countries than in developed countries, but the specific reasons for these differences are not fully clear. While chronic diseases resulting from H. pylori are generally of adult onset, chronic infection usually begins in childhood. In this cross-sectional study (1998-2000), the authors compared prevalences of H. pylori infection among children under age 6 years on both sides of the Rio Grande. Participants included 264 children of women from low-income families who were receiving services at health clinics in Juarez, Mexico, or El Paso, Texas, from April 1998 through October 2000. Data were collected through personal interviews and serologic testing for H. pylori antibodies. The crude odds ratio for H. pylori prevalence among Mexican children as compared with US children was 3.94 (95% confidence interval: 1.72, 9.06). After adjustment for covariates, the odds ratio decreased to 1.70 (95% confidence interval: 0.64, 4.52). The adjustments that produced the greatest reduction in the odds ratio for location were those for household crowding and maternal education. This study identified specific factors that may explain geographic variation in H. pylori prevalence among children.

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, Narino, 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.

We concur overall with Taylor's presentation of what is known about the dynamics of Helicobacter pylori infection in childhood (1). but we would like to emphasize that many more questions than answers remain. Because H. pylori infection is not normally detectable at onset, information about incidence must be inferred from age-specific prevalence or repeated screening for prevalent infection over time (2). The infection does not appear to confer lasting immunity, and currently, there is no way to determine whether uninfected individuals have been infected in the past (3). Thus, the average duration of naturally acquired infections and the proportion that are self-limiting is not known for any population or age group. Few studies have examined the infection status of asymptomatic individuals over time. Prospective studies that observe infection status at frequent time intervals have been particularly rare; thus, detailed information regarding the natural history of the infection, peak age ranges for acquisition, and factors that may differentially influence acquisition, persistence, and recurrence is limited. Taylor speculates that antibiotics and characteristics of H. pylori strains may account for spontaneous elimination of infection in children. Emerging evidence suggests that factors that influence the ecology of the gastric niche or the immune response may also influence persistence. In recent laboratory studies, feeding H. pylori-infected mice algae meals with a high antioxidant concentration induced immune changes associated with reduced bacterial load (4). This finding complements an observed inverse association between consumption of fruits and vegetables and H. pylori prevalence in Colombian children (5). We would expect the age-related peaks in acquisition and elimination to vary considerably across popu