Prevalence of Helicobacter Pylori Antibodies in Normal Children

Helicobacter pylori is a well-known cause of gastritis in children and adults1 and plays a major role in peptic ulcer disease.2 Recently H. pylori has been associated with an increased risk of gastric cancer.3, 4 Previous investigators have found an increasing prevalence of H. pylori infection with increasing age in asymptomatic people around the world.5, 6 Seroprevalence rates in the United States are consistently reported as approaching 60% in adults older than 60 years old.7 An earlier study on H. pylori in children between 3 and 20 years of age in the United States found an overall seroprevalence of 31%.8 This rate was higher than previously thought2 and rivaled seroprevalence rates of children in developing countries where prevalence approaches 70% by teenage years.9, 10 However, the children in that study8 were about to undergo surgical procedures for various medical complaints and may not be representative of normal healthy children. Also in that study sequential sera were not available to determine the actual time of seroconversion. The current study looks at multiple yearly sera from healthy children followed in pediatric well child clinics. The purpose of this study was to assess the age at which asymptomatic children become seropositive for H. pylori and to determine whether seroconversion could be correlated with illness. In addition this study examines whether acquisition of infection occurs uniformly over time. Because hepatitis A is transmitted by the fecal-oral route, we hoped that by comparing seroprevalence rates for the two organisms we might support this mode of transmission for H. pylori. Methods. Serum specimens were obtained from 200 children previously enrolled in one of a number of vaccine trials conducted at Vanderbilt University. Following protocols for other studies these children (between 2 and 11 years of age) had at least four consecutive yearly serum samples between 1980 and 1994. Apart from routine sampling 100 of the children participating in these studies were told to come to our institution for an evaluation and an additional blood drawing if they became ill at any time during the study period. Additionally information about parental education level, number of rooms in the house and family size was obtained for these enrollees. For all enrollees parents completed demographic information from the time of their enrollment in the initial vaccine trials. Information obtained included age, race and gender for all children enrolled. Children with any chronic medical complaints were excluded from the study. These yearly sera from healthy, asymptomatic children were tested for the presence of IgG against H. pylori with an enzyme-linked immunosorbent assay from a commercial kit (Pyloristat®; Biowhittaker, Inc.). This kit has previously been proved in children to be sensitive and specific for the detection of IgG against H. pylori.11 The most recent serum from each child was evaluated and if proved to contain anti-H. pylori IgG, consecutive yearly sera from that child were evaluated until the first IgG-negative serum was found. This was done to determine the age at which the seropositive children were initially infected. Sera from all 200 children were also tested for the presence of IgG antibody against hepatitis A antigen using enzyme-linked immunosorbent assays from a commercial kit (HAVAB EIA®; Abbott Laboratories). This test was performed on the same serum samples described above, and the results of the hepatitis A enzyme-linked immunosorbent assay were compared with the H. pylori results. The study was approved by the Institutional Review Board of Vanderbilt University Medical Center. Kaplan-Meier type survival curves were used to estimate probability of no seroconversion. One minus these estimated values was plotted to illustrate seroconversion rates by age. Time followed was the children's age if they were seronegative and the age at seroconversion if sera were available that indicated a change from seronegative to seropositive. In children whose last sera tested positive, seroconversion age was estimated by assuming a uniform distribution of seroconversion from age 1 year to their present age (i.e. (earliest serology age + 1)/2). Log rank tests were used to compare seroconversion curves between groups. Logistic regression analysis estimating probability of seroconversion indicated consistent results with the seroconversion curve comparisons. Fisher's exact test was used to compare tabulated data (e.g. hepatitis A vs. H. pylori seroconversion). SAS Statistical Software (SAS Institute Inc., Cary, NC) and S-Plus for Windows (StatSci, Seattle, WA) were used for statistical and graphical analysis. Results. The ages of our 200 patients ranged from 2 to 11 years with a mean age of 6.7 years. There were 95 girls and 105 boys; 33.5% of the children were black and 66.5% were white. The parents of one-half of our patients had less than a high school education. Figure 1 shows an increased seroprevalence of H. pylori with increasing age (P = 0.01). There was also a trend toward increased seroprevalence in children whose parents had no formal education beyond high school (P = 0.07). By 10 years of age 34% of children whose parents had less than a high school education were seropositive, whereas only 16% of children with more highly educated parents had seroconverted. There was no significant difference in rates of sero-prevalence between males and females or between black and white children. The overall seroprevalence of IgG against H. pylori in this population was 14%. Figure 1 illustrates the probability of seroconverting vs. age. By 5 years of age the estimated probability of being seropositive was 11% (95% confidence interval, 6 to 15%), and by 10 years the probability was 25% (95% confidence interval, 13 to 36%). Ten of 28 seropositive children were seropositive for H. pylori at the time their first serum was obtained so that their age at last negative serum was estimated by assuming a uniform distribution of age at seroconversion. However, a change in the distributional assumption did not affect the comparisons. Four sera from this group of 200 children were positive for hepatitis A antibody; 2 were also seropositive for H. pylori whereas 2 were seronegative for H. pylori (P = 0.095). Medical histories were available for 12 of the seropositive children. Except for minor childhood illnesses they had all been generally healthy before seroconversion. However, one child did have a gastrointestinal illness during the visit before his positive serology test. Other illnesses for which these children were evaluated included sore throat, conjunctivitis, upper respiratory tract illness, pharyngitis, chickenpox and otitis media. None of the children had symptoms of gastritis or ulcer disease. Discussion. This study showed a significant increase in the seroprevalence of H. pylori infection with increasing age. By checking yearly serum samples on all of our subjects we were able to assess the uniform nature of seroconversion. This result is consistent with other studies in this country as well as in the developing world.5, 8, 10 In fact by the time our healthy patient population reached 10 years of age, nearly 25% of them were infected. This is consistent with the data obtained by Fiedorek et al.8 in children about to undergo surgical procedures. In India where seroprevalence reaches 58% by 19 years of age,10 a large population is infected for a much longer period of time. This chronic infection with H. pylori is believed to contribute to the development of gastric carcinoma.12 Moreover countries such as Japan that have a higher seroprevalence of H. pylori have a higher seroprevalence of H. pylori have a higher rate of gastric cancer.4 We also found parental education level to be inversely related to seroprevalence. This relationship persisted after controlling for age. Since so many of our children came from families with low parental education, our total seroprevalence rate of 14% may be higher than in a more educated population. A previous seroprevalence study8 showed a significant inverse relationship between family income and seropositivity. Similarly adults from low socioeconomic groups have been shown to have a higher rate of H. pylori infection than more affluent adults.13 Children living with caretakers that have less than a high school education are likely to have a low family income and are more likely to be exposed to and infected with H. pylori. In an attempt to explain the route of transmission for H. pylori, we compared its seroprevalence rate with that of hepatitis A in our study population. Seroprevalence for hepatitis A was higher in those children also infected with H. pylori but, perhaps because of the sample size, this difference was not statistically significant. This trend does suggest a fecal-oral route of transmission for H. pylori similar to that of hepatitis A. Further evidence comes from a recent study in which H. pylori was isolated from feces of people with symptomatic infection.14 Others have detected H. pylori in saliva specimens from infected children suggesting another possible route of person-to-person transmission.15 None of the patients in our study had gastrointestinal symptoms accompanying their infection with the exception of one patient who had an episode of gastroenteritis that resolved before his seroconversion. In addition a large number of children will asymptomatically seroconvert in their school age years. Seroreversion is an uncommon event that has been shown to occur only after long periods of time or after treatment16 and was unlikely to have affected a significant number of our subjects. Therefore we speculate that to prevent infection and the resultant morbidity that occurs in adult life, it will be necessary to initiate preventive measures such as vaccination in the preschool years. Preventing H. pylori infection could substantially reduce the incidence of gastritis and peptic ulcer disease in adults. Because gastric infection with H. pylori is associated with an increased risk of gastric carcinoma,4 prevention of infection would presumably decrease gastric cancer rates as well. Acknowledgments. Work was supported in part by Contract NO1-AI-02645, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD, and performed in the National Institutes of Health-funded facilities of the Vanderbilt Clinical Research Center (5MO-IRR 90095). We thank M. J. Blaser, M.D., for his helpful suggestions and review of our data. Felice Adler-Shohet, M.D.; Pamela Palmer, B.A.; George Reed, Ph.D.; Kathryn Edwards, M.D. Division of Infectious Diseases Department of Pediatrics (FAS, PP, KE) Department of Preventive Medicine (GR) Vanderbilt University School of Medicine Nashville, TNFIG. 1: Probability of seroconverting with increasing age in our population......, 95% confidence intervals.