Helicobacter pylori colonization in early life

Helicobacter pylori infection is a major cause of upper gastrointestinal disease throughout the world. Colonization begins in childhood, although little is known about its age of onset, rate, or mode of colonization. Our aim was to identify the age of acquisition of H. pylori colonization in Gambian children. A cohort of 248 Gambian children aged 3 to 45 months was studied at intervals of 3 months for 2 years, using the 13C-urea breath test, specific IgM and specific IgG serology. The prevalence of positive breath tests rose from 19% at 3 months of age to 84% by age 30 months. Elevated specific IgG and IgM antibody levels were associated with positive breath tests, although there was discrepancy between breath test results and serology, particularly IgG serology, during the 1st year of life. Neither IgG nor IgM serology could be validated as reliable diagnostic tools for infant H. pylori colonization compared with the 13C-urea breath test. Reversion to negative breath test, in association with declining specific antibody levels, occurred in 48/248 (20%) of children. On the assumption that the 13C-urea breath test is a reliable index of H. pylori colonization, we conclude that the infection is extremely common from an early age in Gambian children. Transient colonization may occur. Previous studies relying on serodiagnosis may have significantly underestimated the true early prevalence of colonization in the developing world, where the target age for intervention studies is probably early infancy.     

Interpreting the 13C-urea breath test among a large population of young children from a developing country.

The 13C-urea breath test is a noninvasive tool for the diagnosis of gastric Helicobacter pylori infection. However, it has not been validated in young children from the developing world, where infection is very common. 13C urea breath tests were performed on 1532 occasions on 247 Gambian infants and children aged from 3 to 48 mo. The means and variances of the separate sub-populations of 13C enrichment results contained within the overall dataset were estimated by a Genstat procedure using the EM algorithm, thereby identifying a cut-off value to discriminate positive from negative results. To illustrate the appropriateness of this calculated cut-off value, 13C urea breath tests were performed upon a small group of 14 patients aged 6 to 28 mo undergoing diagnostic upper endoscopy. Fixed gastric antral biopsies were examined to identify H. pylori. Two subpopulations were identified within the large dataset. A cut-off value of 5.47 delta per thousand relative to Pee Dee Belemnite limestone above baseline at 30 min identified 95% of the normally distributed negative sub-population and 99.4% of the log normal distributed positive sub-population. Comparison with endoscopic data confirmed that this cut-off value was appropriate for this population, as 7/7 children without H. pylori on their gastric biopsies had negative urea breath tests, and 6/7 children with gastric H. pylori colonization had positive urea breath tests. These findings confirm the value of the urea breath test as a diagnostic tool in young children from developing countries. They also offer a way to calculate the most appropriate cut-off value for use in different populations and the likelihood that it will correctly assign any value into the appropriate sub-population, without the need for endoscopy.     

Helicobacter pylori colonization in infants in a peri-urban community in Karachi, Pakistan

BACKGROUND: The prevalence and incidence of Helicobacter pylori in children in Pakistan is not known. OBJECTIVES: To measure the prevalence and age of acquisition of Helicobacter pylori infection/colonization in infants in a peri-urban community in Karachi, Pakistan. SETTING: Field based epidemiologic study in a peri-urban community in Karachi, Pakistan. METHODS: Infants aged 1 to 3 months were recruited from a birth cohort from the community. C-urea breath test (C-UBT) was performed on recruitment, and the test was repeated at 2, 3, 6, and 9 months of age. RESULTS: One hundred forty-eight infants were recruited and had C-UBT on 319 occasions over a period of 2 years. Two hundred thirty-one=(72%) tests were positive: 80% (49/61) infants at 1 month of age, 79% (33/42) at 2 month of age, 76% (92/121) at 3 month of age, 58% (37/64) at 6 months of age, and 67% (20/30) at 9 months of age. CONCLUSIONS: The study reveals an early colonization/infection of infants and a high prevalence of Helicobacter pylori in a peri-urban community in Karachi, Pakistan.     

Effect of rapid diagnosis of influenza virus type a on the emergency department management of febrile infants and toddlers.

BACKGROUND: Evidence shows that the rapid detection of influenza using an enzyme-linked immunosorbent assay decreases antibiotic use in the treatment of pediatric patients. To our knowledge, the effect on other diagnostic testing in an emergency department (ED) has not been examined. OBJECTIVE: To determine the effect of rapid diagnosis of influenza virus type A on the clinical management of febrile infants and toddlers in a pediatric ED at an urban children's hospital. MATERIALS AND METHODS: A retrospective review of ED records from an electronic database was performed. All children 2 to 24 months of age, with a temperature higher than 39 degrees C who had a positive influenza virus type A test result using an enzyme-linked immunosorbent assay from November 1, 1998, through April 30, 2000 (n = 72), were included in this study. Two groups were compared-those who had positive test results reported before discharge from the ED (early diagnosis) and those who had positive test results after discharge (late diagnosis). RESULTS: Forty-seven patients (65%) were in the early diagnosis group and 25 (35%) in the late diagnosis group. The groups were similar for age, temperature, and triage category. Fewer patients in the early diagnosis group received ceftriaxone sodium compared with those in the late diagnosis group (2% vs 24%, P =.006); there were fewer urinalyses (2% vs 24%, P =.006) and complete blood cell counts performed (17% vs 44%, P =.02). CONCLUSIONS: Rapid confirmation of influenza virus type A infection seems to decrease ancillary tests and antibiotic use in febrile infants and toddlers in the ED. A prospective study with a larger group is needed to confirm these findings.     

Helicobacter pylori infection in children of Texas

BACKGROUND: Acquisition of the Helicobacter pylori infection usually occurs in childhood. The prevalence of infection differs among ethnic groups and in adults is inversely related to the socioeconomic status of the individual's family during childhood. This study investigates the seroprevalence of H. pylori infection in children of different ethnic groups in relation to socioeconomic class and investigates the prevalence of acute H. pylori infection among children who have had recent onset of abdominal pain. METHODS: Serum samples were collected from 797 children, aged 6 months to 18 years, of various socioeconomic and ethnic backgrounds, at a large urban children's hospital. H. pylori status was determined by an anti-H. pylori immunoglobulin (Ig)G enzyme-linked immunosorbent assay (ELISA) validated for pediatric use. To determine the prevalence of acute H. pylori infection, children brought to the emergency center with abdominal symptoms without diarrhea and overt signs of acute abdomen were evaluated with both serology and the 13C-urea breath test. Acute H. pylori was defined as a positive 13C-urea breath test result and negative IgG serology for H. pylori. RESULTS: The overall seroprevalence of H. pylori was 12.2% and increased with age (e.g., 8.3% at 6-11.9 months and 17.9% at 13 years). The prevalence was inversely related to socioeconomic status (6.6%, moderate to high vs. 15%, low socioeconomic status). The difference in seroprevalence among blacks (16.8%), Hispanics (13.3%), and whites (8.3%; P < 0.01) could be accounted for by differences in socioeconomic status. Eighteen percent of children who were evaluated at the emergency center for recent-onset abdominal pain had acute H. pylori infections. CONCLUSIONS: Socioeconomic status, not ethnic group, is the more important risk factor for acquisition of H. pylori infection during childhood. Acute H. pylori infection was a relatively common cause of recent-onset, nonsurgical abdominal pain.     

Antibody response to pneumococcal capsular polysaccharide vaccine in African children

Antibody titers to six pneumococcal polysaccharides were measured by enzyme-linked immunosorbent assay in Gambian children before and 1 month after vaccination with a 23-valent pneumococcal capsular polysaccharide vaccine at the ages of 2, 4, 6 or 9 months or at 5 to 10 years of age. IgG responses to type 1, 3 and 5 polysaccharides were seen in children in all age groups. Responses to type 19F and 23F polysaccharides were seen only in the oldest children. Few children of any age responded to immunization with type 6A pneumococcal polysaccharide. The immune response of Gambian infants to immunization with six components of a polyvalent pneumococcal polysaccharide vaccine was comparable with that described previously in Finnish infants and would be unlikely to provide substantial protection against invasive pneumococcal disease.     

Helicobacter pylori in Gambian children with chronic diarrhoea and malnutrition.

Infection with Helicobacter pylori (formerly Campylobacter pylori) was studied by measuring antibody titres to H pylori in Gambian children. Serological evidence of infection was found in 12 of 82 (15%) infants aged less than 20 months; this increased to 62 of 135 (46%) in those aged 40-60 months. Positive serology was found in 41 of 77 (53%) infants with chronic diarrhoea and malnutrition (mean age 19 months, range 5-36) compared with 18 of 70 (26%) of age matched healthy controls and nearly a quarter (12/49, 24%) of age matched undernourished (marasmic) subjects. These data show that infection with H pylori is common in the Gambia and that in infancy this infection is associated with chronic diarrhoea and malnutrition.     

Evaluation of stool antigen test for Helicobacter pylori infection in asymptomatic children from a developing country using 13C-urea breath test as a standard

OBJECTIVE: Prevalence of asymptomatic Helicobacter pylori infection is very high in infants and children in developing countries. C urea breath test (UBT) is a reliable non-invasive diagnostic test for H. pylori infection in children that avoids invasive endoscopy. We compared a newly introduced H. pylori stool antigen test (with a high sensitivity and specificity in symptomatic children) with UBT in asymptomatic children mostly 1-5 years old, from a population with a high prevalence of infection. METHOD: Eighty six asymptomatic children (42 boys and 44 girls) were tested for H. pylori infection using the UBT and a stool antigen test (HpSA) based on a sandwich enzyme immunoassay for antigen detection. RESULTS: Forty five of the eighty-six (52.3%) children tested positive for H. pylori using the breath test. In 34 of these forty-five children, H. pylori antigen was detected in stool (sensitivity = 75.6%, 95% CI = 63 to 88%). Of the 50 of 86 (58%) children positive by HpSA test, 34 were positive for breath test. Of the 41 children with negative UBT test 25 were negative for stool antigen test (specificity = 61%, 95% CI = 46 to 76%). CONCLUSION: The sensitivity and specificity of the new stool antigen test are lower in asymptomatic children with high H. pylori prevalence rate compared to those reported for children with gastrointestinal symptoms. Its usefulness is limited for diagnosis in an asymptomatic child with H. pylori infection.     

Meningococcal antibody titres in infants of women immunised with meningococcal polysaccharide vaccine during pregnancy.

Seventy five Gambian women were immunised with a single dose of a group A+group C meningococcal polysaccharide vaccine during the last trimester of pregnancy. IgG antibody titres were measured in mothers and in their infants by an enzyme-linked immunosorbent assay (ELISA). All women had a good response to vaccination and maternal antibodies were high at the time of delivery (23.2 micrograms/ml for group A antibodies and 14.3 micrograms/ml for group C antibodies). However, only a proportion of this antibody crossed the placenta; cord blood:maternal antibody ratios were 30% for group A antibody and 44% for group C antibody, respectively. Considerable variability in cord blood:maternal blood ratios was seen between individuals. This could not be related to age, parity, or ethnic group. Mean group A and group C cord blood:maternal blood ratios were lower in women with serological evidence of syphilis than in seronegative women, and diminished transfer of group A antibody was noted in women with active malarial infection of the placenta. Antibody titres declined rapidly in infants and by the age of 3-4 months these had reached control values. Maternal immunisation may give infants some protection against group A and group C meningococcal disease but only during the first few months of life.     

Helicobacter pylori infection and cytotoxic antigen associated gene "A" status in short children.

BACKGROUND: Helicobacter pylori is now an accepted gastroduodenal pathogen and is being investigated for possible implications in nongastroenterological conditions such as growth impairment. Subjects infected by cytotoxic Cag-A positive strains seem more likely to develop serious gastroduodenal diseases but the possible role of Cag-A positive strains in non gastroenterological diseases has not been fully investigated. OBJECTIVE: 1) To evaluate the prevalence of Helicobacter pylori infection and Cag-A positivity in short children compared to auxologically normal children. All the subjects were without gastro-intestinal symptoms and were not obese or significantly underweight. 2) To verify the reliability of the ELISA assay for H. pylori. SUBJECTS: H. pylori infection was assessed in 338 children, 182 auxologically normal and 156 short children, with and without deficiency in growth hormone, by the determination of specific IgG antibody. In 79 subjects (all seropositive and a random sample of seronegative children), 13C-urea breath test and cytotoxic Cag-A positive strains were examined. RESULTS: The overall seroprevalence of H. pylori infection by IgG antibody was 18/156 (11.5%) and 13/182 (7.1%) in short and auxologically normal children respectively. The 13C-urea breath test was positive in 29 children: 17 (10.9%) short and 12 (6.6%) auxologically normal. Western blotting documented infection by cytotoxic Cag-A positive strains in 12/17 (70.6%) and 8/12 (66.6%) of short and auxologically normal children respectively. None of the differences between the two groups were significant. CONCLUSIONS: 1) We found a similar prevalence of H. pylori infection and Cag-A positivity in two large pediatric populations of short or auxologically normal children. Therefore: 1) Our data did not confirm a role of H. pylori infection in short stature in children. 2) We found a high reliability of ELISA assay for the detection of IgG antibodies compared to breath test.     

Chlamydia trachomatis infection in infants: a prospective study.

In a prospective study of Chlamydia trachomatis infection in pregnancy, 198 mothers positive for chlamydial antigen were identified; the infants of 174 were followed for up to six months and C trachomatis was recovered in cell culture from 43 infants (25%). Conjunctivitis occurred in significantly more infants who were positive for C trachomatis (20 of 43, 47%) than in those who were negative (18 of 131, 14%). There were also significantly more lower respiratory tract infections among infants with positive cultures (six of 43, 14%, compared with three of 131, 2%). The chlamydial antigen enzyme linked immunosorbent assay (ELISA) was positive in 61 of 131 infants from whom C trachomatis was not recovered in cell culture. False positive results were usually associated with the isolation of Staphylococcus aureus from samples of pharyngeal aspirate. Our results confirm that C trachomatis infection is a common cause of neonatal conjunctivitis, and respiratory infection in the first few months of life, with an incidence of 8.2/1000 live births. Because the infection is easily treated by oral erythromycin, however, screening during pregnancy is not warranted.     

婴儿沙眼衣原体肺炎病原学和临床的研究

为证实婴儿肺炎中沙眼衣原体的感染，应用ＨｅＬａ－２２９细胞对４９例肺炎婴儿的鼻咽部分泌物进行沙眼衣原体（ＣＴ）培养分离，以特异的单克隆荧光抗体进行鉴定；同时应用酶联免疫吸附试验检测患儿血清中ＣＴ特异性ＩｇＭ。结果其中９例患儿（最小４天，最大８个月）的标本于荧光显微镜下检测到典型ＣＴ包涵体荧光颗粒、并于相应的血清中检测到高效价ＩｇＭ，检出率为１８．４％（９／４９），其中６个月以下婴儿为２３．５％（８／３４）。９例患儿病初均未采用红霉素治疗，病程迁延最长达４４天，平均２０天。经研究证实，ＣＴ确为婴儿肺炎的重要病原体，故对婴儿肺炎，尤其６个月以下婴儿肺炎，应高度注意ＣＴ感染的可能；早期确诊、及时应用ＣＴ敏感药物，方能取得较好疗效。     

Plasma Antibodies to Cow's Milk Are Increased by Early Weaning and Consumption of Unmodified Milk, but Production of Plasma IgA and IgM Cow's Milk Antibodies Is Stimulated Even during Exclusive Breast-Feeding

ABSTRACT. We measured levels of cow's milk-specific (CM) antibodies of immunoglobulin classes G, A and M by enzyme-linked immunosorbent assay in plasma of 198 healthy infants; a variable number of samples taken at birth and at ages of 2, 4, 6, 9, 12 and 28 months were available (altogether 765 samples). The rise in the level of IgG CM antibodies was highest and most rapid in infants exposed to CM formula before the age of 1 month. The level fell by 9 months, but rose again by 12 months. This second rise was attributed to the introduction of dairy milk. Partially breast-fed and fully weaned infants had similar levels of IgG CM antibodies. The levels of IgG CM antibodies were unaffected by the infants' own atopy, their heredity for atopy, and the umbilical serum level of IgG CM antibodies. IgA and IgM CM antibodies were absent at birth. Their levels increased similarly in exclusively breast-fed infants and infants fed CM formula. We conclude that plasma IgG antibodies to cow's milk are increased by early weaning and by consumption of unmodified cow's milk. Production of plasma IgA and IgM antibodies to cow's milk is stimulated even during exclusive breast-feeding.     

Contribution of the 13C-urea breath test to the detection of Helicobacter pylori gastritis in children.

Serology, 13C-urea breath test, histology, Campylobacter-like organism testing, and culture were performed in 95 consecutive children to evaluate the contribution of these tests to the detection of Helicobacter pylori infection. In analyses considering any combination of three positive tests as "gold standard" for diagnosing H pylori infection, 26 children were Helicobacter positive (27%), which is only one patient more than the number of children with only a positive culture. The accuracy of culture was excellent when "any combination of three positive tests" was used as the gold standard (sensitivity 96%, specificity 100%, positive predictive value 100% [false positivity 0%], negative predictive value 99% [false-negative results 1%]). The results of invasive and noninvasive tests were comparable. When culture was considered as "gold standard," the sensitivity of serology and 13C-urea breath test was 96%; the specificity was 96% and 93%, respectively; the positive predictive value was 89% and 83% (false-positive results in 11% and 17%); and the negative predictive value for both was 99% (false-negative results in 1%). It is concluded that culture can be used as gold standard, but that non-invasive tests such as serology and/or 13C-urea breath test can be used to diagnose H pylori infection in children, since each has at least 95% sensitivity and 92% specificity.     

Detection of Helicobacter pylori in the stools of newborn infants

To investigate the transmission route of Helicobacter pylori, stool specimens of 50 infants 3 days old were studied by a stool antigen immunoassay and polymerase chain reaction (PCR). By PCR, H. pylori DNA was detected in 15 neonates (30%). The stool antigen test was positive in only 1 neonate with a positive PCR. The detection rate of H. pylori DNA in neonates was significantly higher for those with mothers having urines positive for anti-H. pylori IgG antibody (60%) than for those mothers with negative urines (17%) (P < 0.01). A follow-up study was done 24 months later on 8 of the 15 infants with positive PCR results, including the infant with a positive stool antigen test. All infants were negative by both PCR and stool antigen test.     

Hepatitis C virus infection in infants whose mothers took street drugs intravenously.

To assess the risk of transmission of hepatitis C virus from mother to infant during pregnancy or at delivery, we measured the antibody to hepatitis C virus (anti-HCV) by an enzyme-linked immunosorbent assay (ELISA) and a recombinant immunoblot assay (RIBA) in serum from 43 infants whose mothers took illicit drugs intravenously. Passively transmitted maternal anti-HCV was detected in 17 (40%) of the 43 infants tested with the ELISA during the first 4 postnatal months. Ten of these initially seropositive infants were followed to 15 months of age or beyond; anti-HCV cleared from nine infants and persisted in one. Among 24 initially seronegative infants, three (12.5%) showed persistent anti-HCV at 6, 11, and 18 months of age, respectively. The remaining two infants were initially tested with ELISA at 6 and 15 months of age; both were transiently seropositive, but anti-HCV disappeared by 12 and 24 months of age, respectively. Among the 17 infants with maternal antibody, nine with ELISA reactions greater than 2.5 optical density units were reactive by RIBA: the eight with weaker reactivity by ELISA were nonreactive by RIBA. When serum samples from the four infants who showed persistent reactivity by ELISA were tested with RIBA, one reacted to both antigens displayed by RIBA (C-100 and 5-1-1), one reacted to the 5-1-1 antigen only, and two were nonreactive. Serum transaminase values were elevated in three of these four infants; all four were also infected with human immunodeficiency virus. The results indicate that vertically transmitted hepatitis C virus may be a cause of hepatitis in infants, especially those coinfected with human immunodeficiency virus. Neonates at risk of hepatitis C virus infection should be monitored beyond 12 months of age. The interpretation of tests for anti-HCV antibody during infancy requires further investigation.     

Influence of age on 13C-urea breath test results in children

BACKGROUND: The 13C-urea breath test for diagnosis of Helicobacter pylori infection has not been validated in infants and young children. The influence of age on the test results was studied by conventional validation against invasive methods and by mathematical estimation in a large pediatric population. METHODS: The breath test was performed in 1499 children aged 2 months to 18 years. After a fasting period of 4 hours or more, 75 mg 13C-urea was ingested with cold apple juice, breath samples were taken at baseline and at 15 and 30 minutes. The distribution of the natural logarithms of the delta-over baseline (DOB) values were calculated, and the optimal cutoff values between positive and negative test results and gray zones with a risk of misclassification more than 10% were determined for both time points. In a subgroup of 149 children results of the breath test were compared with concordant results of histology and rapid urease test; 53 of them were less than 6 years of age. RESULTS: Logarithmic results of 1499 breath tests revealed two normally distributed subgroups with minimal overlap. The calculated optimal cutoff values were 4.7/1000 at 15 minutes and 5.0/1000 at 30 minutes. At 30 minutes, only 2.6% of all results were in the calculated gray zone (2.6-6.5/1000). Age was negatively correlated to DOB values of both negative (r = -0.223) and positive results (r = -0.291; P < 0.001). Breath test-negative and -positive children 6 or less years of age had significantly higher mean DOB values (P < 0.02) and a larger proportion of results within the gray zone than older children. Compared with biopsy-based results, the least discrepancies occurred at a cutoff of 5.0/1000: 0 of 61 infected (sensitivity 100%) and 6 of 88 noninfected children. Because five of the false-positive results were obtained in children less than 6 years of age, specificity and positive predictive values were lower in this age group than in older patients (88.1% vs. 97.8% and 68.8% vs. 98.0%, respectively). CONCLUSIONS: Under the applied conditions, the 13C-urea breath test shows an excellent separation between positive and negative results. Because of some overlap and a strong age effect, definition of a gray zone appears more meaningful than a threshold value. Because infants and young children have a high risk for false-positive breath test results, the values for cutoff and gray zones may have to be adapted. Further validation studies against invasive methods are warranted in this age group.     

The prevalence of Helicobacter pylori serum antibodies in children with recurrent abdominal pain

As part of a large, prospective study we investigated the prevalence Helicobacter pylori serum antibodies in children with recurrent abdominal pain (RAP). All patients suffered from recurrent bouts of abdominal pain for at least 6 months and ranged in age from 6 to 12 years. H. pylori antibodies were detected using an enzyme-linked immunosorbent assay. The prevalence of H. pylori antibodies in the RAP group was compared to that of a control group which consisted predominantly of pre-operative children. None of the control group suffered or had suffered from RAP. Antibodies to H. pylori were found in 7 of 82 (8.5%) RAP patients and in 2 of 39 (5.1%) control children. The latter difference is not significant and suggests that RAP is only rarely caused in children by H. pylori infection.     

Helicobacter pylori, anemia, and iron deficiency: relationships explored among Alaska native children

BACKGROUND: Attempts to understand determinants of anemia and iron deficiency have led researchers to examine the role of Helicobacter pylori infection. The current study assessed determinants of anemia and iron deficiency, including H. pylori, in Alaska Native children. METHODS: In 1999, a population-based survey was conducted among 86 children (67% response rate), mean age of 43.7 months (standard deviation = 16.8 months). Samples of breath, stool, and venous blood were obtained from children for measures of anemia, iron deficiency, H. pylori, fecal blood loss, and current inflammation. Standardized interviews with parents provided information on demographics, illness, and intake of dietary iron, iron-absorption inhibitors, and enhancers. RESULTS: Of the 86 children studied, 17.4% were anemic and 38.6% were iron deficient. Forty-one percent of the cohort had H. pylori-specific IgG antibodies, 86% tested positive by the urea breath test (UBT), and 80% tested positive by the stool antigen test. Presence of H. pylori antibodies emerged as a significant risk factor for anemia and iron deficiency in adjusted analyses controlling for demographic factors, current inflammation, and antibiotic use. In contrast, children with positive UBT or stool antigen results were significantly less likely to have anemia or iron deficiency than those with negative results. CONCLUSIONS: Results from different measures of H. pylori may reflect different stages of infection. Relationships between H. pylori and anemia/iron deficiency may depend on the phase of infection measured, with serologic tests reflecting established H. pylori infection associated with anemia/iron deficiency, and UBT and stool antigen results reflecting an earlier stage of infection.     

Neonatal rotavirus vaccination with RIT 4237 bovine rotavirus vaccine: a preliminary report

We vaccinated 244 newborn infants orally with RIT 4237 bovine rotavirus vaccine or placebo and followed them serologically and clinically for 16 months. Initially 39 of the 119 (33%) vaccine recipients compared with 1 of the 120 placebo recipients seroconverted by enzyme-linked immunosorbent assay-immunoglobulin M. After the first winter rotavirus season, at 7 months of age 55% of the vaccinated infants and 37% of the unvaccinated infants were rotavirus-seropositive by enzyme-linked immunosorbent assay-immunoglobulin G (P less than 0.01, chi square test). At 12 months of age, after a low rotavirus prevalence season, 34% of the vaccinated children and 23% of the unvaccinated children remained seropositive. There were 14 confirmed episodes of rotavirus gastroenteritis in the vaccine group and 10 episodes in the placebo group during the first 16 months. However, only 1 of the episodes in the vaccine group was severe, 4 were moderately severe and 9 were mild, whereas 7 episodes in the placebo group were severe and 3 were moderately severe (P less than 0.001 between groups, Fisher's exact test). There was no clear correlation between vaccine-induced clinical protection and initial serologic response (enzyme-linked immunosorbent assay-immunoglobulin M) to vaccination, but during follow-up severe rotavirus gastroenteritis was more likely to occur in children with no serum rotavirus immunoglobulin G antibody at the time of infection. We conclude at the present stage that neonatal rotavirus vaccination with RIT 4237 vaccine gives no protection against rotavirus infection but appears to modify the severity of gastroenteritis.