Passaro, Douglas J.*; Taylor, David N.†; Gilman, Robert H.‡§; Cabrera, Lilia‡; Parsonnet, Julie*
Chronic infection with Helicobacter pylori is associated with gastric ulcers (1), duodenal ulcers (2), gastric cancer (3), and gastric lymphoma (4). Although diseases associated with H. pylori usually occur in adults, most H. pylori infections are acquired during childhood. The health effects of childhood infection are poorly understood and are difficult to study. Follow-up of patients whose infection onset date could be estimated suggests that acute infection is accompanied by upper gastrointestinal distress (5,6) and temporary hypochlorhydria (6–8). However, these findings are inconsistent; only one well-documented pediatric infection acquisition has been reported (9).
A cohort study in Peru demonstrated that children who developed an incident H. pylori infection had more diarrheal disease in the year after infection than either uninfected or persistently infected children; this effect was most marked in the first 2 months after H. pylori acquisition (10). To determine whether growth rates were also affected by incident H. pylori infection, we performed a nested case-control study using this cohort as our sampling base.
This study was conducted in Pampas de San Juan de Miraflores (Pampas), a community 10 km south of Lima, from January 1, 1995, through September 30, 1997. Of its approximately 40,000 inhabitants (11), 45% are aged 18 years or younger (1995 local census). Most residences are huts with sheet metal roofing. Community children aged 3 years or younger have a mean of six episodes of diarrhea yearly (12). Three population-based serostudies of childhood enteric disease have been undertaken in Pampas since 1995. Pooling subjects—one child, aged 6 months to 12 years, per consenting household—from these three studies yielded a cohort of 345 children (10). Cases were selected from the 32 children (9%) who developed serologic evidence of H. pylori infection during the study period. The 32 controls for the matched-pair analysis were selected from the 259 children (75% of cohort) whose results were persistently seronegative during the study period. Children who had persistently positive or borderline serology test results or who seroreverted were excluded from analysis.
Baseline demographic information and serum specimens were collected from each child. Health interviews, including information about diarrheal disease, were conducted daily. Sera were collected from children every 4 months. Body measurements were recorded monthly on the day of the child's birthday by one of two field workers trained in anthropometrics. Weight (to the nearest 50 g) was determined with spring scales. Height (to the nearest 1 mm) was determined with a locally made wooden platform and a sliding footboard. Field workers measured recumbent length in children younger than 2 years and standing height in children aged 2 years and older. Children with less than 7 anthropometric measurements were excluded from analysis. Informed consent was obtained from the guardian of each participant.
H. pylori Antibody Testing
We tested sera for H. pylori–specific immunoglobulin G using an in-house enzyme-linked immunosorbent assay (4) standardized using control sera from Peruvian adults and children (10). An optical density twice the mean of negative controls indicated H. pylori infection. Infants younger than 6 months were excluded to avoid interference from maternal immunoglobulin G antibody.
Seroconversion was defined as a change from negative to positive between consecutive sera with a 50% or greater increase in optical density (10,13). We based our estimate of the time required to develop H. pylori–specific immunoglobulin G on published assessments of acute infections in three adults (8,14,15) and one child (9). Seroconversion in the child occurred between 5 and 10 weeks after infection, which is consistent with the data from adult infections. We then estimated that, in children, H. pylori immunoglobulin G seroconversion requires 7 weeks (i.e., between 5 and 10 weeks) to occur. The onset date of H. pylori infection was therefore defined as the midpoint between each child's last negative and first positive serum test result, minus 7 weeks.
All anthropometric measurements were smoothed using 3-month moving medians (16). The 2-month growth velocity (percentage gains in height or weight) of each seroconverter was compared with the growth velocities of H. pylori–uninfected controls. Because socioeconomic status is an important independent risk factor both for H. pylori infection and for growth delay, we matched children by size. Children matched for gender and age (within 1 month) were ranked by baseline size (either height or weight, depending on the growth outcome being analyzed). The child closest in baseline size to the case was selected as the control. If two children were of identical (tied) size, the control was selected by coin toss. Height and weight gains were quantified by calculating a “growth index” for each matched pair. The growth index was the percent change in height or weight for the case divided by the percent change in height or weight for the corresponding control. Growth indices for matched pairs were assessed using the sign test. A generalized estimating equation was used to confirm that the relation between growth index and age was adequately represented by a linear model (17); this relation was then evaluated by categorizing age as 0–1, 2, or 3–10 years and applying the χ2 test for linear trend (EpiInfo 6.0, Atlanta, GA) (18). Multivariate regression analysis was used to adjust for the separate effects of sex and age on growth indices.
To determine whether diarrhea was associated with observed effects on growth velocity, diarrhea prevalence rates were calculated for cases and controls. Prevalent diarrhea was defined as a day in which the child's caretaker felt that the child had more or looser stools than normal. Diarrhea prevalence was compared among cases and matched controls using the sign test. All P values are two-tailed.
Twenty-six children with adequate follow-up developed an incident H. pylori infection during the study period (six children had been excluded because of infrequent anthropometric analysis). The median age at seroconversion was 2.4 years (interquartile range, 1.4–5.4 years); 15 (58%) were girls (Fig. 1).
In the two months after H. pylori infection, these 26 seroconverters gained a median of 24% less weight (interquartile range, 63% less to 21% more) than matched controls. H. pylori infection was associated with attenuated weight gain only in children aged 2 years or older (Fig. 1); among children aged 2 years or older, cases gained less weight than matched median controls in 14 (78%) of 18 matched pairs (P = 0.03). We reanalyzed this data using narrower age groups to demonstrate that weight indices decreased progressively with age: among children younger than 2 years, cases gained less weight than median controls in 3 (38%) of 8 matched pairs; among children aged 2 years, cases gained less weight than median controls in 6 (67%) of 9 matched pairs; and among children aged older than 2 years, cases gained less weight than median controls in 8 (89%) of 9 matched pairs (χ2 for trend, 4.7;P = 0.03).
Weight gain was significantly attenuated in girls (13 of 15 matched pairs;P = 0.007) but not boys (4 of 11 matched pairs;P = 0.5). However, since girls tended to seroconvert at an older age than boys, it was not clear whether this differential response to acute H. pylori infection represented a gender effect or an age effect. Regression analysis demonstrated that older age (15% lower growth index per year of age; 95% confidence interval, −18% to −6%) but not female sex (31% lower growth index among girls vs. boys; 95% confidence interval, −64% to +25%) was independently associated with growth attenuation.
Cases gained 31% less height (interquartile range, 58% less to 42% more) than controls. This relation was not consistent (10 of 16 pairs;P = 0.5) and did not vary by age or sex. Cases had more prevalent diarrhea than controls in 7 (64%) of 11 matched pairs (P = 0.5); in 15 other pairs, information was incomplete or no diarrhea was noted. This relation did not vary by age or sex, but no diarrhea was noted by either cases or controls older than 2 years during the study period.
This study suggests that acute H. pylori infection may lead to decreased weight gain soon after infection, and that this effect is seen in children aged 2 years or older. Another study, conducted in Pampas 6 years earlier, also found temporary growth faltering associated with the acute onset of a chronic gastrointestinal infection (cryptosporidiosis) (19,20). Reasons for this finding merit additional study. One theory is that acute H. pylori infection is accompanied by an increase in diarrhea as a direct result of H. pylori (21) or gastric hypochlorhydria with an attendant vulnerability to other enteric pathogens (6,22–31). However, three cohort studies exploring the association between H. pylori infection and diarrhea have yielded conflicting results (10,32,33). Furthermore, while 7 of 11 matched pairs had more days of diarrhea than controls, diarrhea was limited to children aged 2 years or younger, in whom the effect of H. pylori on growth was minimal.
Another theory to explain the decreased weight gain is the development of a systemic inflammatory response to acute H. pylori infection. Persons who develop H. pylori infection have high tissue and serum concentrations of the proinflammatory cytokines interleukin-1β, interleukin-8, and perhaps interleukin-6 (34–39). Furthermore, there is evidence from animal models and humans that surface antigens of H. pylori, such as Lewis blood group antigens, undergo significant alteration in the first year of infection (40–43). The organism might induce a brisk immune response on first infection before settling into the more familiar “quiescent” state associated with the organism's long-term survival in the human stomach.
A Gambian study found significant growth retardation in the first 9 months of life among infants infected by H. pylori (44). However, ours is the first to investigate growth in older children with incident H. pylori infection. Previous investigations of children using prevalent H. pylori infection have yielded mixed results. The only follow-up study of children with prevalent H. pylori infection demonstrated a marked decrease in growth among 7- to 11-year-old girls, but not boys, infected with H. pylori (45). Several cross-sectional studies have shown that H. pylori infection either had no effect on anthropometric indices (46) or a modest effect that did not remain after adjusting for socioeconomic status (47,48).
However, two other cross-sectional studies have indicated that the relation between H. pylori and growth retardation was age-specific. In an Italian study, H. pylori was associated with decreased height-for-age among children aged 8.5 to 14 years, but not among those aged 3 to 8.5 years (49); a population-based study performed in Columbia demonstrated an inverse relation between H. pylori infection and height-for-age in older (but not younger) children (50). The regression analysis from our study suggests that age at infection acquisition is more important than gender in guiding the host response to the infection. Perhaps the acute phase of H. pylori infection is better adapted to immature immune systems. However, our study does not rule out the presence of a gender effect. Furthermore, it is not clear why girls should seroconvert later than boys; this finding warrants confirmation in larger studies of H. pylori acquisition.
The power of this study is limited by the small number of children who seroconverted. We did not detect a significant change in height velocity during the study period; short-term height effects tend to be smaller than short-term weight effects. Because acute H. pylori infection is not well characterized clinically, identifying a large group of acutely infected children is difficult. In future studies, efficiency could be increased by using more sensitive methods for detecting pediatric H. pylori infection, such as the stool antigen or the urea breath test (51–53).
H. pylori infection may be associated with diarrheal disease, anemia, and growth retardation in children (54). Since H. pylori infects nearly 50% of the world's population, and since H. pylori is usually acquired during childhood (55), the health effects of acute infection in children are an important area for research.
William Checkley provided expert suggestion regarding data management and analysis.
1. Graham DY, Lew GM, Klein PD, et al. Effect of treatment of Helicobacter pylori infection on the long-term recurrence of gastric or duodenal ulcer: A randomized, controlled study. Ann Intern Med 1992; 116:705–8.
2. Marshall BJ, Goodwin CS, Warren JR, et al. Prospective double-blind trial of duodenal ulcer relapse after eradication of Helicobacter pylori. Lancet 1988; 2:1437–42.
3. Parsonnet J, Friedman GD, Vandersteen DP, et al. Helicobacter pylori infection and the risk of gastric carcinoma. N Engl J Med 1991; 325:1127–31.
4. Parsonnet J, Hansen S, Rodriguez L, et al. Helicobacter pylori infection and gastric lymphoma. N Engl J Med 1994; 330:1267–71.
5. Ramsey EJ, Carey VK, Peterson WL, et al. Epidemic gastritis with hypochlorhydria. Gastroenterology 1979; 76:1449–57.
6. Morris A, Nicholson G. Experimental and accidental C. pylori infection in humans. In: Blaser MJ, ed. Campylobacter pylori in gastritis and peptic ulcer disease. New York: Igaku-Shoin, 1989, pp 61–72.
7. Morris A, Nicholson G. Ingestion of Campylobacter pyloridis causes gastritis and raised fasting gastric pH. Am J Gastroenterol 1987; 82:192–99.
8. Frommer DJ, Carrick J, Lee A, Hazell SL. Acute presentation of Campylobacter pylori gastritis. Am J Gastroenterol 1988; 83:1168–70.
9. Mitchell JD, Mitchell HM, Tobias VT. Acute Helicobacter pylori infection in an infant, associated with gastric ulceration and serological evidence of intra-familial transmission. Am J Gastroenterol 1992; 87:382–86.
10. Passaro DJ, Taylor DN, Meza R, Cabrera L, Gilman RH, Parsonnet J. Acute Helicobacter pylori infection is associated with diarrheal disease in Peruvian children. Pediatrics 2001; 108:E87.
11. Vigilancia Nutricional: Encuesta de Consumo de Alimentos, Pampas de San Juan. Lima: PRISMA, 1994.
12. Oberhelman RA, Gilman RH, Sheen P, et al. A placebo-controlled trial of Lactobacillus GC to prevent diarrhea in undernourished Peruvian children. J Pediatr 1999; 134:15–20.
13. Taylor DN, Sanchez JL, Smoak BL, DeFraites R. Helicobacter pylori infection in Desert Storm troops. Clin Infect Dis 1997; 25:979–82.
14. Sobala GM, Crabtree JE, Dixon MF, et al. Acute Helicobacter pylori infection: Clinical features, local and systemic immune response, gastric mucosal histology, and gastric juice ascorbic acid concentrations. Gut 1991; 32:1415–18.
15. Morris AJ, Ali MR, Nicholson GI, Perez-Perez GI, Blaser MJ. Long-term follow-up of voluntary ingestion of Helicobacter pylori. Ann Intern Med 1991; 114:662–63.
16. Selvin S. Statistical Analysis of Epidemiological Data. 2nd ed. New York: Oxford University Press, 1996, p. 24.
17. Hastie TJ, Tibshirani RJ. Generalized Additive odels. New York: Chapman and Hall, 1990.
18. Schlesselman JJ. Case-Control Studies: Design, Conduct, Analysis. New York: Oxford University Press, 1982.
19. Checkley W, Epstein LD, Gilman RH, Black RE, Cabrera L, Sterling CR. Effects of Cryptosporidium parvum infection in Peruvian children: Growth faltering and subsequent catch-up growth. Am J Epidemiol 1998; 148:497–506.
20. Checkley W, Gilman RH, Epstein LD, et al. Asymptomatic and symptomatic cryptosporidiosis: Their acute effect on weight gain in Peruvian children. Am J Epidemiol 1997; 145:156–63.
21. Guarino A, Bisceglia M, Canani RB, et al. Enterotoxic effect of the vacuolating toxin produced by Helicobacter pylori in Caco-2 cells. J Infect Dis 1998; 178:1373–78.
22. Evans CA, Gilman RH, Rabbani GH, Salazar G, Akbar A. Gastric acid secretion and enteric infection in Bangladesh. Trans Royal Soc Trop Med Hyg 1997; 91:681–85.
23. Cook CG. Infective gastroenteritis and its relationship to reduced gastric acidity. Scand J Gastroenterol 1985; 111:17–23.
24. Larner AJ, Hamilton MI. Review article: Infective complications of therapeutic gastric acid inhibition. Alimentary Pharmacol Ther 1994; 8:579–84.
25. Peterson WL, Mackowiak PA, Barnett CC, et al. The human gastric bactericidal barrier: Mechanisms of action, relative antibacterial activity, and dietary influences. J Infect Dis 1989; 159:979–83.
26. Jones DGC, Langman MJS, Lawson DH, Vessey MP. Post-marketing surveillance of the safety of cimetidine: Twelve month morbidity report. Q J Med 1985; 54:253–68.
27. Sack GH, Hennessey RN, Mitra RC, et al. Gastric acidity in cholera. Clin Res 1970; 18:682.
28. Giannela R, Broitman SA, Zamcheck N. Salmonella enteritidis: I. Role of reduced gastric secretion in pathogenesis. Am J Dig Dis 1971; 16:1000–06.
29. Waddell WR, Kunz LJ. Association of Salmonella enteritis with operations on the stomach. N Engl J Med 1956; 255:555–59.
30. Coremans G, Harlet L, Van Cutsem E, Vantrappen G. Gastric acid suppression increases the risk of Salmonella infection and septicemia [abstract]. Gastroenterology 1993(suppl); 104:A686.
31. Black RE, Levine M, Clements ML, Hughes TP, Blaser MJ. Experimental Campylobacter jejuni infection in humans. J Infect Dis 1988; 157:472–79.
32. Isenbarger DW, Bodhidatta L, Hoge CW, et al. Prospective study of the incidence of diarrheal disease and Helicobacter pylori infection among children in an orphanage in Thailand. Am J Trop Med Hyg 1988; 59:796–800.
33. Rahman MM, Mahlanabis D, Sarker SA, et al. Helicobacter pylori colonization in infants and young children is not necessarily associated with diarrhea. J Trop Pediatr 1998; 44:283–87.
34. Yakabi K, Ro S, Okazaki R, et al. Water extract of Helicobacter pylori stimulates interleukin-8 secretion by a human gastric epithelial cell line (JR-St) through protein tyrosine phosphorylation. J Gastroenterol Hepatol 2000; 15:263–70.
35. Yamaoka Y, Kodama T, Kita M, Imanishi J, Kashima K, Graham DY. Relation between clinical presentation, Helicobacter pylori density, interleukin 1β and 8 production, and cagA status. Gut 1999; 45:804–11.
36. Harris PR, Smythies LE, Smith PD, Dubois A. Inflammatory cytokine mRNA expression during early and persistent Helicobacter pylori infection in nonhuman primates. J Infect Dis 2000; 181:783–86.
37. Kim JS, Jung HC, Kim JM, Song IS, Kim CY. Interleukin-8 expression by human neutrophils activated by Helicobacter pylori soluble proteins. Scand J Gastroenterol 1998; 33:1249–55.
38. Shimada T, Terano A. Chemokine expression in Helicobacter pylori-infected gastric mucosa. J Gastroenterol 1998; 33:613–17.
39. Furukawa K, Takahashi T, Arai F, Matsushima K, Asakura H. Enhanced mucosal expression of interleukin-6 mRNA but not of interleukin-8 mRNA at the margin of gastric ulcer in Helicobacter pylori-positive gastritis. J Gastroenterol 1998; 33:625–33.
40. Rasko DA, Wilson TJ, Zopf D, Taylor DE. Lewis antigen expression and stability in Helicobacter pylori isolated from serial gastric biopsies. J Infect Dis 2000; 181:1089–95.
41. Wirth H, Yang M, Peek Jr, RM Tham KT, Blaser MJ. Helicobacter Lewis expression is related to the host Lewis phenotype. Gastroenterology 1997; 113:1091–98.
42. Wirth H, Yang M, Peek Jr, RM Hook-Nikanne J, Fried M, Blaser MJ. Phenotypic diversity in Lewis expression of Helicobacter pylori isolates from the same host. J Lab Clin Med 1999; 133:488–500.
43. Wirth H, Yang M, Dubois A, Berg DE, Blaser MJ. Host Lewis phenotype-dependent selection of H. pylori Lewis expression in rhesus monkeys [abstract]. Gut 1998; 43:A26.
44. Dale A, Thomas JE, Darboe MK, Coward WA, Harding M, Weaver LT. Helicobacter pylori infection, gastric acid secretion, and infant growth. J Pediatric Gastroenterol Nutr 1998; 26:393–97.
45. Patel P, Mendall MA, Khulusi S, Northfield TC, Strachan DP. Helicobacter pylori infection in childhood: Risk factors and effect on growth. BMJ 1994; 309:1119–23.
46. Viara D, Menegatti M, Salardi A, et al. Helicobacter pylori and diminished growth in children: Is it simply a marker of short stature? Ital J Gastroenterol Hepatol 1998; 30:129–33.
47. Quinonez JM, Chew F, Torres O, Begue RE. Nutritional status of Helicobacter pylori-infected children in Guatemala as compared with uninfected peers. Am J Trop Med Hygiene 1999; 61:395–98.
48. Oderda G, Palli D, Saieva C, Chiorboli E, Bona G. Short stature and Helicobacter pylori infection in Italian children: Prospective multicentre hospital based case-control study. BMJ 1998; 317:514–15.
49. Perri F, Pastore M, Leandro G, et al. Helicobacter pylori infection and growth delay in older children. Arch Dis Child 1997; 77:46–9.
50. Goodman KJ, Correa P, Tenganá Aux HJ, DeLany JP, Collazos T. Nutritional factors and Helicobacter pylori infection in Columbian children. J Pediatr Gastroenterol Nutr 1997; 25:507–15.
51. Yanez P, la Garza AM, Perez-Perez G, Cabrera L, Munoz O, Torres J. Comparison of invasive and noninvasive methods for the diagnosis and evaluation of eradication of Helicobacter pylori infection in children. Arch Med Res 2000; 31:415–21.
52. Ni YH, Lin JT, Huang SF, Yang JC, Chang MH. Accurate diagnosis of Helicobacter pylori infection by stool antigen test and 6 other currently available tests in children. J Pediatr 2000; 136:823–27.
53. Braden B, Caspary WF. Detection of Helicobacter pylori infection in children: When to perform which test? Ann Med 2001; 33:91–7.
54. Weaver LT. Aspects of Helicobacter pylori infection in the developing and developed world:Helicobacter pylori infection, nutrition, and growth of West African infants. Trans Royal Soc Trop Med Hyg 1995; 89:347–50.
55. Passaro DJ, Parsonnet J: Epidemiology of Helicobacter pylori. In: Correa P, Perez-Perez G, eds. Helicobacter pylori in gastroduodenal lesions: the second decade. New York: Prous Science, 1999.
© 2002 Lippincott Williams & Wilkins, Inc.