International Journal of Infectious Diseases
Volume 14, Issue 7 , Pages e567-e571, July 2010

Is there a link between seropositivity to Helicobacter pylori and hepatitis A virus? A systematic review

Faculty of Medicine, King Saud University, Riyadh, Saudi Arabia

Received 20 April 2009; received in revised form 29 August 2009; accepted 6 September 2009. published online 08 January 2010.

Article Outline

Summary 

Background

Since hepatitis A virus (HAV) is acquired primarily through the fecal–oral pathway, several investigators have used HAV seropositivity as a proxy for exposure to this pathway. This paper is a critical review of the evidence relevant to the association between seropositivity to HAV and Helicobacter pylori, and considers the validity of comparisons for testing the hypothesis that H. pylori spreads by the fecal–oral route.

Materials and methods

: A Medline search identified reports of all types published in the English language literature that were linked to the keywords ‘Campylobacter pylori’, ‘hepatitis A’, or ‘Helicobacter pylori’, cross-referenced with ‘seroepidemiology’, ‘seroprevalence’, or ‘seropositivity’. Studies identified by the search were included in the review if they used specific IgG antibodies to classify the serostatus of subjects for both HAV and H. pylori infection and provided an estimate of the magnitude of the association between HAV and H. pylori or information that permitted calculation of an odds ratio (OR).

Results

Out of the 21 studies identified, 15 met the inclusion criteria. The studies showed ORs for an association of HAV and H. pylori that ranged from 0.81 to 8.4. After adjustment for potential confounders, ORs shifted toward the null. They also showed that HAV seroprevalence is lower than H. pylori seroprevalence in early life and then becomes higher in later life. Thus in most populations, the trends cross over at some point.

Conclusion

The observed associations between the two infections are generally overestimated by the confounding effects of age and socio-economic status-related factors, and when these factors are controlled, the association becomes weak. Moreover, HAV infection elicits a long-term antibody response, while H. pylori infection does not. Consequently, serostatus comparison does not constitute a convincing test of the fecal–oral transmission hypothesis for H. pylori.

Keywords: Helicobacter pylori, Hepatitis A, Seroepidemiology, Seroprevalence, Seropositivity

 

Back to Article Outline

Introduction 

Helicobacter pylori is one of the most common bacterial infections in humans.1 It is known to cause chronic gastritis and peptic ulcers and to increase the risk of gastric cancer. Although this infection appears to be transmitted from person-to-person, the precise route of transmission from one person to another is controversial, and it is not known if other modes of transmission are involved. Varying degrees of evidence suggesting fecal–oral, oral–oral, gastric–oral, waterborne, and zoonotic transmission have been reported.2 Many epidemiological features of H. pylori infection suggest fecal–oral spread, therefore many studies have sought evidence of fecal–oral transmission.3 Since hepatitis A virus (HAV) is acquired primarily through the fecal–oral pathway,4 several investigators have used HAV seropositivity as a proxy for exposure to this pathway. This has led to the exploration of the possible association between H. pylori and HAV as a test of the hypothesis that H. pylori is transmitted by the fecal–oral route.5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17 To date, all studies of this association have been based on comparisons of seroprevalence. However, differences in immunological responses between H. pylori and HAV infections impact the interpretation of seropositivity. HAV antibodies are believed to persist indefinitely. Thus, HAV IgG seropositivity reflects prior exposure to infections, and someone who is seronegative has most likely never been infected. In contrast, the natural immune response to H. pylori infection does not confer lasting immunity. Thus H. pylori seropositivity may reflect a current or cleared infection, while seronegativity does not rule out prior infection. While HAV serostatus accurately measures lifetime exposure to the infection, H. pylori serostatus does not. In this paper, the evidence relevant to the association between seropositivities to HAV and H. pylori has been systematically reviewed and the validity of comparison for testing the hypothesis that H. pylori spreads by the fecal–oral pathway has been assessed.

Back to Article Outline

Methods 

A Medline search identified reports of all types published in the English language literature that were linked to the key words ‘Campylobacter pylori’ (as H. pylori was named previously), ‘hepatitis A’, or ‘H. pylori’, cross-referenced with ‘seroepidemiology’, ‘seroprevalence’, or ‘seropositivity’. The bibliography of each identified paper was consulted to locate additional relevant publications. Studies identified by the search were included in the review if they used specific IgG antibodies to classify the serostatus of subjects for both HAV and H. pylori infection and provided an estimate of the magnitude of the association between HAV and H. pylori or information that permitted calculation of an odds ratio (OR). The selected studies were examined to assess the adequacy of the sample size, the potential for selection bias, misclassification, and confounding factors, as well as consideration of effect modification. Along with the examination of the evidence from these studies, the comparability of immune response to the two infections was considered.

Back to Article Outline

Results 

The search identified 21 studies (Table 1) that had examined seropositivity to both H. pylori and HAV. Six studies were excluded;5, 6, 9, 15, 18, 19 five of them did not provide information required for a measure of association5, 6, 15, 18, 19 and the sixth was excluded because IgM was used to classify HAV serostatus.9 Study characteristics of the 15 included studies are presented in Table 2.

Table 1. All the studies identified in the search
First author, year [Ref.]OR presented in reportOR calculated from reported dataInsufficient data for ORUsed serological test to determine:
Anti-HAV IgGAnti-HP IgG
Excluded studies:
al-Moagel, 19905 √Not mentioned√
Graham et al., 19916 √√√
Okuda et al., 19969 Used IgM instead√
Sakamoto, 199715 √√√
Lee et al., 200419 √√√
Chen et al., 200318 √√√
Included studies:
Hazell et al., 19947√ √√
Hyams et al., 19958√ √√
Webb et al., 199610 √ √√
Furuta et al., 199711√ √√
Luzza et al., 199712√ √√
Pretolani et al., 199713√ √√
Rudi et al., 199714 √ √√
Sathar et al., 199716√ √√
Stroffolini et al., 199817√ √√
Malaty et al., 200325 √ √√
Tosun et al., 200424 √ √√
Egemen et al., 200623 √ √√
Nurgalieva et al., 200522 √ √√
Lin et al., 200021√ √√
Moreira et al., 200520√ √√

OR, odds ratio; HAV, hepatitis A virus; HP, Helicobacter pylori.

Table 2. Summary of the methodology and the results of the included studies
Author, year [Ref.]PopulationNo. of serosurveyed subjectsOR (95% CI) of the association between seropositivity of HP and HAV
CrudeAdjusted only for ageAdjusted for age and other factors
Rudi et al., 1997 [14]16–73-year-olds; hospital staff, Heidelberg, Germany4344.7 (2.9–7.7)
Stroffolini et al., 1998 [17]17–24-year-olds; male Italian military students16591.7 (1.3–2.2)1.3 (1.0–1.7)
Sathar et al., 1997 [16]0–13-year-olds; random sample of black South African children681 1.0 (0.66–1.58)
Furuta et al., 1997 [11]Healthy subjects attending clinic for checkups, Japan10432.0 (1.5–2.8)0.87
Hyams et al., 1995 [8]17–50-year-olds; males randomly selected from 1000 US military personnelNot mentioned 1.9a
Hazell et al., 1994 [7]Random sample from rural and urban areas, China15012.1 (1.7–2.7)
Luzza et al., 1997 [12]From rural town, southern Italy4665.6 (3–10)2.5 (1.2–5)
Webb et al., 1996 [10]18–65-year-olds; male volunteers from Stoke-on-Trent, UK4672.1 (1.4–3.1)
Pretolani et al., 1997 [13]20–85-year-olds; random sample of healthy subjects from the Republic of San Marino, Italy15283.2 (2.6–4.1) 2.0 (1.3–3.3)b
Malaty et al., 2003 [25]Survey of adults and children6148.4 (5.3–13.3)
Tosun et al., 2003 [24]2–16-year-olds; random sample from Anatolia, Turkey905.1 (1.8–14.9)2.1
Egemen et al., 2006 [23]1–18-year-olds; cluster sample of asymptomatic children from Izmir, Turkey1040.81 (0.34–0.89)0.95 (0.4–2.21)
Nurgalieva et al., 2005 [22]10–60-year-olds; visitors to the emergency room, Almaty, Kazakhstan2884.5 (2.06–9.83)
Lin et al., 2000 [21]A random sample of school teachers1045.8 (1.4–38.6) 7.3 (1.4–38.6)c
Moreira, 2005 [20]1–12-year-olds; cross-sectional survey of pediatric cases admitted to hospital1212.8 (1.3–5.97)2.3 (1.02–5.03)

OR, odds ratio; CI, confidence interval; HP, Helicobacter pylori; HAV, hepatitis A virus.

aAdjusted for age and ethnicity.

bAdjusted for age and length of schooling.

cAdjusted for age, sex and area of residence.

The size of the study populations ranged from 99 to 1659 subjects. Of the 13 studies, eight recruited their study participants from community settings, while the other five selected participants from hospital staff, military personnel, visitors to emergency room, or subjects attending health checkups (Table 2).

Age-related prevalence trends for each infection were evaluated graphically (Figure 1) in 10 of the published reports.7, 10, 11, 12, 13, 14, 16, 20, 21, 22 In general, these studies show that HAV seroprevalence is lower than H. pylori seroprevalence in early life and then becomes higher than H. pylori in later life. Thus in most populations, the trends cross over at some point (Figure 1).

Thirteen of the reviewed reports presented either unadjusted ORs or data from which to calculate them,7, 10, 11, 12, 13, 14, 17, 20, 21, 22, 23, 24, 25 while the other two presented only ORs adjusted for other variables.8, 16 All unadjusted ORs ranged between 0.81 and 8.4 with the lower 95% confidence limits between 0.34 and 3.0. Of the 15 reports, the authors of 14 of them observed that seropositivity for both H. pylori and HAV increased with age. Ten of the 15 reports presented an OR adjusted at least for age.8, 11, 12, 13, 16, 17, 20, 21, 23, 25 In seven of the eight studies that presented both unadjusted and age-adjusted ORs, the OR decreased upon adjustment for age. Age-adjusted ORs ranged from 0.87 to 7.3.11, 12, 13, 17, 20, 21, 23, 25 Three reports presented ORs adjusted for age and other factors including sex, education, ethnicity, area of residence, and number of siblings.8, 13, 23

Back to Article Outline

Discussion 

From this review, it becomes clear that after adjustment for confounders, the associations between the two infections are severely weakened. A methodological flaw of over-estimation of the association between these two infections by ignoring the effect of other variables was observed in some of these articles. In this review, a few limitations have become apparent, with an equal emphasis on epidemiological and methodological issues, which limit the appropriateness of comparing H. pylori and HAV seropositivity as evidence of a fecal–oral route for H. pylori transmission.

Differences in the epidemiology of the two diseases 

Differences in immune response 

In all of the studies that were identified that examined the association between H. pylori and HAV, the classification of H. pylori and HAV infection status depended on the identification of IgG antibodies specific to the pathogen. Anti-HAV IgG is long-lived, thus once formed it is generally detectable for life.26, 27, 28, 29 In contrast, several studies have shown that H. pylori IgG titers decline following elimination of the infection, often to undetectable levels within a year or two.30, 31, 32, 33 Therefore, serostatus at a given time-point for the two infections may correspond to distinct experiences with the pathogens and thus is not a valid means of comparing the exposure histories in individuals. H. pylori seropositivity primarily captures individuals with current or recently cleared infection and will be influenced by factors that determine persistence of the infection, such as antibiotic treatment and susceptibility factors, while HAV seropositivity captures individuals who have been infected at any time in the past. This difference in immunological response is a plausible explanation for the lack of parallelism in the pattern of the two infections seen in Figure 1 and suggests that the degree of concordance in serostatus may not accurately reflect whether the two infections share the same mode of transmission.

Multiple transmission routes 

Evidence suggests that H. pylori may be transmitted through multiple transmission routes2 and that the dominant pathway may depend on cultural and environmental conditions. Thus the strength of association between H. pylori and HAV infection may vary depending on the relative importance of the fecal–oral pathway for H. pylori in a given population.

HAV vaccine administration 

Currently, two types of hepatitis A vaccine are commercially available: human pooled immune serum globulin (ISG)28 and inactivated hepatitis A vaccine.4 Targeted populations for hepatitis A vaccination include staff and children attending day care centers when outbreaks occur; the household members of an infected person; travelers to endemic areas; and armed forces personnel. If the study group includes individuals who have been vaccinated against HAV, this would lead to misclassification of exposure to HAV. Subjects who have received standard doses of ISG will show a negative HAV IgG regardless of exposure to HAV because they are protected from being infected with the virus. The immunological response to the inactivated vaccine is very similar to that of a naturally acquired infection.28, 34 Current serological tests cannot differentiate antibodies developed in response to this vaccine from antibodies developed in response to HAV infection.28 Thus, some individuals who are seropositive for HAV due to the vaccination may not have been exposed to the HAV transmission pathway.

Infectivity 

If the infective dose and other factors that influence infectivity through the fecal–oral route differ for H. pylori and HAV, then HAV infection may be an invalid marker of the relevant fecal–oral exposure.

Methodological limitations 

Potential selection bias 

Selection bias is a particular concern in studies that are not population-based.8, 11, 14, 17 In these studies, subjects were selected from those attending clinic checkups,11 hospital staff,14 or military personnel.8, 17 Clinic patients may take antibiotics more frequently on average than community-based groups. Thus they may have cleared a prior H. pylori infection and no longer have detectable antibodies. In such populations, the association between exposure to H. pylori and HAV may be underestimated. On the other hand, hospital staff and military personnel may have increased exposure to both infections, and in these groups, the association between the infections may be overestimated.

Confounding 

Seropositivity for both HAV and H. pylori increases with age, thus any measures of their association should be adjusted for age. Incomplete control of potential confounders was a general problem among the reviewed studies, as only three of them controlled for factors beyond age. Likely confounders such as crowding and other indicators of socio-economic status were not considered in most of the studies. Since both infections occur more frequently in poor socio-economic conditions, measures of association that are not adjusted for these factors are likely to be overestimated.

Back to Article Outline

Conclusions 

In conclusion, the observed associations between the two infections are generally overestimated by confounding factors related to age and socio-economic status, and when these factors are controlled the association is weakened. Given the fact that these two infections have different immune responses, the serostatus comparison does not constitute a convincing test of the fecal–oral transmission hypothesis for H. pylori.

Conflict of interest: The author declares that there is no conflict of interest in this study.

Funding: This study was fully funded by the author.

Back to Article Outline

References 

  1. Pounder RE, Ng D. The prevalence of Helicobacter pylori infection in different countries. Aliment Pharmacol Ther. 1995;9(Suppl 2):33–39
  2. Goodman KJ, Correa P. The transmission of Helicobacter pylori. A critical review of the evidence. Int J Epidemiol. 1995;24:875–887
  3. Stone MA. Transmission of Helicobacter pylori. Postgrad Med J. 1999;75:198–200
  4. Viral hepatitis A. In: Benenson AS, editor. Control of communicable diseases manual.; 1995, p. 217-20.
  5. al-Moagel MA, Evans DG, Abdulghani ME, Adam E, Evans DJ, Malaty HM, et al. Prevalence of Helicobacter (formerly Campylobacter) pylori infection in Saudi Arabia, and comparison of those with and without upper gastrointestinal symptoms. Am J Gastroenterol. 1990;85:944–948
  6. Graham DY, Adam E, Reddy GT, Agarwal JP, Agarwal R, Evans DJ, et al. Seroepidemiology of Helicobacter pylori infection in India. Comparison of developing and developed countries. Dig Dis Sci. 1991;36:1084–1088
  7. Hazell SL, Mitchell HM, Hedges M, Shi X, Hu PJ, Li YY, et al. Hepatitis A and evidence against the community dissemination of Helicobacter pylori via feces. J Infect Dis. 1994;170:686–689
  8. Hyams KC, Taylor DN, Gray GC, Knowles JB, Hawkins R, Malone JD. The risk of Helicobacter pylori infection among U.S. military personnel deployed outside the United States. Am J Trop Med Hyg. 1995;52:109–112
  9. Okuda M, Tanaka T, Nakamoto H, Katai A, Nakata H, Koike M, et al. Helicobacter pylori infection in childhood: H. pylori isolation rate in gastric juice in relation to positive serum antibody rates. J Gastroenterol. 1996;31(Suppl 9):6–8
  10. Webb PM, Knight T, Newell DG, Elder JB, Forman D. Helicobacter pylori transmission: evidence from a comparison with hepatitis A virus. Eur J Gastroenterol Hepatol. 1996;8:439–441
  11. Furuta T, Kamata T, Takashima M, Futami H, Arai H, Hanai H, et al. Study of transmission routes of Helicobacter pylori in relation to seroprevalence of hepatitis A virus. J Clin Microbiol. 1997;35:1891–1893
  12. Luzza F, Imeneo M, Maletta M, Paluccio G, Giancotti A, Perticone F, et al. Seroepidemiology of Helicobacter pylori infection and hepatitis A in a rural area: evidence against a common mode of transmission. Gut. 1997;41:164–168
  13. Pretolani S, Stroffolini T, Rapicetta M, Bonvicini F, Baldini L, Megraud F, et al. Seroprevalence of hepatitis A virus and Helicobacter pylori infections in the general population of a developed European country (the San Marino study): evidence for similar pattern of spread. Eur J Gastroenterol Hepatol. 1997;9:1081–1084
  14. Rudi J, Toppe H, Marx N, Zuna I, Theilmann L, Stremmel W, et al. Risk of infection with Helicobacter pylori and hepatitis A virus in different groups of hospital workers. Am J Gastroenterol. 1997;92:258–262
  15. Sakamoto K. Investigation of infection routes of Helicobacter pylori in health checks of residents by random sampling. Kurume Med J. 1997;44:273–280
  16. Sathar MA, Gouws E, Simjee AE, Mayat AM. Seroepidemiological study of Helicobacter pylori infection in South African children. Trans R Soc Trop Med Hyg. 1997;91:393–395
  17. Stroffolini T, Rosmini F, Ferrigno L, Fortini M, D’Amelio R, Matricardi PM. Prevalence of Helicobacter pylori infection in a cohort of Italian military students. Epidemiol Infect. 1998;120:151–155
  18. Chen LK, Hwang SJ, Wu TC, Chu CH, Shaw CK. Helicobacter pylori and hepatitis A virus infection in school-aged children on two isolated neighborhood islands in Taiwan. Helicobacter. 2003;8:168–172
  19. Lee IH, Ni YH, Chang MH. Stable seroprevalence of Helicobacter pylori infection in children during 1989-1999 in Taipei, Taiwan. Pediatr Int 2004; 46:311-4.
  20. Moreira ED, Nassri VB, Santos RS, Matos JF, de Carvalho WA, Silvani CS, et al. Association of Helicobacter pylori infection and giardiasis: results from a study of surrogate markers for fecal exposure among children. World J Gastroenterol. 2005;11:2759–2763
  21. Egemen A, Yilmaz O, Akil I, Altuglu I. Evaluation of association between hepatitis A and Helicobacter pylori infections and routes of transmission. Turk J Pediatr. 2006;48:135–139
  22. Malaty HM, Tanaka E, Kumagai T, Ota H, Kiyosawa K, Graham DY, et al. Seroepidemiology of Helicobacter pylori and hepatitis A virus and the mode of transmission of infection: a 9-year cohort study in rural Japan. Clin Infect Dis. 2003;37:1067–1072
  23. Lin DB, Tsai TP, Yang CC, Wang HM, Nieh WT, Ling UP, et al. Association between seropositivity of antibodies against hepatitis A virus and Helicobacter pylori. Am J Trop Med Hyg. 2000;63:189–191
  24. Nurgalieva ZZ, Malaty HM, Graham DY, Almuchambetova R, Machmudova A, Kapsultanova D, et al. Helicobacter pylori infection in Kazakhstan: effect of water source and household hygiene. Am J Trop Med Hyg. 2002;67:201–206
  25. Tosun SY, Kasirga E, Ertan P, Aksu S. Evidence against the fecal-oral route of transmission for Helicobacter pylori infection in childhood. Med Sci Monit. 2003;9:CR489–CR492
  26. Plotkin SA, Orenstein WA, Hepatitis . A vaccine. In:  Feinstone SM,  Gust ID editor. Vaccines. Philadelphia: W. B. Saunders Co.; 1999;p. 650–671
  27. Stites DP, Terr AI, Parslow TG. Medical immunology. 9th ed. Stamford, CT: Appleton & Lange; 1997.
  28. Stapleton JT. Host immune response to hepatitis A virus. J Infect Dis. 1995;171(Suppl 1):S9–14
  29. Prevention of hepatitis A through active or passive immunization: recommendations of the Advisory Committee on Immunization Practices (ACIP). MMWR Recomm Rep 1996; 45:1-30.
  30. Cutler AF, Prasad VM. Long-term follow-up of Helicobacter pylori serology after successful eradication. Am J Gastroenterol. 1996;91:85–88
  31. Cutler AF, Prasad VM, Santogade P. Four-year trends in Helicobacter pylori IgG serology following successful eradication. Am J Med. 1998;105:18–20
  32. Kosunen TU. Antibody titres in Helicobacter pylori infection: implications in the follow-up of antimicrobial therapy. Ann Med. 1995;27:605–607
  33. Wang WM, Chen CY, Jan CM, Chen LT, Perng DS, Lin SR, et al. Long-term follow-up and serological study after triple therapy of Helicobacter pylori-associated duodenal ulcer. Am J Gastroenterol. 1994;89:1793–1796
  34. Sandman L, Davidson M, Krugman S. Inactivated hepatitis A vaccine: a safety and immunogenicity study in health professionals. J Infect Dis. 1995;171(Suppl 1):S2–S50

PII: S1201-9712(09)00361-0

doi:10.1016/j.ijid.2009.09.003

International Journal of Infectious Diseases
Volume 14, Issue 7 , Pages e567-e571, July 2010

Article Tools

* Image Usage & Resolution