Epidemiology of Helicobacter pylori Infection


Reprint requests to: Michael Bruce, Arctic Investigations Program, Centers for Disease Control and Prevention, 4055 Tudor Centre Drive, Anchorage, Alaska 99508, USA. E-mail: zwa8@cdc.gov


This review summarizes studies on the epidemiology of Helicobacter pylori published in peer-reviewed journals between April 2007 and March 2008. Infection with H. pylori often occurs in childhood, and once established, can persist lifelong if untreated. Prevalence of H. pylori infection is higher in developing countries when compared to developed countries, and can vary by ethnicity, place of birth, and socioeconomic factors even among persons living in the same country. Prevalence of infection is decreasing in many countries due to improvements in sanitation and living standards and the relatively recent movement of populations from rural to urban settings; however, post-treatment recurrence rates of H. pylori infection remain high in developing countries, and in given populations within developed countries. In addition, a number of recent studies have begun to explore the possible link between childhood infection with H. pylori and protection against asthma and allergy.

Several reviews published in the last year have focused on the epidemiology of Helicobacter pylori infection [1–6]; however, this article presents an analysis of original studies published during this period. The majority of studies used either serologic testing or urea breath test (UBT), or histologic testing to determine H. pylori positivity. Stool antigen testing or polymerase chain reaction (PCR) testing was used in a minority of studies. The majority of epidemiologic studies were performed in adults; however, many studies were performed in children and some in particular subgroups such as pregnant women [7], teachers [8,9], and students [10].

Prevalence, Risk Factors, Acquisition, and Transmission

Studies on the prevalence of H. pylori infection were published from 19 different countries: India [11], Saudi Arabia [10], Bulgaria [12,13], China [14,15], Latvia [16], Tanzania (Pemba Island, Zanzibar) [7], Taiwan [9], USA (Alaska) [8,17], Canada [18], Lebanon [19], Japan [20], Malaysia [21], South Africa (Venda region, Limpopo) [22], Turkey [23–25], Russia [26], South Korea [27], Iceland, Estonia, and Sweden [28], adding valuable knowledge from countries with different socioeconomic, cultural, and racial groups. Increasing H. pylori prevalence with age was demonstrated in numerous studies and is due largely to a birth cohort effect rather than late acquisition of infection.

The traditional approach to the epidemiology of H. pylori infection in different geographic regions of the world has been to divide the world into developed (low prevalence) and developing (high prevalence) countries; however, this division is becoming less clear due to the rapid rise of socioeconomic level in subpopulations in many developing countries and to the increasing use of antimicrobials for infection with H. pylori and other bacteria in both developed and developing countries. Many of the recently published prevalence studies were population-based and looked at changes in prevalence over time; others were hospital- or clinic-based studies performed often over shorter time periods. We focused on prevalence studies among healthy persons without gastrointestinal (GI) illness because prevalence among persons with GI illness is covered in another article in this issue of Helicobacter. The overall trend of decreasing prevalence of H. pylori infection in countries of variable socioeconomic development was evident in both categories of publications.

Population-based Prevalence Studies Among Healthy Persons (Without GI Illness)

Overall H. pylori prevalence among the many studies varied from a low of 11% to a high of 69%[28]. The link between low socioeconomic level and higher prevalence (and high socioeconomic level and lower prevalence) of infection was demonstrated in several studies; however, an overall decline in prevalence of H. pylori was noted in both high and low socioeconomic status countries. In a population-based seroprevalence study of 1471 healthy adults and children living in China (Guangzhou city), sera were collected from March to August 2003 and tested by enzyme-linked immunosorbent assay (ELISA). Overall, 47% of persons tested seropositive; the age-standardized seroprevalence was 49.3% while it was 62.5% in a previous study performed in 1993, representing a 13.2% decline in age-standardized seroprevalence over the 10-year period. Seroprevalence among children aged 1–5 years decreased by 11.4% (30.8% in 1993 to 19.4% in 2003, p < .017) [14]. A decrease in H. pylori prevalence, particularly among children, could result in a decrease in acquisition, transmission, and overall infection in a given population.

A seroprevalence (ELISA) study performed by Fujimoto et al. among healthy children and adults residing in three areas of Japan found that the age-adjusted H. pylori seroprevalence declined from 68.4% in 1993 to 52.5% in 2002 (p < .01). Prevalence increased with age in all three regions. The seroprevalence was higher in children with H. pylori-positive mothers (21.6%) versus H. pylori-negative mothers (3.2%) [29]. A study of school children in Russia (St Petersburg) showed a trend of overall declining H. pylori seroprevalence from 44% in 1995 to 13% in 2005. Crude and age-adjusted odds ratios (OR) for infection in children demonstrated a correlation between higher parental education level and lower H. pylori seropositivity (OR 1.8, p = .06). No association was found between H. pylori seropositivity and other risk factors such as sex, type of dwelling, income, or number of people living at home [26]. In a population-based, nationwide seroprevalence study of 15,916 healthy persons age ≥ 16 years living in South Korea, sera were collected from January to December 2005 and tested by ELISA [27]. Overall H. pylori seropositivity was 59.6%. A comparison to a previous study demonstrates a decrease by 7% from 1998 (66.9%) to 2005 (59.6%). Seroprevalence was the highest in the following groups: persons aged 50–59, males, low income groups, persons with a low level of education, and among persons living in rural areas.

In addition to the two studies cited above that demonstrate a relationship between parental education level and H. pylori prevalence, a recent population-based study performed in Canada (Ontario) by Naja et al. [18] demonstrated that H. pylori seropositivity was lower in persons with a higher education level. Overall seroprevalence was 23.1%. Prevalence was found to increase with age and number of siblings, and was found to be associated with being nonwhite and born outside of Canada. In addition, a recent study from Lebanon showed that among 87 H. pylori-positive children, 75 (86%) were from low-income families; 12 were from high-income families (14%) [19].

Prevalence among teachers and students was evaluated in several studies published over the past year. These studies demonstrated that H. pylori prevalence was not higher in these groups compared to persons of similar age in their population. A study by Lynn et al. among non-Indigenous Alaska teachers demonstrated a seropositivity rate of 24%[8]. Age older than 40 years was associated with infection; however, other established risk factors such as water source, residence in a developing country, or education level were not associated with H. pylori seropositivity. Similarly, in a study by Lin et al. in Taiwan, H. pylori seropositivity was not found to be associated with previously documented risk factors such as water source or parental education level. Seropositivity among 253 teachers was 45.1%, and among 1950 students age 9–15 years was 11.0–12.3%[9]. A study performed by Almadi et al. looking at H. pylori prevalence among 120 medical students in Saudi Arabia found that 42 (35%) had active H. pylori infection by UBT. No difference in H. pylori prevalence was found between preclinical (no patient contact) and clinical (patient contact) students [10]. This prevalence rate is low in comparison to the previously documented rate of 70% among adults in Saudi Arabia [30].

Other studies demonstrated unexpected results such as the study by Farag et al. examining 857 pregnant women aged 20–40 years living on Pemba Island, Zanzibar, Tanzania. They showed that a low proportion of pregnant women (17.5%) had an active H. pylori infection (UBT) [7]. Typical prevalence rates among a variety of groups on the African mainland are ≥ 80%; prevalence data from pregnant woman typically reflect H. pylori prevalence among the general population [2,31,32]. While there is no clear explanation for these unexpectedly low prevalence rates, H. pylori has infected humankind for millennia and it is possible that the original inhabitants of Pemba Island arrived uninfected and remained so until relatively recently due to their geographic isolation from the mainland.

Prevalence studies published over the past year demonstrate variable levels of H. pylori prevalence among healthy persons. High H. pylori prevalence was associated with a variety of risk factors such as increasing age, low socioeconomic status, low education level, poor sanitation, rural residence, and birth in a developing country. An alarming finding from reviewing the last year's literature on prevalence is the particularly high H. pylori prevalence rate among children. Despite this high prevalence, a declining trend in prevalence over time in both children and adults living in developed and developing countries was noted which will hopefully continue in the future as the standard of living and hygiene improve around the world.

H. pylori, Asthma, and Allergy

The rapid rise in allergic disorders and asthma in the developed world is likely due in large part to exposures to environmental agents such as pollen, particulates, and airborne pollutants. Over the past decade, a number of studies have looked at whether infections during early childhood could prevent or diminish atopy and asthma. Infections with particular bacteria or viruses in childhood could result in an enhanced host cell-mediated immune response (Th1 type) causing an increase in production of particular cytokines (such as interleukin (IL)-10), resulting in an inhibitory effect on Th2 cells and an overall down-regulation of the immune response (inhibition of inflammation) thus preventing hyper-reactive states (the hygiene hypothesis) [33–35].

A number of studies published in the past year looked specifically at immune response in relation to infection with H. pylori such as a study by Oderda et al. that found that among H. pylori-infected children, the prevalence of allergy was significantly higher in children with lower cytokine expression in the gastric mucosa [36], and also demonstrated a rise in IL-10 production associated with H. pylori infection. In a study performed by Janson et al. seroprevalence of IgG antibody to less than three of seven infectious agents studied was found to be an independent risk factor for atopy, allergic asthma, and allergic rhinitis [37]. Data from this study are consistent with the theory that cumulative infections protect against atopy and respiratory allergies [38–40].

Seiskari et al. evaluated allergic sensitization and microbial load in children living in Finland and Russian Karelia and determined that microbial antibodies were more frequently found in Russian than in Finnish children and Russian children had a lower risk of allergic sensitization. Seventy-three percent of Russian children and 5% of Finnish children had antibodies to H. pylori [41].

Chen et al. evaluated associations between H. pylori status and a previous history of asthma/allergy using data from 7663 adults who participated in the Third National Health and Nutrition Examination Survey (NHANES). The presence of cagA-positive H. pylori strains was inversely related with ever having asthma. They also determined that H. pylori colonization (especially with cagA-positive strains) was inversely associated with currently or ever having a diagnosis of allergic rhinitis, particularly of childhood onset. Inverse associations were also found between H. pylori colonization and sensitization to pollens and molds and the presence of allergy symptoms in the previous year [42]. This study provides evidence that colonization with H. pylori, particularly cagA-positive strains, may reduce the risk of allergy and asthma and poses intriguing questions such as “Could the disappearance of H. pylori in developed countries be related to the increase in atopy and asthma that has been observed?” Further studies are needed to determine the mechanism whereby this is occurring and to confirm these interesting observations.

H. pylori, Water, Sanitation, and the Environment

Sanitation and water quality most certainly play a role in early acquisition and transmission of H. pylori. The United Nations recently declared 2008 as the international year of sanitation to raise awareness of sanitation issues around the world, and to accelerate progress toward the millennium development goal target to reduce by half the proportion of the 2.6 billion people without access to basic sanitation by 2015.

Past studies have demonstrated H. pylori prevalence to be associated with household hygiene, water use, and sanitary conditions. Other studies have found no association between sanitation/water source and H. pylori prevalence. Not surprisingly, prevalence studies published over the past year found mixed results. A study performed in South India by Ahmed et al. found that H. pylori prevalence was higher among people who drank water from wells (92%) compared to tap water drinkers (75%) [11]. Prevalence was higher among people with a low clean water index (CWI, 88%) versus people with a high CWI (33%). Low CWI was defined as: 1, never boiling water before drinking, 2, bathing less than once a week, and 3, storing and reusing water. H. pylori prevalence was also found to be higher (86 versus 70%) among people with low compared to high socioeconomic status [11]. However, in prevalence studies performed by Lin et al. in Taiwan [9] and Lynn et al. in Alaska [8], water source did not correlate with H. pylori positivity among teachers/students. One likely reason that numerous prevalence studies have failed to find an association between water source/sanitary conditions and H. pylori positivity is the fact that these studies ask about current water use and current sanitary conditions among study participants when in fact, infection with this organism likely occurs in childhood (before age 10). Another possible reason for this lack of association is due to the fact that many previous studies failed to control for other confounders or real effectors such as low socioeconomic status, crowding, poor nutrition, or other factors related to poverty.

Recent concerns over the presence of H. pylori in the environment, particularly the water supply, have resulted in a number of studies looking at the viability of the organism in water. To date, H. pylori has been extremely difficult to culture from water and PCR methods have been used to detect the presence of H. pylori DNA. Three studies over the past year have looked at methods of H. pylori detection and viability of the organism in water over time. A study by Queralt et al. looked at the survival of H. pylori in mineral water over a 3-week time period using culture and PCR methods. H. pylori could no longer be cultured after 5 days; cell viability decreased up to day 14, and at day 21 all cell membranes were damaged. In addition, they noted that after day 3, there was a conversion from spiral to coccoid form. H. pylori DNA was detectable by PCR at day 21 and as far out as 3 months. The authors concluded that H. pylori should be considered a waterborne pathogen. H. pylori can survive in water, but loses its culturability and bacillary morphology rapidly. However, it remains viable in a coccoid form for long periods of time. PCR methods appear to overestimate H. pylori presence in water while culture methods underestimate it [43].

Moreno et al. evaluated survival of H. pylori in chlorinated water and found that H. pylori converts to a coccoid form with exposure to chlorine. Viable cells (by culture) were detected 3 hours after exposure to chlorine but not after 24 hours. This study suggests that H. pylori could survive disinfection practices normally used in drinking water in the viable but nonculturable form which would allow them to reach final consumption points and at the same time be undetectable by culture methods [44]. A study by Sen et al. evaluated a quantitative PCR technique for detection of H. pylori in drinking water. The new quantitative molecular method used in this study, which involved development of an internal control for evaluation and standardization of this quantitative PCR assay, is a start at developing accurate and robust screening tools for detection of H. pylori DNA in water at levels as low as 5–10 cells per liter [45].

Recurrence of H. pylori infection

Recurrence of H. pylori infection after successful eradication is uncommon in developed countries, but frequent in developing countries [46]. Most cases of recurrence in developed countries are thought to be due to recrudescence (infection with the same strain); whereas in developing countries, most cases of recurrence are believed to be due to reinfection (infection with a new strain) [6]. Over the past year, three articles focusing on recurrence of H. pylori infection over time were published. A study by Thong-Ngam et al. looked at recurrent infection among adults in Thailand over a 4-year follow-up period. Participants were treated for H. pylori infection and documented to be negative by UBT at enrollment. At 4 years, the recurrence rate was 13.5% (5.4% at 1 year, 8.11% at 2 years, 10.81% at 3 years, 13.5% at 4 years). In addition, this study looked at risk factors for recurrence of infection; no risk factors for recurrence were detected [47]. A study performed in Bangladesh by Ahmad et al. looked at H. pylori recurrence 6 years post-cure among 41 remaining participants from a previous reinfection study (n = 90) and determined a 30% reinfection rate at 6 years (72 months); reinfectees were found to be more likely to experience duodenal ulcer relapse [48]. This reinfection rate is lower than previously published data from the first 2 years of the study which showed a reinfection rate of approximately 18% per year [49].

Sheu et al. in Taiwan followed 359 participants (who had been initially cured of their infection) over a 3-year period and performed UBT each year to document H. pylori recurrence. Dental assessments were conducted by the same dentist at enrollment and at 1 year. Participants were divided into a “dental disease group” (DDG) and a “no dental disease group” (NDDG). Persons with documented periodontal disease or dental caries were placed into the DDG and persons without dental disease into the NDDG. At year 1, the H. pylori recurrence rate was 13.2% in the DDG verses 3.5% in the NDDG (p < .001). The same trend in recurrence rates was observed for year 2 (18.4% DDG verses 2.8% NDDG) and year 3 (20% DDG verses 3.8% NDDG). Low socioeconomic status was determined to be an independent risk factor for recurrence (on multivariable analysis) only at year 3. The authors concluded that the presence of dental disease could predispose to H. pylori recurrence [50]. Previous studies have shown that the oral cavity can serve as a niche for H. pylori [51,52]. Other studies have shown a lower H. pylori eradication rate among persons documented to be PCR positive from dental plaque/saliva when compared to oral H. pylori-negative cases [53]. Further studies are needed to further elucidate the role of dental hygiene and H. pylori infection and recurrence.

Conflicts of interest

The authors have declared no conflicts of interest.