Epidemiology of Helicobacter pylori Infection and Public Health Implications

Abstract

This review summarizes studies on the epidemiology and public health implications of Helicobacter pylori published in peer-reviewed journals from April 2010 through March 2011. Prevalence rates vary widely between different geographical regions and ethnic groups. An interesting study from the USA identified the degree of African ancestry as an independent predictor of H. pylori infection. Two studies have demonstrated early childhood as the period of transmission of infection and identified an infected sibling as an important risk factor. An oral–oral route of spread has been substantiated with several studies showing the presence of H. pylori in the oral cavity. Studies have shown the presence of H. pylori in drinking water and the role of poor living conditions and sanitation in H. pylori infection, supporting an oral–fecal route of spread. Screening for H. pylori as a gastric cancer prescreening strategy has been described in Japan, and the importance of H. pylori eradication as a gastric cancer–prevention strategy has now been further emphasized in Japanese guidelines. Two studies have shown a decrease in the burden of dyspepsia and peptic ulcer disease with H. pylori eradication.

This article presents a review of the literature concerning the epidemiology and public health implications of Helicobacter pylori infection published from April 2010 till March 2011. The authors searched PubMed and Embase using MeSH terms “Helicobacter infections/epidemiology” and “Helicobacter infections/prevention and control”, and repeated the PubMed search independently using MeSH term “Helicobacter” alone and using the set operator AND with the terms “Epidemiology”, “Prevalence”, “Incidence”, “Transmission”, “Risk Factors”, “Prevention and Control” or “Environment”. The identified literature is summarized below by subtopic: prevalence; incidence; transmission; risk factors; and public health implications.

Prevalence

Serology was the most common method of diagnosis used in these studies, but several studies were endoscopy-based and diagnoses were then made by rapid urease test (RUT), culture, immuno-histochemistry or histology. A few studies utilized the urea breath test (UBT) or the stool antigen test (SAT). In an interesting and important study, Gong et al. [1] compared the accuracy of serological testing to histological diagnosis in a gastric cancer screening field survey in northern China. They found that serological testing using a commercially available kit and utilizing the recommended cut-off level underestimated the prevalence of H. pylori by almost 30% and emphasized the need for local validation of serological tests.

We have grouped the studies according to the different geographical regions for ease of reference (Table 1). There were 10 studies from the Asia-Pacific region on adults [2–11] and 4 on children [12–15]. Li et al. [2] reported a well-conducted study which was carried out in Shanghai as part of a systematic investigation of gastrointestinal disease in China. Using a multistage, stratified sampling method, they recorded a H. pylori prevalence of 73.3% (2310/3151) by serological testing for all subjects and 71.7% (733/1022) by endoscopy for subjects who agreed for the procedure. In large endoscopy-based studies from Korea [3], Vietnam [4], and Turkey [5], H. pylori was detected from 50–70% of the population studied. Tsukanov et al. [6] in one of the few studies from eastern Siberia recorded inordinately high rate of H. pylori infection, exceeding 90% for both “Europoid” (European descent) and “Mongoloid” (Asian descent) populations. Among selected subpopulations, Ullah et al. [7] reported a high H. pylori prevalence of 77.3% among a group of Bangladesh fish handlers, while Rahim et al. [8] in a study of aborigines in the Northeastern part of Malaysia reported a prevalence rate of 19%.

Table 1.

Studies reporting prevalence of Helicobacter pylori infection, published between April 2010 and March 2011, tabulated according to different geographical regions

Author	Population studied	Diagnostic test performed	Number of subjects or patients	H. pylori prevalence (%)
Asia-Pacific
Li et al. [2]	Random sample of residents in Shanghai, China (aged 18–80 years old)	Serology	1022	71.7
Nam et al. [3]	Healthy screening population in Korea	Rapid urease test (RUT)	10102	50.8
Nguyen et al. [4]	Patients undergoing gastroscopy in Hanoi and Ho Chi Minh, Vietnam (aged 14–86)	Serology, RUT, culture, histology, immuno-histochemistry	270	65.6
Ozdil et al. [5]	Consecutive dyspeptic patients undergoing gastroscopy in Turkey	Histology	3301	71.3
Tsukanov et al. [6]	Consecutive patients referred for dyspepsia in Eastern Siberia	Serology, RUT, histology	689 “Europoid”	93.6
			1440 “Mongoloid”	94.3
Ullah et al. [7]	Fish handlers and non-fish handlers in Bangladesh	Serology	163 fish handlers	77.3
			72 non-fish handlers	37.5
Rahim et al. [8]	Aborigines living in northeastern Penin sular Malaysia	Serology	480	19
Pandeya et al. [9]	Australian adults	Serology	1355	15.5
Fraser et al. [10]	Teenage females in high schools in Auckland, New Zealand	Serology	386 Pacific Island	49.0
			120 Maori	26.7
			162 Asian	24.7
			124 European	13.7
Al Faleh et al. [11]	Random sample of students aged 16–18 from 3 regions in Saudi Arabia	Serology	1157	47
Ozen et al. [12]	School children in Istanbul, Turkey	Serology	473	34
Thankachan et al. [13]	Children aged 6–10 in Bangalore, India	Urea breath test (UBT)	543	79
Cherian et al. [14]	African refugee children from resettlement in Australia	Stool antigen test (SAT)	163	84.0
Abdollahi et al. [15]	Children aged 3–18 with reflux symptoms in Iran	RUT, Histology	263	22.4
Africa
Tanih et al. [16]	Consecutive dyspeptic patients at a hospital in Port Elizabeth, South Africa	Culture, PCR	254	66.1
Joutei et al. [17]	Dyspeptic patients in Morocco	Histology	755	69
Aje et al. [18]	Patients with dyspepsia and controls in Nigeria	Serology, SAT	46 dyspeptics 46 controls	67.4
			46 controls	78.3–91.3
Hestvik et al. [19]	Healthy children aged 0–12 in Kampala, Uganda	SAT	427	44.3
S. America
Dattoli et al. [20]	Children aged 4–11 in northeastern Brazil	Serology	1104	28.7
Miranda et al. [21]	Children attending a public hospital in Sao Paulo, Brazil	Serology	326	35.6
Janjetic et al. [22]	Children aged 4–16 with upper gastrointestinal symptoms	UBT	395	24.3
Egorov et al. [23]	Children in poor suburbs of Quito, Ecuador.	SAT	124	61
Araf et al. [24]	Adolescents in a public school in Sao Paolo, Brazil (aged 10–16)	UBT	194	40.7
Ortega et al. [25]	Dyspeptic patients in Chile	RUT	5664	78
Fialho et al. [26]	Poor urban community in northeastern Brazil	UBT in children, serology in adults	570 members of 128 households	66.0
Europe
Vendt et al. [27]	Children from ambulatory admission in Estonia	Serology	363	27
Katsanos et al. [28]	Retrospective consecutive Albanian and Greek patients undergoing gastroscopy in western Balkans	Histology	101 Albanian	54
			101 Greek	34
N. America
Epplein et al. [29]	African Americans and whites from the Southern Community Cohort Study	Serology	689	79
Sonnenberg et al. [30]	Central laboratory database of gastric biopsies, USA	Histology	78985	7.5
McJunkin et al. [31]	Patients who underwent endoscopy in Charleston, USA	Histology	251	6.8
Cardenas et al. [32]	Subjects of all ages at U.S.- Mexico border	SAT	386	38.2

Pandeya et al. [9] in an Australian study of community controls of a nationwide study on esophageal cancer recorded a H. pylori prevalence rate of 15.5% in a study population of mainly white subjects. Fraser et al. [10] showed significant differences in H. pylori prevalence between Pacific Island (49.0%) vs. Maori (26.7%) and Asian (24.7%) vs. European adolescents (13.7%). Several studies on children and adolescents in Asia showed prevalence rates ranging from 20% to 84% [12–15]. Overall, as expected, the H. pylori prevalence rates from the Asia-Pacific region were high except among the white population of Australia and New Zealand. The prevalence of H. pylori infection was generally lower among children except for the one study from India [13] and another looking at African refugee children from resettlement in Western Australia [14].

Four studies were reported from Africa [16–19]. Studies from Africa recorded high H. pylori prevalence rates ranging from 41.3% to 91.3% [16–19]. There were seven studies that reported H. pylori prevalence from South America [20–26]. Four of these studies were on children [20–23]. The study by Dattoli et al. [20], a continuation of previous studies on diarrheal diseases in a town in northeastern Brazil, reported a H. pylori seroprevalence of 28.7%. Several risk factors for H. pylori infection were identified in the study and will be discussed in a later section. The other three studies on children [21–23] reported H. pylori prevalence rates ranging from 24.3% to 61.0%.

There were few studies from Europe [27,28] and North America [29–32]. In an important and interesting study from USA, Epplein et al. [29] reported a high H. pylori prevalence rate of 79.0% among a subpopulation of poor Americans (predominantly blacks) with a direct correlation of high H. pylori prevalence to the low, moderate, and high “African” ancestry. In contrast, Sonnenberg et al. [30] in a review of a huge database of all biopsies collected in a central laboratory in the USA reported a H. pylori prevalence of only 7.5%.

Several studies have focused on specific disease groups to determine the possible relationship with H. pylori infection [33–38] (Table 2). Kirchner et al. [33] did not find a significant difference in H. pylori seroprevalence between liver cirrhotic and noncirrhotic patients. Senbanjo et al. [34] compared the seroprevalence of H. pylori between children with and without sickle cell disease and found the prevalence to be high in both. High prevalence of H. pylori infection was seen among morbidly obese patients undergoing bariatric surgery (85.5%) [35] and patients with myelodysplasia (75.3%) [36]. On the other hand, an inverse relationship with HIV infection was noted in a study from Brazil [37]. This marked disparity has been observed previously [39–41], but the reason for it remains unclear. Schimke et al. [38] reported H. pylori seroprevalence of 62.0% among a cohort of patients with type 2 diabetes mellitus.

Table 2.

Helicobacter pylori prevalence among special disease groups

Author	Population studied	Diagnostic test	Number of patients	H. pylori prevalence (%)
Kirchner et al. [33]	Cirrhotic patients and patients with chronic hepatitis without cirrhosis in Germany	Histology in cirrhotics serology in noncirrhotics	110 cirrhotics	69
			44 noncirrhotics	63
Senbanjo et al. [34]	Children with sickle cell disease (SCD) and controls in Nigeria	Serology	118 SCD	67.8
			118 controls	63.6
Al Akwaa [35]	Morbidly obese patients undergoing bariatric surgery in Saudi Arabia	Histology	62	85.5
Diamantidis et al. [36]	Patients with myelodysplastic syndromes in Greece	Serology, urea breath test, rapid urease test (RUT), histology	73	57.5–80.4
Fialho et al. [37]	HIV-positive patients and HIV-negative controls undergoing gastroscopy for dys pepsia in Northeastern Brazil	Histology, RUT	113 HIV-positive	37.2
			141 HIV-negative	75.2
Schimke et al. [38]	Patients with type 2 diabetes mellitus in Australia	Serology	1179	62.0

Two studies looked at time trend differences [31,42]. Nakajima et al. [42] studied subjects who went for annual health check at their hospital and reported a drop in H. pylori seroprevalence from 70% to 50% over a 17-year period (1988–2005) and along with this, a decline in the prevalence of peptic ulcer disease (PUD) and gastric cancer. In an endoscopy-based study from the USA with relatively small numbers, McJunkin et al. [31] also reported a dramatic drop in H. pylori prevalence (from 65.8% to 6.8%) and PUD (from 38.8% to 5.6%) over an 11-year period.

Incidence

There was only one study reporting on incidence of H. pylori infection. In this study by Muhsen et al. [43]., a cohort of Israeli Arab children at preschool age was tested for H. pylori infection using SAT and the test was repeated at school age. The prevalence of H. pylori infection was 49.7% and 58.9% at preschool age and school age, respectively. Among children tested in both examinations, there were fourteen new H. pylori infections among seventy previously uninfected children (20%) over a 4-year period, giving an annual incidence of 5%.

Transmission

Transmission of H. pylori is still not entirely clarified, but human-to-human spread through oral–oral or fecal–oral route is thought to be the most plausible.

Several studies looked at the spread of H. pylori infection between siblings [20,26,43–45]. Two of these were well-conducted cohort follow-up studies [43,44]. In the study by Muhsen et al. [43], Israeli Arab children aged 3–5 from three villages in northern Israel were followed up for 3–4 years. Having H. pylori-infected sibling was identified as an independent risk factor for both “early” and “persistent” H. pylori infection as well as late acquisition of the infection. In a second study, Cervantes et al. [44] reported that persistent H. pylori infection in older siblings always preceded infection in younger siblings and that the former was an important predictor of H. pylori infection in the latter, especially if the age difference was <3 years.

Other studies were basically cross-sectional studies and also showed infected siblings and mothers, overcrowding and poor social conditions as risk factors for H. pylori infection in children [20,26]. Siblings of young patients with gastric cancer were also found to have a higher prevalence of H. pylori infection than controls supporting spread between siblings [45]. Infected siblings appear to be an important reservoir of H. pylori infection in children.

Several studies showed the presence of H. pylori in saliva, dental plaques, oral cavity, and tonsillar tissue as well as in the esophagus [46–52]. These studies lend weight to an oral–oral route of spread of H. pylori infection. The presence of H. pylori in oral cavity is more frequent in seropositive subjects [46], and several studies from Brazil have consistently showed an association between gastric H. pylori infection and the presence of this bacterium in the oral cavity [47–49]. Moreover, the bacterium identified in the samples of the different sites within a given subject among all patients in one study [48] and in up to 89% in another study [49] were of identical genotype. The association is reinforced by a recent meta-analysis [53] where the prevalence of H. pylori infection in the oral cavity in gastric H. pylori-positive patients was significantly higher (45.0%) than that in gastric H. pylori-negative patients (23.9%) (OR = 3.61, p < 0.0001). In addition, it was reported that the eradication rate of H. pylori from the stomach (85.8%) is much higher than from the oral cavity (5.7%) (OR = 55.59, p < 0.00001), raising concerns that H. pylori in the oral cavity could be a source of re-infection following successful gastric eradication.

A study reported the presence of H. pylori in tonsillar tissue of up to 48% of patients who underwent tonsillectomy [54]. However, this study utilized RUT which may yield false-positive results because of the presence of other urease-producing organisms in a polymicrobial environment such as the tonsillar tissues. In a separate study [55], H. pylori was not detected at all using fluorescence in situ hybridization and polymerase chain reaction–DNA hybridization assay (PCR–DEIA) in the adenotonsillar tissue of a cohort of children who underwent adenoidectomy or tonsillectomy with a H. pylori prevalence of 39%, suggesting that adenotonsillar tissue does not constitute an extragastric reservoir for H. pylori.

H. pylori could be cultured from rectal fluid and terminal ileal fluid in the setting of rapid intestinal transit supporting a fecal–oral route of transmission [56]. Al Sulami et al. [57] reported for the first time the occurrence of H. pylori in treated drinking water (2.0% of total isolates) in Basra, Iraq. In another study from Pakistan, Samra et al. [58], using a PCR assay targeting virulence genes found H. pylori in 40% of samples of drinking water. Linke et al. [59] introduced a highly sensitive and specific test for the detection of H. pylori in drinking water biofilms utilizing real-time PCR method. However, as detection of H. pylori DNA may not represent the presence of viable bacterium, the true significance of a positive test remains uncertain and requires further studies. Presence of viable H. pylori in drinking water, if confirmed, would be an important source of transmission, pointing to a fecal–oral route of spread.

Risk Factors

Poor Living Conditions

In a study from Brazil, Dattoli et al. [20] reported increased H. pylori infection with a larger number of siblings, nursery schooling, and housing in a street without paved roads and without flushed toilets indicating impoverished living conditions associated with poorer sanitation and overcrowding to be risk factors for H. pylori infection. Similarly, Fialho et al. [26] demonstrated the number of people per room and number of children in the household as independent risk factors for H. pylori infection. Using a statistical inference model, Strebel et al. [60] found “more than three children living in the household”, “more persons living per m² than average”, “home situated at main road” and “using well water” to be strongly associated with H. pylori infection.

Infected Siblings

Several studies [20,26,43–45] consistently supported infected siblings as a risk factor for H. pylori infection and these have been discussed earlier.

Race

Some studies examined the effect of race on H. pylori infection. Epplein et al. [29] recruited low-income African American and white patients into a large prospective study involving twelve southeastern states of the USA. Prevalence rates were inordinately high for both groups compared with known published prevalence rates among white Americans [61]. Interestingly, the amount of African ancestry using “ancestry informative genetic markers” predicted the prevalence of H. pylori with the highest African ancestry correlating with the highest H. pylori prevalence rates after adjustment for education, socioeconomic, and other environmental factors. This finding points to a possible genetic susceptibility to H. pylori infection. Fraser et al. [10] in a study on iron deficiency in New Zealand showed a difference in H. pylori prevalence according to ethnicity, being highest among Pacific Island students followed by Maori and Asian students, and lowest in European students. This confirms earlier observations made by Fraser et al. among different ethnic groups in New Zealand [62]. On the other hand, Muhsen et al. [43] found that among Arab Israelis living in three villages in northern Israel, H. pylori prevalence rate correlated with the socioeconomic status of the village, although ethnically they were all the same. Pandeya et al. [9] also observed differences in H. pylori prevalence between individuals born in Australia and New Zealand compared with those born overseas, the rate being lower in the former.

Public Health Implications

The epidemiological studies published in the past year support the evidence that H. pylori is most common in impoverished areas with overcrowding and poor sanitation. Transmission occurs during childhood through an oral–oral or a fecal–oral route. Dattoli et al. [20]. demonstrated this very well in their study on risk factors, and Cervantes et al. [44], for example, identified early childhood with transmission between siblings as an important mode of transmission of infection. Public health measures should be targeted to alleviate poor living conditions which will in turn result in decreased transmission and reduction of the reservoir of infection.

There is conflicting data on the association of H. pylori infection with anemia. Some studies did not find any associations [22,24] while others did [10,63]. The association between H. pylori infection and anemia was addressed in recent review articles [64,65]. H. pylori infection has been reported to negatively impact child growth in one study [23], but overall data continue to show a lack of such association as pointed out in a review article [66]. Nevertheless, this is of great concern particularly in high prevalence areas as it may impact significantly on the well-being of a community or population.

There were two articles that looked at the outcome of H. pylori eradication and the development of gastric cancer, which is the most serious outcome of H. pylori infection. Kosunen et al. [67] in a large longitudinal cohort follow-up study for 10 years noted a marked decline in gastric cancer incidence following H. pylori eradication. In a second study from Japan, Take et al. [68] in another cohort follow-up study showed that gastric cancer developed at a rate of 0.30% per year even after H. pylori eradication. This indicates as we have known before that once pre-malignant changes have already developed, a “point of no-return” is reached.

In Japan, annual screening gastroscopy for gastric cancer has been implemented for a long time. Mizuno et al. [69] published an important paper which showed that pre-screening high-risk individuals in the population with serum pepsinogen and H. pylori serology can identify those with high risk of developing gastric cancer who can then undergo gastroscopy. In this population-based cohort study, participants were followed up for a total of 9 years and the incident cases of gastric cancer were recorded. Those with H. pylori and atrophy had an 11-fold increased risk of developing gastric cancer, but the highest risk was with those with absent H. pylori but presence of atrophic gastritis indicating a group with longstanding severe gastritis from which H. pylori disappeared.

Several review papers addressed the issue of prevention and elimination of gastric cancer in Japan. Asaka et al. [70] in a review paper on “strategies on eliminating gastric cancer” proposed gastric cancer screening by simultaneous measurement of serum pepsinogen and H. pylori antibody as described earlier by Mizuno et al., combined with eradication of H. pylori in all individuals at risk. He commented that the proposed strategy would prevent about 150,000 deaths from gastric cancer during the 5 years after its adoption and would probably reduce the incidence of gastric cancer by more than 80–90% within 10 years. In another review paper, Asaka et al. [71] reported on a study carried out by the Japan Gast Study Group which showed in a randomized study the effect of H. pylori eradication for prevention of recurrent gastric cancers following endoscopic mucosal resection. Shiota et al. [72]. discussed the most recent update on the Japanese Society for Helicobacter Research guidelines in 2009 [73] which has emphasized the importance of H. pylori eradication in preventing gastric cancer. The most important revision was the recommendation that all H. pylori-infected subjects be treated and eliminated regardless of clinical outcome. H. pylori eradication for all infected subjects will prevent not only H. pylori related diseases but also the spread of bacterium in future.

Harvey et al. [74] in the Bristol Helicobacter project found that the effect of H. pylori eradication was cost beneficial. Eradication of H. pylori infection in the community gives cumulative long-term benefit, with a continued reduction in the development of dyspepsia severe enough to require a consultation with a general practitioner up to at least 7 years. The cost savings resulting from this aspect of a community H. pylori eradication program, in addition to the other theoretical benefits, make such programs worthy of serious consideration, particularly in populations with a high prevalence of H. pylori infection.

Of public health interest too is a study by Feinstein et al. [75] who analyzed hospital discharge data from 1998–2005 in the USA using the Nationwide Inpatient Sample database. The overall peptic ulcer disease (PUD) hospitalization rate declined from 71.1 to 56.8 per 100,000 population from 1998 to 2005. At the same time, the H. pylori-related hospitalization rates also decreased from 35.9 to 19.2 per 100,000 population. The authors suggested that the decline in PUD hospitalization was because of the decline in H. pylori related complications.