Prevalence, risk factors, and virulence genes of Helicobacter pylori among dyspeptic patients in two different gastric cancer risk regions of Thailand

Gastric cancer risk is varied among different regions of Thailand. We examined the characteristics of Helicobacter pylori infection in two regions of Thailand. The H. pylori status of 273 dyspeptic patients (136 from the South and 137 from the North; a low and high incidence of gastric cancer region, respectively) was evaluated, and virulence genotypes (cagA, vacA, hrgA and jhp0562-positive/β-(1,3)galT) were determined. The overall H. pylori infection rate was 34.1% (93/273). The prevalence was higher in the North than in the South (50.4% vs. 17.6%, P <0.001) and was significantly higher among individuals with the following characteristics: low income, birthplace in the Northeast or North regions, agricultural employment, or consumption of alcohol or unboiling water. Among these socio-demographic determinants, region was an independent risk factor for H. pylori infection (odds ratio = 6.37). Patients including both H. pylori infected and uninfected cases who lived in the North had significantly more severe histological scores than those in the South. In contrast, among H. pylori-positive cases, patients in the South had significantly more severe histological scores than those in the North. Of the 74 strains cultured, 56.8% carried Western-type cagA, with a higher proportion in the South than in the North (76.2% vs. 49.1%, P = 0.05). In disagreement with the current consensus, patients infected with the Western-type cagA strains had more severe inflammation scores in the antrum than those infected with the East Asian-type cagA strains (P = 0.027). Moreover, Western-type cagA strains induced more severe histological scores in patients from the South than those of either genotype from the North. Other virulence genes had no influence on histological scores. The incidence of gastric cancer in Thailand was different among regions and corresponded to differences in the prevalence of H. pylori infection. More careful follow-up for patients in the South will be required, even if they are infected with H. pylori carrying Western-type cagA.

Introduction For more than a century, it was believed that the presence of bacteria in the human stomach came from digested food contaminants, until the identification of a gram-negative bacteria, Helicobacter pylori, which is a causative agent of various severe gastroduodenal diseases, including gastric cancer [1]. Accordingly, the risk of developing gastric cancer is reduced following H. pylori eradication, and eradication should be performed before there is irreversible damage done to the patient [2].
Several environmental factors are associated with an increased risk of H. pylori infection, and these factors are found among those living in developing countries and in lower socioeconomic groups in the developed world, probably due to poor hygiene standards and crowded households [3]. A previous study showed that improved hygiene conditions have association with the prevalence of H. pylori infection in many parts of North America and Europe [4]. H. pylori infection rates also decreased in those born after 1950, simultaneously with a rapid change in sanitary conditions and in the standard of living. These changes include the development of a clean public water systems in Japan following World War II [5]. In addition to bacterial and host factors, there is a positive correlation between gastric cancer risk and smoking habits [6] and alcohol consumption [7], in addition to the fermented food [8], salt [9], and nitrates [10,11].
It is believed that multistep and multifactorial precancerous processes that are associated with persistent chronic mucosal inflammation and the direct action of H. pylori virulence factors determine the severity of disease and long-term patient outcome [12]. It has been reported that individuals infected with cagA-positive H. pylori strains have an increased risk of peptic ulcer disease, atrophic gastritis, intestinal metaplasia, and gastric cancer, compared to individuals who are infected with cagA-negative strains [13,14]. The second repeat of the C-terminal region of CagA sequences has a difference between East Asian-type CagA and Western-type CagA. East Asian-type CagA has a higher binding affinity for the Src homology-2 domain-containing phosphatase 2 (SHP2), resulting in a greater risk of peptic ulcer development and/or gastric cancer when compared to its Western counterpart [15][16][17][18]. In addition, examination of the sequence of the 300 bp region upstream of the first EPIYA (Glu-Pro-Ile-Tyr-Ala) motif in the C-terminal region of CagA, denoted hereafter as the pre-EPIYA region, revealed that a 39-bp deletion is present in most strains isolated from East Asia, but is absent from most strains from Western countries (denoted as the no deletion type) [19]. A previous study also showed that the EPIYA sequence in the EPIYA-B motif is more highly associated with gastric cancer than the EPIYT sequence is, which might be due to differences in its affinity for phosphoinositol 3-kinase and in its induction of protein kinase B [20]. Finally, a conserved sequence motif of 16 amino acids (FPLXRXXXVXDLSKVG) in the C-terminal region of CagA is a prerequisite for the CagA-SHP-2 interaction, and 11 of the 16 amino acids in the CagA multimerization (CM) motif are well conserved between Western-(containing multiple CM motifs) and East Asiantype (containing a single CM motif) CagA strains [21]; therefore, the type and number of CM motifs may influence the potential of individual CagA proteins to multimerize in host cells [22].
cagA status is linked to the vacA genotype, which causes variations in the vacuolating activity of different H. pylori strains. The s1m1i1-type strains produce vacuolating toxin (VacA) and are more likely to be associated with gastric cancer than the s2m2i2 type, which is a nonvacuolating type [23]. Other regions, such as a deletion (d) [24,25] and a polymorphic site in the 3 0 end of vacA, denoted as the c region, have been reported to be better predictors of disease severity than either the s or m region [26]. Additionally, there have been one report that the presence of hrgA is increased among gastric cancer patients in East Asia [27], although the subsequent studies from other group could not confirm the association [28,29]. There were also several reports that the presence of jhp0562 and β- (1,3)galT is associated with the development of peptic ulcers [30,31].
Thailand is a country in Southeast Asia with an H. pylori infection rate ranging from 54.1% to 76.1% [32]; however, the age-standardized incidence rate (ASR) of gastric cancer has been reported to be 3.1/100,000, which is relatively low among Asian countries (data available from the International Agency for Research on Cancer; GLOBOCAN2012, http://globocan.iarc.fr/). Importantly, gastric cancer risk in Thailand is varied among different regions. The North has the highest incidence rate (ASR: 7.9 for males and 5.2 for females); in contrast, the South has the lowest incidence rate (ASR: 2.0 for males and 1.4 for females) in Chiang Mai and Songkhla province, respectively [33]. It is important to analyze factors affecting the different gastric cancer risk between these two regions. In this study, we examined the prevalence of H. pylori infection and also identified and analyzed environmental factors in both regions. Furthermore, we collected H. pylori strains from both regions and evaluated the virulence factors of those strains.

Study population
We performed a cross sectional study in dyspeptic patients who were living in Muang Ranong city, Ranong province, which is in the South of Thailand, on August 28, 2015. From January 19-20, 2016, we also recruited dyspeptic patients from two cities in the Chiang Rai province, which is in the North: Chiang Kong and Chiang Saen (Fig 1). The project was announced by local hospital officers and all the participants were living in the community. Questionnaire information was gathered before an endoscopy. Experienced endoscopists collected three gastric biopsy specimens during each endoscopy session: two samples from the antrum, approximately 2-3 cm from the pyloric ring, and one sample from the greater curvature of the corpus. To minimize potential bias, we used the same experienced pathologist (TU), who has also performed histological evaluations of patients from Nepal, Myanmar, Vietnam, Bhutan, the Dominican Republic, and Indonesia [34][35][36][37][38][39][40]. Biopsy specimens for bacterial culture were immediately placed in transport media and were put in a mobile freezer at -20˚C and were transported to Thammasat University Hospital, Thailand within a day of collection and were kept at -80˚C. Then, we transported the biopsy specimens with dry ice to the Central Laboratory in Oita University Faculty of Medicine, Japan. All biopsy specimens could be kept frozen during the transportation and were directed to culture within 12 hours upon arrival of the specimens. Overall, although the total duration time from collecting biopsy specimens in Thailand to starting culture in Japan varied (average 20 days), duration time with the temperature above -80˚C was less than 24 hours. Two antral specimens were used for the H. pylori culture and histological examinations. One corporal specimen was used for histological examinations. Peptic ulcers were identified through endoscopy. Written informed consent was obtained from all participants, and the study protocol was approved by the Ethics Committee of Faculty of Medicine, Thammasat University (Pathum Thani, Thailand) and Oita University Faculty of Medicine (Yufu, Japan).

H. pylori infection status
H. pylori infection status was diagnosed based on culture and histology, which was confirmed by immunohistochemistry (IHC), as the tests could directly check for the presence of H. pylori. For the H. pylori culture, one antral biopsy specimen was homogenized and directly inoculated onto H. pylori selective media (Nissui Pharmaceutical co., LTD, Tokyo, Japan). The plates were incubated for up to 7 days at 37˚C under microaerophilic conditions (10% O 2 , 5% CO 2 and 85% N 2 ). The H. pylori like colony were sub-cultured on Mueller Hinton II Agar medium (Becton Dickinson, NJ, USA) supplemented with 10% horse blood without antibiotics (Nippon Biotest Laboratories Inc., Tokyo, Japan) and confirmation by gram staining (spiral red rod) [41]. Isolated strains were stored at -80˚C in Brucella Broth (Becton, Dickinson and Company, Sparks, MD, USA) containing 10% dimethyl sulfoxide and 10% horse serum.
For histology, one antral and one corporal specimen from each patient was examined. Biopsy materials were fixed in 10% buffered formalin and embedded in paraffin. Serial sections were stained with hematoxylin and eosin, as well as with May and Giemsa stains. The degree of inflammation, neutrophil activity, atrophy, intestinal metaplasia, and bacterial density were classified into four grades, in accordance with the updated Sydney system: 0, 'normal'; 1, 'mild'; 2, 'moderate'; and 3, 'marked' [42]. Samples with a grade of 1 or higher of atrophy were considered atrophy-positive. In addition, the gastritis stage was assessed based on the topographic location (antrum or corpus) of atrophy, in accordance with the Operative Link on Gastritis Assessment (OLGA) system [43].

Interluekin-8 production from gastric epithelial cell infected with H. pylori strains
The human adenocarcinoma gastric cell line (AGS) were seeded into 24-well plates at a density of 5×10 4 cells per well and incubated in RPMI 1640 (Nacalai Tesque, Inc., Kyoto, Japan) supplement with 10% heat-inactivated fetal bovine serum (Gibco, Grand Island, NY, USA) at 37˚C in humidified incubator containing 5% CO 2 for 2 days before infected with H. pylori strains. We selected 20 strains; 5 from the North with Western-type cagA, 5 from the North with East-Asian-type cagA, 5 from the South with Western-type cagA, 5 from the South with East-Asian-type cagA. As control, we used 2 reference strains (ATCC26695 with Western-type cagA and TN2 with East-Asian-type cagA) and mock-infection. All the H. pylori strains were cultured on Brucella agar supplement with 10% horse serum plate for 24 hours at 37˚C before infection (Becton, Dickson and company, Sparks, MD, USA). The AGS cells were co-cultured with H. pylori at an MOI of 100 and incubated at 37˚C in humidified incubator containing 5% CO 2 for 24 hours. The medium was removed for measuring the IL-8 production by using Human IL-8 ELISA kit (eBioscience, Inc., Science Center Dr., San Diego, CA, USA) following the manufacturer's instruction. All samples were measured in duplicate in three independent experiments.

Statistical analysis
Discrete variables were tested using a chi-square test (the prevalence of H. pylori, demographics and sanitation status, and genotypes vs. regions or diagnosis). The difference of histologic score and IL-8 levels between genotypes were tested using the Mann-Whitney U test. A multivariate logistic regression model was used to calculate the odds ratios (OR) of the clinical outcomes. All determinants with P values less than 0.10 were entered together into the full logistic regression model, and the model was reduced by excluding variables with P values greater than 0.10. OR and a 95% confidence interval (CI) was used to estimate the risk. A P value less than 0.05 was accepted as statistically significant. The SPSS statistical software package version 18.0 (SPSS, Inc., Chicago, IL) was used for all statistical analyses.

Prevalence of H. pylori infection
We initially registered 340 dyspeptic patients (170 each from the North and South). Among them a total of 306 patients went to the hospital on the scheduled day and were recruited consecutively: Ranong in the South (n = 158) and Chiang Kong (n = 102) and Chiang Saen (n = 46) in the North. We excluded 33 patients: 8 patients had missing histology information, and 25 patients had histories of H. pylori eradication. Therefore, the total study population was 273 patients (265 with gastritis and 8 with peptic ulcers). The population consisted of 136 patients from the South (93 females and 43 males; mean age of 52.0 ± 14.7 years; age range of 15-82 years) and 137 patients from the North (85 females and 52 males; mean age of 56.4 ± 13.6 years; age range of 21-88 years).
The prevalence of H. pylori infection was the highest when samples were examined with a histological test (32.3%, P <0.001), while it was 27.1% by bacterial culture (Table 1). We determined that the patient was H. pylori-positive if culture and/or histology confirmed by IHC had positive results; the final prevalence was determined to be 34.1%. The prevalence in the North was higher than that in the South (50.4% vs. 17.6%, P <0.001) ( Table 1). As there was no difference in the prevalence between the two cities in the North, Chiang Kong and Chiang Saen (50.5% vs. 50.0%, P = 0.95), we combined both cities in the subsequent analyses.

Demographics, sanitation, H. pylori, and region
The OR was calculated for H. pylori infection rate (Table 2 and S2 Table). Subjects who were born in the Northeast and North regions had significantly higher H. pylori infection rate than those born in the South (OR = 10.5 and OR = 5.34, respectively). The prevalence of H. pylori infection among subjects who had low income or consumed boiling water or alcohol consumption was significantly higher than the prevalence among subjects who had high economic status (OR = 3.00), boiling their drinking water (OR = 1.93), or did not consume alcohol (OR = 2.21), respectively. Agriculture was an occupation with the highest risk of H. pylori infection (OR = 3.92). In contrast, there were no statistically significant relationships between H. pylori infection rate and gender, ethnicity, marital status, body mass index, consumption of fried and fermented diet, or smoking habits. When we entered all determinants with P values of less than 0.10 by bivariate analysis (age, sex, region, income status, source of drinking water, occupation, and alcohol consumption) into the full logistic regression model, subjects who  Table 3 shows the demographics and sanitation status in both regions. Individuals from the North had a significantly higher mean age and lower economic status than individuals from the South (P = 0.01 and P <0.001, respectively). There were significant differences in ethnic composition and obesity of subjects in both regions. Unboiling water and alcohol consumption was higher among subjects who were living in the North than among those from the South (for both, P <0.001). There was no statistical difference seen with differing gender, marital status, consumption of fried or fermented food, or smoking habits in either region.

Gastric mucosal status and region
First, we examined the histology of patients including both H. pylori positive and negative cases. Histological findings scored for atrophy showed that 184 patients' antrum biopsies (67.4%) were grade 0; 77 (28.2%) were grade 1, 11 (4.0%) were grade 2, and none were grade 3. Additionally, 243 patients' corpus biopsies (89.0%) were grade 0, 25 (9.2%) were grade 1, 3 (1.1%) were grade 2, and only 2 (0.7%) were grade 3. When the biopsies with grade 1 or more were considered to be atrophy-positive, 88 patients (32.4%) had antral atrophy, and 30 patients (11.0%) had corporal atrophy. OLGA scores showed 179 biopsies (65.6%) were stage 0, 79 (29.0%) were stage I, 11 (4.0%) were stage II and only 3 (1.1%) were stage III, while stage IV was not found. Histological scores broken down by region (including both H. pylori positive and negative patients) are shown in S3 Table and S1 Fig. Patients who were living in the North had significantly higher histological scores in the antrum and OLGA scores than those from the South had. Moreover, activity, inflammation, and bacterial density scores in the corpus were significantly higher in patients who were living in the North than in those living in the South.
Next, we took the H. pylori infection into account. As expected, all histological scores, aside from intestinal metaplasia, were significantly higher in patients positive for H. pylori than in those who were negative for H. pylori (S3 Table and S2 Fig). There was a high prevalence of H. pylori infection in the North and low prevalence in the South, suggesting that H. pylori infection could partially explain the different incidence rates of gastric cancer among individuals from various regions in Thailand.

H. pylori genotype and gastric mucosal status
We focused on bacterial factors in order to determine whether there were different genotypes in the two regions. We were able to obtain 74 strains by bacterial culture: 53 from the North and 21 from the South (Table 5). In total, all strains contained cagA, and the Western-type cagA was predominant (56.8% of all strains). One strain each of AB, ABB, and ABBB type was regarded as Western-type cagA based on the sequence similarity of B segments with Westerntype cagA. Overall, 40.5% of strains possessed the East Asian-type cagA. Two strains were cagA-positive by PCR; however, the cagA genotype for these two strains was undetermined by sequencing, since the sequence primer pair did not amplify cagA.
Surprisingly, patients infected with strains carrying Western-type cagA had significantly higher inflammation scores than did patients infected with East Asian-type cagA (Table 4). This result was inconsistent with the current consensus. This statistical difference was still apparent, even after using multivariate analysis that adjusted for age, sex, location, and birthplace (OR = 3.98, 95% CI = 1.18 to 13.48, P = 0.026). The distribution of the cagA genotype by  Table 5). There was difference of the distribution of pre-EPIYA among both regions (P = 0.002). There was no difference in histological scores between individuals infected with strains carrying Western-type cagA and East Asiantype cagA within the same region. However, among patients infected with Western-type cagA strains, patients living in the South had significantly higher activity, inflammation scores in the antrum, and atrophy scores in the corpus than did patients in the North (Table 4, Fig 3). There was no such difference among patients infected with East Asian-type cagA strains between these two regions. We also analyzed the histological scores between Western-type cagA strains from patients living in the South and East Asian-type cagA strains from patients living in the North to confirm whether H. pylori strains from the South were more virulent than were strains from the North (Fig 4). Surprisingly, patients infected with Western-type cagA strains from the South had significantly higher activity, inflammation scores in the antrum, and atrophy scores in the corpus than did patients who were infected with East Asian-type cagA from the North (1.25 [1] vs. 0.72 [1], P = 0.007, 1.93 [2] vs. 1.24 [1], P < 0.001 and 0.5 [0] vs. 0. 16 [0], P = 0.046, respectively). We assume that H. pylori strains from the South were more virulent than those from the North. We further analyzed cagA in detail, including the pre-EPIYA motif, the EPIYA motif, and the CM motif. Sequence analyses of pre-EPIYA are shown in Table 5. The no-deletion type (typically observed in Western countries) was the predominant type. All Western-type cagA strains had the no-deletion type, while East Asian-type cagA strains had either the no-deletion (4/30, 13.3%), 6-bp (7/30, 23.3%), 18-bp (14/30, 46.7%), or 39-bp deletion type (5/30, 16.7%). The EPIYA motifs in these strains were also evaluated ( Table 6). In total, 224 EPIYA motifs were found in the 72 CagA proteins. The most common 5-amino acid EPIYA motif was EPIYA (200/224, 89.2%), in agreement with our previous studies in Japan and with 4,534 CagA protein sequences from three data sources [48,49]. The proportion of strains with EPIYT in the North was significantly higher than that in the South (19/160 [11.9%] vs. 0/63 [0.0%], P = 0.001). Importantly, all Western-type cagA strains from the South had the EPIYA sequence at the B motif, which was reported to be more highly associated with gastric cancer than was the EPIYT sequence [20].
The CM motif is the 16-amino acid sequence that is present in the C-terminal EPIYA-C or EPIYA-D segment. The conserved amino acid sequence motif is FPLXRXXXVXDLSKVG [22]. The typical Western-type cagA CM motif located within the EPIYA-C segments and/or immediately distal to the last repeat of the EPIYA-C segment contains the peptide sequence FPLKRHDKVDDLSKVG, while typical East Asian-type cagA carries only one CM motif (FPLRRSAAVNDLSKVG) [22,48]. The CM motif sequences of strains from Thailand are shown in Table 7 and Fig 5. All 30 East Asian-type cagA strains and 39 of the Western-type cagA strains (AB, ABB, and ABBB were excluded because they only contained one or no CM motif) were included. Among East Asian-type cagA strains, 80% (4/5) of strains from the South had CM motif sequences typically observed in East Asian strains, while the typical motif sequences were rare among strains from the North (1/25, 4.0%) ( Table 7, Fig 5A and 5B). Less than 30% of the 1 st and 2 nd motif sequences of Western-type cagA contained typical Westerntype CM motif sequences (11/39 (28.2%) and 8/39 (20.5%), respectively). There were reports that the sequence patterns of pre-EPIYA, the CM motif, and EPIYA sequences were related to pathophysiological activities of CagA [19,20,22]; however, we could not find an association between histological scores and their sequences patterns.
All of the vacA-carrying strains were of the s1 genotype. As for the m region, the vacA m1 genotype was predominant (68.9%) ( Table 5). i1, c1, and d1 were the predominant genotypes at their respective regions (98.6%, 55.4%, and 75.7%). All strains with the s1m1 genotypes (68.9%) also contained the i1 genotype. After examining all the vacA subtypes, it was determined that the s1m1i1c1d1 and s1m1i1c1c2d1 genotypes were predominant (32.4% and 24.3%, respectively). However, there was no association between vacA subtypes and histological scores or regions. The majority of isolates carried the hrgA gene (73.0%). jhp0562-positive/ β- (1,3)galT-negative and jhp0562/β-(1,3)galT-double positive were the predominant types (44.6% and 41.9%, respectively). There was no association between the status of either gene and histological scores or regions.

IL-8 levels produced from AGS cells infected with H. pylori
AGS cells were co-cultured with various H. pylori strains. As shown in S3 Fig, the South H. pylori strains had IL-8 production slightly higher than the North strains, but the difference were not statistically significant (average IL-8 levels: 2,371 vs. 2,246 pg/mL, P = 0.76). In contrast with our expectation, within the South strains; the East-Asian-type cagA strains tended to induce higher IL-8 production than Western-type cagA strains (2,602 vs. 2,139 pg/mL, P = 0.076). However, there was no such difference among two types of cagA strains in the North (2,274 vs. 2,217 pg/mL, P = 0.92).

Discussion
In this study, we examined the H. pylori infection status in the North and South of Thailand. One important finding was that the prevalence of H. pylori infection was significantly higher H. pylori in two regions of Thailand in subjects who were living in the North (50.4%) than in those living in the South (17.6%), which might explain the variation in gastric cancer risk in Thailand, where the incidence of gastric cancer is approximately three times higher in the North than in the South [33]. However, our findings showed that patients in the South had more severe histological scores than did those in the North among H. pylori-infected patients. Therefore, we further examined the virulence genotypes of H. pylori. Surprisingly, histological analysis showed that inflammation scores in the antrum were significantly higher in subjects infected with Western-type cagA strains than the scores were in patients infected with East Asian-type cagA strains, even after being adjusted for age, sex, location, and birthplace. Moreover, our findings showed that Western-type cagA strains from the South could induce more severe activity, inflammation in the antrum, and atrophy in the corpus than could either of the cagA-genotype strains from the North. These results suggested that H. pylori-infected patients from the South might have a higher risk of developing gastric cancer. There are several possible explanations for our unexpected data. First, H. pylori from the South has a higher proportion of strains with EPIYA at the B motif than does H. pylori from the North; this difference might be associated with its virulence. An analysis of Western-type cagA revealed evidence that strains with EPIYA sequences of the EPIYA-B motif were significantly more associated with gastric cancer than were those with EPIYT sequences [20]. Secondly, strains in the South might have unknown virulence gene(s) other than cagA/vacA. We only examined several putative virulence factors. Unexpected data in this study are not unique, and our previous study showed that Western- type cagA in Nepalese strains was more virulent than was East Asian-type cagA [40]. In addition, our previous study using IHC showed that most of the Mongolian population that was infected with H. pylori did not react with an anti-East Asian-type CagA-specific antibody [50], which has immunoreactivity only against the East Asian-type CagA strains and not against Western-type CagA strains [44]. These data suggested that in Mongolia, where the incidence of gastric cancer death is the highest worldwide, most strains were either cagA-negative or Western-type cagA. Therefore, it will be impossible to explain the virulence only by the genotypes of cagA alone or even by the combination of cagA with well-known virulence factors such as vacA. Since H. pylori contains approximately 1,500 genes, other unknown virulence factors of H. pylori might play important roles in its pathogenesis in Thailand. Further studies are necessary to search for new virulence genes using next-generation sequencing. Thirdly, factors other than bacterial factors play important roles for promoting pathological differences between the North and the South.We examined the ability to induce IL-8 from gastric cells and there was no definite relationship on IL-8 production between the cagA genotypes and the location. However, since IL-8 is only one of the pathological markers and further studies will be necessary by measuring other pathological markers to confirm the roles of bacterial factors on the pathogenesis. Finally, the most possible explanation is that subjects from the South may have increased gastric cancer risk after H. pylori infection because of an interaction between host genotypes and bacterial virulence. For example, the highest gastric cancer risk in a Western country is in individuals with a combination of the high-expression type of interleukin-1 alleles and colonization with vacA s1-type H. pylori, probably due to amplification of mucosal inflammatory response and inhibition of acid production [51].
Also recent studies show a mismatch of host (individual) ancestry versus bacterial (H. pylori) ancestry can lead to more severe histology. A study in Colombia showed that African H. pylori ancestry was relatively benign in humans of African ancestry but was deleterious in individuals with substantial Amerindian ancestry, suggesting that coevolution likely modulated disease risk [52]. In this study, we did not examine host factors in detail (e.g., single nucleotides polymorphisms of inflammatory cytokines), and these factors might explain the different rates of incidence of gastric cancer among H. pylori-infected patients in the North and South.
Here, we showed that several socio-demographic factors were associated with an increasing prevalence of H. pylori infection. A subject with a low income was significantly more likely to be infected with H. pylori than those with a high economic status. Additionally, the North had a greater number of subjects with low economic status than did the South. These data are in agreement with the fact that with the exception of Bangkok the total monthly income per household in 2013 in the North (19,267 THB, which is approximately 849.9 USD in February 2017) was lower than the income in the South (27,504 THB) (data available at Statistical yearbook Thailand 2013, http://web.nso.go.th). Age, birthplace in the North, unboiling water consumption, being a farmer, and greater alcohol consumption were risk factors for greater prevalence of H. pylori infection in Thai populations, although the region is the most important factor, based on multivariate analysis. Interestingly, the age groups of 40-49 and 50-59 had the highest risk for H. pylori infection; however the prevalence of H. pylori infection was decreased in patients over 60 which might not be associated with the timing of the improvement of sanitation in Thailand. Further study will be necessary for clarify the age related risk for H. pylori infection taking the previous infection in elder ages into account. Some other factors might also become risk factors for gastric cancer after further examination. Our previous study showed that Chinese individuals living in Thailand have a high gastric cancer risk [16]; however, individuals of Chinese ethnicity were not examined in this study. Although there was no association between fermented food consumption and regional differences, a previous study reported that a variety of dietary habits appear to greatly influence pan-gastritis frequency in Thailand, irrespective of H. pylori infection rate [53]. For example, the North is a mountainous area, and individuals consume fermented, salty sauces with hot and spicy foods daily, which are dietary habits that induce gastritis; on the other hand, people in the South, a coastal region, have seafood, beans, fresh vegetables, and fruits as their common dishes [53]. Similarly, gastric cancer is extremely common in individuals from the mountainous regions of Central and South America and is associated with dietary habits [54].
This study has several limitation. First, study population in this study came from dyspeptic patients; therefore the specifically was not the general population. However, our small number of preliminary data for volunteers without dyspeptic symptoms in each region showed the similar prevalence of H. pylori infection to that of dyspeptic patients (Subsomwong P et al., unpublished data). Further population-based study using the large number of participants will be necessary to confirm our current data. Finally, we only obtained samples from two cities in the North and one city in the South. Therefore, our results might not be generalizable across Thailand. Further investigation is necessary to obtain samples from other cities in both regions, as well as from other regions throughout the country. These studies are now in progress.

Conclusion
The prevalence of H. pylori infection in Thailand is associated with the incidence of gastric cancer. Importantly, Western-type cagA strains were even more virulent in patients from the South than were strains with either cagA genotype in patients from the North. Therefore, a more careful follow-up for patients infected with H. pylori in the South is required, even if these patients are infected with Western-type cagA strains.