Seroprevalence of Helicobacter pylori in Korea: A multicenter, nationwide study conducted in 2015 and 2016

Abstract Background The Korean College of Helicobacter and Upper Gastrointestinal Research has studied Helicobacter pylori (H. pylori) prevalence since 1998 and found a dynamic change in its prevalence in Korea. The aim of this study was to determine the recent H. pylori prevalence rate and compare it with that of previous studies according to socioeconomic variables. Methods We planned to enroll 4920 asymptomatic Korean adults from 21 centers according to the population distribution of seven geographic areas (Seoul, Gyeonggi, Gangwon, Chungcheong, Kyungsang, Cholla, and Jeju). We centrally collected serum and tested H. pylori serum IgG using a chemiluminescent enzyme immunoassay. Results We analyzed 4917 samples (4917/4920 = 99.9%) from January 2015 to December 2016. After excluding equivocal serologic results, the H. pylori seropositivity rate was 51.0% (2414/4734). We verified a decrease in H. pylori seroprevalence compared with previous studies performed in 1998, 2005, and 2011 (P < .0001). The H. pylori seroprevalence rate differed by area: Cholla (59.5%), Chungcheong (59.2%), Kyungsang (55.1%), Jeju (54.4%), Gangwon (49.1%), Seoul (47.4%), and Gyeonggi (44.6%). The rate was higher in those older than 40 years (38.1% in those aged 30‐39 years and 57.7% in those aged 40‐49 years) and was lower in city residents than in noncity residents at all ages. Conclusions Helicobacter pylori seroprevalence in Korea is decreasing and may vary according to population characteristics. This trend should be considered to inform H. pylori‐related policies.


| INTRODUCTION
Helicobacter pylori (H. pylori) has infected more than half of the world's population 1 and is an important cause of gastric cancer, mucosalassociated lymphoid tissue lymphoma, and peptic ulcer disease. 2 Chronic infection with H. pylori is strongly associated with gastric cancer 3 the highest incidence of which is observed in Korea, Mongolia, Japan, and China. 4 Eradication of H. pylori has thus been attempted in China and Japan to reduce gastric cancer levels. 5,6 Accordingly, determination of the H. pylori prevalence of normal asymptomatic participants is crucial for the establishment of national health policies in these Eastern Asia countries. Nationwide studies of H. pylori prevalence were performed in Korea in 1998Korea in , 2005, and 2011. [7][8][9] Although these studies obtained data from a large number of participants, they had several no-  Informed consent was obtained, and a questionnaire on socioeconomic status was administered by a physician or nurse. The questionnaire included family history of gastric cancer, family income, education status, and habitation pattern in preschool, school, and posthigh school periods. Family history of gastric cancer was confined to parents, siblings, and children. Family income was divided into low (<US $3000 per month), medium (US $3000-10,000 per month), and high (>US $10,000 per month). Education status was divided into low (middle school graduate or less), medium (high school graduate or university dropout), and high (university graduate or more). We investigated two aspects of habitation status, geographic area, and type of residence in terms of city or noncity during each life period.

K E Y W O R D S
Helicobacter pylori, Korea, prevalence 71%, respectively. 10 We used only positive (≥1.10 U/mL) or negative (<0.9 U/mL) results, and equivocal results (0.9-1.09 U/mL) were excluded from the analysis.

| Statistics
A sample size of 4920 was calculated to obtain a two-sided 95% confidence interval with a width equal to 0.028, assuming a H. pylori infection rate of 55% based on a previous study. 9 We distributed par- We only analyzed the results of asymptomatic participants from previous studies. A significance level of P < .05 was applied to all analyses except for multiple comparisons. We used the Bonferroni correction to calculate the P values for multiple comparisons.

| Seropositivity of participants and comparison with the results from 1998, 2005, and 2011
We enrolled 4963 asymptomatic participants from 21 centers, and 4917 samples were found to be suitable for the H. pylori IgG test ( Figure 1).

| Risk factors for H. pylori seropositivity
Age, body mass index (BMI), geographic area, and education level were significantly associated with H. pylori seropositivity ( Table 2).

| Impact of habitation according to life period on H. pylori seropositivity
We investigated H. pylori seropositivity according to habitation type (city vs noncity) and life period ( obesity and H. pylori infection is controversial, and the causality of these associations has not been proven. 2,21,22 We also investigated the correlation between BMI and residence style after high school (city vs noncity) but observed no significant correlation between the two (data not shown).
We found no sex differences in the prevalence of H. pylori (OR 1.00, 95% CI: 0.88-1.15), but a recent meta-analysis of 169 studies reported that male sex was associated with a higher prevalence of H. pylori (OR 1.12, 95% CI: 1.09-1. 15). 23 Our current study was not designed to determine sex differences and several confounding factors were not assessed (such as smoking history, urinary tract infection history in women, and sex hormones).
Our study had some limitations of note. First, there may have been a selection bias. We enrolled asymptomatic participants from tertiary hospitals or their health screening centers. Thus, we may have enrolled more participants with a higher socioeconomic status. Second, there could have been a recall bias. We considered the habitation status of our study patients but this is a relatively subjective parameter and might be recalled incorrectly. Third, although this was a prospective study, the enrollment period was relatively long. However, most participants (91%) were enrolled in 2015.
In conclusion, our current multicenter, nationwide study found a decrease in H. pylori seroprevalence in South Korea and a difference in the seroprevalence rate according to geographic area and habitation type. Our findings could be useful as future baseline data or to inform H. pylori-related policies in Korea. We compared groups 1-4 by considering the relatively large participant number. A P value less than .008 was considered significant due to multiple comparisons. Group 1 vs 2, P < .001; group 1 vs 3, P < .001; group 1 vs 4, P < .001; group 2 vs 3, P = .618; group 2 vs 4, P = .164; group 3 vs 4, P = .044.
T A B L E 3 Helicobacter pylori seropositivity according to habitation type and life period