Higher Proportions of a Healthy Gastric Mucosa in Healthy Japanese Adults with Later Birth Year: Analysis of 41,957 Participants

Purpose: In Japan, most gastric cancers are associated with gastric mucosal atrophy caused by chronic infection with Helicobacter pylori (H. pylori). To recognize the condition of the gastric mucosa and to determe the infection status of H. pylori are important for predicting the individual risk of gastric cancer. The present study aimed to determine the proportion of Japanese adults with a healthy gastric mucosa (without H. pylori infection) among 12 birth-year groups encompassing 1935 to 1990. Methods: The gastric mucosa was classied as healthy or having gastritis based on routine double-contrast upper gastrointestinal barium X-ray radiopgraphy examination (UGI-XR). The participants were 41,957 healthy Japanese adults. Serum or urine H. pylori antibody levels were also assessed. Results: In total, 25,424 participants had a healthy mucosa without a history of H. pylori eradication. The proportions of participants with a healthy mucosa by birth year were 19.8% (57/288), 27.1% (306/1,128), 32.4% (569/1,756), 37.6% (1,808/4,811), 49.2% (3,207/6,522), 60.1% (3,966/6,550), 71.2% (5,224/7,342), 77.2% (5,114/6,624), 80.6% (3,342/4,149), 85.0% (1,404/1,652), 85.3% (302/354), and 94.7% (125/132) in 1935, 1940, 1945, 1950, 1955, 1960, 1965, 1970, 1975, 1980, 1985, and after 1990, respectively (p for trend < 0.01). All participants with a healthy mucosa showed negative results in H. pylori antibody tests. Conclusion: The proportion of participants with a normal gastric mucosa increased linearly with birth years. Prevalence of a morphologically healthy gastric mucosa may have been increasing, in parallel with prevalence of H. pylori infection has been decreasing.


Introduction
In Japan, double-contrast upper gastrointestinal barium X-ray (UGI-XR) has been used for organized gastric cancer screening based on national policy since 1983. In 2005, Japanese guidelines for gastric cancer screening highlighted evidence suggesting that UGI-XR can help to reduce gastric cancer death [Hamashima et al. 2008]. After its discovery in the stomach (Warren and Marchall.1983), Helicobacter pylori (H. pylori) was determined to be a carcinogen leading to gastric cancer (WHO 1994). In Japan, most gastric cancers are associated with chronic infection and atrophy caused by H. pylori (Uemura et al. 2001). The prevalence of gastric cancer is very low in those with H. pylori-negative healthy mucosa (Matsuo et al.2011), while one-sixth to one-fourteenth of those with H. pylori infection have gastric cancer until 85 years of age (Kawai et al. 2021). Therefore, when judging the need for further examination after gastric cancer screening with UGI-XR, it is important to recognize the condition of the gastric mucosa individually, as this can aid in predicting the risk of gastric cancer. Thus, diagnosis of the background gastric mucosa has become a proactive strategy (Nakajima 2019).
Since Asaka et al. (1992) reported the prevalence of H. pylori infection in the Japanese population based on serum levels of H. pylori antibodies, there have been many reports on the prevalence of H. pylori infection. These studies reported H. pylori prevalence is decreasing in Japan and consequently age-adjusted incidence of gastric cancer is decreasing. However, most studies only used H. pylori antibodies in the serum or urine as the method of detection (Wang et al. 2017). Few epidemiological studies have observed difference in morphological gastric mucosal status using UGI-XR depending on birth years to date.
The present study aimed to determine the proportion of Japanese adults with a healthy gastric mucosa among 12 birthyear groups encompassing 1935 to 1990. The subjects were 41,957 participants who underwent UGI-XR in a health check center.

Ethics
All authors had access to the study data and reviewed and approved the nal manuscript. Our study was approved by the Ethics Committee of Ibarakiken Medical Center (2020-001) on March 31, 2020. This cross-sectional study was conducted in accordance with the World Medical Association Declaration of Helsinki, and informed consent was obtained from the participants. Before the participants underwent UGI-XR and/or endoscopic examination at our institute, separate informed consent was obtained; moreover, they were provided with the choice to opt out of the study. The participants were also informed that they had the right to refuse the use of their data for this study at any time. For easy communication, our contact information including, address, phone number, and email address were provided along with the informed consent. The same information was displayed on the notice board at the entrance of the institute and where the medical examination was conducted. The choice to opt out was approved by the ethics committee.

Assessments
Ibarakiken Medical Center is a health check-up center. Our institute assesses the gastric mucosal background in patients who undergo double-contrast screening (UGI-XR) and determines the need for endoscopy. UGI-XR was performed using a digital radiographic system with an X-ray image intensi er (I.I.-DR) (DREX-AA28M2/I, Toshiba) between 2007 and 2015, and with the X-ray at panel detector (FPD) (DIAVISTA, Hitachi) between 2016 and 2019. UGI-XR was performed using two or three types of barium with 210% W/V barium contrast, 5 g of gas-forming agent, and 130 mL total volume. Among 14 UGI-XR images taken daily, six double-contrast images were selected to assess the gastric mucosal background. These included a frontal image in the supine position, a right anterior oblique view in the near supine position, a left anterior oblique view in the near supine position, a frontal view in the prone position with the head down, a right anterior oblique view in the supine standing position for the upper body, and an anterior oblique view in the near supine position ("barium-divided" image).
The mucosa was considered healthy when the areae gastricae and folds were smooth in all mucosal surfaces of the stomach (Fig. 2a, Fig. 3a Radiographs were interpreted according to consensus decisions. The gastric mucosa was assessed rst, followed by H. pylori antibody results. Statistical analyses were performed using EZR (Kanda 2013) or R (http://cran.r-project.org), including the Cochran-Armitage test.

Results
In total, 26,073 participants were found to have a healthy mucosa, and 649 individuals had a history of H. pylori eradication. Finally, data were analyzed for 25,424 participants with a healthy gastric mucosa.  1935,1940,1945,1950,1955,1960,1965,1970,1975,1980,1985, and after 1990, respectively (Table 1, Figure 4). The proportion of participants with a healthy mucosa increased linearly with birth year (p < 0.001). No H. pylori antibodies were detected among participants with a healthy mucosa. Two types of gastritis were known before the discovery of H. pylori infection: type A gastritis, which is characterized by diffuse, atrophic gastritis, except in the antrum, with high gastrin levels and autoantibodies in gastric parietal cells; and type B, which refers to diffuse gastritis that spreads from the antrum with advancing age (Strickland and Mackay 1973). Type B gastritis is now considered to be associated with persistent H. pylori infection, while type A gastritis is classi ed as autoimmune gastritis. Since we diagnosed a healthy gastric mucosa rst in the diagnostic procedure, the presence of autoimmune gastritis exerted no effect on the results. Several reports have suggested that the prevalence of H. pylori infection depends on the environment during childhood, especially the condition of drinking water (Klein 1991;Mendall 1992: Olmos 2000. In Japan and the Ibaraki prefecture, the water supply increased between 1960 and 1990 (https://www.mhlw.go.jp/content/000624219.pdf; http://mobile.pref.ibaraki.jp/hokenfukushi/seiei/suido/seiei/suido/02abot/documents/h25ibarakikennosuidou.pdf).
Taken together these results and the results from our study, there may be a relationship between the decline of H. pyloriassociated gastritis and the water supply.
In Japan, organized gastric cancer screening using UGI-XR for residents over 40 years of age was o cially launched in 1983. Although we did not analyze whether H. pylori-associated gastritis played a role in the pathogenesis of gastric cancer, gastric cancer screening was performed for gastritis in a majority of participants. Wang et al. (2015) reported that gastric cancer mortality rates in the Japanese population decreased in younger birth cohorts. Indeed, the ageadjusted incidence rate and mortality rate in Japan have decreased since 1975 (  This study has several limitations. First, only participants from one institution were assessed. However, the participants were from the whole Ibaraki area and the sample size is relatively large. Sampling bias may have exerted little effect on the results in this study. Second, the diagnostic accuracy of UGI-XR might be a problem. In this study, all subjects with a healthy gastric mucosa showed negative results in the serum or urine H. pylori antibody test, which seemed to reinforce the accuracy of UGI-XR diagnosis. Kamada et al. (2015) reported the prevalence of H. pylori-associated gastritis based on histological ndings of biopsy specimens between the 1970s, 1990s, and 2010s. The prevalence of H. pyloriassociated gastritis was the highest in the 1970s group and the lowest in the 2010s group. The proportion of participants with gastritis in the 1990s group using their biopsy specimens in the study were similar to that of our results using UGI-XR images. Thus, we think the diagnoses of a healthy mucosa and gastritis in our study made by UGI-XR may be valid.
Finally, additional endoscopic examination results or histological results of the biopsied samples were not used.
Endoscopic examination, especially diagnosis of H. pylori infection using biopsy samples, does not necessarily have high sensitivity (Kato et al. 2019). Thus, we assert the accuracy of UGI-XR for assessing the gastric mucosal background.
In summary, our study assessed differences in the proportions of participants with a healthy mucosa using UGI-XR depending on birth years. The ndings indicate that prevalence of a morphologically healthy gastric mucosa has been increasing in parallel with decreasing prevalence of H. pylori infection, which may make it necessary to reconsider screening procedures to account for the changes in the prevalence of H. pylori infection.
Declarations Figure 1 Study ow chart  Proportion of participants with a healthy gastric mucosa (%). The X-axis shows the birth year, and the Y-axis shows the proportion of participants with a healthy gastritc mucosa. Horizontal lines show point estimates of the proportion, and vertical lines represent their 95% con dence intervals. The proportion of individuals with a healthy gastric mucosa increased with birth year (p < 0.001)