Gastric cancer is one of the most common malignant tumors. In 2020, it was estimated that there were 1.09 million new gastric cancer cases and 0.77 million deaths from gastric cancer all over the world. Among all the new gastric cancer cases, more than 40% occurred in China. Meanwhile, gastric cancer caused about 12.4% of all cancer-related deaths, making it the third leading cause of cancer-related deaths in China (1).

Gastric cancer can be classified into two categories according to anatomical subsites: cardia gastric cancer and non-cardia gastric cancer (2–4). Due to different epidemiological characteristics and distinct pathogeneses, cardia and non-cardia gastric cancer are treated as two different diseases. Non-cardia gastric cancer is more common than cardia gastric cancer. In 2018, non-cardia gastric cancer accounted for up to 82% (0.85/1.03 million) of all gastric cancer cases around the world (2, 5). Regarding pathogeneses, Helicobacter pylori (H. pylori) has been proven to be one of the most important risk factors for non-cardia gastric cancer, with approximately 90% of non-cardia gastric cancer cases attributable to H. pylori infection worldwide in 2018 (5, 6), whereas no association was found between cardia gastric cancer and H. pylori infection (7, 8).

Given the strong association between non-cardia gastric cancer and H. pylori infection, reducing H. pylori prevalence has been listed as an important primary prevention strategy for gastric cancer prevention (9–11). Reliable estimation of H. pylori prevalence in non-cardia gastric cancer may be essential to the control and prevention of non-cardia gastric cancer, policy-making, and health resource allocation. To date, lots of studies conducted in China have reported the prevalence of H. pylori in non-cardia gastric cancer. However, the prevalence varied greatly across studies (12, 13). To the best of our knowledge, there is no study pooling the prevalence of H. pylori in non-cardia gastric cancer at the national level. Therefore, the primary objective of this meta-analysis was to estimate the pooled prevalence of H. pylori in non-cardia gastric cancer in China. Additionally, we also explored potential causes of heterogeneity in the reported prevalence.

Based on 5324 non-cardia gastric cancer patients from 55 studies, which covered more than half of provinces and autonomous regions in China, we conducted a meta-analysis to estimate the prevalence of H. pylori in non-cardia gastric cancer patients at the national level for the first time. We reported a comprehensive estimate of H. pylori prevalence in non-cardia gastric cancer patients in China at 66.5% (95%CI: 62%-71%). Moreover, substantial geographic variations were noted in the prevalence of H. pylori, which was the highest in Northwest China and the lowest in North China.

The results of our study indicated that the pooled prevalence of H. pylori in non-cardia gastric cancer was 66.5% (95%CI: 62%-71%) in China. The prevalence of H. pylori in non-cardia gastric cancer in different countries varied greatly, such as 38.5% in the United States, 57% in Spain, 79.5% in South Korea, and 91.8% in Japan (4, 21–23). This difference might be explained by the following reasons: socio-economic status, dietary habits, and racial disparities. Suerbaum S et al. have mentioned that there was a strong inverse correlation between H. pylori prevalence and socio-economic status. Populations with lower socio-economic status were more likely to be infected with H. pylori (24, 25). Lower social status was commonly accompanied by crowded living conditions and poor hygienic conditions, which might increase the risk of acquisition and transmission of H. pylori (24, 26). Several previous studies have revealed that the consumption of high-salt foods (such as pickles and preserved products) might be associated with an increased chance of H. pylori infection (27–29). Tsugane S et al. speculated that a high-salt diet might induce gastric mucosa damage and destroy mucosal barrier. These changes in gastric mucosa might lead to an increase in H. pylori infection (29). Data based on National Health and Nutrition Examination Surveys of the United States have also shown that racial disparities played a certain role in the prevalence of H. pylori. Compared with H. pylori prevalence in whites, the prevalence in African Americans was higher (30).

Several previous studies, especially in Japan, reported that the H. pylori prevalence in gastric cancer was as high as 99% when strict criteria were used to diagnose H. pylori infection (31–33). Some other studies in Japan and South Korea used multiple testing methods to diagnose H. pylori infection, which also showed that the H. pylori prevalence in gastric cancer patients was above 90% (34–36). The H. pylori prevalence of these studies was much higher than the pooled H. pylori prevalence in our study. There were several possible reasons for this difference. First, these previous studies used strict criteria or multiple testing methods to diagnose H. pylori infection, which could greatly increase the sensitivity of detection. Moreover, most participants of these studies were cases of gastric cancer, while the participants of our study were non-cardia gastric cancer. Finally, most of these studies were conducted in Japan or South Korea. The H. pylori prevalence varied in different countries and regions.

The results of our study indicated a significant heterogeneity in the prevalence of H. pylori in non-cardia gastric cancer across different geographical regions in China. The highest H. pylori prevalence was noted in Northwest China at 78.9%, which was consistent with the highest incidence of gastric cancer in this region reported by the 2016 Chinese Cancer Registry Annual Report (37). Meanwhile, in line with our findings showing that the lowest H. pylori prevalence in non-cardia gastric cancer was noted in North China, Zhang WD et al. have also mentioned that the lowest H. pylori prevalence among the general population was in North China (38). The socio-economic status might be the main factor accounting for this geographic region difference. Unbalanced socio-economic development in different regions would lead to differences in living environment, education level, hygiene conditions, as well as water source (26, 39–41). The combined effect of the above factors may have resulted in the variation of H. pylori prevalence in different geographic regions.

The prevalence of H. pylori in non-cardia gastric cancer also showed significant variations across different provinces, and the highest H. pylori prevalence was noted in Heilongjiang Province at 90%. However, the estimate of Heilongjiang Province was based on only 10 cases from one study, leading to a wide range of its 95% CI. Furthermore, there were another seven provinces where the estimates of H. pylori prevalence were also based on only one study with the sample size ranging from 38 to 238. At the provincial level, part of the estimates in our study were estimated in the basis of a relatively small sample size, which might result in unstable results. Therefore, the results should be interpreted with caution, and additional studies with a larger sample size are needed to confirm our findings in the future.

The prevalence of H. pylori in the general population and non-cardia gastric cancer cases from different geographic regions was not completely consistent. From January 2002 to June 2004, Zhang WD et al. conducted an epidemiological survey on the H. pylori prevalence among 26,341 general people from 19 provinces and autonomous regions in China. The study showed that the H. pylori prevalence in the general population from different geographic regions ranged from high to low as: 66.26% in Central China, 59.16% in Eastern China, 58.27% in Western China, 50.08% in Southern China, and 46.84% in Northern China. In terms of provinces, the H. pylori prevalence among general population in Tibet was the highest at 84.62%, while the H. pylori prevalence in Guangdong was the lowest at 42.01% (38). Our results indicated that the geographic region with the lowest H. pylori prevalence in non-cardia gastric cancer cases was noted in North China, which was in line with the previous survey conducted by Zhang WD et al. The possible reason for the difference in the distribution of H. pylori between general population and cases of non-cardia gastric cancer was that non-cardia gastric cancer was caused by a combination of various risk factors including H. pylori infection, dietary habits, ethnicity, smoking, radiation exposure, family history, etc. H. pylori infection is one of the most important factors, but not the only one (42).

Type of sample might have an influence on the estimates of prevalence of H. pylori in non-cardia gastric cancer. The results of subgroup analysis revealed significant heterogeneity across different types of samples, with the highest prevalence of H. pylori detected in breath (78.7%). However, the results of meta-regression indicated that the differences in H. pylori prevalence among different sample types were not significant. The disparity in results might be attributed to the small number of the studies included. This meta-analysis only included four studies with 278 cases that detected the H. pylori prevalence in breath. Spineli LM et al. have mentioned that in order to obtain robust results, at least 10 studies should be included for each covariate in meta-regression implementation. In order to further confirm our finding, more studies detecting the H. pylori prevalence in breath will be needed in the future (43). In line with our findings, Liao YQ et al. also mentioned that the prevalence of H. pylori detected in breath would be higher than that in other samples. This might be partly because H. pylori infection detected in breath could reflect the H. pylori infection status in the whole gastrointestinal tract. Other pathogenic microorganisms in the gastrointestinal tract that could produce urease might cause false positive in the detection, which might increase the positive rate of H. pylori (44, 45).

The results of subgroup analysis and meta-regression both showed that the prevalence of H. pylori in non-cardia gastric cancer was not significantly associated with different H. pylori testing methods. This finding was in agreement with a recent study estimating the prevalence of H. pylori in cases with gastrointestinal diseases other than gastric cancer, which showed that the H. pylori prevalence detected by different H. pylori testing methods were generally similar (46). In our study, all H. pylori detected in breath was tested using 14C urea breath test. However, type of sample was significantly associated with H. pylori prevalence, while the relationship between testing method and H. pylori prevalence was not significant. Spineli LM et al. have revealed that when the subgroups contained fewer studies, the subgroup analysis could be underpowered to test the relationship between variables (43). Borenstein M et al. also indicated that one of the key factors driving the precision of subgroup analysis was the number of studies (47). Compared with type of sample, H. pylori testing method contained more subgroups, resulting in less studies in each subgroup. This may be the possible reason that the results of the above two subgroup analyses were different.

The studies included in our meta-analysis used several single testing methods to detect H. pylori infection among non-cardia gastric cancer cases. However, several guidelines revealed that one single testing method could not be considered as the gold standard for H. pylori detection (48–50). These commonly used testing methods all have some disadvantages. For example, IgG would remain in the blood for months or years even after H. pylori was eradicated. As such, antibody-based tests (e.g., ELISA) could not distinguish between current and past infections (51, 52). Moreover, ELISA is less accurate than 14C urea breath test and the cut-off values need a local validation (48, 53). As for 14C urea breath test, several factors such as atrophy, bismuth, proton pump inhibitor (PPI), and antibiotics may lead to false-negative, and it also should be validated locally (48, 54). Several guidelines have recommended the combination of single testing methods (e.g., combination of a validated serology and urea breath test), which could improve the accuracy of the detection (48–50). However, in our meta-analysis, most included studies (43/55, 78.2%) used a single testing method to detect H. pylori, which might lead to bias in the pooled H. pylori prevalence among non-cardia gastric cancer cases in China. Because H. pylori infection tends to clear as non-cardia gastric cancer progresses and the detection of past H. pylori infection is difficult, especially in retrospective studies (19/55, 34.5%), the pooled H. pylori prevalence in Chinese non-cardia gastric cancer cases may be underestimated (35, 55, 56).

The results of our study also suggested that there was no correlation between the diagnosis period and the prevalence of H. pylori in non-cardia gastric cancer. Previous meta-analysis conducted by Qiao J et al. showed that the prevalence of H. pylori in cases with gastric cancer in China declined from 1996 to 2015 (57). However, no decreasing trend of H. pylori prevalence in non-cardia gastric cancer was observed in our study. This difference might be attributed to the different study population. The previous study included cases with gastric cancer, while the included cases in our study were limited as cases with non-cardia gastric cancer.

The “test and treat” strategy for H. pylori infection has been shown to be cost-effective in some western countries (e.g., the United Kingdom and the United States) (58, 59). Japan has included H. pylori eradication therapy into the coverage of national medical insurance, becoming the first country to implement universal H. pylori “test and treat” strategy worldwide (60). Even with high prevalence of H. pylori infection and gastric cancer, whether to implement the “test and treat” strategy for H. pylori infection among the general population in China was still controversial. Several factors such as cost-effectiveness, personal willingness, usage of antibiotic, gastric cancer incidence, and prevalence of H. pylori infection would affect the implementation of this strategy (59, 61). Our study pooled the H. pylori prevalence in non-cardia gastric cancer and explored the influence factors of pooled prevalence, which might have certain public health significance in providing evidence for the control and prevention of non-cardia gastric cancer, as well as for policy making and health resource allocation.

The strengths of this meta-analysis included the following items. To the best of our knowledge, this was the first meta-analysis that estimated the pooled prevalence of H. pylori in non-cardia gastric cancer in China. An overview of current studies about H. pylori prevalence in non-cardia gastric cancer conducted in China was presented in our study. Moreover, our study has included a large number of studies. Based on a large sample size of 5324 cases from 55 studies, we were able to obtain a relatively robust estimate of the pooled H. pylori prevalence in non-cardia gastric cancer.

There were several limitations in our meta-analysis. Even though subgroup analysis and meta-regression were performed to minimize the heterogeneity across the included studies, significant heterogeneity still could be observed in subgroup analysis. The factors included in our study could not well explain the heterogeneity, which might affect the generalizability of our results. Moreover, some important factors (e.g., dietary habit, drinking, and gender) could not be extracted from the included studies, which might have potential influence on the heterogeneity. Another limitation was that some estimates in our study (e.g., H. pylori prevalence in different provinces) were calculated based on the small number of cases. Therefore, these results should be interpreted with caution, and more studies are needed to further confirm these results in the future. Another limitation of our study was that single testing methods used in the included studies for detecting H. pylori infection all had some limitations, and the results could not be completely accurate, which may lead to a certain bias in the pooled H. pylori prevalence. Finally, significant publication bias may result in an underestimation of pooled H. pylori prevalence.

This meta-analysis presented an overview of H. pylori prevalence in non-cardia gastric cancer in China. In conclusion, our study estimated the pooled prevalence of H. pylori in non-cardia gastric cancer was 66.5% (95%CI: 62%-71%) in China. Variation in H. pylori prevalence across different geographical regions was statistically significant, with the highest H. pylori prevalence (78.9%) in Northwest China and the lowest (53.1%) in North China. Type of sample might be associated with H. pylori prevalence, and further studies are needed to confirm this finding. A large proportion of non-cardia gastric cancers was associated with H. pylori infection, emphasizing the potential benefits of H. pylori eradication for reducing the disease burden of non-cardia gastric cancer. Our study might have certain public health significance in providing evidence for the control and prevention of non-cardia gastric cancer, as well as for policy making and health resource allocation.

The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding author.

Study concepts: FH; Study design: FH; Data acquisition: FX and YW; Quality control of data and algorithms: YL and ZW; Data analysis and interpretation: YL and DL; Manuscript preparation: YL, FX, and YW; Manuscript editing: YL; Manuscript review: ZW, DL, and FH. All authors contributed to the article and approved the submitted version.

This study was supported by the Doctoral Start-up Foundation of Guizhou Medical University [No. J (2020)65], National Natural Science Foundation of China Incubation Program, Guizhou Medical University [No. 20NSP061], the First-Class Discipline Construction Project in Guizhou Province Public Health and Preventive Medicine [No.2017 (85)], Foundation for the Establishment of Postdoctoral Mobile Station in Public Health and Preventive Medicine, Guizhou Medical University [41202020204], and Guizhou Basic Research (Science and Technology Fund) Project [ZK (2022) General 373]. All funding parties did not have any role in the design of the study or in the explanation of the data.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.