The prevalence of gastric intestinal metaplasia in the Saudi population: a single-tertiary center retrospective study

Introduction

Gastric cancer (GC) ranks as the fifth most common cancer and the third leading cause of cancer-related mortality worldwide. Its incidence varies geographically, with half of the emerging cases reported in developing countries (1). According to the International Agency for Research on Cancer, Saudi Arabia reported an incidence rate of 2.7 per 100,000 individuals in 2020, ranking the 15^th globally, with a mortality rate of 2.1 per 100,000 (12^th globally) (2). Due to its low survival rate, prevention remains imperative. Despite significant advancement in GC management, early detection and prevention are still the most effective strategies (3). Unfortunately, early GC is often asymptomatic, leading to delayed diagnosis in two-thirds of patients until advanced stages, limiting treatment options (4).

Gastric intestinal metaplasia (GIM) is characterized by the transformation of gastric mucosa into an intestinal phenotype, which can occur in any gastric region (5). This process reflects a chronic injury-repair response (6). GIM arises from persistent chronic inflammation triggered by various factors, including the Helicobacter pylori (H. pylori) infection (7,8). As a critical precancerous step, GIM can progress into dysplasia and ultimately adenocarcinoma (9), following the well-established Correa cascade a multistep, multifactorial pathway from chronic gastritis to GC (10). The risk of GC is directly correlated with the severity, duration, and extent of GIM (11). with studies reporting a 10- to 40-fold increased risk in affected individuals (12). Endoscopic surveillance plays a pivotal role in detecting GIM and assessing its severity (8), significantly improving survival rates (13,14). Advanced endoscopic modalities such as image-enhanced endoscopy, chromoendoscopy, magnified endoscopy, and artificial intelligence have enhanced the identification of precancerous lesions and tumor depth of invasion (15).

Histopathological diagnosis of GIM relies on multiple diagnostic protocols. The updated Sydney system provides as standardized endoscopic and histologic sampling methodology, grading H. pylori infection, gastric atrophy, and GIM as none, mild, moderate and severe (16). Further staging systems include The Operative Link for Gastritis Assessment (OLGA) (17), which classifies gastric atrophy into stages 0–IV, and the Operative Link for Gastric Intestinal Metaplasia (OLGIM) (18), a Sydney-based scoring method that identifies patients at highest risk of GIM progression to GC. Additionally, a histologic classification system has been proposed, categorizing GIM into three subtypes based on mucin expression patterns and morphologic features (19): types I (complete, goblet cells secreting sialomucins, columnar and/or Paneth cells are present), II (incomplete, lacks columnar and/or Paneth cells) and III (like incomplete type but the predominantly expressing sulfomucins rather than sialomucins.

GC can be classified by anatomical origin into cardia and non-cardia subtypes, each with distinct etiologies: cardia GC associates with gastroesophageal reflux disease (GERD) and obesity, while non-cardia GC primarily links to H. pylori infection (20). Growing recognition of H. pylori carcinogenic role has spurred global eradication recommendations (21), though concerns remain about its efficacy in established GIM, where chronic inflammation and genetic instability may cause irreversible damage, raising questions about whether eradication prevents cancer or merely reduces risk (22,23). While some East Asian countries have successfully implemented population-wide screening and endoscopic surveillance programs, significantly reducing GC incidence (14,24), a consensus global standard for GIM surveillance remains undefined.

This study aimed to assess the prevalence of GIM and its associated risk factors among patients who undergoing upper endoscopy procedures or gastric resection at King Abdulaziz University Hospital (KAUH), in the western region of Saudi Arabia. Our findings provide real-life evidence-based insights into the epidemiology of GIM in the Saudi population. We present this article in accordance with the STROBE reporting checklist (available at https://jgo.amegroups.com/article/view/10.21037/jgo-2025-1-1037/rc).

Methods

Study design and data collection

This is a single-center, retrospective study analyzed data extracted from the medical database between 2016 and 2022 at KAUH, at Makkah province. As a quaternary-care and major referral center for the Western Region of Saudi Arabia, KAUH service a diverse patient population. Adult patients who underwent gastric resection or upper gastrointestinal endoscopy with gastric biopsy for various indications during the study period were included. Records were excluded if they had missing key clinical data, unavailable histopathological analysis, upper endoscopy performed without gastric biopsy, unclear documentation regarding surgical procedures and patients with extragastric malignancy. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This study was conducted in compliance with all national and international laws and regulations. This study received the approval of the Research Ethics Committee of Faculty of Medicine, King Abdulaziz University, Ministry of Higher Education (approval No. HA-02-J-008). Individual consent for this study was waived due to the retrospective nature of the study and the use of de-identified data.

Data on patients’ demographics, risk factors (H. pylori infection status, coexisting dysplasia, or malignancy), and indications for endoscopy were retrospectively collected from the hospital’s electronic medical records. Histopathological findings were obtained from the pathology department at KAUH, including the presence of GIM, the coexistence of dysplasia or malignancy, and H. pylori infection status. Histopathological assessment was performed by experienced gastrointestinal pathologists. Notably, at our endoscopy unit, endoscopists are adherent to a standardized protocol [the updated Sydney system (24)] for gastric biopsies during upper endoscopy including biopsy sites and sampling strategy. To ensure patient confidentiality, all identifiers were anonymized during data extraction and analysis.

Statistical analysis

Data were analyzed using SPSS program version 27 (IBM^®). Chi-squared test (χ²) was used to assess and describe the relation between two categorized variables. Shapiro-Wilk test was used to test the normality of the study sample. Independent Sample t-test analysis was applied to determine the difference between two groups regarding parametric variables. Nonparametric variables were assessed using the Mann-Whitney test to determine the statistical difference between two groups and the Kruskal-Wallis test to determine the statistical difference between more than two groups. Univariable and multivariable logistic regression analyses were used to test the predictors of the binary outcome variables, and a P value of less than 0.05 was considered statistically significant. Quantitative data were presented as mean and standard deviation [mean ± standard deviation (SD)].

Results

Patient demographics and characteristics

A total of 2,103 patients who underwent upper endoscopy procedures or gastric resection at KAUH from January 2016 to December 2022 were included in this study. The cohort comprised adult patients with mean age of 45.99±16.72 years and a female predominance (54.9 % female vs. 45.1% male). Most participants were Saudi nationals (73%, n=1,535), while non-Saudis accounted for 27% (n=568).

The samples comprised 1,608 (76.5%) from random gastric biopsies and 495 (23.5%) from surgical gastric resections. Aside from bariatric surgery (24.4%), there were multiple indications for endoscopic procedures including dyspepsia (22.4%), gastritis (10.9%), GERD (6.6%), gastrointestinal bleeding (3.9%), unexplained iron deficiency anemia (4.3.%), non-specific abdominal pain (3.7%), dysphagia (2%) and other indications (20.9%).

Among 2,074 patients with available H. pylori testing results, 547 (26.4%) were positive for infection. Histopathological examination revealed dysplasia in 42 patients (2.0%) and malignancy in 46 patients (2.2%). The malignant cases comprised of 23 intestinal-type adenocarcinomas; 20 diffuse-type carcinomas, and 3 cases for other subtypes, as detailed in Table 1.

Prevalence and risk factors of GIM

GIM was identified in 119 patients (5.7% of the cohort), with notable demographic and clinical characteristics. The GIM group had a mean age of 55.83±15.88 years, significantly older than the non-GIM patients (45.25±16.55 years, P<0.001), demonstrating a clear age correlation. The gender distribution was nearly equal (59% female vs. 60% male). 78 Saudi patients (65.5%) had GIM, nationality did not significantly affect GIM status. Most cases were detected through endoscopic biopsies (110 cases, 92.4%), with significantly fewer found in surgical resections (9 cases, 7.6%, P<0.001). Common clinical indications for these procedures included for dyspepsia, gastritis, anemia, gastrointestinal bleeding, and other indications (Table 2). Gastric dysplasia was found in 6 patients, while GC was present in eleven patients (Figure 1). H. pylori infection was reported in 53 patients (Table 2). The analysis revealed several important risk factors and clinical association. H. pylori infection was significantly more prevalent in GIM patients (44.9%) compared to non-GIM patients (25%, P value <0.001), representing a 1.8-fold increased risk.

Figure 1 Flow diagram for analysis, with a total number of enrolled patients, number of patients for GIM status and number of patients to estimate the status of H. pylori among the GIM group and the No-GIM group. GIM, gastric intestinal metaplasia; H. pylori, Helicobacter pylori.

Gastrointestinal bleeding was significantly higher in the GIM group with a P value of 0.045. Notably, GIM was strongly associated with more advanced gastric pathology, including GC (11 cases, 9.2%, P value <0.001) and dysplasia (6 cases, 5.0%, P value <0.001). These findings suggest that GIM serves as an important clinical marker for elevated risk of malignant progression in gastric mucosa (Table 2, Figure 2).

Figure 2 (A) The comparison between the number of patients who suffered from cancer in the GIM group and No-GIM group; in the No-GIM group a total of 36 patients (1.84%) suffered from cancer out of 1,953 patients; in the GIM group a total of 11 patients (9.24%) suffered from cancer out of 119 patients. Using the univariate analysis, we found that the percentage of patients who suffered from cancer was significantly higher in the GIM group than in the No-GIM group (P value of 0.001). (B) Comparison between the number of different types of cancer in the GIM group and the No-GIM group; a total of 47 patients suffered gastric cancer of those 16 patients in the No-GIM group suffered intestinal cancer compared to 7 patients in the GIM; 16 patients in the No-GIM suffered diffused cancer compared to 4 patients in the GIM group; 3 patients in the No-GIM suffered other types of cancer; data for one patient was missing. GIM, gastric intestinal metaplasia.

Discussion

The global incidence of GC varies significantly, making early detection or premalignant lesions like GIM crucial to improving outcomes particularly in the era of advanced endoscopic diagnostic modalities. Our study provides the first comprehensive evaluation of GIM prevalence in western Saudi Arabia, where GC rates are notably higher than other regions (2). Our study identified GIM in 5.7% (119/2,103) and GC in 2.2% of patients. These prevalence rates are notably lower than those reported in a national screening study by Zacharakis et al. (25), which found 16% GIM and 4.4% GC prevalence among 1,080 asymptomatic individuals undergoing endoscopic screening in central Saudi Arabia. This discrepancy likely reflects our younger patient population (mean age 45.99 years) and different study design, as 26% of our cohort underwent bariatric surgery rather than pure cancer screening. The reference study specifically targeted an older asymptomatic population (>65% were over 56 years) in a screening context. Overall, the prevalence of GIM in preoperative endoscopic biopsy specimens among candidates for bariatric surgery has been reported to range from 2.4% to 2.7% (25,26). Moreover, supporting evidence from a recent study conducted in central Saudi Arabia demonstrated a very low prevalence of GIM among young adults (18–45 years) presenting with upper gastrointestinal symptoms, estimated at approximately 1.3% (25/1,823) (27). On a global scale, the prevalence of GIM has been estimated at 17.5% [95% confidence interval (CI): 14.6–20.8%] in a recently published meta-analysis of 20 studies including 57,263 patients with great differences across continents, suggesting a strong association with geographical factors (28) which may explain the discrepancy between these global estimates and our findings. In the current study, the presence of non-Saudi nationality (about 27%) in the recruited data may represent a confounding factor limiting the generalizability of the study results. Nonetheless, non-Saudi nationals represent a significant component of the Saudi population which is well established within the whole kingdom [reaching 44.4% of the whole population in the latest estimates from the General Authority for Statistics (GASTAT) in mid-2024 (29)], receiving care within the Saudi healthcare system and sharing similar environmental and healthcare exposure. With significant effect on real-life medical practice. In addition, gender and nationality did not differ between the study groups (group with GIM and the other non-GIM group) which may nullify this confounding factor. On the other hand, age, presence of coexisting dysplasia, presence of coexisting cancer, status of H. pylori, and procedures’ indications showed a significant difference between both groups.

Our analysis revealed a significant age-dependent association with GIM, with affected patients being substantially older than those without GIM (mean age 55.83±15.88 vs. 45.25±16.55 years; P<0.001). This finding aligns with established epidemiological patterns demonstrating a progressive increase in GIM prevalence with advancing age, ranging from approximately 2% in younger populations to over 16% in elderly individuals (26). The strong age correlation supports the hypothesis that GIM development represents an acquired, time-dependent process in gastric mucosal transformation, which may subsequently progress to GC.

Our findings demonstrate a significant association between H. pylori infection and GIM. While H. pylori was present in both groups, its prevalence was nearly twice as high in GIM patients compared to those without GIM (44.9% vs. 25.1%; P<0.001) (Table 2). This goes in line with the previous large study by Huang et al. (30), which demonstrated that H. pylori positive samples showed significantly higher prevalence of GIM (29.9%) compared with those without H. pylori (11.0%). This strong correlation aligns with established literature demonstrating H. pylori’s role in the stepwise progression from gastritis to metaplasia and ultimately GC (27). These results underscore the importance of comprehensive H. pylori assessment in clinical practice. While current infection status is important, obtaining a thorough history of previous infections is equally critical, as chronic H. pylori exposure has been shown to have particularly detrimental effects on gastric mucosa and significantly impacts long-term prognosis. This highlights the need for both active infection screening and detailed patient history-taking when evaluating patients at risk for premalignant gastric lesions.

Our study revealed a strong correlation between GIM and more advanced gastric pathology. Patients with GIM demonstrated significantly higher rates of both GC (9.2% vs. 1.8%; P=0.001) and dysplasia (5.0% vs. 0.8%; P=0.001) compared to those without GIM (Figure 2). This progression likely reflects the combined effects of two key factors: the advanced mean age of GIM patients and their substantially higher prevalence of H. pylori infection (44.9% vs. 25.1%). Histologically, adenocarcinoma predominated among our GC cases, with intestinal-type (50%) being more common than diffuse-type (43.5%) in GIM-positive patients. This distribution aligns with established patterns observed in Arab populations (25,28), supporting the regional consistency of GC subtypes. The predominance of intestinal-type adenocarcinoma in GIM patients further reinforces the recognized metaplasia-dysplasia-carcinoma sequence in gastric carcinogenesis. This clearly outline major clinical implications of our findings demonstrating the necessity for adopting risk stratification programs, endoscopic surveillance strategies aiming for GC prevention in Saudi Arabia.

Several limitations should be considered when interpreting our findings. Being a single-center retrospectively designed study, has its limitation on causal inference, with potential selection and referral bias. We were also unable to access complete historical data regarding prior H. pylori infections, which may influence GC risk. The histopathological analysis was constrained by unavailable data on: (I) precise GC anatomical classification (cardia vs. non-cardia); and (II) GIM subtyping (complete, incomplete, or mixed variants)—important prognostic factors in gastric carcinogenesis. Additionally, due to the retrospective design, formal inter-observer variability analysis of the histopathological finding was not consistently available. Furthermore, we could not account for established GC risk factors including smoking history, dietary patterns, or socioeconomic status, as these variables fell beyond the scope of our primary research objectives. Lastly, the presence of non-Saudi nationality within the recruited data may represent a confounding factor limiting its generalizability.

On the other hand, our study has several important strengths that enhance the validity and significance of its findings. As the first investigation of its kind conducted in western Saudi Arabia, this research provides valuable baseline data on GIM prevalence in a previously understudied region. The study’s robust methodology included a large patient cohort (2,103 individuals) followed over a 6-year period [2016–2022], ensuring sufficient statistical power and longitudinal perspective. A particular strength of our work is the comprehensive inclusion of both endoscopic biopsy and surgical resection specimens, which provides a more complete clinical picture than studies limited to one specimen type. This approach allowed us to identify significant differences in GIM detection rates between diagnostic and therapeutic procedures. Additionally, the study duration has involved the COVID-19-pandemic period in which most of the formal national registries are not available including the age-specific incidence of GC.

Conclusions

This five-year study provides the first comprehensive evaluation of GIM prevalence in western Saudi Arabia among patients undergoing upper endoscopy or gastric resection, revealing GIM in 5.7% of cases and demonstrating significant associations with GC (9.2% in GIM patients vs. 1.8% in non-GIM). Our findings identify advanced age (mean 55.8 years in GIM patients) and H. pylori infection (44.9% prevalence) as key risk factors while highlighting the current lack of standardized GIM surveillance protocols in clinical practice. These findings support the established association between GIM and known risk factors for gastric malignancy and highlight variability in current surveillance practices. These results underscore the need for evidence-based guidelines targeting high-risk populations particularly older adults and H. pylori-positive individuals to improve early detection and prevention of GC through systematic monitoring of premalignant lesions. While the retrospective and single-center nature of the study limits generalizability, the results provide important regional data that may help inform risk stratification and clinical decision-making. Future research should focus on developing regional risk-stratification models and optimizing surveillance intervals to enhance clinical outcomes and to guide evidence-based approaches for early detection and prevention of GC in high-risk populations.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://jgo.amegroups.com/article/view/10.21037/jgo-2025-1-1037/rc

Data Sharing Statement: Available at https://jgo.amegroups.com/article/view/10.21037/jgo-2025-1-1037/dss

Peer Review File: Available at https://jgo.amegroups.com/article/view/10.21037/jgo-2025-1-1037/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jgo.amegroups.com/article/view/10.21037/jgo-2025-1-1037/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This study was conducted in compliance with all national and international laws and regulations. This study received the approval of the Research Ethics Committee of Faculty of Medicine, King Abdul-Aziz University, Ministry of Higher Education (approval No. HA-02-J-008). Individual consent for this study was waived due to the retrospective nature of the study and the use of de-identified data.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Iwu CD, Iwu-Jaja CJ. Gastric Cancer Epidemiology: Current Trend and Future Direction. Hygiene 2023;3:256-268.
2. Alghamdi IG. Epidemiology of gastric cancer in Saudi Arabia from 2004 to 2017. Mol Clin Oncol 2023;19:93. [Crossref] [PubMed]
3. Arnold M, Rutherford MJ, Bardot A, et al. Progress in cancer survival, mortality, and incidence in seven high-income countries 1995-2014 (ICBP SURVMARK-2): a population-based study. Lancet Oncol 2019;20:1493-505. [Crossref] [PubMed]
4. Akbari M, Tabrizi R, Kardeh S, et al. Gastric cancer in patients with gastric atrophy and intestinal metaplasia: A systematic review and meta-analysis. PLoS One 2019;14:e0219865. [Crossref] [PubMed]
5. Sugano K, Moss SF, Kuipers EJ. Gastric Intestinal Metaplasia: Real Culprit or Innocent Bystander as a Precancerous Condition for Gastric Cancer? Gastroenterology 2023;165:1352-1366.e1. [Crossref] [PubMed]
6. Graham DY, Rugge M, Genta RM. Diagnosis: gastric intestinal metaplasia - what to do next? Curr Opin Gastroenterol 2019;35:535-43. [Crossref] [PubMed]
7. Jencks DS, Adam JD, Borum ML, et al. Overview of Current Concepts in Gastric Intestinal Metaplasia and Gastric Cancer. Gastroenterol Hepatol (N Y) 2018;14:92-101.
8. Huang RJ, Choi AY, Truong CD, et al. Diagnosis and Management of Gastric Intestinal Metaplasia: Current Status and Future Directions. Gut Liver 2019;13:596-603. [Crossref] [PubMed]
9. Li D, Bautista MC, Jiang SF, et al. Risks and Predictors of Gastric Adenocarcinoma in Patients with Gastric Intestinal Metaplasia and Dysplasia: A Population-Based Study. Am J Gastroenterol 2016;111:1104-13. [Crossref] [PubMed]
10. Correa P, Haenszel W, Cuello C, et al. A model for gastric cancer epidemiology. Lancet 1975;2:58-60. [Crossref]
11. Pittayanon R, Rerknimitr R, Klaikaew N, et al. The risk of gastric cancer in patients with gastric intestinal metaplasia in 5-year follow-up. Aliment Pharmacol Ther 2017;46:40-5. [Crossref] [PubMed]
12. Spence AD, Cardwell CR, McMenamin ÚC, et al. Adenocarcinoma risk in gastric atrophy and intestinal metaplasia: a systematic review. BMC Gastroenterol 2017;17:157. [Crossref] [PubMed]
13. Asaka M, Mabe K. Strategies for eliminating death from gastric cancer in Japan. Proc Jpn Acad Ser B Phys Biol Sci 2014;90:251-8. [Crossref] [PubMed]
14. Kaneko E, Nakamura T, Umeda N, et al. Outcome of gastric carcinoma detected by gastric mass survey in Japan. Gut 1977;18:626-30. [Crossref] [PubMed]
15. Martins BC, Moura RN, Kum AST, et al. Endoscopic Imaging for the Diagnosis of Neoplastic and Pre-Neoplastic Conditions of the Stomach. Cancers (Basel) 2023;15:2445. [Crossref] [PubMed]
16. Dixon MF, Genta RM, Yardley JH, et al. Classification and grading of gastritis. The updated Sydney System. International Workshop on the Histopathology of Gastritis, Houston 1994. Am J Surg Pathol 1996;20:1161-81.
17. Rugge M, Meggio A, Pennelli G, et al. Gastritis staging in clinical practice: the OLGA staging system. Gut 2007;56:631-6. [Crossref] [PubMed]
18. Capelle LG, de Vries AC, Haringsma J, et al. The staging of gastritis with the OLGA system by using intestinal metaplasia as an accurate alternative for atrophic gastritis. Gastrointest Endosc 2010;71:1150-8. [Crossref] [PubMed]
19. Filipe MI, Potet F, Bogomoletz WV, et al. Incomplete sulphomucin-secreting intestinal metaplasia for gastric cancer. Preliminary data from a prospective study from three centres. Gut 1985;26:1319-26.
20. Arnold M, Ferlay J, van Berge Henegouwen MI, et al. Global burden of oesophageal and gastric cancer by histology and subsite in 2018. Gut 2020;69:1564-71. [Crossref] [PubMed]
21. Plummer M, de Martel C, Vignat J, et al. Global burden of cancers attributable to infections in 2012: a synthetic analysis. Lancet Glob Health 2016;4:e609-16. [Crossref] [PubMed]
22. Graham DY. Helicobacter pylori update: gastric cancer, reliable therapy, and possible benefits. Gastroenterology 2015;148:719-31.e3. [Crossref] [PubMed]
23. Liu KS, Wong IO, Leung WK. Helicobacter pylori associated gastric intestinal metaplasia: Treatment and surveillance. World J Gastroenterol 2016;22:1311-20. [Crossref] [PubMed]
24. Dixon MF, Genta RM, Yardley JH, et al. Classification and grading of gastritis. The updated Sydney System. International Workshop on the Histopathology of Gastritis, Houston 1994. Am J Surg Pathol 1996;20:1161-81.
25. Cheng YL, Elli EF. Management of gastric intestinal metaplasia in patients undergoing routine endoscopy before bariatric surgery. Updates Surg 2022;74:1383-8. [Crossref] [PubMed]
26. Wang S, Wang Q, Xu L, et al. Beware Pathological Findings of the Stomach in Patients Undergoing Bariatric Surgery: a Systematic Review and Meta-analysis. Obes Surg 2021;31:337-42. [Crossref] [PubMed]
27. Albadrani A, Zacharakis G, Alqahtani MS, et al. A Longitudinal Study of Premalignant Gastric Lesions and Early Onset Gastric Cancer Among Young Adults in Central Saudi Arabia. Curr Oncol 2025;32:428. [Crossref] [PubMed]
28. Soroorikia S, Kazeminia M, Qaderi K, et al. Global prevalence of gastric intestinal metaplasia: a systematic review and meta-analysis. Syst Rev 2024;13:247. [Crossref] [PubMed]
29. Arabia GAfSGokoS. population estimate. In: 2024. Available onlne:
30. Huang RJ, Ende AR, Singla A, et al. Prevalence, risk factors, and surveillance patterns for gastric intestinal metaplasia among patients undergoing upper endoscopy with biopsy. Gastrointest Endosc 2020;91:70-77.e1. [Crossref] [PubMed]