The mean age of the patients was 43.90 ± 15.60 years with extremes of 4 and 80 years. The 40 - 59 years age group was the most represented (41.94%). Among the 310 patients included in this study, there was a predominance of women with a male-to-female ratio of 0.83. Patients from Bariba ethnicity were the most represented (29.03%), followed by Fon and related (20.97%). The majority of the patients lived in urban areas (70.65%) (Table 1).

Epigastric pain was the most frequent reason for requesting upper gastrointestinal endoscopy (53.23%) followed by gastroesophageal reflux (18.39%). Erythematous gastropathy represented 70.35% of lesions observed in upper gastrointestinal endoscopy, followed by erosive gastritis (15.03%) (Table 2).

a: Caustic ingestion, Anemia, Upper gastrointestinal hemorrhage, Dysphagia or Odynophagia, Periumbilical pain, Left hypochondrium pain, Diffuse abdominal pain, Vomiting, Weight loss. b: Ulcerated gastritis, Large fundial folds, Sessile polyp.

Based on Sydney’s parameters assessment, among the 310 patients with chronic gastritis included in this study, 215 (69.35%) had moderate degree of inflammation, 251 (80.97%) had cellular activity and 171 (55.16%) had glandular atrophy. In addition, intestinal metaplasia was present in 52 patients (16.77%) and lymphoid follicles in 28 patients (9.03%). Dysplasia was observed in 38 patients (12.26%). Helicobacter pylori was identified histologically in 114 patients (36.77%) (Table 3).

In bivariate analysis, socio-demographic characteristics, notably gender and age, were not statistically associated with the presence of dysplasia with p-values of 0.921 and 0.399 respectively. Endoscopically, erythematous gastropathy was statistically associated with the presence of dysplasia (p-value = 0.005). Still, patients with erythematous gastropathy were less likely to progress to gastric dysplasia [0.13 - 0.73] than those with other endoscopic aspects. Histologically, Sydney’s parameters including the degree of inflammation, the presence of cellular activity and the presence of glandular atrophy were statistically associated with the presence of gastric dysplasia with p-values of 0.005, 0.110 and 0.158 respectively. Indeed, patients with moderate or severe inflammation were 3.02 and 5.47 times more likely to progress to gastric dysplasia, respectively. Patients with glandular atrophy were 2.95 times more likely to progress to gastric dysplasia. In addition, intestinal metaplasia was statistically associated with the presence of gastric dysplasia (p-value < 0.001). Patients with intestinal metaplasia were 8.33 times more likely to progress to gastric dysplasia. In contrast, the presence of Helicobacter pylori and the presence of lymphoid follicles were not statistically associated with progression to gastric dysplasia with p-values of 0.813 and 0.387 respectively (Table 4).

In multivariate analysis, Sydney’s parameters, notably the degree of inflammation, the presence of cellular activity and the presence of glandular atrophy were not statistically associated with the presence of gastric dysplasia with p-values of 0.066, 0.879 and 0.153 respectively. On the other hand, the presence of erythematous gastropathy in upper gastrointestinal endoscopy was statistically associated with the presence of gastric dysplasia (p-value = 0.042). However, patients with erythematous gastropathy were less likely to progress to gastric dysplasia [0.04 - 0.77] than those with other endoscopic aspects. Furthermore, the presence of intestinal metaplasia in histopathology check up was statistically associated with progression to gastric dysplasia (p-value < 0.001). Patients with intestinal metaplasia were 6.77 times more likely to progress to gastric dysplasia (Table 5).

In this study, the frequency of chronic gastritis was 91.45%. This high prevalence is close to the 93.1% found by Konaté et al. [6] in Mali, and the 98.2% found by Bamba et al. in Senegal in 2017 [4]. Prevalences also close to ours, namely 85.8%, 83.44% and 81% were respectively reported by Essadik et al. [10] in Morocco in 2013, Darré et al. [5] in Togo in 2010 and Doh et al. [23] in Senegal in 2016. On the other hand, a lower prevalence of 53% was reported in the French population by Potet et al. [27]. This disparity could be explained by the higher prevalence of Helicobacter pylori in developing countries [9].

- Age

In this study, the mean age of the patients was 43.90 years. Similar findings to ours were reported by Koura et al. [22] in 2017 in Burkina-Faso and Itoudi et al. [28] in 2014 in Gabon which were 43.2 and 44 years respectively. However, our finding was slightly higher than that of Doh et al. [23] in Senegal in 2016 (39.9 years) and Konaté et al. [6] in Mali in 2007 (38.14 years). On the other hand, it was lower than that of Bamba et al. in Senegal in 2017 (45 years), Darré et al. in Togo (49.3 years) and Ray-Offor et al. [24] in 2017 in Nigeria (47.1 years). In France, Potet et al. [27] reported a mean age of 53 years, which is much higher than ours and those for most African studies. This difference could be attributed to the old age of the Western population compared to African populations.

- Gender

In this study, 54.52% of the patients were female. This female predominance was also reported by Koura et al. [22] in Burkina Faso (53.3%), Bamba et al. [4] in Senegal (61%), Essadik et al. in Morocco (53.1%) and Du et al. [29] in China (51.2%). Several other African studies have also reported this female predominance [6] [23] [28]. In contrast, a male predominance was reported by Darré et al. [5] in Togo, Ray-Offor et al. [24] in Nigeria, Attia et al. [30], and Diomandé et al. [31] in Côte d’Ivoire where the male-to-female ratios were 1.4, 1.2, 2 and 1.3 respectively. Anyway, the main cause of chronic gastritis is Helicobacter Pylori infection. Indeed, gender is not a risk factor for this infection. Rather, it is favored by individual characteristics, hygiene and socioeconomic conditions [32].

- Area of residence

In this study, 70.65% of the patients lived in an urban area in Parakou. This finding is close to the 77.8% reported by Ankouane et al. [33] in 2013 in urban areas in Cameroon. Actually, patients living in urban areas in Parakou have easier access to the CHUD-B/A. This could justify their predominance in our study.

In this study, epigastric pain was the most frequent reason for requesting upper gastrointestinal endoscopy (53.23%) followed by gastroesophageal reflux (18.39%). In the study of Koura et al. [22] in Burkina-Faso, epigastric pain was also the most frequent reason for requesting endoscopy with a finding very close to ours (52.8%). Bagny et al. [34] in Togo in 2011 reported a frequency of epigastric pain of 48.8%. Bamba et al. [4] in 2017 in Senegal found a much higher frequency of epigastric pain (91%) followed by dyspepsia (22%). Indeed, epigastric pain is the most encountered symptom because it is a clinical manifestation of the inflammatory reaction taking place within the gastric mucosa. In fact, in the classic semiological tetrad of the local inflammatory syndrome, there is redness (erythematous aspect in endoscopy), heat, swelling and pain manifested here by epigastric pain. In this study, erythematous gastropathy represented 70.35% of the lesions found in upper gastrointestinal endoscopy, followed by erosive gastritis (15.03%). Bamba et al. [4] in Senegal in 2017 reported a higher prevalence of erosive gastropathy of 90%. A lower prevalence of erythematous gastropathy was found by Bagny et al. [34] in Togo in 2011 (36.2%). This difference could be due to the inclusion in these studies of patients with symptoms suggestive of gastric involvement. In addition, 5.81% of lesions in our study were gastric ulcerations against 7% in Ray-Offor et al. [24] in Nigeria and 13.79% Bentahar et al. [35] in Algeria in 2016. Micronodular aspect which would be an indicator of Hp infection in endoscopy [36] was present in our study in 6.45% of cases against 4% in Bamba et al. in Senegal. This fluctuation in proportions may be related to the different sample sizes.

- Degree of inflammation

The degree of inflammation was moderate according to Sydney classification, in most cases (69.35%). This predominance of moderate degree of inflammation was also reported in Togo (71%) and Senegal (52.7%) [4] [5].

- Cellular activity

Cellular activity (presence of neutrophils) was noted in 80.97% of cases. This cellular activity was medium according to Sydney classification in most cases (61.75%). Similar proportions of cellular activity were reported in Côte d’Ivoire (81.4%) and Togo (83.81%) [5] [30]. The high proportion of active chronic gastritis in this study and in most African countries should be related to the pathogenicity of Helicobacter pylori. Several studies have shown that cellular activity was significantly associated with Helicobacter pylori infection [28] [37].

- Glandular atrophy

In our study, glandular atrophy was observed in 55.16% of cases. It was present, according to the Sydney classification, at moderate intensity in 63.74%. This prevalence of glandular atrophy in this study is close to that found by Udoh et al. [38] in Nigeria in 2009 (53%). Ankouane et al. [39] in Cameroon in 2014 and Darré et al. in Togo in 2010 found much higher prevalence (74.7% and 83%). Koura et al. [22] in Burkina-Faso in 2017, in contrast, noted a lower prevalence than in this study (26.8%). This is also the observation in Senegal where studies conducted noted very low prevalence of glandular atrophy, less than 20% [4] [23].

- Helicobacter pylori

Histologically, Helicobacter pylori infection was reported in 36.77% of chronic gastritis cases. It was present in 59.65% with low density. This prevalence of Helicobacter pylori is much lower than those reported in Mali by Konaté et al. [6] in 2007 (89.4%). The same is true for Jmaa et al. [40] in Tunisia in 2005, Essadik et al. [10] in 2011 in Morocco, Doh et al. [23] in 2016 in Senegal who found respective prevalence of 87.8%, 69.2% and 75.6% largely higher than ours. Lower prevalence than the latter but still higher than ours were found by Koura et al. [22] (58.3%) in Burkina-Faso in 2017 and Bagny et al. [34] (53.4%) in Togo in 2011. This low prevalence in our study could be explained in several ways. First, the retrospective nature of the study certainly took into account cases of chronic gastritis treated either by the physician or self-medication with antibiotics or traditional treatment (phytotherapy) which is very frequent in our setting. The most represented patients in the study belonged to the 40-59 years age group, whereas Helicobacter pylori infection is much more frequent in young patients. The absence of the use of special stains such as modified Giemsa that allow better visualization of Helicobacter pylori could also explain this prevalence. In industrialized countries, the prevalence is generally low. Sethi et al. [41] in 2011 in Canada and Verma et al. [42] in 2009 in the United States reported 37.9% and 23.7% by histological examination of gastric biopsy respectively. This low prevalence could be explained by the high socioeconomic level of developed countries. In addition, in the study of Sethi et al. in Canada, it was a specific population without functional signs and the biopsies were performed in a preoperative context. Kpossou et al. [43] in southern Benin in 2018 noted a prevalence of 34.7%. This result is lower than ours. The difference could be explained by the Helicobacter pylori diagnosis method used. Indeed, these authors used an indirect method (carbon-14 urea breath test), unlike our study where the diagnosis was made by histopathological check up of gastric biopsies, which is the conventional reference method for the detection of Helicobacter pylori.

- Intestinal metaplasia

The prevalence of intestinal metaplasia in our study was 16.77%. Udoh et al. in Nigeria [38] reported a prevalence of metaplasia consistent with ours (16.6%). This prevalence was higher than that reported by Koura et al. [22] in Burkina-Faso (13.5%), Bamba et al. [4] in Senegal in 2017 (11.5%) and that of Ankouane in Cameroon in 2014 (6.3%). Darré et al. [5] in Togo in 2010 found a higher prevalence (54.25%).

- Dysplasia

The prevalence of dysplasia in our study was 12.26. Doh et al. [23] in Senegal in 2016 found a prevalence close to ours (11.1%). In contrast, it was very low in Bamba et al. in 2017 (1.7%) who had a larger population [4]. Very low findings were also reported in Nigeria by Ray-Offor [24] in 2017 (5.8%) and Jemilehoun et al. [44] (2.4%) in an original research published in 2020. These discrepancies in the prevalence of precancerous lesions could be explained by the fact that their distribution varies from country to country [18] [39] [40]. Indeed, the environmental factors and socioeconomic conditions that determine the high-risk groups for gastric cancer vary from country to country, and within the same country differ between regions [18]. In very high-risk groups such as Japan and China, and some high-risk groups (Blacks, Hispanics, and in the United States of America, gastric precancerous condition rates are generally very high [39] [40] [45] [46] [47]. These differences in the prevalence of precancerous conditions may also be related to the different sample sizes and the site of the gastric biopsy.

- Erythematous gastropathy on endoscopy

This study has evidence that an erythematous gastropathy in upper gastrointestinal endoscopy would be a predictive factor of precancerous lesions absence in histopathology check up (p-value = 0.042). Indeed, precancerous lesions that occur after several decades are characterized by cellular and architectural changes that are reflected in the gastric mucosa by an atypical aspect and relief. Dysplasia then presents itself endoscopically as flat depressed or polypoid lesions [48]. The erythematous aspect visualized in endoscopy is the result of chronic inflammation of the mucosa generally caused by Helicobacter pylori, and marks the evolutionary cascade beginning [49] [50]. The long-term evolution would be the occurrence of pre-cancerous lesions. However, to date there is no clear correlation between endoscopic findings and histopathology outcomes. Many studies have demonstrated the low correlation between histopathology outcomes and endoscopic findings [51] [52]. But others also investigate possible associations. For example, an Iranian study has shown that abnormal endoscopic findings (both ulcerative and non-ulcerative) were associated with precancerous lesions [53].

- Intestinal metaplasia

Patients with histopathological evidence of intestinal metaplasia (p < 0.001) during chronic gastritis were more likely to develop precancerous lesions in this study. Indeed, gastric carcinogenesis is a multistep process that involves, in most of cases, a progression from normal mucosa to chronic gastritis (chronic inflammation of the gastric mucosa), to glandular atrophy (loss of gastric glands), to intestinal metaplasia (substitution of gastric epithelium by intestinal epithelium), to dysplasia and then to gastric cancer. This sequence of events can last for several years and has been referred to as Correa’s cascade of multistep gastric carcinogenesis [54]. Intestinal metaplasia is considered a precancerous condition [48]. This is the rationale for this association. Indeed, well-established evidence points to an association between intestinal metaplasia and intestinal-type gastric cancer [55]. It is a lesion with a high risk of dysplasia and therefore of malignancy. In fact, it directly precedes dysplasia in Correa cascade [54]. Its extent and topography would be reliable indicators of cancer risk [56]. Given this high risk of dysplasia, guidelines have been proposed for patients with intestinal metaplasia. They suggest metaplasia promoting factors elimination, notably Helicobacter pylori, and endoscopic surveillance for patients with intestinal metaplasia who are at increased risk for gastric cancer due to their ethnicity or family history. Optimal surveillance intervals have not been widely studied and must be individualized. For individuals with severe OLGA and OLGIM scores (stage III/IV), without dysplasia endoscopic surveillance with histopathological check up of gastric biopsies every 3 years is suggested. For individuals with a family history of cancer added to the previous conditions, an even more intense 1 - 2 year surveillance endoscopy has been suggested [57].