Study populations
A total of 186 subjects who came to the 971 hospital for a screening gastroscopy as part of a routine health check-up, were prospectively enrolled in this study from January 2020 to October 2020. The subjects were divided into CAG and CNAG according to endoscopic pathology. Exclusion criteria were a history of H. pylori eradication, previous gastric surgery, treatment with proton pump inhibitors or H2 blockers within the previous 2 weeks, serious systemic disease, pregnancy, a history of cancer of any type, and those taking anti-secretory or anti-coagulant drugs. A fasting blood sample was obtained from each patient before endoscopy.
Serological measurements and endoscopic and histopathological examinations
Details on the serological measurements and endoscopic and his[1]topathological examination procedures were previously described ( 18,19 ). Serum PGI, PGII, and G-17 concentrations in morning fasting blood samples were measured using enzyme-linked immunosorbent assays (ELISAs; Pepsinogen I ELISA; Pepsinogen II ELISA and Gastrin-17 ELISA kit, Snibe Diagnostic, shenzhen, China), serum anti-H. pylori IgG were measured using colloidal gold immunochromatography assay (GICA, anti-H. pylori kit, HUIAN, shenzhen, China) .
Gastroscopy with sampling of gastric biopsies was defined as the gold standard for definitive diagnosis. Endoscopic biopsies were obtained from the antrum and corpus, all along the greater curvature (one biopsy from both sites). Biopsy specimens were routinely fixed in neutral formalin and processed in paraffin. Tissue sections were stained with HE, Alcian blue and modified Giemsa (H pylori stain) methods.
Statistical analyses
SPSS 19.0 was used to carry out statistical analysis. The normal distribution measurement data is expressed as mean±standard deviation, the average between the two groups is compared by t test, and the counting data is expressed by rate or percentage, and the comparison is χ2 test.
H. pylori serological antibody results was classified as negative and positive. The other biomarkers (PGI, PGII, the PGI/II ratio, G-17) were categorized according to quartiles of their distributions in the study cohort. For the cross-sectional analysis, odds ratios (ORs) with 95% confidence intervals (95% confidence intervals (CIs)) were calculated using logistic regression.
In the risk prediction modeling analysis, receiver operator characteristic curves with corresponding C statistics (area under the curve, AUC) based on logistic models were used to measure the discriminatory performance of combination of predictors where the pathology diagnosis was considered the “gold standard”. A P value<0.05 (two-sided) was considered statistically significant.