Serum gastric biomarkers and metabolic syndrome 2017-2021 dataset

This study dataset was derived from a cross-sectional investigation conducted at the Health Management Center of the Third Xiangya Hospital, Central South University between May 2017 and June 2021. Using standardized protocols, we collected multidimensional health data from 24,635 examinees aged 18-80 years. Data acquisition strictly adhered to the "Expert Consensus on Basic Health Examination Items" published by the Chinese Medical Association. All physical measurements were performed using uniformly calibrated medical equipment (including stadiometers, electronic sphygmomanometers, etc.), while biochemical parameters were analyzed using Hitachi fully automated biochemical analyzers.Serum gastric function biomarkers (PGⅠ, PGⅡ, and G-17) were quantified using ELISA kits (Biohit, Finland; Catalog numbers: 601010.01, 601020.02, and 601035 respectively) following manufacturer's instructions with standard curve quantification. Blood samples were processed within 30 minutes post-collection and stored at -80°C until analysis.The dataset comprises two core data files:"Clinical_Characteristics.xlsx" contains complete records of all 24,635 participants, with each row representing an individual subject. Column variables encompass demographic characteristics (age [years], sex, smoking/alcohol consumption [categorical variables]), anthropometric measurements (BMI [kg/m²], waist circumference [cm], blood pressure [mmHg]), and laboratory parameters (fasting glucose [mmol/L], HbA1c [%], lipid profile [mmol/L])."Gastric_Biomarkers.csv" stores gastric biomarker measurements (PGⅠ/PGⅡ [μg/L], G-17 [pmol/L]) and Helicobacter pylori infection status (determined by ¹³C/¹⁴C urea breath test).Following rigorous quality control procedures that excluded duplicates and outliers, the primary missing data involved medication history fields (∼4.1%) and lifestyle questionnaires (2.3%), which were addressed using multiple imputation methods. Measurement errors were controlled within the following ranges: ±3 mmHg for blood pressure, <3% coefficient of variation (CV) for biochemical analyses, and <12% inter-plate variation for ELISA assays.Metabolic syndrome (MetS) was diagnosed according to International Diabetes Federation criteria, with grouping implemented through derived variables. Statistical analyses were performed using R software (version 4.0.5; https://www.r-project.org/). Continuous variables were analyzed using Student's t-tests (normal distribution) or Mann-Whitney U tests (non-normal distribution), while categorical variables were evaluated with χ² tests. We constructed three-tiered logistic regression models with progressive covariate adjustment: Model 1 (age/sex-adjusted), Model 2 (additional adjustment for comorbidities and lifestyle factors), and Model 3 (further inclusion of anthropometric indicators). All data were stored in universal formats (Excel/CSV) to ensure compatibility with major statistical software packages including SPSS and SAS.