Development and internal validation of a routine metabolic screening model for hyperuricemia among gas station workers in Guangxi, China

Background： Hyperuricemia is a common metabolic abnormality in occupational health examinations and is closely associated with cardiovascular disease, chronic kidney disease, and various metabolic disorders. Gas station workers may face an elevated risk of uric acid abnormality due to occupational characteristics such as shift work, prolonged standing, and irregular lifestyles. Current identification of hyperuricemia relies mainly on serum uric acid measurements, and there remains a lack of simple screening tools based on routine metabolic indicators that are applicable to gas station workers. Methods： We conducted a cross-sectional study of 902 gas station workers in a region of Guangxi, China. Data on demographic characteristics, lifestyle factors, occupational factors, anthropometric measurements, and routine metabolic laboratory indicators were collected through questionnaire surveys, physical examinations, and laboratory tests. Hyperuricemia was defined as serum uric acid > 420 μmol/L in men, > 360 μmol/L in women, or a self‑reported previous diagnosis by a physician. To develop a routine metabolic screening model suitable for occupational health settings, the main model excluded serum uric acid, urea, creatinine, estimated glomerular filtration rate, and urinary parameters. Participants were randomly divided into training and validation sets at a 7:3 ratio stratified by hyperuricemia status. Nineteen candidate predictors were included in the training set. Least absolute shrinkage and selection operator (LASSO) regression was used for variable selection, and the selected variables were incorporated into a multivariable logistic regression model to construct a prediction model and nomogram. Model discrimination was evaluated using the area under the receiver operating characteristic curve (AUC), calibration was assessed using Brier score and bootstrap calibration curves, clinical net benefit was evaluated using decision curve analysis, and risk stratification ability was assessed by grouping predicted probabilities into quartiles. Results： Among the 902 participants, 232 had hyperuricemia, yielding a prevalence of 25.7%. Under the lambda.1se criterion, LASSO regression identified four predictors: sex, waist circumference, triglycerides, and high‑density lipoprotein cholesterol (HDL‑C). Multivariable logistic regression showed that waist circumference and triglycerides were positively associated with hyperuricemia, while HDL‑C was negatively associated; the association for male sex showed an increasing trend but did not reach statistical significance after multivariable adjustment. The final logistic regression equation was: logit(P) = −6.1767 + 0.4134 × Male + 0.0742 × Waist circumference + 0.1477 × Triglycerides − 1.1388 × HDL‑C. The AUC was 0.773 (95% CI: 0.733–0.814) in the training set and 0.815 (95% CI: 0.757–0.874) in the validation set, indicating good discriminative ability. The Brier scores were 0.155 and 0.143 for the training and validation sets, respectively. Bootstrap internal validation showed a corrected C‑index of 0.768, a calibration slope of 0.972, and a mean absolute error of 0.011, suggesting stable calibration. Decision curve analysis demonstrated that the model provided net benefit over both "treat all" and "treat none" strategies within certain threshold probability ranges. In the validation set, hyperuricemia prevalence increased from 5.9% in the lowest quartile to 58.8% in the highest quartile (P for trend < 0.001), confirming good risk stratification ability. Conclusions： We developed a routine metabolic screening model for hyperuricemia using readily available occupational health examination indicators. The final model includes four easily obtainable variables—sex, waist circumference, triglycerides, and HDL‑C—and demonstrates good discrimination, calibration, and risk stratification. The model does not rely on serum uric acid or renal function indicators and may serve as an auxiliary tool for initial screening and risk stratification of hyperuricemia in occupational health settings. Due to the cross‑sectional design, the predicted probabilities reflect current hyperuricemia status rather than future risk; external validation in other regions and occupational populations, as well as prospective studies, are warranted.