Value of serum uric acid in a risk prediction model for postmenopausal osteoporosis

Abstract

Objective: To investigate the potential role of serum uric acid (UA) in risk stratification for postmenopausal osteoporosis (PMOP) and to establish an accessible risk prediction model that can aid in early screening and diagnosis.

Methods: We retrospectively enrolled 295 postmenopausal women who underwent dual-energy X-ray absorptiometry (DXA) at Zhuhai Hospital affiliated with Jinan University from July 2021 to July 2023. Participants were divided into a PMOP group (T-score < -2.5; n = 125) and a control group (T-score ≥ -2.5; n = 170). Clinical and laboratory data were collected, including markers of inflammation, renal function, and uric acid levels. Univariable and multivariable logistic regression analyses identified independent risk factors for PMOP. A nomogram was constructed based on the final logistic regression model and evaluated for discrimination and calibration using receiver operating characteristic (ROC) curves, calibration curves, and the concordance index (C-index).

Results: The PMOP group exhibited significantly higher mean values of age, alkaline phosphatase (ALP), neutrophil count (NEU), monocyte count (MO), monocyte-to-lymphocyte ratio (MLR), and the systemic immune-inflammation index (SII), while demonstrating significantly lower lymphocyte counts (LYM), height, OSTA scores, and albumin (ALB). Serum UA values were slightly lower in the PMOP group than in the control group. Multivariable logistic regression yielded a prediction model incorporating ALB, ALP, MLR, and UA. The area under the ROC curve (AUC) for this model was 0.781 (95% CI: 0.682-0.879). The calibration curve aligned well with the ideal reference line, and the C-index was 0.779 (95% CI: 0.728-0.831).

Conclusion: Serum uric acid may have a contributory role in risk stratification for PMOP when combined with key clinical and laboratory markers. This nomogram-based model demonstrates moderate predictive performance; future large-scale multicenter prospective cohorts are warranted to validate these findings and to refine the model by accounting for potential confounding factors such as medication use, dietary intake, and lifestyle habits.

Keywords: Bone mineral density; Nomogram; Postmenopausal osteoporosis; Risk prediction model; Serum uric acid.

Fig. 1

Nomogram of the predictive model

Fig. 2

Calibration curve. This calibration curve assesses the consistency between predicted and actual probabilities of PMOP in our risk prediction model. The y-axis represents the observed (actual) incidence of PMOP, while the x-axis represents the predicted incidence of PMOP. The “Ideal” line serves as a reference line for perfect calibration. The “Apparent” line reflects the fit between the predicted and observed values, and the “Bias-corrected” line indicates the fit after internal correction. The closer either the “Apparent” line or “Bias-corrected” line is to the “Ideal” line, the better the agreement between the predicted and observed values. Conversely, greater deviation from the “Ideal” line indicates poorer consistency between the model’s predictions and actual outcomes

Fig. 3

Receiver operating characteristic curve. This ROC curve evaluates the performance of our PMOP risk prediction model. The y-axis represents the true positive rate (TPR), while the x-axis represents the false positive rate (FPR), which is calculated as 1—specificity. The area under the red ROC curve indicates the model’s predictive power; a larger area under the curve (AUC) signifies higher accuracy of the model in distinguishing between individuals with and without PMOP