Development and internal validation of a risk model for hyperuricemia among people living with HIV in Hangzhou, China: a retrospective longitudinal cohort study.

BACKGROUND

The aim of this study was to identify independent risk factors for the time to hyperuricemia (HUA) in PLWH and develop an HUA risk model based on a retrospective study in Hangzhou, China.

METHODS

PLWH treated at Hangzhou Xixi Hospital between July 2016 and December 2024 were randomly assigned to training and validation cohorts in a 7:3 ratio. Independent risk factors associated with the occurrence of HUA were identified using multivariable Cox regression analysis. Characteristic variables for constructing the HUA risk prediction model were selected through the least absolute shrinkage and selection operator (LASSO) method, and the model was visualized using a nomogram. The model's performance was assessed through time-independent receiver operating characteristic (ROC) curve, calibration curve, concordance index (C-index), and decision curve analysis (DCA) to evaluate its discriminatory ability, calibration, and clinical applicability. A generalized linear mixed model (GLMM) was used to investigate the effects of different ART regimens on the SUA levels over time among PLWH.

RESULTS

A total of 631 participants were included in this study, with 439 assigned to the training set and 192 to the validation set. The final model for predicting HUA risk in PLWH incorporated two factors: baseline serum uric acid (SUA) levels and baseline CD4+ T-cell count. The area under the ROC curve (AUC) values for 1-, 3-, and 5-year HUA risks were 0.72, 0.74, and 0.75 in the training set, and 0.66, 0.67, and 0.70 in the internal validation sample, respectively. The calibration curve demonstrated strong agreement between predicted and observed outcomes. Both the C-index and DCA confirmed the nomogram's superior predictive performance. Furthermore, the prevalence of hypercholesterolemia and abnormal eGFR differed significantly between the defined high- and low-risk groups. Comparing with PLWH receiving the B/F/TAF regimens, there was a significant decrease in SUA values over time in PLWH receiving EFV-containing ART regimens.

CONCLUSIONS

We established and validated a HUA-specific nomogram for predicting the risk of HUA in PLWH. This model provides clinicians with a practical tool for the early-stage identification of PLWH at high risk of HUA, a capability that is highly significant for guiding clinical treatment.