Neutrophil-to-lymphocyte ratio predicts inpatient gout recurrence: a large-scale multicenter retrospective cohort with machine-learning validation.

Background: The neutrophil-to-lymphocyte ratio (NLR) is an accessible marker of systemic inflammation. However, its prognostic value for inpatient gout recurrence, particularly in comparison with traditional biomarkers, remains unclear. This study aims to investigate the association of NLR with inpatient gout recurrence, and compare its performance with traditional markers. Methods: In this international, multicenter retrospective cohort study, hospitalized patients with gout were enrolled from the GoutRe cohort (China, 2010-2025) and MIMIC-IV cohort (USA, 2008-2019). Restricted cubic spline, Cox regression and competing risk models were deployed to visualize and assess the association of NLR with inpatient gout recurrence risk. Model performance was evaluated using the C-statistic, net reclassification improvement, and decision curve analysis. Multiple machine learning algorithms were employed for external validation. Results: Among 7,603 patients (GoutRe: 5,584; MIMIC-IV: 2,019), elevated NLR (>2.69) was independently associated with a higher inpatient gout recurrence risk (GoutRe: HR = 2.05; MIMIC-IV: HR = 2.84; both P < 0.001). NLR correlated with systemic inflammation, comorbidities, and use of diuretics/β-blockers. It outperformed serum uric acid (UA) and C-reactive protein (CRP) in predicting inpatient gout recurrence (AUC: 0.62 vs. 0.59 and 0.61, respectively), with improved accuracy when combined with UA (AUC = 0.65, P < 0.01). Predictive value remained consistent across subgroups, including those with normal UA, no tophus, and ongoing anti-inflammatory or urate-lowering therapy. Machine learning models, particularly XGBoost, confirmed NLR's predictive strength. Incorporation of NLR into baseline models improved discrimination and reclassification. Decision curve analysis showed greater net clinical benefit with NLR-based models. Biological plausibility analysis revealed that elevated NLR reflected neutrophilia and lymphopenia, indicative of systemic inflammation during the intercritical period. Conclusions: Elevated NLR is a robust, accessible biomarker independently associated with inpatient gout recurrence. Its integration into clinical risk models enhances prediction accuracy and supports personalized inpatient gout recurrence prevention strategies.