Association between the uric acid-to-high-density lipoprotein cholesterol ratio and tendon healing following arthroscopic rotator cuff repair: a retrospective MRI-based cohort study

To investigate the association between the preoperative serum uric acid-to-high-density lipoprotein cholesterol ratio (UHR) and MRI-assessed tendon healing following arthroscopic rotator cuff repair. Cohort study. This retrospective cohort study included 158 patients with small- to medium-sized full-thickness rotator cuff tears who underwent arthroscopic rotator cuff repair (ARCR). Preoperative data included demographic characteristics, laboratory markers, serum uric acid (UA), high-density lipoprotein cholesterol (HDL-C), calculated UHR (UA/HDL-C), and imaging parameters including Patte stage, Goutallier grade, tangent sign, supraspinatus occupation ratio, and acromiohumeral distance. Postoperative tendon integrity was assessed using the Sugaya classification. Sugaya types I–II were defined as complete structural healing, whereas Sugaya types III–V were defined as incomplete structural healing. Because Sugaya III represents tendon thinning with preserved continuity, it was not considered equivalent to frank re-tear; true re-tear was defined as Sugaya types IV–V. Intergroup comparisons, Spearman correlation analysis, logistic regression, and receiver operating characteristic (ROC) analysis were performed to evaluate the exploratory association between UHR and postoperative structural healing. Sugaya type I was observed in 54 patients, type II in 75, type III in 26, type IV in 2, and type V in 1. Overall, 129 patients (81.6%) showed complete structural healing (Sugaya I–II), whereas 29 patients (18.4%) showed incomplete structural healing (Sugaya III–V). True re-tear (Sugaya IV–V) was rare and occurred in 3 patients. UHR showed a weak positive correlation with ordinal Sugaya grade (Spearman ρ = 0.213, P = 0.007). Compared with the complete-healing group, the incomplete-healing group had higher BMI, more severe Patte stage, a higher proportion of Goutallier grade ≥III, and lower supraspinatus occupation ratio (all P < 0.05), whereas UHR did not differ significantly between groups ( P = 0.434). Continuous UHR was not significantly associated with incomplete structural healing in either unadjusted or multivariable-adjusted logistic regression (unadjusted OR = 1.01, P = 0.557; adjusted OR = 0.99, P = 0.604, per 10-unit increase). ROC analysis showed limited discrimination for UHR (AUC = 0.547; cutoff = 209.56; sensitivity = 86.2%; specificity = 36.4%). In this retrospective MRI-based cohort, preoperative UHR showed only a weak association with ordinal postoperative Sugaya grade and demonstrated limited discriminatory and independent predictive value for incomplete structural healing. BMI and imaging-based tendon/muscle quality parameters appeared more informative. UHR should be interpreted as an exploratory adjunctive marker of metabolic–inflammatory status rather than a clinically validated predictor.