This study is an observational study conducted at the Department of Orthopedics, the First Hospital of Jilin University. The study was approved by the Ethics Committee of First Hospital of Jilin University (2022-085) and followed the principles of the Declaration of Helsinki. All participants provided written informed consent, and patient privacy was strictly safeguarded. The study included patients who underwent hip arthroplasty in our department between September 2018 and September 2023. The patients who met the specified inclusion criteria were included, and those who met the exclusion criteria were excluded. Inclusion criteria: a. underwent hip arthroplasty (total hip arthroplasty or hemiarthroplasty); b. consent to participate in the study. Exclusion criteria: a. with severe hepatic diseases; b. with severe renal diseases; c. loss to follow-up; d. unavailable data; e. died during the follow-up (Figure 1). All surgeries were performed by the senior author (Dongsong Li) or in his presence and direction.

Baseline demographics (age, BMI, sex, residence, smoke and alcoholism history), comorbidities (osteoporosis, hypertension, and so on), fracture history, surgical variables (diagnosis, surgical procedures, and anesthesia), and laboratory parameters including hemoglobin (Hb) (g/L), international normalized ratio (INR), blood glucose (GLU) (mmol/L), albumin (ALB) (g/L) and UA (µmol/L), were extracted from electronic medical records. Abnormal electrocardiograms or radiographs were adjudicated by two independent specialists. Charlson comorbidity index (CCI) was calculated to comprehensively conclude the preoperative comorbid condition (15). Hyperuricemia was defined as UA >420 μmol/L for both males and females according to clinical guidelines.

Patients were prospectively followed for one year. The primary outcome was functional independence, defined as a Functional Independence Measure (FIM) score ≥108 at 1-year follow-up (16). The FIM is a therapist-administered 18-item scale assessing self-care, mobility, and cognition, with total scores ranging from 18 (complete dependence) to 126 (complete independence). Secondary outcomes included: Daily activity level measured by the UCLA Activity Scale (UCLAAS); Patient-Reported Satisfaction (PRS) using Likert-scale questionnaires (1 = "Very dissatisfied” to 5 = "Very satisfied”): Surgery Satisfaction Score (SSS); Pain Satisfaction Score (PSS); Activity Satisfaction Score (ASS); Health Satisfaction Score (HSS); Life Satisfaction Score (LSS).

Continuous variables were presented as mean ± standard deviation, whereas categorical variables were reported as count (percentage). The distribution of continuous variables was evaluated using the Shapiro–Wilk test. Group comparisons utilized the Student's t-test (normal) or the Wilcoxon rank-sum test (non-normal). Categorical variables were analyzed via chi-square/Fisher's exact tests. Restricted cubic splines (RCS) with four knots analyzed nonlinear relationships between UA levels and outcomes based on multiple regression models, and the cutoff values of UA were identified according to the RCS (4). The outcomes of patients with different UA groups were compared. Lastly, multivariable logistic and linear regression (adjusted for age, BMI, and CCI ≥4) quantified UA's association with outcome. The significance level was set at p < 0.05, and Bonferroni correction was used to adjust the p-values to reduce the impact of multiple tests. All analyses were performed using R software version 4.2.2 (R Foundation for Statistical Computing, Vienna, Austria).

Finally, this study enrolled 468 patients following hip arthroplasty, stratified into functional independence (n = 393, FIM ≥108) and non-independence (n = 75) groups at 1-year follow-up. The baseline characteristics of patients included in this study are summarized in Table 1. Significant differences were observed in age and comorbidity burden between groups. Patients with functional independence were younger (58.6 ± 12.9 vs. 63.7 ± 13.2 years, p < 0.001) and had lower rates of Charlson Comorbidity Index ≥3 (8.65% vs. 18.67%, p = 0.009) compared to the non-independence group. No significant differences were noted in sex (54.5% vs. 56.0% female, p = 0.289), BMI (24.2 ± 3.9 vs. 23.8 ± 3.9 kg/m², p = 0.805), or lifestyle factors (smoking: 44.0% vs. 50.7%, p = 0.728). Laboratory parameters revealed higher hemoglobin levels (132.2 ± 19.4 vs. 124.7 ± 21.5 g/L, p = 0.003) and lower INR (0.96 ± 0.08 vs. 1.00 ± 0.09, p < 0.001) in the independence group 1. Continuous UA levels did not differ significantly between groups (336.4 ± 100.95 vs. 353.63 ± 159.31 μmol/L, p = 0.852) while the patients with independence have significantly high rates of HUA (19.34% vs. 32.00%, p = 0.014).

RCS analysis revealed a U-shaped relationship between serum uric acid (UA) and functional independence (Figure 2). RCS was established based on three models, including univariate models, models adjusted for age and sex, and models adjusted for the variables with significant differences in Table 1 (age, CCI, INR, and Hb). For the primary outcome, the risk of functional dependence increased at both extremes of UA levels compared to intermediate levels, and this nonlinear pattern persisted across all multivariable models. Similar nonlinear U-shape trends were observed for secondary outcomes in UCLAAS, PSS, and ASS. To better explore the relationships, we further divided the populations according to the cutoff values of UA determined by RCS models: low UA: <259.77 μmol/L; middle UA: ≥259.77 μmol/L and <436.32 μmol/L; high UA: ≥436.32 μmol/L. Moreover, according to the quantiles of UA, the patients were also grouped into Q1–Q4 groups (low values of UA to high levels of UA).

Dichotomous (HUA vs. normal), tiered (low/mid/high UA), and quartile-based (Q1-Q4) analyses consistently demonstrated worse functional outcomes at UA extremes (Figure 3). The baseline characteristics of patients grouped by different UA levels were summarized in Supplementary Table S1–S3. HUA was associated with lower functional independence rates and lower SSS scores compared with normal individuals (all p < 0.05). Similar to the nonlinear analysis, patients with low UA and high UA may have significantly lower functional independence rates compared with normal individuals. For the secondary outcomes, the adults with high UA may have lower UCLAAS compared with those with middle UA. Surprisingly, patients with moderate UA seemed to have lower ASS scores. In the four-tier system, compared with UA Q3, the patients with UA Q1 and Q4 may have significantly higher nonfunctional independence rates. Similarly, UA Q3 may relate to lower PSS and ASS scores, which seems paradoxical compared with the primary outcome analysis. Sex-stratified analyses confirmed that the U-shaped relationship persisted in both sexes (Supplementary Tables S4, S5). In males, low UA and high UA were linked to higher non-functional independence rates and lower UCLAAS. In females, tiered (low/mid/high UA) and quartile-based (Q1-Q4) UA were associated with PSS and ASS scores, whereas middle UA conferred the most favorable outcomes.

To avoid the bias caused by co-factors, we further established multivariate models to explore the associations between UA and outcomes. We established three kinds of models for FIM: univariate models, multivariate models adjusted for the factors selected by stepwise method (INR, Hb, age, residence, and reason for surgery), and multivariate models adjusted for factors with significance in univariate models (INR, Hb, age, CCI, and osteoporosis, Supplementary Table S6). The results of multivariable models for FIM were summarized in Table 2. UA as a continuous variable showed no significant association with the primary outcome (OR: 1.001, 95% CI 0.999–1.003; p = 0.224) in univariate models, while showing significant associations in multivariate models 1 and 2. HUA and extreme UA levels (Low/High UA, Q1/Q4) exhibited clinically meaningful associations with adverse outcomes (all p < 0.05) in all models.

The results of multivariate models for secondary outcomes were summarized in Table 3. For UCLAAS, continuous UA, HUA, high UA, and UA Q4 were all significantly associated with lower UCLAAS scores, while the low UA and UA Q1 were not significant. For other outcomes, low UA and UA Q1 were estimated to relate to higher ASS scores significantly (all p < 0.05).