The area and width of the gluteus maximus, gluteus medius/minimus and iliopsoas muscles are associated with the two-year prognosis of patients undergoing hip arthroscopy for femoroacetabular impingement syndrome

Abstract

Purpose

This study aims to establish critical thresholds for preoperative muscle area and width that predict two-year postoperative outcomes and to explore the correlation between baseline muscle area and two-year postoperative recovery in patients undergoing hip arthroscopy for femoroacetabular impingement syndrome (FAIS).

Study design

Retrospective Cohort Study, Level IV.

Methods

A cohort of 74 patients treated between March 2021 and March 2022 was analyzed. Preoperative MRI measurements were obtained for the cross-sectional area (CSA) and width of the iliopsoas, gluteus maximus, and gluteus medius/minimus muscles. Two years postoperatively, clinical outcomes including the International Hip Outcome Tool-12 (iHOT-12) and the modified Harris Hip Score (mHHS) were assessed. We evaluated correlations between preoperative muscle areas and postoperative outcomes using Pearson correlation coefficients. Multivariate linear regression models were constructed to identify independent associations between muscle CSA/width and postoperative outcomes. Additionally, receiver operating characteristic (ROC) curve analyses were used to explore potential relationships and thresholds between muscle area and the Patient Acceptable Symptom State (PASS) and the Minimal Clinically Important Difference (MCID).

Results

ROC analyses demonstrated that pre-G-med/min width and CSA predicted several two-year endpoints, with notable performance for iHOT-12-PASS (width AUC 0.68, P = 0.0070; CSA AUC 0.66, P = 0.020) and mHHS-PASS (width AUC 0.85, P = 0.0030; CSA AUC 0.80, P = 0.0090); gluteus maximus (width AUC 0.67, P = 0.012; CSA AUC 0.67, P = 0.013) and iliopsoas (width AUC 0.69, P = 0.0060; CSA AUC 0.72, P = 0.0010) can predict iHOT-12-PASS. Correlation analyses revealed small-to-moderate positive relationships between preoperative gluteus medius/minimus width and CSA with postoperative mHHS and iHOT-12 scores. In multivariate models, Pre-G-med/min CSA remained a significant positive predictor of 2-year mHHS (β = 0.33, P = 0.034), while male sex emerged as a significant positive predictor for 2-year iHOT-12 (β = 0.35, P = 0.030 in the cross-sectional area model and β = 0.28, P = 0.049 in the width model).

Conclusion

Preoperative widths and cross-sectional areas of the gluteus medius/minimus, gluteus maximus, and iliopsoas muscles may be associated with, and to a limited extent help predict, the clinical outcomes at two years postoperatively in patients with FAIS.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12891-026-10096-5.

Introduction

Femoroacetabular impingement syndrome (FAIS) is a morphological abnormality of the hip joint that is commonly observed in young and active adults [1, 9, 26]. Surgical intervention is a commonly employed treatment for FAIS. Existing evidence demonstrates that, in the short term, surgical procedures can lead to significant improvements in joint function and a reduction in pain, thereby enhancing overall clinical outcomes [6, 9].

Previous studies have reported continuous changes in muscle area following surgery for FAIS and confirmed a potential relationship between these changes and improvements in subjective symptoms [28, 29]. An increase in muscle area postoperatively may indicate better subjective scoring [29]; However, there are still limited reports on how to use more clinically applicable measurement methods to predict postoperative outcomes and muscle recovery. As part of the rehabilitation protocol, limited weight-bearing and avoidance of excessive range of motion (ROM) typically commence postoperatively, which may lead to potential muscle atrophy and loss of lower limb strength [3]. A greater baseline hip muscle mass may effectively compensate for this muscle loss, thereby influencing clinical outcomes. It remains to be determined whether a cut-off value for baseline muscle mass exists that can reliably predict clinical prognosis and serve as a predictor of clinical outcomes.

Therefore, the primary aim of this study is to establish critical thresholds for baseline muscle area and width that can predict two-year clinical outcomes in patients with femoroacetabular impingement syndrome (FAIS). The secondary aim is to investigate the relationships between baseline muscle area and two-year postoperative outcomes. We hypothesize that in patients undergoing hip arthroscopy for FAIS, preoperative muscle cross-sectional area (CSA) and muscle width will serve as predictors of two-year postoperative clinical outcomes, with higher preoperative muscle quality exhibiting a positive correlation with two-year postoperative recovery.

Methods

After approval from the Institutional Review Board, these data were continuously reviewed at our institution from March 2021 to March 2022. Patients diagnosed with femoroacetabular impingement syndrome (FAIS) based on clinical symptoms and imaging findings, who failed to respond to conservative treatment and subsequently underwent hip arthroscopy, were included for eligibility assessment. The exclusion criteria were as follows: incomplete MRI data, history of previous hip surgery, and concomitant hip diseases, including Tonnis grade > 1, avascular necrosis, osteoid osteoma, synovial chondromatosis, pigmented villonodular synovitis, and developmental dysplasia of the hip (DDH).

Surgical technique

The primary hip arthroscopy was performed by a fellowship-trained surgeon (Y.X.) with over 10 years of experience in hip arthroscopy using the standard supine approach described in previous studies [22]. Following anesthesia and traction, the anterolateral, anterior, and mid-anterior portals were established. A thorough examination of the central compartment was conducted through capsulotomy between the portals to assess the acetabular rim, acetabular labrum, articular cartilage, and ligaments. After the examination, synovectomy was performed on inflamed synovium. Acetabuloplasty was carried out for pincer deformity. If a labral tear was observed, labral repair was performed using a multi-suture anchor technique if the remaining fibrous tissues were sufficient to restore the sealing function of the labrum. If the remaining fibrous tissue was of poor quality or insufficient to restore the sealing function, labral reconstruction was performed using an autologous iliotibial band. Additionally, chondroplasty was performed for partial-thickness chondral lesions and cartilage flaps. After treating the central compartment, the arthroscope was moved to the peripheral compartment, where femoroplasty was performed to address cam deformities. Finally, the capsule was routinely repaired.

Rehabilitation protocol

All patients followed the same instruction of rehabilitation. Isometric contraction and passive range-of-motion exercises were allowed during the first 2 days following surgery. Partial weight-bearing exercises to restore range of motion and regular gait were permitted, allowing for partial weight bearing from day 3 to week 3. Full weightbearing walking was permitted at week 4. Full weightbearing muscle strength exercises and dynamic balance exercises were permitted at week 6. Patients could gradually resume activities and return to sports based on their tolerance.

Data collection

We collected patients’ demographic characteristics, including age at the time of surgery, sex, affected side, height, weight, body mass index (BMI), and symptom duration. Preoperative radiographic evaluations were performed to obtain the alpha angle (AA) on the Dunn view, the lateral center-edge angle (LCEA), and the Tonnis grade on the anteroposterior (AP) view, as described in previous studies [13, 17, 19].

Patient-reported outcomes (PROs), including the Visual Analog Scale (VAS) for pain, the modified Harris Hip Score (mHHS), the International Hip Outcome Tool-12 (iHOT-12), and the Tegner Activity Scale, were used to assess hip function and activity level [4, 7, 10]. The VAS, mHHS, and iHOT-12 were typically collected via questionnaires preoperatively and again during follow-up, at least two years postoperatively, through telephone interviews or outpatient visits. Patient acceptable symptom state (PASS) thresholds used were defined in published studies: iHOT-12 = 69.1, mHHS = 74 [5, 21]. MCID cutoff values for VAS, mHHS, and iHOT-12 were calculated using a distribution-based method proposed by Norman et al. [20] with the MCID cutoff set to half the standard deviation of the preoperative scores in the cohort. We then computed the proportion of patients achieving MCID and PASS.

MRI measurement

All enrolled patients underwent MRI examinations preoperatively using a 3.0 T scanner (Magnetom Trio with Tim system, Siemens Healthcare, Erlangen, Germany). Fat-saturated proton density (FSPD) sequences and T2-weighted sequences were performed in the axial, coronal, and oblique sagittal planes. The widest circumference of the femoral head was observed on the corresponding coronal images [8], and the CSA of the muscles was measured on the axial slices. As shown in Fig. 1, we measured the CSA of the iliopsoas (IP), gluteus medius/minimus complex (G-med/min), and gluteus maximus (G-max). Muscle width is defined as the thickness at the midpoint of the longest axis of the muscle in the maximum cross-section of the femoral head.

Fig. 1.

Measurements of hip muscle cross-sectional area (CSA). The shaded region denotes area, the dashed line denotes width, and the solid line denotes length. IP - Iliopsoas; G-med/min - Gluteus Medius/Minimus Complex; G-max - Gluteus Maximus

Statistical analysis

All data were analyzed using IBM SPSS Statistics for Windows, version 27.0. Continuous variables are presented as mean ± standard deviation (range). Categorical variables are presented as counts (percentages). The relationship between muscle area and the achievement of the two-year postoperative PASS and MCID was evaluated using ROC curve analysis, with AUC values approaching 1 indicating excellent predictive ability. The normality of continuous variables was assessed using the Shapiro-Wilk test before Pearson correlation. Pearson correlation analysis was used for normally distributed variables, whereas Spearman correlation analysis was used for non-normally distributed variables. Correlation analyses were conducted to examine the associations between two-year postoperative outcomes and baseline muscle area and width. A multivariable linear regression model was constructed to identify potential factors influencing postoperative PROs. Candidate variables for the multivariable linear regression models were selected a priori on the basis of clinical relevance and previous literature, and included sex, age, body mass index, symptom duration, and the corresponding preoperative muscle cross-sectional area or width.

Results

Intrarater Reliability was assessed using a one-way mixed effects intraclass correlation coefficient model, demonstrating excellent reliability for a sample of 15 MRI measurements reviewed by one rater, with an ICC(1,2) of 0.957 (95% confidence interval, 0.879–0.985; P < 0.001). Interrater reliability was determined using a 2-way mixed effect intraclass coefficient model and was determined to be excellent for a sample of 15 MRI measurements reviewed by 2 raters, with ICC(3,2) of 0.961 (95% confidence interval, 0.890–0.987; P< 0.001).

As shown in Fig. 2, a total of 74 patients were included in the study cohort. The mean follow-up duration was 28.08 ± 2.37 months, and 73.3% of patients completed the follow-up. The baseline characteristics were listed in Table 1. The mean AA was 62.74 ± 7.23 degrees, and the mean LCEA was 34.10 ± 6.56 degrees.

Fig. 2.

Flow chart of patient selection

Table 1.

Patients’ characteristics

	Mean ± SD (range) or No. of hips (%)
Gender
Male	27 (36.50)
Female	47 (63.50)
Age(x ± s, y)	37.35 ± 9.51
Side
left	49 (66.20)
right	25 (33.80)
BMI(x ± s, kg/m²)	22.71 ± 2.92
Duration of symptoms, months	26.61 ± 24.15
AA, deg	62.74 ± 7.23
LCEA, deg	34.10 ± 6.56
Tonnis grade
0	28 (37.80)
1	46 (62.20)
Follow-up period, months	28.08 ± 2.37

Pre-G-med/min width significantly predicted iHOT-12-PASS (AUC = 0.68, P = 0.0070; cutoff = 24.85) and Pre-G-med/min CSA also predicted iHOT-12-PASS (AUC = 0.66, P = 0.02; cutoff = 925.80), while for mHHS-PASS the same Pre-G-med/min width showed significance (AUC = 0.85, P = 0.0030; cutoff = 752.00) and Pre-G-med/min CSA likewise predicted mHHS-PASS (AUC = 0.80, P = 0.0090; cutoff = 752.00); Pre-IP width significantly predicted iHOT-12-PASS (AUC = 0.69, P = 0.0060; cutoff = 20.95) and mHHS-PASS (AUC = 0.77, P = 0.018; cutoff = 19.80), Pre-IP CSA predicted iHOT-12-PASS (AUC = 0.72, P = 0.0010; cutoff = 521.50), and Pre-IP width and Pre-IP CSA also reached significance for iHOT-12-PASS (AUC = 0.67, P = 0.013; cutoff = 29.25; AUC = 0.67, P = 0.012; cutoff = 3484.00). All other variables did not reach statistical significance (P > 0.05)(Table 2).

Table 2.

Predictive performance of preoperative hip muscle area and width for achieving two-year MCID and PASS (ROC analysis). Pre-G-med/min, Preoperative gluteus medius/minimus. Pre-IP, Preoperative iliopsoas. Pre-G-max, Preoperative gluteus maximus

	iHOT-12-MCID	mHHS-MCID	iHOT-12-PASS	mHHS-PASS
Pre-G-med/min width	0.45(0.28–0.61)	0.51(0.35–0.69)	0.68(0.56–0.81)	0.85(0.72–0.98)
P Value	0.55	0.91	0.0070	0.0030
Pre-G-med/min CSA	0.55(0.38–0.72)	0.66(0.51–0.80)	0.66(0.53–0.78)	0.80(0.63–0.97)
P Value	0.56	0.052	0.020	0.0090
Pre-G-max width	0.49(0.34–0.65)	0.45(0.30–0.59)	0.67(0.55–0.80)	0.66(0.54–0.78)
P Value	0.92	0.51	0.012	0.18
Pre-G-max CSA	0.48(0.33–0.64)	0.57(0.41–0.73)	0.67(0.55–0.79)	0.63(0.47–0.80)
P Value	0.84	0.38	0.013	0.26
Pre-IP width	0.51(0.33–0.68)	0.52(0.37–0.68)	0.69(0.57–0.81)	0.77(0.64–0.91)
P Value	0.96	0.80	0.0060	0.018
Pre-IP CSA	0.52(0.36–0.69)	0.59(0.44–0.73)	0.72(0.61–0.84)	0.69(0.51–0.88)
P Value	0.77	0.28	0.0010	0.093

Bold indicates a statistically significant difference (p < 0.05)

As shown in Table 3 The correlation analysis revealed that preoperative muscle morphology, including the width and CSA of specific muscles, was significantly associated with postoperative outcomes in patients with FAIS. For the postoperative modified Harris Hip Score, the preoperative gluteus medius/minimus width (r = 0.27, P = 0.023) and CSA (r = 0.26, P = 0.024) demonstrated significant positive correlations. Similarly, for the postoperative International Hip Outcome Tool-12, significant correlations were observed with the preoperative gluteus medius/minimus width (r = 0.30, P = 0.010), iliopsoas width (r = 0.27, P = 0.018), gluteus medius/minimus CSA (r = 0.24, P = 0.037), and gluteus maximus CSA (r = 0.27, P = 0.021).

Table 3.

Correlation between preoperative muscle morphology (Width and CSA) and postoperative outcomes in patients with FAIS. Post-mHHS: modified Harris Hip Score at 2 years postoperatively. Post-iHOT-12, iHOT-12 score at 2 years postoperatively. Pre-G-med/min, Preoperative gluteus medius/minimus. Pre-IP, Preoperative iliopsoas. Pre-G-max, Preoperative gluteus maximus

	r	P
Post-mHHS
Pre-G-med/min width	0.27	0.023
Pre-G-med/min CSA	0.26	0.024
Post-iHOT-12
Pre-G-med/min width	0.30	0.010
Pre-IP width	0.27	0.018
Pre-G-med/min CSA	0.24	0.037
Pre-G-max CSA	0.27	0.021

Bold indicates a statistically significant difference (p < 0.05)

As shown in Table 4, after adjustment, Pre-G-med/min CSA demonstrated a significant positive association with 2-year postoperative mHHS (β = 0.33, P = 0.034). All other variables did not reach statistical significance (P > 0.05).

Table 4.

Multivariate linear regression of baseline factors and muscle cross-sectional area on 2-year postoperative mHHS. Pre-G-med/min, Preoperative gluteus medius/minimus. Pre-IP, Preoperative iliopsoas. Pre-G-max, Preoperative gluteus maximus

	Beta	P
Sex	0.15	0.34
Age	0.063	0.63
BMI	-0.13	0.34
AA	0.17	0.17
LCEA	-0.12	0.33
Pre-G-med/min CSA	0.33	0.034
Pre-G-max CSA	-0.26	0.10
Pre-IP CSA	0.068	0.66

Bold indicates a statistically significant difference (p<0.05)

As shown in Table 5, after adjustment, male demonstrated a significant positive association with 2-year postoperative iHOT-12 (β = 0.35, P = 0.030). All other variables did not reach statistical significance (P > 0.05).

Table 5.

Multivariate linear regression of baseline factors and muscle cross-sectional area on 2-year postoperative iHOT-12. Pre-G-med/min, Preoperative gluteus medius/minimus. Pre-IP, Preoperative iliopsoas. Pre-G-max, Preoperative gluteus maximus

	Beta	P
Sex	0.35	0.030
Age	-0.058	0.64
BMI	-0.19	0.15
AA	0.012	0.92
LCEA	-0.21	0.089
Pre-G-med/min CSA	0.17	0.27
Pre-G-max CSA	-0.13	0.41
Pre-IP CSA	0.12	0.42

Bold indicates a statistically significant difference (p<0.05)

As shown in Table 6, All variables did not reach statistical significance (P > 0.05).

Table 6.

Multivariate linear regression of baseline factors and muscle width on 2-year postoperative mHHS. Pre-G-med/min, Preoperative gluteus medius/minimus. Pre-IP, Preoperative iliopsoas. Pre-G-max, Preoperative gluteus maximus

	Beta	P
Sex	0.079	0.59
Age	0.018	0.89
BMI	-0.075	0.57
AA	0.15	0.24
LCEA	-0.093	0.45
Pre-G-med/min width	0.0010	0.99
Pre-G-max width	0.24	0.089
Pre-IP width	0.082	0.50

Bold indicates a statistically significant difference (p<0.05)

As shown in Table 7, after adjustment, male demonstrated a significant positive association with 2-year postoperative iHOT-12 (β = 0.28, P = 0.049). All other variables did not reach statistical significance (P > 0.05).

Table 7.

Multivariate linear regression of baseline factors and muscle width on 2-year postoperative iHOT-12. Pre-G-med/min, Preoperative gluteus medius/minimus. Pre-IP, Preoperative iliopsoas. Pre-G-max, Preoperative gluteus maximus

	Beta	P
Sex	0.28	0.049
Age	-0.066	0.57
BMI	-0.19	0.13
AA	-0.027	0.82
LCEA	-0.21	0.071
Pre-G-med/min width	0.18	0.22
Pre-G-max width	0.16	0.24
Pre-IP width	0.048	0.67

Bold indicates a statistically significant difference (p<0.05)

Discussion

The primary findings of this study include the identification of the relationship between preoperative muscle volume/area and postoperative clinical outcomes in patients with FAIS, as well as the establishment of specific cutoff values for muscle volume/area and width predictive of postoperative outcomes.

In this study, MRI was utilized to measure the cross-sectional area (CSA) of muscles around the hip joint, a reliable and reproducible method for quantifying muscle area and reflecting muscle strength [14, 18, 27]. Our research exhibit that there exists a threshold value for preoperative muscle area and width that can effectively predict clinical outcomes. Malloy et al. found that, compared to the asymptomatic contralateral hip joint, patients with unilateral FAIS had significantly reduced muscle CSA (both rectus femoris and gluteal muscles) in the symptomatic hip. Interestingly, CSA was positively correlated with preoperative pain levels, which was interpreted as a compensatory change [15]. In patients with FAIS, a larger preoperative muscle CSA often reflects higher levels of physical engagement and is a result of the body’s active compensation for pain [24]. Thus, despite more pronounced pain, these patients often possess a solid foundation of physical activity, and higher preoperative muscle reserves positively contribute to postoperative rehabilitation. It has been reported that the average time to resume daily activities exceeds three months [2, 11]. Most patients do not recover to their preoperative level of physical participation within a year, those with better muscle strength may expedite this process. Conversely, patients with smaller preoperative muscle CSA may experience muscle atrophy due to long-term disuse or more severe functional impairments. Even if surgical correction addresses anatomical issues, the effectiveness of postoperative rehabilitation may be limited, which could be a significant reason for variations in prognosis. This further emphasizes the importance of preoperative rehabilitation—through targeted muscle strengthening and gait optimization. Preoperative rehabilitation not only enhances the potential for postoperative recovery but also improves compensatory patterns, establishing a solid foundation for postoperative balance restoration [23].

A biomechanical study demonstrated that after hip arthroscopy, FAIS patients exhibited a significant increase in the dynamic range of motion (ROM) of the sagittal plane during single-leg squats. The greater changes in sagittal plane ROM during squatting were significantly correlated with changes in clinical scores in FAIS patients [12]. This change is highly dependent on preoperative muscle strength [16]. Studies have shown that CSA is positively correlated with strength, and a decrease in CSA may account for up to 90% of the loss in hip joint strength. Since muscle CSA directly determines the strength reserve of the muscles surrounding the hip joint, we have reason to believe it plays a crucial role not only in postoperative functional recovery but also as a key indicator for predicting the long-term clinical prognosis of patients.

This study found that the gluteus medius CSA and its width are significantly associated with postoperative outcomes across multiple statistical analyses. This association may reflect the biomechanical roles of the gluteus medius in hip stability, gait compensation, and distribution of muscle strength [25], potentially influencing the trajectory of postoperative functional recovery. However, Several aspects of these findings should be interpreted conservatively. The discriminative ability of the ROC models was modest, with most area-under-the-curve values falling within the 0.6–0.7 range, indicating limited rather than strong predictive performance. Likewise, the correlations between preoperative muscle morphology and postoperative outcomes were small to moderate in magnitude (r ≈ 0.24–0.30), and statistical significance should not be equated with clinical importance. In the multivariable analysis, only the gluteus medius/minimus cross-sectional area remained an independent predictor, so the broader associations observed in the univariable and ROC analyses are best regarded as hypothesis-generating. The cutoff values derived from the ROC analyses are intended only as preliminary references that might assist preoperative risk stratification, patient counseling, and the targeting of prehabilitation; they require external validation in larger, multicenter cohorts before they can be adopted in routine clinical practice.

In the multivariable models, male sex emerged as an independent positive predictor of the 2-year iHOT-12 but was not independently associated with the mHHS. This finding should be interpreted cautiously. Possible explanations include sex-related differences in baseline symptom severity and self-reported function. Because the iHOT-12 weights quality-of-life and psychosocial domains more heavily than the mHHS, sex differences in these domains may partly account for the discordant results across the two instruments. Given the modest sample size and the exploratory nature of this analysis, the role of sex warrants confirmation in larger, prospectively designed cohorts.

Limitations

While our study is the first to propose threshold values for muscle area related to two-year outcomes in FAIS patients, some limitations must be acknowledged. First, due to the retrospective nature of the study, selection bias is unavoidable. the follow-up rate was 73.3%, and the possibility of attrition bias arising from missing data could not be fully excluded. Second, the effects of fat infiltration on surgical prognosis were not considered. Third, the patients included in this study were selected from a single surgeon’s cohort, hence caution should be exercised when generalizing the results to the broader FAIS population. Forth, potentially important confounders, including rehabilitation protocol compliance, preoperative physical activity level, objective muscle strength, and sports participation, were not available and could not be incorporated into the models. Finally, the discriminative ability of the ROC models was modest, with most area-under-the-curve values in the 0.6–0.7 range, and multiple correlation and ROC analyses were performed without correction for multiple comparisons; the present findings are therefore exploratory and require confirmation in larger, prospective, multicenter studies.

Conclusion

Preoperative widths and cross-sectional areas of the gluteus medius/minimus, gluteus maximus, and iliopsoas muscles may be associated with, and to a limited extent help predict, the clinical outcomes at two years postoperatively in patients with FAIS.

Supplementary Information

Supplementary Material 1. ^{(12.2KB, docx)}

Abbreviations

FAIS

Femoroacetabular impingement syndrome

CSA

Cross-sectional area

MRI

Magnetic resonance imaging

iHOT-12

International Hip Outcome Tool-12

mHHS

modified Harris Hip Score

PASS

Patient Acceptable Symptom State

MCID

Minimal Clinically Important Difference

ROC

Receiver operating characteristic

AUC

Area under the curve

VAS

Visual Analog Scale

BMI

Body mass index

AA

Alpha angle

LCEA

Lateral center-edge angle

ROM

Range of motion

G-med/min

Gluteus medius/minimus complex

G-max

Gluteus maximus

IP

Iliopsoas

PRO

Patient-reported outcome

ICC

Intraclass correlation coefficient

Authors’ contributions

1. Conception and design of the study: Yuang Hao, Xin Miao, Tong-Chuan He, Yichuan Zhu2. Acquisition of data: Yuang Hao, Yichuan Zhu3. Analyses of data: Yuang Hao4. Drafting the work: Yuang Hao 5. Revising it critically for important intellectual content: Xin Miao and Yan Xu.6. contributions during the revision stage: Jinwoo Kim7. Final approval of the version to be published: Yuang Hao, Yao Pan, Tong-Chuan He, Xin Miao, Jinwoo Kim and Yan Xu.

Funding

National Key Research and Development Program of China (Key Technologies Research and Development Program, 2023YFC2410803), National Key Research and Development Program of China (Key Technologies Research and Development Program, 2024YFC3606704), Peking University Third Hospital (BYSYZD2023011), Peking University Third Hospital Fund for Interdisciplinary Research(BYSYJC2024026),National Natural Science Foundation of China (82372496), Natural Science Foundation of Beijing Municipality (F252057) funded this study.

Data availability

All relevant data supporting the conclusions are included within the article and tables. The datasets used and/or analyzed during the current study available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study was approved by the Ethics Committee of Peking University Third Hospital (No. M2019193). Informed consent was obtained from all individual participants included in the study. All methods were performed in accordance with the guidelines and regulations of the Ethics Committee of Peking University Third Hospital, and the study adhered to the Declaration of Helsinki.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.