BACKGROUND:Immune checkpoint inhibitor (ICI)-associated myocarditis is an uncommon but life-threatening complication. Management is particularly challenging when high-dose corticosteroids are contraindicated and in the presence of clinical-biomarker dissociation. CASE SUMMARY:A 69-year-old man presented with cardiogenic shock and ventricular tachycardia 3 weeks after his first dose of pembrolizumab for lung adenocarcinoma. Despite corticosteroid constraints owing to pulmonary tuberculosis and the complexity of clinical decision-making during tapering with rebound troponin T elevation, the patient achieved favorable long-term survival. DISCUSSION:This case demonstrates a successful management strategy for fulminant ICI-associated myocarditis under the dual constraints of limited corticosteroid use and persistent biomarker elevation, offering valuable clinical insights. TAKE-HOME MESSAGES:Rebound elevation of troponin T levels may indicate myocardial remodeling rather than active injury, supporting corticosteroid tapering in clinically stable patients. In patients with ICI-associated myocarditis and contraindications to corticosteroids, early steroid tapering combined with alternative immunosuppressive therapy may still achieve favorable outcomes.
BACKGROUND:The interplay between mitochondrial dysfunction and immune infiltration in heart failure with preserved ejection fraction (HFpEF) remains poorly understood. This study aimed to elucidate this relationship and identify key regulatory genes. METHODS:We integrated bioinformatics analyses of human HFpEF datasets (GSE108904, GSE126062) with experimental validation. Differential expression analysis, weighted gene co-expression network analysis (WGCNA), and the MitoCarta3.0 database were used to identify hub mitochondrial genes. Immune cell infiltration was assessed, and key findings were validated in peripheral blood mononuclear cells (PBMCs) from HFpEF patients and in mouse/cell models of the disease. RESULTS:Two hub mitochondrial genes, CHCHD1 and EFHD1, were associated with HFpEF. Immune profiling revealed increased macrophage infiltration in HFpEF, which was negatively correlated with the expression of both hub genes. Experimental validation confirmed a significant decrease in EFHD1 expression in HFpEF patient PBMCs, HFpEF mouse heart tissues, and a cellular inflammation model. Critically, EFHD1 expression showed a significant negative correlation with clinical indicators of heart failure (NT-proBNP and E/e' ratio). CONCLUSION:Our integrated analysis reveals CHCHD1 and EFHD1 as key mitochondrial genes linking mitochondrial dysfunction to immune dysregulation in HFpEF, offering new insights into its molecular mechanisms and potential therapeutic targets.
BACKGROUND:Coronary artery bypass grafting (CABG) is associated with a significant long-term risk of major adverse cardiovascular and cerebrovascular events (MACCE). The triglyceride-glucose (TyG) index, a surrogate marker of insulin resistance, has demonstrated predictive value for cardiovascular outcomes but does not account for the heterogeneity of obesity phenotypes. Integrating the TyG index with measures of visceral adiposity may improve risk stratification in patients after CABG. OBJECTIVES:This study aimed to systematically evaluate and compare the ability of four novel TyG‑adiposity indices (TyG‑CVAI, TyG‑BRI, TyG‑LAP, and TyG‑ABSI) to predict long‑term MACCE following CABG, and to explore the potential mediating role of systemic inflammation in this association. METHODS:In a retrospective cohort of 2547 consecutive patients undergoing CABG, the four TyG‑adiposity indices were calculated from baseline clinical and laboratory data. The primary endpoint was the incidence of MACCE. Associations were assessed using multivariable Cox regression, restricted cubic splines for dose‑response relationships, and mediation analysis. Model performance was evaluated using Uno's C‑index, time‑dependent ROC curves, and decision curve analysis. RESULTS:Over a median follow‑up of 4.25 years, 667 MACCE events occurred. All TyG‑adiposity indices were significantly higher in patients who experienced MACCE. After full adjustment, each index independently predicted increased MACCE risk, with strong dose‑response relationships (all P for trend <0.001). TyG‑BRI demonstrated the highest discriminatory ability (Uno's C‑index 0.709, 95 % CI 0.681-0.737), outperforming TyG‑CVAI (0.685), TyG‑LAP (0.651), TyG‑ABSI (0.587), and the TyG index alone (0.576). C‑reactive protein mediated 12.26 %-24.20 % of the association between the indices and MACCE risk. CONCLUSION:TyG‑adiposity composites, particularly TyG‑BRI, are independently associated with long‑term MACCE risk after CABG and provide incremental predictive value over the TyG index or simple anthropometric measures. The association is partly mediated by systemic inflammation. These indices may help refine risk stratification across metabolic phenotypes and identify high‑risk patients for intensified management.
INTRODUCTION:Ceramides and chronic kidney disease (CKD) independently associate with coronary microvascular dysfunction (CMD), but their combined association remains unestablished, warranting investigation. METHODS:This cross-sectional analysis included coronary artery disease (CAD) patients from the CRUISE-MET trial (NCT06383208) who underwent coronary angiography and ceramide profiling measurement. The microvascular resistance was assessed in all patients using the angiography-derived microcirculatory resistance (AMR). CKD was classified based on both estimated glomerular filtration rate and urine albumin-to-creatinine ratio. The combined associations and statistical mediation of ceramides and renal dysfunction on CMD was examined, with sensitivity analyses to ensure the robustness. RESULTS:In this analysis of 794 CAD patients, CKD patients demonstrated significantly higher AMR values (2.56 ± 0.51 vs. 2.46 ± 0.48, p = 0.003) and CMD prevalence (57.1% vs. 41.7%, p < 0.001). Three ceramide species - Cer(18:1/16:0), Cer(18:1/18:0), and Cer(18:1/24:1) - showed strong positive correlations with AMR (all p < 0.05). LASSO regression identified 14 factors associated with CMD, with E/e', CKD, and Cer(18:1/16:0) emerging as top correlates. Cer(18:1/16:0) demonstrated superior predictive performance (area under the curve [AUC] = 0.609, 95% CI: 0.570-0.648, p < 0.001), with its addition to baseline models significantly improving reclassification (δAUC = 0.03, p = 0.010; cNRI>0 = 0.23, p < 0.001; integrated discrimination improvement = 0.03, p < 0.001). Notably, patients with concurrent CKD and elevated Cer(18:1/16:0) exhibited significantly higher odds of CMD (OR = 3.74, 95% CI: 2.39-5.84). Statistical mediation analysis suggested that Cer(18:1/16:0) mediated 13.8% of the association between CKD and CMD. CONCLUSION:Ceramides, particularly Cer(18:1/16:0), partially mediated the statistical association between CKD and CMD. These findings highlight the potential role of sphingolipid metabolism in cardiorenal pathophysiology.
Background:Blood pressure variability (BPV) is recognized as an independent risk factor for cardiovascular morbidity and mortality. However, findings regarding the association between the coefficient of variation (CV) of 24-h blood pressure and mortality in patients with acute myocardial infarction (AMI) remain inconsistent. The aim of this study was to examine the association between 24-h CV and the risk of in-hospital, 30-day, 1-year, and 3-year mortality in AMI patients. Methods:This retrospective cohort study included AMI patients who were admitted to the intensive care unit (ICU) from the MIMIC-IV 2.0 database. The CV was calculated from blood pressure measurements taken during the first 24 h. Patients were divided into three groups according to tertiles of 24-h blood pressure CV. The relationship between 24-h CV and the risk of all-cause mortality in AMI patients was analyzed using logistic regression, Cox proportional hazards regression, and Kaplan-Meier survival analyses. Results:A total of 1291 eligible AMI patients were included. Compared with Q1 (lowest CV), Q3 (highest 24-h systolic blood pressure [SBP]-CV) was significantly associated with an increased risk of in-hospital (odds ratio [OR]: 3.333, 95% confidence interval [CI]: 1.725-6.439), 30-day (hazard ratio [HR]: 1.868, 95% CI: 1.201-2.904), and 1-year (HR: 1.420, 95% CI: 1.058-1.907) mortality. Both Q2 and Q3 of 24-h diastolic blood pressure (DBP)-CV were significantly associated with an increased risk of in-hospital, 30-day, 1-year, and 3-year mortality. Kaplan-Meier survival curves demonstrated that patients in Q3 had significantly lower 30-day, 1-year, and 3-year survival probabilities than those in other tertiles for both SBP and DBP. Conclusions:Our findings suggest that 24-h SBP-CV and DBP-CV are significantly associated with both short- and long-term outcomes in AMI patients. BPV may serve as a prognostic marker for all-cause mortality in AMI patients.
BACKGROUND:Coronary microvascular dysfunction (CMD) is a key mechanism in angina with nonobstructivecoronary arteries (ANOCA). While psychological stress affects cardiovascular health,the link between specific psychological symptoms and CMD in ANOCA patients remains unclear. AIMS:This study aimed to investigate the association of anxiety, depression, and sleep disorder with CMD assessed by invasive coronary physiology in ANOCA patients. METHODS:A total of 102 ANOCA patients underwent coronary angiography and invasive coronary function testing including thermodilution-derived index of microcirculatory resistance (IMR) andcoronary flow reserve. CMD was defined as coronary flow reserve <2.5 or IMR ≥25. Anxiety, depression,and sleep quality were assessed using the Generalized Anxiety Disorder-7, Patient HealthQuestionnaire-9, Pittsburgh Sleep Quality Index scales, respectively. RESULTS:Among the 102 enrolled ANOCA patients, 50.0% exhibited anxiety symptoms, 30.4%showed depression symptoms, and 67.6% reported sleep disorders. Patients with abnormal IMR hadsignificantly higher prevalence of anxiety (72.9% vs. 29.6%; P <0.001), depression (43.8% vs. 18.5%;P = 0.009), and sleep disorders (81.2% vs. 55.6%; P = 0.006) than those without. Multivariable analysisidentified anxiety as an independent predictor of CMD with abnormal IMR (odds ratio, 4.84; 95%confidence interval, 1.98-12.36; P <0.001). For CMD defined by abnormal coronary flow reserve, sleepdisorder showed a borderline association. Restricted cubic spline analyses showed that dose-dependentCMD risk increases with worsening psychological symptoms, especially for IMR-defined CMD. CONCLUSIONS:Anxiety, depression, and sleep disorders are prevalent and dose-dependently associatedwith CMD in ANOCA, particularly elevated IMR.
BACKGROUND:Heart failure with preserved ejection fraction (HFpEF) constitutes over 50% of heart failure cases but lacks disease-modifying therapies. The pathophysiological role of eicosapentaenoic acid (EPA) in HFpEF remains undefined. METHODS:Integrated lipidomics was conducted across HFpEF discovery and validation cohorts. "Two-hit" murine HFpEF model combining high-fat diet (HFD) and 0.5 g/l-NAME was established to recapitulate human metabolic-inflammatory pathology. EPA's efficacy was evaluated through prophylactic/therapeutic interventions (160/320 mg/kg/day, human-equivalent 2/4 g/day). Mechanistic studies integrated transcriptomics, molecular docking, triggering receptor expressed on myeloid cells 2 (TREM2) knockout, and siRNA silencing. RESULTS:Plasma EPA deficiency correlated with diastolic dysfunction severity and conferred incremental diagnostic value. High-dose EPA (4 g/day equivalent) prevented/reversed diastolic impairment and apoptosis in HFpEF mice. EPA rescued impaired efferocytosis through dual modulation of TREM2, concurrently enhancing functional transmembrane receptor expression while suppressing pathological ectodomain shedding. TREM2 ablation attenuated EPA-mediated benefits on diastolic function and efferocytosis. CONCLUSION:Our work identifies plasma EPA depletion as a potential biomarker for risk stratification and delineates the EPA-TREM2-efferocytosis axis as a putative therapeutic mechanism for HFpEF, suggesting the potential of EPA as a theranostic candidate.
BACKGROUND:Whether fractional flow reserve (FFR)-guided revascularization improves outcomes in atherosclerotic renal artery stenosis (ARAS) remains uncertain. OBJECTIVES:This study will evaluate whether the benefit of FFR-guided stenting sustained at 1 year including renal outcomes. METHODS:FAIR pilot is an investigator-initiated, multicentre, open-label trial with blinded endpoint adjudication. Adults with ARAS were randomized 1:1 to FFR-guided vs angiography-guided groups. Stenting was performed in the FFR-guided group when FFR <0.80, and it was performed for all in the angiography-guided group regardless of FFR. The percentage changes in ambulatory daytime mean systolic blood pressure (DMSBP), composite index of antihypertensive medicines (CIAHM), and renal outcomes were analyzed for 1 year. RESULTS:Among all 107 randomized patients, 1-year follow-up was achieved in 50 (93%) in the FFR-guided group and 51 (96%) in the angiography-guided group. At 1 year, DMSBP, CIAHM, and renal outcomes were similar between the groups, while stenting rate was lower with FFR guidance (46% vs 100%, P < 0.01). Patients with FFR <0.80 who underwent stenting experienced the greatest blood pressure reductions (median [IQR] ΔDMSBP: 13 [4, 21] mm Hg at 3 months; 8 [3, 16] mm Hg at 6 months, 7 [0, 16] mm Hg at 12 months; all paired P < 0.05) and sustained declines in CIAHM (2 [0, 5] at 3, 6, and 12 months; all paired P < 0.05). Significant 12-month estimated glomerular filtration rate improvement after stenting was only observed when FFR <0.80, with greater percentage eGFR improvement under quantile regression analysis. CONCLUSIONS:At the 1-year follow-up, FFR-guided renal artery stenting reduced stent use without loss of blood pressure control or renal safety compared with angiography guidance.
Background:Coronary artery disease (CAD) remains a fundamental etiology of morbidity and mortality among patients receiving dialysis, a cohort known to have a particularly poor prognosis. While traditional CAD risk factors, such as hyperlipidemia, hypertension, and diabetes, remain noteworthy, these factors do not fully capture the complexity of the disease in dialysis patients. Therefore, innovative and robust biomarkers that can refine risk stratification and enhance prognostic precision in this vulnerable population are imperative. Hence, this study aimed to evaluate the use of the red cell distribution width-to-albumin ratio (RAR) to predict clinical outcomes, particularly in dialysis patients with CAD. Methods:We analyzed data from a multicenter cohort of 1128 CAD patients on dialysis who were enrolled between January 2015 and June 2021. Patient stratification into tertiles was performed based on RARs. The primary endpoints were all-cause mortality and cardiovascular mortality. The secondary endpoint was the occurrence of major adverse cardiovascular events (MACEs), comprising non-fatal myocardial infarction, non-fatal stroke, and other cardiovascular events. Results:The median follow-up duration was 20.9 months 378 (33.5%) patients experienced all-cause mortality, 261 (19.1%) cardiovascular mortality, and 485 (43.0%) MACEs. In multivariable Cox proportional hazards regression, patients in the highest RAR tertile demonstrated remarkably escalated risks of all-cause mortality (hazard ratio (HR): 1.592, 95% confidence interval (CI): 1.212-2.092), cardiovascular mortality (HR: 1.503, 95% CI: 1.086-2.080), and MACEs (HR: 1.452, 95% CI: 1.141-1.846) compared with those in the lowest tertile. Moreover, restricted cubic spline analysis revealed a linear, dose-dependent association between elevated RAR and an increased risk of these adverse outcomes. The integration of the RAR into existing risk models, specifically the Global Registry of Acute Coronary Events and Gensini scores, noticeably amplified the predictive performance, as demonstrated by notable enhancements in both net reclassification improvement and integrated discrimination improvement. Conclusions:An elevated RAR serves as an independent predictor of poor outcomes in patients with CAD undergoing dialysis. Clinical Trial Registration:This study was registered at ClinicalTrials.gov (Identifier: NCT05841082; https://clinicaltrials.gov/study/NCT05841082).
Cardiovascular-Kidney-Metabolic (CKM) syndrome is associated with increased risks of cardiovascular mortality and target organ failure. Despite optimal lipid-lowering therapies, patients remain at substantial residual risk for adverse cardio-renal outcomes. Conventional cholesterol-centric models do not fully account for this inter-organ pathology, limiting precision risk assessment and intervention. Emerging evidence suggests that ceramides, key intermediates in sphingolipid metabolism, may function as circulating signaling molecules involved in inter-organ communication. Recent studies have identified several putative ceramide-responsive receptors through which ceramides may contribute to lipotoxic injury in remote organs. They have been associated with pathological processes including insulin resistance, renal microvascular injury, and atherosclerosis, particularly in the context of impaired renal function. Based on a systematic literature search, this review summarizes the cross-organ pathogenic mechanisms of ceramides in CKM syndrome. We evaluate the role of ceramide-based risk scores (e.g., CERT1) in risk stratification and critically discuss the opportunities and challenges associated with emerging therapeutic strategies.
BACKGROUND AND AIMS:The optimal therapy for patients with atherosclerotic renal artery stenosis (ARAS) remains unresolved. This study compared the efficacy of renal fractional flow reserve (FFR)-guided revascularization and traditional angiography-guided revascularization. METHODS:In total, 101 patients with ARAS and hypertension were randomly assigned to either the FFR-guided or angiography-guided group (ClinicalTrials.gov identifier: NCT05732077). Stenting was performed in the angiography-guided group regardless of FFR, whereas stenting was only performed in the FFR-guided group for patients with FFR < 0.80. The primary endpoints were the percentage changes in ambulatory daytime mean systolic blood pressure (DMSBP) and composite index of antihypertensive medicines (CIAHM) after 3 months. RESULTS:The percentage changes in DMSBP (4% [-2%, 11%] vs 4% [-3%, 10%]; P = .97) and CIAHM (0% [0%, 3%] vs 1% [0%, 4%]; P = .33) did not differ between groups. However, the rate of stenting was significantly lower in the FFR-guided group (46.0% vs 100.0%, P < .01). Moreover, compared with the findings in patients with FFR ≥ 0.80 who did not receive stenting, stenting was beneficial in patients with FFR < 0.80 (adjusted mean DMSBP reduction, 6.2 [95% confidence interval {CI}, 0.6-11.9] mmHg; mean CIAHM reduction, 3.1 [95% CI, 1.5-4.7]), but not in those with FFR ≥ 0.80 (1.4 [95% CI, -4.5-7.2] mmHg, and 0.7 [95% CI, -1.1-2.5], respectively). CONCLUSIONS:FFR-guided revascularization significantly reduced unnecessary stenting compared with angiography-guided revascularization. Both blood pressure and antihypertensive medication usage decreased significantly after stenting in patients with FFR < 0.80.
We developed a NIR-II fluorescent nanoprobe, DCy-BOH@BSA-HA, for targeted identification of atherosclerotic (AS) plaques. The probe remains quiescent initially but is dually activated by the plaque microenvironment-specifically elevated ROS and viscosity-while (hyaluronic acid) HA mediates CD44 targeting. This strategy enabled precise NIR-II fluorescence imaging of aortic plaques in a murine AS model.
The coronary artery calcium (CAC) score is a crucial diagnostic indicator, yet its prediction from coronary angiography remains inaccurate. As the primary technique for guiding intervention, angiography analyzes dynamic, multi-view X-ray videos of blood flow, which poses significant processing challenges and leads to persistently low prediction accuracy. In this study, we propose a multimodal Coronary Calcification Vision-Language Model (CoCaViLM) fusing angiography video and diagnostic texts. The token compression module condenses the information of an entire video into a single token, thereby achieving a substantial reduction in token count while preserving the integrity of the original video content. To introduce medical prior knowledge and enhance learning, we insert the video embeddings into designated placeholder positions within the text prompt, forming an interleaved vision–language representation. We evaluated it on a dataset consisting of X-ray angiography sequences paired with diagnostic texts, the proposed approach achieved a correlation of 0.5368, outperforming the current best video–text models.
INTRODUCTION:This study aims to investigate the impact of different double-stent methods on the structure and mechanics of coronary bifurcation lesions, providing reference indicators for clinicians in selecting an appropriate interventional procedure. METHODS:Three-dimensional reconstruction of coronary Computed Tomography Angiography (CTA) image data of a patient with coronary bifurcation disease was performed. Two types of double-stent (Cullotte and Crush) procedures were simulated, and their effects were evaluated using Finite element analysis. Intravascular Ultrasound (IVUS) validation and retrospective clinical analysis were performed to support computational findings. RESULTS:The stress distribution following the Cullotte stent was concentrated in the SB, whereas the stress after the Crush procedure was localized at the overlap with the proximal main vessel three-layer stent. Compared with the Crush procedure, the Culotte approach resulted in a lower percentage of double-stent malapposition, better dilation of vascular stenosis, and less narrowing of the SB stent, suggesting a more favorable clinical outcome. IVUS validation and retrospective clinical analysis were performed to support computational findings. DISCUSSION:Culotte stenting resulted in better stent-vessel conformity and more favorable stress distribution. The findings support FEA as a valuable tool in procedural planning. CONCLUSION:The findings suggest that the Culotte technique may offer mechanical advantages over the Crush technique, potentially improving long-term clinical outcomes. These results emphasize the role of computational modeling in optimizing interventional strategies.
INTRODUCTION:The objective of this research was to explore the possible link between markers of liver fibrosis and survival rates in a group of adults who have been diagnosed with both chronic kidney disease (CKD) and coronary artery disease (CAD). METHODS:The National Health and Nutrition Examination Survey (NHANES) data (1999-2018) for participants with both CAD and CKD were analyzed. The fibrosis-4 index (FIB-4), Nonalcoholic Fatty Liver Score (NFS), Forns index, and aspartate aminotransferase/alanine aminotransferase (AST/ALT) ratio were identified as crucial biomarkers. All-cause and cardiovascular disease (CVD) mortality were primary outcomes, assessed using Cox models, Kaplan-Meier curves, and receiver operating characteristic (ROC) analysis. RESULTS:A total of 1,192 CKD and CAD patients were included. The Cox regression analysis revealed substantial correlations between elevated FIB-4, NFS, Forns index, and AST/ALT levels and a heightened risk of all-cause (hazard ratio [HR]: 1.188, 95% confidence interval [CI]: 1.108-1.274; HR: 1.145, 95% CI: 1.069-1.227; HR: 1.142, 95% CI: 1.081-1.201; HR: 1.316, 95% CI: 1.056-1.639, respectively) and CVD mortality (HR: 1.133, 95% CI: 1.007-1.275; HR: 1.155, 95% CI: 1.024-1.303; HR: 1.208, 95% CI: 1.109-1.316 and HR: 1.636, 95% CI: 1.203-2.224, respectively). The ROC analysis indicated comparable predictive accuracy for all three biomarkers, with AST/ALT showing slightly superior performance. CONCLUSION:Liver fibrosis markers, including AST/ALT, NFS, Forns index and FIB-4, are significant mortality predictors in CAD-CKD patients. The AST/ALT ratio, being easily measurable, may serve as an effective predictive tool for risk stratification in this population.
OBJECTIVES:Intramural hematoma (IMH) is a serious aortic condition characterized by the presence of a contained hematoma within the aortic media. However, the animal model with a high incidence of IMH was lacking, and the specific pathological characteristics of IMH have not been thoroughly characterized. METHODS AND RESULTS:We conducted an experimental study using 4-week-old male, 4-week-old female, and 8-week-old male C57BL/6J mice. These mice were subjected to angiotensin II infusion at a rate of 1000 ng/kg/min for a period of 4 days. In situ imaging was performed, and aorta was harvested and serially sectioned. Histological staining and immunostaining techniques were employed, and the subcellular structure was examined using transmission electron microscopy. Our findings revealed that 4-week-old male mice exhibited a higher susceptibility to angiotensin II-induced IMH, characterized by more circumferential appearances and larger affected areas. Furthermore, IMH was more likely to occur in the upper segment of the descending aorta rather than the lower segment. Within the IMH, older fibrinous thrombus was predominantly observed near the adventitia, while younger red thrombus was more prevalent near the lumen. Additionally, platelet activation and degranulation were observed, along with fibrin cross-linking and thrombus organization, indicating a potential relationship between platelet activation and the progression of IMH. CONCLUSION:Our study demonstrated that angiotensin II infusion promoted the development of IMH during the early stages, especially in juvenile mice. Furthermore, the presence of platelet activation and thrombus organization suggested their potential involvement in the progression of IMH.
Insulin resistance, represented by increased triglyceride-glucose (TyG) index levels, shows interplay with visceral obesity and together promotes cardiovascular diseases and mortality. However, significant controversies exist regarding whether modified TyG indices, such as TyG-BMI, TyG-WC, and TyG-WHtR, outperform the TyG index in predicting cardiovascular outcomes. We aimed to explore whether there was a synergistic effect of a body shape index (ABSI), a better parameter reflecting visceral obesity, and the TyG index on cardiovascular mortality. We analyzed data from the National Health and Nutrition Examination Survey (NHANES) 2001–2018 of 17,329 individuals. The associations of the TyG index and ABSI with cardiovascular mortality were investigated via Cox regression analysis and restricted cubic splines. Receiver operating characteristic (ROC) curve analysis was performed to compare the predictive value. Mediation analysis was used to explore the potential mediator. A total of 673 (3.9