Abstract
Background
The connection between low-density lipoprotein cholesterol (LDL-C) levels and postoperative outcomes following cardiac valve surgery (CVS) remains controversial. Therefore, this research aimed to study the connection of LDL-C levels with the incidence of adverse clinical outcomes in patients undergoing CVS.
Methods
1,304 patients undergoing first-time CVS were identified from the MIMIC-IV database. Participants were split into two groups based on admission LDL-C levels: a low LDL-C group (< 1.4 mmol/L, n = 215) as well as a high LDL-C one (≥ 1.4 mmol/L, n = 1,089). The main endpoint was 30-day all-cause mortality (ACM), with secondary endpoints including in-hospital, 90-, 180- and 365-day ACM rates. The connection between preoperative LDL-C levels and clinical outcomes was evaluated via survival analysis, mediation analysis, and subgroup analyses.
Results
Multivariable Cox regression analysis demonstrated preoperative LDL-C level served as an independent protective factor against 30-day ACM following cardiac valve surgery (adjusted HR: 0.594; 95% CI: 0.368–0.960). In stratified analysis, patients in the higher LDL-C tertile exhibited a remarkably lower mortality risk (HR: 0.285; 95% CI: 0.134–0.604). Consistent subgroup analyses validated these findings’ robustness across all clinically relevant subgroups. Notably, mediation analysis provided mechanistic evidence that the mortality-increasing effect of lower LDL-C levels may be partially mediated through activation of systemic inflammation.
Conclusion
Lower preoperative LDL-C levels are independently related with higher 30-day mortality following cardiac valve surgery, potentially mediated by the activation of inflammatory pathways.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12944-025-02855-5.
Keywords: Low-density lipoprotein cholesterol, All-cause mortality, Cardiovascular disease, Cardiac valve surgery
Introduction
Valvular heart disease (VHD) represents a clinically significant manifestation of cardiovascular pathology with a substantial global disease burden. Epidemiological studies demonstrate marked geographic variation in prevalence, ranging from 2.5% in the general US adult population [1] to 51% among elderly cohorts in the United Kingdom [2]. The disease typically advances from asymptomatic valvular dysfunction to severe heart failure, significantly affecting patients’ functional status and quality of life.
While medical management may provide symptomatic relief and potentially retard disease progression in early stages [3], valvular surgery remains the cornerstone of definitive therapy for advanced VHD, particularly in cases of hemodynamic compromise. Contemporary data from China reveal an estimated 25 million patients requiring surgical intervention for VHD, with 77,077 valve procedures performed in 2021 alone - constituting 27.7% of all cardiac surgeries nationally [4]. Despite technical advancements, postoperative complications remain prevalent, which points to the importance of improved risk stratification as well as perioperative management in this high-risk surgical population [4].
Substantial evidence has demonstrated LDL-C acts as a causal risk factor for ASCVD [5, 6]. Building on current evidence, international guidelines suggest strict control of plasma LDL-C in ASCVD patients, with high-risk patients advised to achieve LDL-C levels < 1.4 mmol/L [7, 8]. This consensus has established LDL-C as a crucial therapeutic target in CVD prevention and management. Statin-based lipid-lowering therapy currently represents the cornerstone treatment strategy for primary and secondary ASCVD prevention [9]. While LDL-C’s role in VHD is still incompletely characterized, emerging evidence suggests that atherosclerotic processes affect not only coronary arteries but also contribute to valvular structural and functional abnormalities [10, 11]. Notably, aortic valve disease—the most prevalent form of VHD—demonstrates significant correlations between fibrocalcific progression and circulating LDL-C levels [12]. Mechanistically, oxidized LDL (ox-LDL) deposition in valvular tissue activates inflammatory cascades that exacerbate valvular damage [13]. These pathophysiological insights highlight the clinical imperative for optimized perioperative management, particularly through developing early risk stratification systems. Investigating preoperative LDL-C levels may therefore yield novel prognostic markers and therapeutic targets for surgical VHD patients.
The “lower is better” paradigm of LDL-C management is widely accepted in cardiovascular medicine [14], but there is still debate about the best therapeutic targets [15]. This debate stems from two conflicting lines of evidence: on one hand, elevated LDL-C levels accelerate atherosclerosis progression and increase cardiovascular event risk [11]. However, new research indicates that very low LDL-C (< 1.0 mmol/L) may be linked to higher ACM, suggesting a possible U-curve connection between LDL-C and mortality [16].
In order to ascertain the ideal therapeutic range for LDL-C, this study examines the connection of LDL-C levels with unfavorable outcomes in patients following valvular heart surgery. Additionally, this paper sought to explore the mediating factors underlying this relationship.
Materials and methods
Study design and population
The research is retrospective cohort research. The MIMIC-IV database, version 2.2, provided all of the data used in this investigation. A single-center, publicly available collection of deidentified, structured health data from critically ill patients hospitalized to the BIDMC in Boston, Massachusetts, from 2008 to 2019 is called the MIMIC-IV database. It contains thorough clinical data from 50,920 patients hospitalized to BIDMC’s ICU throughout the designated time frame. Informed permission was not required since the patient data was deidentified.
Based on the 9th (ICD-9) as well as 10th (ICD-10) versions, 6,020 patients who had heart valve surgery during their initial hospital stay were chosen from the MIMIC-IV database. 1,304 individuals were encompassed in the research after patients with incomplete covariate data (n = 4,716) and those who had previously had cardiac valve surgery were eliminated (Figure S1). Structured Query Language on the Navicat (version 15.0.12) platform was used to retrieve data from these patients’ first day of hospitalization from MIMIC IV 2.2.
Data collection
The extracted data encompassed demographic information, basic vital signs, comorbidities, medication history, basic laboratory parameters, as well as pre-treatment scoring systems. Demographic data included age, gender, as well as ethnicity. Comorbidities encompassed CHD, CHF, arrhythmias, hypertension, diabetes, liver diseases, smoking, and alcohol abuse. The ICD-9/ICD-10 diagnostic codes covered a variety of hepatic conditions and complications, including liver diseases, but not liver cancer. The history of medication was the use of ACEI/ARB and statins. The laboratory parameters that were measured within 24 h of admission were LDL-C, WBC, blood glucose, PLT, TG, HDL-C, TBIL and mechanical ventilation time. Also, the severity of the disease in surgical ICU patients was measured using the SAPS II and OASIS.
Outcomes
The 30-day ACM was the main outcome of this study. The secondary endpoints comprised in-hospital, 90-, 180- and 365-day ACM. The time frame of observation was the time of admission of the patient in the hospital until death. The data on dates of death were retrieved from the MIMIC-IV 2.2 database.
Calculation of inflammatory biomarkers
Regarding indicator selection, the data quality assessment revealed substantial missing rates (> 40%) for traditional inflammatory biomarkers (encompassing C-reactive protein, procalcitonin, as well as interleukin-6) in the MIMIC-IV database, rendering them unsuitable for large-scale analyses. To ensure methodological rigor and reproducibility, the data strictly adhered to MIMIC-IV utilization guidelines and prioritized inflammatory indicators meeting the following criteria: (1) validated associations with infectious outcomes in a peer-reviewed MIMIC-IV-based study [17]; (2) Existing clinical evidence supports its clinical efficacy in compliance with evidence-based medicine standards [18]. Finally, several inflammatory markers were included for mediation analysis: SIRI, SII, PLR, MLR and NLR. The specific calculation methods are as follows: SIRI = neutrophil count×monocyte count/lymphocyte count [19]; SII = platelet count×neutrophil count/lymphocyte count [20]; PLR = platelet count/lymphocyte count [21]; MLR = monocyte count/lymphocyte count [22]; NLR = neutrophil count/lymphocyte count [23].
Statistical analysis
Continuous variables are presented as mean ± SD (normally distributed) or median (interquartile range) (non-normally distributed). Frequencies (percentages) were leveraged to express categorical variables. The relevant statistical methods were used for between-group comparisons: Pearson’s chi-square test or Fisher’s exact test was leveraged to evaluate categorical variables, while one-way ANOVA or the Kruskal-Wallis H-test were leveraged to examine continuous variables.
Participants were split into two groups according to ASCVD-derived lipid management targets (< 1.4 mmol/L and ≥ 1.4 mmol/L). Univariate and multivariate Cox proportional hazards models were leveraged to explore the connection of LDL-C with the outcomes, including ACM at in-hospital, 30, 90, 180 and 365-day intervals. Model 1 was age as well as sex adjusted. Model 2 was also adjusted to TG, use of medications (statin), diabetes, hepatic impairment, smoking, and BMI. The results were displayed in the form of HRs with 95% CI. Besides, LDL-C was estimated as continuous in order to enhance the strength of the results. KM survival curves were built to compare the endpoint event rates, and statistical significance was measured using log-rank tests.
Mediation analysis was employed causal to examine the inflammatory biomarker-mediated pathways linking preoperative LDL-C to 30-day ACM. In this framework, inflammatory biomarkers were formally tested as potential mediators using a counterfactual approach, with LDL-C as the exposure variable and 30-day mortality as the primary outcome (Figure S2). Covariate adjustments included age, sex, TG, medication use (statin), diabetes, hepatic impairment, Smoking, BMI. Total, direct, and indirect effects were quantified through a parametric regression-based counterfactual framework. The direct effect quantifies LDL-C’s mortality risk independent of inflammatory pathways, while the indirect effect reflects mortality risk attributable to LDL-C-induced inflammatory dysregulation. Mediation proportion was computed as [indirect effect/(direct effect + indirect effect)]×100% (reported only when indirect effects reached statistical significance).
Furthermore, subgroup analyses based on different factors like age, sex, CHD, hypertension, diabetes mellitus, hepatic impairment and medication use (statin) were performed to gain deeper insights into the connection between LDL-C and ACM at 30-day. R software (3.6.1) as well as SPSS software (22.0) were leveraged to perform all statistical analyses. The statistical significance was taken as two-tailed P < 0.05.
Results
Baseline characteristics
6,020 patients undergoing cardiac valve surgery were initially screened. After excluding 4,716 patients who underwent reoperation or lacked preoperative lipid profile data, 1,304 patients were encompassed in the final statistical analysis. Of the total cohort, 40 (3.1%) patients succumbed within the first 30 days of follow-up, while 1264 (96.9%) demonstrated 30-day survival.
Table 1 indicated that the low LDL-C group were older (P < 0.05), with more diabetes, less albumin, hemoglobin, and HDL-C and higher SAPS II scores (all P < 0.01). Moreover, they were more likely to use statins and were shown to have more in-hospital mortality (IHM) and 30, 90, 180 and 365-day mortality.
Table 1.
Baseline characteristics of cardiac valve surgery patients stratified by LDL-C levels
| Variables | low LDL-C group (< 1.4 mmol/L), n = 215 | high LDL-C group (≥ 1.4 mmol/L), n = 1089 | P value |
|---|---|---|---|
| LDL-C, mmol/L | 1.09 ± 0.26 | 2.56 ± 0.86 | < 0.01 |
| Age, y | 72.20 ± 11.57 | 70.00 ± 12.99 | 0.02 |
| Female, % | 77 (35.81) | 439 (40.31) | 0.22 |
| BMI, kg/m2 | 28.85 ± 6.58 | 28.42 ± 5.99 | 0.39 |
| Past medical history | |||
| Hypertension, % | 155 (72.09) | 817 (75.02) | 0.37 |
| Diabetes, % | 74 (34.42) | 260 (23.88) | < 0.01 |
| CHD, % | 90 (41.86) | 403 (37.01) | 0.18 |
| Hepatic Impairment, % | 30 (13.95) | 132 (12.12) | 0.46 |
| Smoking, % | 14 (6.51) | 53 (4.87) | 0.32 |
| Alcoholism, % | 14 (6.51) | 58 (5.33) | 0.49 |
| laboratory indicators | |||
| Glucose, mmol/L | 7.53 ± 2.94 | 7.02 ± 2.02 | 0.02 |
| Albumin, g/L | 3.69 ± 0.63 | 3.90 ± 0.60 | < 0.01 |
| Hemoglobin, g/dL | 9.81 ± 1.54 | 10.12 ± 1.55 | < 0.01 |
| Platelets, 10^9/L | 182.33 ± 84.76 | 175.28 ± 87.84 | 0.28 |
| WBC, 10^9/L | 12.47 ± 9.84 | 12.74 ± 6.35 | 0.60 |
| Total bilirubin, mg/dL | 0.98 ± 0.91 | 0.98 ± 0.96 | 0.97 |
| TG, mmol/L | 1.37 ± 0.82 | 1.51 ± 1.11 | 0.09 |
| HDL-C, mmol/L | 2.02 ± 0.59 | 3.01 ± 0.73 | < 0.01 |
| Mechanical ventilation time, d | 2.81 ± 3.43 | 2.43 ± 3.08 | 0.11 |
| Vital signs | |||
| Mbp, mmHg | 73.17 ± 7.03 | 74.22 ± 7.09 | 0.05 |
| Temperature, ℃ | 36.74 ± 0.47 | 36.70 ± 0.46 | 0.37 |
| Los in Hospital, (d) | 12.55 ± 9.27 | 11.21 ± 9.88 | 0.07 |
| Los in Icu, (d) | 4.26 ± 5.11 | 3.69 ± 4.95 | 0.13 |
| SAPS II | 39.99 ± 13.17 | 36.75 ± 12.06 | < 0.01 |
| OASIS | 31.23 ± 8.65 | 29.87 ± 7.89 | 0.02 |
| Drug therapy | |||
| ACEI/ARB,% | 133 (61.86) | 610 (56.01) | 0.11 |
| Statins, % | 201 (93.49) | 879 (80.72) | < 0.01 |
| Clinical outcomes | |||
| In-hospital mortality, (%) | 17 (7.91) | 28 (2.57) | < 0.01 |
| 30-day mortality, (%) | 16 (7.44) | 24 (2.20) | < 0.01 |
| 90-day mortality, (%) | 23 (10.70) | 42 (3.86) | < 0.01 |
| 180-day mortality, (%) | 28 (13.02) | 71 (6.52) | < 0.01 |
| 365-day mortality, (%) | 37 (17.21) | 101 (9.27) | < 0.01 |
LDL-C Low-density lipoprotein cholesterol, ACEI Angiotensin-converting enzyme inhibitors, ARB Angiotensin receptor blocker, BMI Body mass index, MBP Mean blood pressure, CHD Coronary heart disease, WBC White blood cell count, TG Triglyceride, HDL-C High-density lipoprotein cholesterol, SAPS II Simplified Acute Physiology Score II, OASIS Oxford Acute Severity of Illness Score
LDL-C and clinical outcomes
Contemporary clinical guidelines recommend maintaining LDL-C below 1.4 mmol/L for high-risk populations [7, 8]. But no consensus exists regarding optimal LDL-C targets in VHD research. However, no consensus exists regarding optimal LDL-C targets in VHD management. Emerging basic science investigations have revealed shared pathophysiological mechanisms between valvular calcification and atherosclerosis, where lipid deposition drives pathological calcification through activation of osteogenic differentiation signaling pathways in valvular interstitial cells [10, 11]. Based on these considerations, this study divided the research population into two groups: LDL-C < 1.4 mmol/L (n = 215) as well as ≥ 1.4 mmol/L (n = 1,089).
Table 2 demonstrates patients in the higher LDL-C group exhibited remarkably lower 30-day ACM (adjusted HR 0.285, 95% CI 0.134–0.604; P = 0.001) in the fully adjusted model. This protective connection was still consistent across all follow-up periods: IHM (HR 0.320, 95% CI 0.158–0.649; P = 0.002), 90-day mortality (HR 0.321, 95% CI 0.176–0.585; P < 0.001), 180-day mortality (HR 0.492, 95% CI 0.297–0.815; P = 0.006), and 365-day mortality (HR 0.475, 95% CI 0.308–0.733; P < 0.001). These Cox regression results were further supported by KM survival analysis (log-rank P < 0.001; Fig. 1), demonstrating robust agreement between both analytical methods.
Table 2.
Association between LDL-C groups and postoperative mortality
| LDL-C Level, mmol/L | Unadjusted | Model 1 | Model 2 | |||
|---|---|---|---|---|---|---|
| HR (95% CIs) |
P value | HR (95% CIs) |
P value | HR (95% CIs) |
P value | |
| In-hospital mortality | ||||||
| < 1.4 mmol/L | 1.000 | 1.000 | 1.000 | |||
| ≥ 1.4 mmol/L | 0.364 (0.197,0.671) | 0.001 | 0.345 (0.186,0.640) | < 0.001 | 0.320 (0.158,0.649) | 0.002 |
| 30-Day all-cause mortality | ||||||
| < 1.4 mmol/L | 1.000 | 1.000 | 1.000 | |||
| ≥ 1.4 mmol/L | 0.288 (0.153,0.542) | < 0.001 | 0.299 (0.158,0.563) | < 0.001 | 0.285 (0.134,0.604) | 0.001 |
| 90-Day all-cause mortality | ||||||
| < 1.4 mmol/L | 1.000 | 1.000 | 1.000 | |||
| ≥ 1.4 mmol/L | 0.346 (0.208,0.575) | < 0.001 | 0.371 (0.223,0.618) | < 0.001 | 0.321 (0.176,0.585) | < 0.001 |
| 180-Day all-cause mortality | ||||||
| < 1.4 mmol/L | 1.000 | 1.000 | 1.000 | |||
| ≥ 1.4 mmol/L | 0.475 (0.306,0.735) | < 0.001 | 0.506 (0.327,0.785) | 0.002 | 0.492 (0.297,0.815) | 0.006 |
| 365-Day all-cause mortality | ||||||
| < 1.4 mmol/L | 1.000 | 1.000 | 1.000 | |||
| ≥ 1.4 mmol/L | 0.507 (0.348,0.739) | < 0.001 | 0.540 (0.370,0.787) | 0.001 | 0.475 (0.308,0.733) | < 0.001 |
Cox proportional hazards regression analysis of LDL-C as a grouping variable for all-cause mortality at in-hospital, 30-day, 90-day, 180-day and 365-day follow-up periods
Model 1 was adjusted for age, gender
Model 2 was further adjusted for TG, medications (statin), diabetes, hepatic impairment, Smoking, BMI
LDL-C Low-Density Lipoprotein Cholesterol, TG Triglyceride, BMI Body Mass Index
Fig. 1.
Kaplan–Meier survival analysis curves for all-cause mortality. Kaplan–Meier curves of 30-day (A), 90-day (B), 180-day (C) and 365-day (D) all-cause mortality stratified by LDL-C level. LDL-C, Low-density lipoprotein cholesterol
While analyses of LDL-C as a continuous variable revealed no remarkable associations with 90-day (P = 0.060) or 180-day (P = 0.166) mortality, significant protective relationships persisted for 30-day ACM (adjusted HR 0.594, 95% CI 0.368–0.960; P = 0.033), IHM (adjusted HR 0.632, 95% CI 0.410–0.973; P = 0.037) and 365-day mortality (adjusted HR 0.785, 95% CI 0.619–0.995; P = 0.045) in the fully adjusted model (Table 3).
Table 3.
Association between continuous LDL-C and postoperative mortality
| Unadjusted | Model 1 | Model 2 | ||||
|---|---|---|---|---|---|---|
| HR (95% CIs) |
P value | HR (95% CIs) |
P value | HR (95% CIs) |
P value | |
| In-hospital mortality | 0.699 (0.503,0.973) | 0.034 | 0.701 (0.496,0.989) | 0.043 | 0.632 (0.410,0.973) | 0.037 |
| 30-Day all-cause mortality | 0.617 (0.421,0.906) | 0.014 | 0.639 (0.431,0.948) | 0.026 | 0.594 (0.368,0.960) | 0.033 |
| 90-Day all-cause mortality | 0.664 (0.494,0.891) | 0.006 | 0.724 (0.534,0.983) | 0.038 | 0.703 (0.486,1.016) | 0.060 |
| 180-Day all-cause mortality | 0.768 (0.612,0.964) | 0.023 | 0.836 (0.662,1.057) | 0.135 | 0.824 (0.627,1.083) | 0.166 |
| 365-Day all-cause mortality | 0.755 (0.622,0.916) | 0.004 | 0.812 (0.666,0.991) | 0.041 | 0.785 (0.619,0.995) | 0.045 |
Cox proportional hazards regression analysis of LDL-C as a continuous variable for all-cause mortality at in-hospital, 30-day, 90-day, 180-day and 365-day follow-up periods
Model 1 was adjusted for age, gender
Model 2 was further adjusted for TG, medications (statin), diabetes, hepatic impairment, Smoking, BMI
LDL-C Low-Density Lipoprotein Cholesterol, TG Triglyceride, BMI Body Mass Index
Mediation analysis
A previous epidemiological study has identified a remarkable inverse connection of LDL-C levels with infection-related hospitalization risk [24]. Notably, another multiple large-scale cohort study has demonstrated a potential J/U-shaped connection of LDL-C concentrations with infection susceptibility, suggesting excessively low LDL-C may impair innate immune defenses [18]. However, direct mechanistic evidence supporting this hypothesis remains scarce. To close the critical knowledge gap, this study did formal mediation analyses to evaluate the proposed theoretical model in which postoperative systemic inflammatory response mediates the connection of preoperative LDL-C levels with postoperative infection risk.
Mediation analysis revealed statistically significant mediating effects of NLR (95% CI [-0.008, -0.002]), MLR (95% CI [-0.004, -0.002]), PLR (95% CI [-0.003, -0.001]), SII (95% CI [-0.006, -0.002]) as well as SIRI (95% CI [-0.006, -0.002]) in the LDL-C to 30-day mortality pathway. These mediation effects remained statistically significant after full covariate adjustment (Table 4), indicating that lower LDL-C levels may exert detrimental effects on 30-day ACM by systemic inflammation.
Table 4.
Mediation analysis between LDL-C and 30-day all-cause mortality
| Dependent variable | Mediation variable | β (Bootstrap 95% CI) | Mediation proportion | |
|---|---|---|---|---|
| Crude | Adjusted | |||
| 30-day all-cause mortality | NLR | |||
| Total effect | -0.012(-0.007, -0.017) | -0.010(-0.004, -0.016) | ||
| Direct effect | -0.007(-0.002, -0.012) | -0.006(-0.001, -0.011) | ||
| Indirect effect | -0.005(-0.002, -0.008) | -0.004(-0.002, -0.008) | 39.6% | |
| MLR | ||||
| Total effect | -0.018(-0.011, -0.025) | -0.018(-0.011, -0.025) | ||
| Direct effect | -0.014(-0.008, -0.020) | -0.015(-0.008, -0.022) | ||
| Indirect effect | -0.004(-0.003, -0.005) | -0.003(-0.002, -0.004) | 17.7% | |
| PLR | ||||
| Total effect | -0.018(-0.011, -0.025) | -0.017(-0.010, -0.024) | ||
| Direct effect | -0.016(-0.009, -0.023) | -0.015(-0.009, -0.021) | ||
| Indirect effect | -0.002(-0.001, -0.003) | -0.002(-0.001, -0.003) | 10.3% | |
| SII | ||||
| Total effect | -0.012(-0.007, -0.017) | -0.010(-0.004, -0.016) | ||
| Direct effect | -0.007(-0.002, -0.012) | -0.006(-0.001, -0.011) | ||
| Indirect effect | -0.005(-0.002, -0.008) | -0.004(-0.002, -0.006) | 42.5% | |
| SIRI | ||||
| Total effect | -0.012(-0.007, -0.017) | -0.010(-0.004, -0.016) | ||
| Direct effect | -0.008(-0.003, -0.013) | -0.006(-0.001, -0.011) | ||
| Indirect effect | -0.004(-0.002, -0.006) | -0.004(-0.002, -0.006) | 38.3% | |
Adjusted for age, gender, TG, medications (statin), diabetes, hepatic impairment, Smoking, BMI
NLR Neutrophil-to-lymphocyte ratio, MLR Monocyte-to-lymphocyte ratio, PLR Platelet-to-lymphocyte ratio, SII Systemic immune-inflammation index, SIRI Systemic inflammatory response index, TG Triglyceride, BMI Body Mass Index
Analysis of nonlinear relationships
RCS analysis was applied to further examine the connection of continuous LDL-C levels with 30-day ACM. Based on the RCS curve analysis, the result revealed no ‘J’- or ‘U’-shaped connection of LDL-C with 30-day ACM (P = 0.147) (Fig. 2).
Fig. 2.
Restricted cubic splines (RCS) Analysis of the Associations Between 30-day All-Cause Mortality and the LDL-C Level. Adjusted for age, gender, TG, medications (statin), diabetes, hepatic impairment, Smoking, BMI. LDL-C, Low-Density Lipoprotein Cholesterol; TG, triglyceride; BMI, Body Mass Index
Subgroup analysis
Subgroup analyses stratified by sex, age, hypertension, diabetes, CHD, hepatic impairment, and statin use revealed consistent connection of LDL-C with 30-day ACM across all groups (Fig. 3; all P-interaction > 0.05), supporting result robustness in cardiac valve surgery populations.
Fig. 3.
Subgroup Analysis of the Associations Between 30-day All-Cause Mortality and the LDL-C Level
Discussion
Valve surgery constitutes an essential intervention in cardiac surgical practice. The rising volume of valvular procedures has made the early identification of postoperative risk factors crucial for optimizing clinical outcomes. While LDL-C is an established risk factor for ASCVD [25] historically governed by the “lower is better” paradigm [14, 26], emerging evidence has challenged this conventional perspective [27]. This study demonstrates that VHD patients with preoperative LDL-C < 1.4 mmol/L experienced remarkably higher 30-day postoperative mortality than those with levels ≥ 1.4 mmol/L. The findings emphasize the necessity of disease-specific lipid management in VHD populations, challenging the universal applicability of current ASCVD-based LDL-C targets for surgical patients.
The precise mechanisms underlying the connection of LDL-C levels with postoperative mortality in cardiac valve surgery patients have not been fully explained. Since postoperative inflammation is a critical factor impacting the prognosis of these patients [28], inflammation was identified as a mediating factor between LDL-C and clinical outcomes. From a mechanism perspective, LDL-C serves as the primary cholesterol transport vehicle in human plasma, facilitating the delivery of sterols to peripheral tissues for membrane synthesis and steroidogenesis [29]. Importantly, circulating LDL-C concentrations function as a sensitive biomarker of systemic nutritional status, with hypocholesterolemia frequently reflecting catabolic states and protein-energy malnutrition [30]. When LDL-C levels are abnormally elevated, they deposit within arterial walls, triggering oxidative modification and inflammatory responses that promote the initiation and progression of atherosclerosis [31, 32]. Beyond this established role, emerging evidence reveals that LDL-C possesses significant anti-inflammatory properties through its capacity to bind pathogen-associated molecular patterns (PAMPs), thereby directly neutralizing bacterial endotoxins (e.g., lipopolysaccharide) and viral particles, thereby inhibiting their inflammatory effects [33]. Furthermore, LDL-C enhances macrophage phagocytic capacity and stimulates antimicrobial peptide secretion [34], mediating indirect anti-infective effects. Consequently, the findings suggest that the attenuation of these crucial anti-inflammatory and immunomodulatory functions in low LDL-C patients may substantially contribute to the observed increase in postoperative mortality.
Beyond its potential anti-inflammatory properties, the connection of low LDL-C levels with hypoalbuminemia suggests a remarkable link with poor nutritional status – a well-established predictor of adverse outcomes in cardiac valve surgery [30]. Clinical evidence consistently demonstrates that preoperative malnutrition significantly increases postoperative complication and mortality rates in these patients [35]. Mechanistically, low LDL-C levels reflect malnutrition and are typically accompanied by hypoalbuminemia [30]. Impaired albumin synthesis reduces plasma colloid osmotic pressure, disrupting Starling forces and promoting fluid extravasation into interstitial spaces, resulting in edema and microcirculatory dysfunction [36]. Second, decreased apolipoprotein B/E (ApoB/ApoE) concentrations in LDL particles impair transendothelial transport of growth factors (e.g., VEGF, PDGF), markedly inhibiting granulation tissue neovascularization [37]. Furthermore, malnutrition may potentiate chronic inflammatory states and oxidative stress, potentially impairing myocardial recovery and long-term survival [38]. These synergistic mechanisms may collectively explain the significantly reduced 30-day survival rate observed in patients with low LDL-C.
Furthermore, this research focuses on preoperative metabolic risk factors. In addition to the influence of early postoperative inflammation on prognosis after cardiac surgery, prior research has confirmed that certain early-postoperative biomarkers, like troponin T, are strong predictors of outcomes following cardiac surgery [39]. Although the MIMIC-IV database employed in the research lacks systematic troponin T measurements, precluding an analysis of the relationship between acute myocardial injury and preoperative metabolic status, the findings demonstrate that preoperative LDL level is a robust predictor independent of conventional risk factors. This suggests that long-term preoperative metabolic disturbances and postoperative acute events may influence patient recovery via distinct pathophysiological pathways. Future prospective should incorporate both preoperative LDL and postoperative biomarkers like troponin T to better clarify their individual and synergistic prognostic value.
However, in patients undergoing cardiac valve surgery, a low preoperative LDL-C level is unlikely to be a direct contributor to mortality. Instead, this research posit that it is more likely a biomarker reflecting greater baseline disease severity, a malnutrition-inflammation state, and compromised host defense. These patients inherently exhibit poor physiological reserve and surgical tolerance, rendering them more susceptible to postoperative complications like infections, impaired wound healing, and subsequent infection-induced multi-organ failure, which collectively contribute to elevated early mortality.
In clinical practice, preoperative LDL-C level can be a simple, cost-effective tool for risk stratification. An unexpectedly low LDL-C value should alert clinicians to a potentially high-risk individual, necessitating a more comprehensive preoperative evaluation (e.g., assessing nutritional status and inflammatory markers) and prompting the development of a more meticulous perioperative management strategy. This may include intensified nutritional support, improved cardiac function, and vigilant monitoring for infection indicators. The findings should not be interpreted to mean that an LDL-C level below 1.4 mmol/L is inherently detrimental, thereby warranting the discontinuation of statin therapy or intentional elevation of cholesterol levels. Conversely, they highlight that for this high-risk subset of patients with low LDL-C, the clinical focus should shift toward improving overall nutritional status and controlling systemic inflammation, rather than solely targeting a further reduction in the numerical cholesterol value. Furthermore, a key objective for future research is to identify the optimal range that balances the risks associated with hypercholesterolemia against the potential hazards of compromised physiological resilience.
Notably, the connection between low LDL-C and 30-day mortality was still consistent regardless of statin use. Consequently, greater caution is warranted when considering lipid-lowering therapy in patients undergoing cardiac valve surgery, irrespective of their statin treatment history. This consideration is particularly critical in patients with hepatic impairment. Despite these complex interactions affecting LDL-C metabolism [40], the data confirm that a measured low LDL-C level retains its prognostic value and should not be ignored. Blindly pursuing further LDL-C reduction in such complex patients could increase the possibility of negative outcomes.
Strengths and limitations
The analysis leveraged the MIMIC-IV database, a publicly available and de-identified critical care repository. Its large scale and detailed clinical data enabled the inclusion of over 15,000 ICU stays, ensuring sufficient statistical power to investigate even uncommon clinical events. To enhance the generalizability of the findings, stratified analyses were performed across multiple subgroups, with adjustments made for potential confounding variables. However, the research has some limitations. To begin with, as a single-center retrospective analysis using MIMIC-IV data, potential selection bias may affect generalizability, requiring validation via multicenter prospective research. Second, some subgroup analyses were underpowered due to small sample sizes. Third, the mechanisms behind reduced LDL-C need further investigation. Finally, the assessment was restricted to admission LDL-C. It did not capture perioperative fluctuations or incorporate key postoperative biomarkers such as troponin T. This feature may have constrained the comprehensive evaluation of postoperative risk in the model. Therefore, future longitudinal research is needed to confirm the findings.
Conclusion
The research indicates lower preoperative LDL-C is significantly related with higher 30-day mortality after cardiac valve surgery, with inflammatory mediators accounting for this association. It provides a novel theoretical foundation for developing individualized treatment strategies for cardiac valve surgery patients. For the patients with low LDL-C, clinical focus should shift towards improving overall nutritional status and controlling systemic inflammation rather than solely targeting further reduction of cholesterol levels.
Supplementary Information
Acknowledgements
The National Science Foundation of China and the National Key Research and Development Program of China funded the research. The authors are grateful to the foundation for its support.
Abbreviations
- ASCVD
Atherosclerotic cardiovascular disease
- BIDMC
Beth Israel Deaconess Medical Center
- ICD-9
9th Revision of the International Classification of Diseases
- ICD-10
10th Revision of the International Classification of Diseases
- KM
Kaplan-Meier
- MIMIC-IV
Medical Information Mart for Intensive Care IV
- ox-LDL
Oxidized low-density lipoprotein cholesterol
- PLT
Platelets
- SD
Standard deviation
- SIRI
Systemic inflammation response index
- TBIL
Total bilirubin
- TG
Triglycerides
- WBC
White blood cell
Authors’ contributions
LML and ZQN contributed to discussion as well as reviewed & edited manuscript. RYH analyzed data and wrote the manuscript. SYL and XLOY revised the manuscript. QYL, QYW, SFL and XYH collected data. QYL and QYW finished the results’ visualization. All writers have read as well as approved the manuscript’s final version. LML takes responsibility for the work’s integrity as a whole.
Funding
The National Natural Science Foundation of China (Grant Number: 82270308) as well as the National Key Research and Development Program of China (Grant Number: 2023ZD0504405-1) provided funding for this study.
Data availability
The corresponding author can deliver the datasets employed and analyzed in this work after reasonable request.
Declarations
Ethics approval and consent to participate
MIMIC-IV (Version 2.2) was the source of the data. Patients’ privacy in MIMIC-IV was safeguarded and identity information was hidden. Therefore, the institutional ethics committee did not require any further consent processes.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Ruiyuan Huang, Siyi Liu and Xiaolan Ouyang contributed equally to this work.
Contributor Information
Zhiqiang Nie, Email: niezhiqiang@gdph.org.cn.
Liming Lei, Email: leiliming@gdph.org.cn.
References
- 1.Nkomo VT, Gardin JM, Skelton TN, Gottdiener JS, Scott CG, Enriquez-Sarano M. Burden of valvular heart diseases: a population-based study. Lancet. 2006;368:1005–11. [DOI] [PubMed] [Google Scholar]
- 2.d’Arcy JL, Coffey S, Loudon MA, Kennedy A, Pearson-Stuttard J, Birks J, Frangou E, Farmer AJ, Mant D, Wilson J, et al. Large-scale community echocardiographic screening reveals a major burden of undiagnosed valvular heart disease in older people: the oxvalve population cohort study. Eur Heart J. 2016;37:3515–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Coisne A, Lancellotti P, Habib G, Garbi M, Dahl JS, Barbanti M, Vannan MA, Vassiliou VS, Dudek D, Chioncel O, et al. ACC/AHA and ESC/EACTS guidelines for the management of valvular heart diseases: JACC guideline comparison. J Am Coll Cardiol. 2023;82:721–34. [DOI] [PubMed] [Google Scholar]
- 4.Yang Y, Wang Z, Chen Z, Wang X, Zhang L, Li S, Zheng C, Kang Y, Jiang L, Zhu Z, Gao R. Current status and etiology of valvular heart disease in china: a population-based survey. BMC Cardiovasc Disord. 2021;21:339. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Arnett DK, Blumenthal RS, Albert MA, Buroker AB, Goldberger ZD, Hahn EJ, Himmelfarb CD, Khera A, Lloyd-Jones D, McEvoy JW, et al. 2019 ACC/AHA guideline on the primary prevention of cardiovascular disease: A report of the American college of Cardiology/American heart association task force on clinical practice guidelines. Circulation. 2019;140:e596–646. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Knuuti J, Wijns W, Saraste A, Capodanno D, Barbato E, Funck-Brentano C, Prescott E, Storey RF, Deaton C, Cuisset T, et al. 2019 ESC guidelines for the diagnosis and management of chronic coronary syndromes. Eur Heart J. 2020;41:407–77. [DOI] [PubMed] [Google Scholar]
- 7.Mach F, Baigent C, Catapano AL, Koskinas KC, Casula M, Badimon L, Chapman MJ, De Backer GG, Delgado V, Ference BA, et al. 2019 ESC/EAS guidelines for the management of dyslipidaemias: lipid modification to reduce cardiovascular risk. Eur Heart J. 2020;41:111–88. [DOI] [PubMed] [Google Scholar]
- 8.National Cholesterol Education Program Expert Panel on Detection E. Treatment of high blood cholesterol in A: third report of the National cholesterol education program (NCEP) expert panel on Detection, Evaluation, and treatment of high blood cholesterol in adults (Adult treatment panel III) final report. Circulation. 2002;106:3143–421. [PubMed] [Google Scholar]
- 9.Cieza A, Causey K, Kamenov K, Hanson SW, Chatterji S, Vos T. Global estimates of the need for rehabilitation based on the global burden of disease study 2019: a systematic analysis for the global burden of disease study 2019. Lancet. 2021;396:2006–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.New SE, Aikawa E. Molecular imaging insights into early inflammatory stages of arterial and aortic valve calcification. Circ Res. 2011;108:1381–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Peeters F, Meex SJR, Dweck MR, Aikawa E, Crijns H, Schurgers LJ, Kietselaer B. Calcific aortic valve stenosis: hard disease in the heart: A biomolecular approach towards diagnosis and treatment. Eur Heart J. 2018;39:2618–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Guo C, Liu X, Mei Z, Chang M, Li J, Wang B, Ji W, Zhang M, Zhang M, Zhang C, An G. AMBP protects against aortic valve calcification by inhibiting ERK1/2 and JNK pathways mediated by FHL3. Theranostics. 2025;15:4398–415. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Kraler S, Garg V, Akhmedov A. Calcific aortic valve disease: novel insights into nitric oxide signalling. Eur Heart J. 2022;43:1665–7. [DOI] [PubMed] [Google Scholar]
- 14.Mortensen MB, Dzaye O, Botker HE, Jensen JM, Maeng M, Bentzon JF, Kanstrup H, Sorensen HT, Leipsic J, Blankstein R, et al. Low-Density lipoprotein cholesterol is predominantly associated with atherosclerotic cardiovascular disease events in patients with evidence of coronary atherosclerosis: the Western Denmark heart registry. Circulation. 2023;147:1053–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Sakuma M, Iimuro S, Shinozaki T, Kimura T, Nakagawa Y, Ozaki Y, Iwata H, Miyauchi K, Daida H, Suwa S, et al. Optimal target of LDL cholesterol level for Statin treatment: challenges to monotonic relationship with cardiovascular events. BMC Med. 2022;20:441. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Rong S, Li B, Chen L, Sun Y, Du Y, Liu B, Robinson JG, Bao W. Association of Low-Density lipoprotein cholesterol levels with more than 20-Year risk of cardiovascular and All-Cause mortality in the general population. J Am Heart Assoc. 2022;11:e023690. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Yang Z, Li Y, Guo T, Yang M, Chen Y, Gao Y. The effect of inflammatory markers on mortality in patients with acute myocardial infarction. Sci Rep. 2025;15:14514. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Chen ZM, Gu HQ, Mo JL, Yang KX, Jiang YY, Yang X, Wang CJ, Xu J, Meng X, Jiang Y, et al. U-shaped association between low-density lipoprotein cholesterol levels and risk of all-cause mortality mediated by post-stroke infection in acute ischemic stroke. Sci Bull (Beijing). 2023;68:1327–35. [DOI] [PubMed] [Google Scholar]
- 19.Valibeygi A, Fardaei M, Niknejad S. Association between stroke and systemic inflammation response index (SIRI): a National health and nutrition examination survey (NHANES) study 2015–2020. BMJ Neurol Open. 2025;7:e000718. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Gao C, Gao S, Zhao R, Shen P, Zhu X, Yang Y, Duan C, Wang Y, Ni H, Zhou L, et al. Association between systemic immune-inflammation index and cardiovascular-kidney-metabolic syndrome. Sci Rep. 2024;14:19151. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Oylumlu M, Yildiz A, Oylumlu M, Yuksel M, Polat N, Bilik MZ, Akyuz A, Aydin M, Acet H, Soydinc S. Platelet-to-lymphocyte ratio is a predictor of in-hospital mortality patients with acute coronary syndrome. Anatol J Cardiol. 2015;15:277–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Wu Y, Xiang HJ, Yuan M. Association of monocyte-lymphocyte ratio and myocardial infarction in the U.S. Population with diabetes. Front Cardiovasc Med. 2024;11:1432838. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Zhou E, Wu J, Zhou X, Yin Y. The neutrophil-lymphocyte ratio predicts all-cause and cardiovascular mortality among U.S. Adults with rheumatoid arthritis: results from NHANES 1999–2020. Front Immunol. 2023;14:1309835. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Trinder M, Walley KR, Boyd JH, Brunham LR. Causal inference for genetically determined levels of High-Density lipoprotein cholesterol and risk of infectious disease. Arterioscler Thromb Vasc Biol. 2020;40:267–78. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Chen L, Chen S, Bai X, Su M, He L, Li G, He G, Yang Y, Zhang X, Cui J, et al. Low-Density lipoprotein Cholesterol, cardiovascular disease Risk, and mortality in China. JAMA Netw Open. 2024;7:e2422558. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Navarese EP, Robinson JG, Kowalewski M, Kolodziejczak M, Andreotti F, Bliden K, Tantry U, Kubica J, Raggi P, Gurbel PA. Association between baseline LDL-C level and total and cardiovascular mortality after LDL-C lowering: A systematic review and Meta-analysis. JAMA. 2018;319:1566–79. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Siniscalchi C, Meschi T, Di Micco P, Imbalzano E, Hernandez-Blasco L, Mahe I, Fernandez-Reyes JL, Garcia-Ortega A, Verhamme P, Alfonso-Megido J, et al. Low-Density lipoprotein cholesterol levels and bleeding risk in venous thromboembolism. JAMA Netw Open. 2025;8:e259467. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Heo RH, Wang MK, Meyre PB, Birchenough L, Park L, Vuong K, Devereaux PJ, Blum S, Lindahl B, Stone G, Conen D. Associations of inflammatory biomarkers with the risk of morbidity and mortality after cardiac surgery: A systematic review and Meta-analysis. Can J Cardiol. 2023;39:1686–94. [DOI] [PubMed] [Google Scholar]
- 29.Browning KL, Lind TK, Maric S, Malekkhaiat-Haffner S, Fredrikson GN, Bengtsson E, Malmsten M, Cardenas M. Human lipoproteins at model cell membranes: effect of lipoprotein class on lipid exchange. Sci Rep. 2017;7:7478. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Lu YW, Lu SF, Chou RH, Wu PS, Ku YC, Kuo CS, Chang CC, Tsai YL, Wu CH, Huang PH. Lipid paradox in patients with acute myocardial infarction: potential impact of malnutrition. Clin Nutr. 2019;38:2311–8. [DOI] [PubMed] [Google Scholar]
- 31.Bentzon JF, Otsuka F, Virmani R, Falk E. Mechanisms of plaque formation and rupture. Circ Res. 2014;114:1852–66. [DOI] [PubMed] [Google Scholar]
- 32.Won KB, Park GM, Yang YJ, Ann SH, Kim YG, Yang DH, Kang JW, Lim TH, Kim HK, Choe J, et al. Independent role of low-density lipoprotein cholesterol in subclinical coronary atherosclerosis in the absence of traditional cardiovascular risk factors. Eur Heart J Cardiovasc Imaging. 2019;20:866–72. [DOI] [PubMed] [Google Scholar]
- 33.Akula MK, Shi M, Jiang Z, Foster CE, Miao D, Li AS, Zhang X, Gavin RM, Forde SD, Germain G, et al. Control of the innate immune response by the mevalonate pathway. Nat Immunol. 2016;17:922–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.York AG, studiiams KJ, Argus JP, Zhou QD, Brar G, Vergnes L, Gray EE, Zhen A, Wu NC, Yamada DH, et al. Limiting cholesterol biosynthetic flux spontaneously engages type I IFN signaling. Cell. 2015;163:1716–29. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Gurbak I, Guner A, Guler A, Sahin AA, Celik O, Uzun F, Onan B, Erturk M. Prognostic influence of objective nutritional indexes on mortality after surgical aortic valve replacement in elderly patients with severe aortic stenosis (from the nutrition-SAVR trial). J Card Surg. 2021;36:1872–81. [DOI] [PubMed] [Google Scholar]
- 36.Gyenge CC, Tenstad O, Wiig H. In vivo determination of steric and electrostatic exclusion of albumin in rat skin and skeletal muscle. J Physiol. 2003;552:907–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Avraham-Davidi I, Ely Y, Pham VN, Castranova D, Grunspan M, Malkinson G, Gibbs-Bar L, Mayseless O, Allmog G, Lo B, et al. ApoB-containing lipoproteins regulate angiogenesis by modulating expression of VEGF receptor 1. Nat Med. 2012;18:967–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Banuls C, de Maranon AM, Veses S, Castro-Vega I, Lopez-Domenech S, Salom-Vendrell C, Orden S, Alvarez A, Rocha M, Victor VM, Hernandez-Mijares A. Malnutrition impairs mitochondrial function and leukocyte activation. Nutr J. 2019;18:89. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Duchnowski P, Smigielski W. Usefulness of myocardial damage biomarkers in predicting cardiogenic shock in patients undergoing heart valve surgery. Kardiol Pol. 2024;82:423–6. [DOI] [PubMed] [Google Scholar]
- 40.Spolitu S, Okamoto H, Dai W, Zadroga JA, Wittchen ES, Gromada J, Ozcan L. Hepatic glucagon signaling regulates PCSK9 and Low-Density lipoprotein cholesterol. Circ Res. 2019;124:38–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The corresponding author can deliver the datasets employed and analyzed in this work after reasonable request.