The Lactate Dehydrogenase-to-Albumin Ratio (LAR) Predicts Mortality in Sepsis-Induced Myocardial Injury

The Lactate Dehydrogenase-to-Albumin Ratio (LAR) Predicts Mortality in Sepsis-Induced Myocardial Injury | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article The Lactate Dehydrogenase-to-Albumin Ratio (LAR) Predicts Mortality in Sepsis-Induced Myocardial Injury Nixiang Jiang, Jinwei Dai, Yang Zhou This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8189180/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Background Sepsis-induced myocardial injury (SIMI), a fatal complication seen in 13%-65% of septic patients with mortality exceeding 50% in severe cases, is characterized by profound metabolic and inflammatory dysregulation.While the Sequential Organ Failure Assessment (SOFA) score is commonly used to assess disease severity in these patients, there remains a need for biomarkers that more directly reflect the metabolic and inflammatory components of SIMI.The lactate dehydrogenase-to-albumin ratio (LAR), which integrates markers of cellular damage and systemic inflammation, has shown promise in sepsis but its prognostic utility specifically in SIMI remains unexplored. Method This retrospective cohort study analyzed data from the MIMIC-IV database involving 4,692 ICU patients with sepsis-induced myocardial injury (SIMI). The primary endpoint was 28-day all-cause mortality. The association was assessed using Cox regression, with restricted cubic splines examining nonlinearity. Subgroup analyses were performed by age, sex, and diabetes. Machine learning models (random forest, XGBoost, decision tree) were developed to validate LAR's predictive value, with performance evaluated by receiver operating characteristic (ROC) curves and decision curve analysis. Results The Random Survival Forest (RSF) model incorporating LAR achieved an AUC of 0.775 (95% CI: 0.752–0.803) for predicting 28-day ICU mortality, outperforming the SOFA score alone (AUC = 0.698, P 0.05), underscoring its robustness as a risk stratifier. Furthermore, decision curve analysis confirmed the clinical utility of the model, demonstrating a superior net benefit across a wide range of risk thresholds. Conclusion The lactate dehydrogenase-to-albumin ratio (LAR) is an independent predictor of 28-day mortality in sepsis-induced myocardial injury. Incorporating LAR into clinical risk stratification could improve early identification of high-risk patients. Health sciences/Biomarkers Health sciences/Cardiology Health sciences/Diseases Health sciences/Medical research Health sciences/Risk factors Lactate Dehydrogenase-to-Albumin Ratio Sepsis Machine Learning Mortality Prediction Boruta Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 1. Introduction Sepsis has long been at the forefront of global health challenges, manifesting as life-threatening organ dysfunction caused by a dysregulated host response to infection 1 .SIMI is a serious, often reversible complication, 2–4 yet it carries a poor prognosis 1 , 5 .With an incidence of 13–65% and mortality exceeding 50% in severe cases, better prognostic tools are urgently needed. Lactate dehydrogenase converts pyruvate to lactate and is a common enzyme in energy metabolism 6 . Studies have observed that the energy metabolism of cardiomyocytes undergoes a significant shift from oxidative phosphorylation to glycolysis in the early stages of sepsis 7 , 8 . Immune cells meet their energy demands by increasing glycolysis, thereby rapidly responding to infection 9 , 10 . However, in the later stages of sepsis, failure to restore oxidative phosphorylation (OXPHOS) in a timely manner may lead to a persistent pro-inflammatory state and myocardial injury 8 .The pathophysiology of SIMI thus creates a clear rationale for biomarkers that integrate metabolic and inflammatory signals.Notably, recent research demonstrates that macrophages can directly link metabolic and immune states by recycling phagocytosed bacteria to fuel immunometabolic responses. Lactate dehydrogenase is involved in this metabolic process 11 . Albumin is an important indicator reflecting systemic inflammation and nutritional status in critically ill patients 12 . Previous studies have found that LAR is an important prognostic indicator in sepsis-associated acute kidney injury (AKI) 13 . Some scholars have also observed the role of metabolic reprogramming in cardiac repair 14 , 15 .Therefore, we hypothesize that LAR, serving as a novel biomarker integrating metabolic stress and systemic inflammation, is an independent predictor of 28-day mortality specifically in patients with SIMI. While LAR has demonstrated prognostic value in other sepsis-related conditions such as acute kidney injury 13 , its association with outcomes in the SIMI population remains uninvestigated. To bridge this knowledge gap, we conducted a large-scale retrospective cohort study utilizing the MIMIC-IV database. We aim to comprehensively evaluate this association using both traditional statistical models and advanced machine learning algorithms, with the goal of validating LAR's utility in improving risk stratification for this high-risk population. 2. Methods 2.1 Data Source The data used in this study were obtained from MIMIC-IV (3.1), a large database containing clinical information for ICU patients at Beth Israel Deaconess Medical Center from 2008 to 2019. After screening 31,910 ICU-diagnosed sepsis patients and applying exclusion criteria, 4,692 patients with sepsis-induced myocardial injury were included in the final analysis. The BIDMC Institutional Review Board approved a waiver of informed consent. The author (JW.D) obtained access to the database (certification number: 62317039). The establishment of this database was approved by the Massachusetts Institute of Technology and Beth Israel Deaconess Medical Center. 2.2 Inclusion and exclusion criteria Inclusion Criteria : Patients aged ≥ 18 years.Meeting the sepsis 3.0 criteria and SIMI criteria 16 . Exclusion Criteria : Patients with an ICU stay of less than 24 h. Missing LDH and ALBUMIN in the first laboratory test. Suffering from heart failure, myocardial infarction and other serious heart diseases. For patients with multiple ICU admissions, only data from the first hospitalization were included. Data specification: For all laboratory tests, the most extreme value (i.e., the maximum for LDH, cTnT, lactate; the minimum for albumin) recorded within the first 24 hours of ICU stay was utilized in the analysis. 2.3 Data collection Data extraction was performed using the pgAdmin software. The collected patient characteristics included the following: basic demographic information such as age, gender, and weight. Information on comorbidities was extracted based on the International Classification of Diseases (ICD) coding system, including hypertension, type 2 diabetes, type 1 diabetes, heart failure, malignancy, chronic kidney disease (CKD), stroke, pneumonia, and septic shock. Vital signs (heart rate, systolic blood pressure (SBP), respiratory rate, arterial oxygen saturation (SaO2)), and laboratory tests [lactate dehydrogenase (LDH), troponin T (cTnT), red blood cell count (RBC), white blood cell count (WBC), platelet count (PLT), serum sodium, serum potassium, serum calcium, anion gap, pH, partial pressure of carbon dioxide (PaCO2), arterial oxygen partial pressure (PaO2), lactate, international normalized ratio (INR) of prothrombin time, total bilirubin, aspartate aminotransferase (AST), blood urea nitrogen, serum creatinine, and serum glucose] were included. Vital signs (heart rate, systolic blood pressure (SBP), respiratory rate, arterial oxygen saturation (SaO2)), and laboratory tests [lactate dehydrogenase (LDH), troponin T (cTnT), red blood cell count (RBC), white blood cell count (WBC), platelet count (PLT), serum sodium, serum potassium, serum calcium, anion gap, pH, partial pressure of carbon dioxide (PaCO2), arterial oxygen partial pressure (PaO2), lactate, international normalized ratio (INR) of prothrombin time, total bilirubin, aspartate aminotransferase (AST), blood urea nitrogen, serum creatinine, and serum glucose] were included. Interventions included mechanical ventilation and renal replacement therapy (RRT). SOFA score was used to evaluate disease severity. LAR was calculated as an index using the following formula: LAR = (admission lactate dehydrogenase (IU/L)) / (albumin (g/L)) 17 . 2.4 Diagnosis of SIMI The first cTnT result after ICU admission was extracted, and the worst value of the day was used for SIMI assessment. The 99th percentile of the upper reference limit for troponin T in this center is 0.01 ng/mL, and SIMI was defined as cTnT > 0.01 ng/mL The first cTnT result after ICU admission was extracted, and the worst value of the day was used for SIMI assessment. The 99th percentile of the upper reference limit for troponin T in this center is 0.01 ng/mL, and SIMI was defined as cTnT > 0.01 ng/mL 16, 18, 19 . 2.5 Outcome The primary outcome was 28-day -ICU mortality.This was defined as the all-cause mortality occurring within a 28-day period starting from the time of ICU admission. Patients who were discharged from the hospital alive before day 28 were considered survivors (i.e., right-censored at their discharge date). The outcome was determined by extracting the dod (date of death) and discharge time stamps from the MIMIC-IV database. The survival time was calculated as the difference between ICU admission time and either the date of death or the date of hospital discharge, whichever occurred first within the 28-day window 20 . 2.6 Statistical analysis Missing data were handled using multiple imputation, while variables with a missing rate greater than 20% were excluded 21 Multicollinearity of variables was assessed using the variance inflation factor (VIF), and variables with a VIF greater than 5 were excluded. Patients were divided into four groups 21 based on the quartiles of LAR levels. Continuous variables with a normal or approximately normal distribution were expressed as mean ± SD and analyzed using analysis of variance (ANOVA). For non-normally distributed variables, Kruskal–Wallis tests were used for analysis. Categorical variables were expressed as numbers and percentages and analyzed using the χ² test or Fisher's exact test. Kaplan–Meier survival curves were used to compare the 28-day survival rates among the four patient groups. We also used the Cox proportional hazards model to evaluate the hazard ratio (HR) and 95% confidence interval (95% CI) for event occurrence. Two models were constructed: Model I without covariate adjustments and Model II adjusted for age, weight, heart rate, respiratory rate, systolic blood pressure, gender, and SOFA score. Time-dependent ROC curves were used to evaluate LAR and other continuous variables. A p-value of < 0.05 was considered statistically significant. All statistical analyses were performed using RStudio (version 4.4.1). 2.6.1 Restricted cubic splines We controlled for covariates (age, weight, heart rate, SBP, respiratory rate, sex, and SOFA scores) and collected data on LAR and outcome variables. The potential non-linear relationship between changes in LAR and survival rates was examined using the rcssci R package, which automatically selects knots for the analysis. 2.6.2 Subgroup analysis Subgroup analyses were performed based on age, sex, and diabetes status. Multivariable analysis was adjusted for age, weight, heart rate, SBP, respiratory rate, sex, and SOFA scores. Cox regression analysis was conducted for each subgroup, and the results were visually presented using a forest plot. 2.7 Establishment and validation of the prediction models The Boruta algorithm was employed to identify the most informative predictors and reduce the risk of overfitting. As a wrapper method built around a random forest classifier, Boruta iteratively compares the importance of each real feature with that of randomly permuted “shadow features.” During each iteration, the algorithm computes the Z-score (mean decrease accuracy) of every variable and marks a feature as “confirmed” if its Z-score consistently exceeds the maximum Z-score among shadow features, or as “rejected” if it is consistently lower. The algorithm was repeated 500 times to ensure stability, and only the confirmed features were retained for downstream model construction. Meanwhile, the predictive importance of the lactate dehydrogenase-to-albumin ratio (LAR) was specifically evaluated to determine its independent contribution to the outcome. The dataset was randomly divided into a training set and a validation set at a ratio of 7:3. Machine learning survival models were then constructed using the mlr3proba framework in R, including RSF, extreme gradient boosting survival model (XGBoost), and the decision tree survival model (Rpart). RSF was implemented as a non-parametric ensemble approach based on bootstrap aggregation with the log-rank splitting rule and 1,000 trees to ensure robustness. XGBoost optimized the Cox partial likelihood loss function to model nonlinear associations and high-dimensional interactions. The Rpart survival tree provided interpretability by identifying hierarchical relationships among predictors. Hyperparameters were tuned through nested fivefold cross-validation within the training set to optimize predictive accuracy while avoiding overfitting. Model performance was evaluated using time-dependent ROC curves, with AUC at 28 days serving as the primary metric for discrimination. Calibration curves were plotted to assess the agreement between predicted and observed survival probabilities. DCA was further performed to evaluate the net clinical benefit across a range of threshold probabilities, quantifying the practical value of LAR and the final model in clinical decision-making. Finally, feature importance was interpreted through SHapley Additive exPlanations (SHAP) values, which provided an intuitive visualization of each predictor’s contribution to the model output. 3. Results 3.1 Baseline Characteristics A total of 4,692 patients were included, and they were divided into four groups based on LAR quartiles ( Figure 1 ). Table 1 presents the baseline characteristics of the patients. Groups with higher LAR levels exhibited higher heart rates, respiratory rates, white blood cell counts, serum potassium, serum glucose, aspartate aminotransferase (AST), and SOFA scores, along with lower platelet counts and oxygen saturation levels. As LAR quartiles increased, the anion gap widened, and pH values decreased, indicating a higher prevalence of metabolic dysregulation in these patients. Table 1. LAR quartile baseline table characteristics LAR Overall LAR Quartile 1 LAR Quartile 2 LAR Quartile 3 LAR Quartile 4 n 4692 1173 1172 1174 1173 age (mean (SD)) 68.32 (15.46) 70.40 (14.40) 70.28 (14.59) 68.05 (15.76) 64.55 (16.27) abps (mean (SD)) 119.10 (30.13) 122.12 (28.23) 121.05 (28.47) 120.30 (35.48) 112.92 (26.73) hr (mean (SD)) 93.09 (22.42) 89.20 (21.69) 91.58 (21.93) 95.52 (21.82) 96.04 (23.51) rr (mean (SD)) 20.85 (6.66) 20.03 (6.36) 20.52 (5.74) 21.36 (6.57) 21.48 (7.72) wbc (mean (SD)) 14.58 (11.29) 12.12 (6.92) 14.24 (9.45) 15.27 (11.60) 16.69 (15.09) rbc (mean (SD)) 3.54 (0.85) 3.47 (0.79) 3.51 (0.83) 3.56 (0.85) 3.61 (0.90) platele (mean (SD)) 202.20 (116.46) 209.04 (110.86) 209.03 (123.11) 201.20 (117.83) 189.54 (112.69) sodium (mean (SD)) 138.23 (6.45) 137.79 (6.11) 138.36 (6.35) 138.65 (6.83) 138.10 (6.48) potassium (mean (SD)) 4.38 (0.89) 4.31 (0.82) 4.31 (0.83) 4.31 (0.85) 4.57 (1.04) anion_gap (median [IQR]) 16.00 [13.00, 19.00] 15.00 [13.00, 18.00] 15.00 [13.00, 18.00] 16.00 [13.00, 19.00] 17.00 [14.00, 21.00] ph (median [IQR]) 7.35 [7.27, 7.41] 7.36 [7.30, 7.42] 7.36 [7.29, 7.42] 7.35 [7.27, 7.41] 7.31 [7.22, 7.39] pco2 (median [IQR]) 41.00 [34.00, 48.00] 41.00 [35.00, 48.00] 41.00 [34.00, 48.00] 40.50 [34.00, 48.00] 40.00 [34.00, 48.00] lactate (median [IQR]) 1.90 [1.30, 3.20] 1.60 [1.10, 2.40] 1.70 [1.30, 2.80] 2.00 [1.40, 3.30] 2.70 [1.60, 4.90] spo2 (mean (SD)) 96.25 (4.88) 96.58 (4.25) 96.54 (4.36) 96.09 (4.72) 95.81 (5.97) inrpt (median [IQR]) 1.30 [1.20, 1.70] 1.30 [1.10, 1.60] 1.30 [1.20, 1.60] 1.30 [1.20, 1.70] 1.50 [1.20, 1.90] ast (median [IQR]) 59.00 [30.00, 162.00] 30.00 [20.00, 49.00] 47.00 [27.00, 84.00] 73.00 [39.00, 155.00] 279.00 [90.00, 904.00] glucose (mean (SD)) 168.94 (96.00) 160.45 (93.74) 163.97 (90.14) 169.41 (91.20) 181.93 (106.73) sofa (mean (SD)) 7.55 (3.82) 6.58 (3.46) 7.02 (3.50) 7.59 (3.70) 8.99 (4.13) gender = M (%) 2807 (59.8) 722 (61.6) 682 (58.2) 698 (59.5) 705 (60.1) t1dm = 1 (%) 50 (1.1) 13 (1.1) 9 (0.8) 12 (1.0) 16 (1.4) t2dm = 1 (%) 1543 (32.9) 455 (38.8) 405 (34.6) 360 (30.7) 323 (27.5) ckd = 1 (%) 969 (20.7) 268 (22.8) 245 (20.9) 228 (19.4) 228 (19.4) htn = 1 (%) 2265 (48.3) 603 (51.4) 607 (51.8) 576 (49.1) 479 (40.8) hld = 1 (%) 1604 (34.2) 448 (38.2) 433 (36.9) 384 (32.7) 339 (28.9) 3.2 Clinical outcomes The 28-day ICU mortality rate among SIMI patients was higher in groups with elevated LAR levels ( Figure 2 ). In the Cox regression analysis, using the Q1 group as the reference, the ICU mortality risk in the Q3 and Q4 groups significantly increased in both Model I and Model II. The hazard ratios (HR) for the Q3 and Q4 groups were 1.638 (95% CI: 1.404-1.911, p < 0.001) and 2.273 (95% CI: 1.953-2.644, p < 0.001), respectively ( Table 2, Figure 2 ). Kaplan–Meier analysis demonstrated a clear and significant association between LAR quartiles and ICU survival, with patients in the fourth quartile (Q4) showing a significantly increased risk of mortality. The Log-rank test revealed a p-value < 0.001, indicating a significant difference between the survival curves of the different quartiles( Figure 3 ). Table 2. 28-day ICU mortality LAR Unadjusted HR (95% CI) Unadjusted P value Adjusted HR (95% CI) Adjusted P value Q1 Reference Reference Q2 1.192 (1.013-1.403) <0.05 1.150 (0.977-1.354) 0.0938 Q3 1.708 (1.466-1.991) <0.001 1.638 (1.404-1.911) <0.001 Q4 2.471 (2.135-2.860) <0.001 2.273 (1.953-2.644) <0.001 3.3 Restricted cubic spline The RCS analysis was conducted using the rcssci package, adjusting for the effects of age, weight, heart rate, systolic blood pressure, respiratory rate, sex, and SOFA score. The RCS analysis of 28-day ICU mortality ( Figure 4A ) revealed a U-shaped association between LAR (log2) and the risk of death. The turning point of the RCS curve was approximately at LAR（log2） = 4.297 ( Figure 4B ). According to the Cox proportional hazards regression model, the turning point was located in Quartile 1, where the risk of death was minimal, consistent with the RCS results. 3.4 Subgroup analysis As shown in Figure 5 , the 28-day ICU mortality subgroup analysis is presented. We adjusted for age, sex, and comorbidities. In the subgroup analysis, it was found that in the subgroups of age >65, age <65, male, female, and diabetic status, groups with higher LAR levels consistently exhibited higher mortality rates. Regardless of whether confounding variables were adjusted, the high LAR group was significantly associated with higher mortality risk across all subgroups. This highlights the potential of LAR as an independent risk predictor. In the unadjusted model, the hazard ratio (HR) for LAR_groupQ4 was particularly significant in the age ≥65 group and the diabetes group, suggesting that elderly and diabetic patients are more sensitive to high LAR levels. Regarding sex, the risk associated with LAR_groupQ4 was slightly higher in males compared to females. After adjustment in the model, the significance of the risk associated with LAR_groupQ4 remained across all subgroups, particularly in the elderly and diabetic groups, further confirming the independent association between LAR and mortality risk. Following adjustment, the influence of sex on LAR was reduced, suggesting that the observed gender differences were likely driven by confounding variables (HR > 1, P < 0.001 in each subgroup). 3.5 Boruta Algorithm Figure 5 illustrates the features selected by the Boruta algorithm. Variables in the red region were identified as important features, including LAR, age, SOFA score, lactate, and others. Figure S1 provides the scoring details of each variable in the Boruta analysis. 3.6 Model Performance Figure 6 shows the ROC curves of various models, with AUC values indicating model performance. The AUC for the RSF model was 0.7745, for XGBoost was 0.7176, and for the decision tree model was 0.6863, indicating that the RSF model performed the best in predicting mortality risk. Higher LAR levels were associated with increased mortality risk (HR > 1, p < 0.05). The machine learning models incorporating LAR demonstrated good predictive performance. DCA confirmed the net benefit of incorporating LAR into mortality prediction, particularly in the high-risk quartiles (Figure S2). Figure 6 demonstrates the performance of LAR compared to other continuous laboratory indicators, with LAR showing a higher AUC value than the others. 4. Discussion This retrospective cohort study of 4,692 ICU patients with SIMI confirms that the lactate dehydrogenase-to-albumin ratio (LAR) is a reliable predictor of 28-day mortality. This is supported by a strong, dose-response relationship, where the highest LAR quartile was associated with a greater than two-fold increase in mortality risk (adjusted HR: 2.273). After controlling for confounding factors, the results remained consistent in the subgroup analysis, demonstrating the robustness of our findings. To date, this study represents the first exploration of the relationship between LAR and adverse outcomes in patients with SIMI 22 . When compared with existing literature, our findings both confirm and extend previous knowledge. Patients with sepsis-induced myocardial injury (SIMI) exhibit a high degree of complexity 23 ，Current research has paid limited attention to the exploration of metabolic conditions in SIMI patients 22 . Recent studies have shown that metabolic reprogramming plays an important role in cardiac repair. Previous research has also observed a significant shift in myocardial energy metabolism from oxidative phosphorylation to glycolysis during the early stages of sepsis 14, 24 . Immune cells adapt to energy demands by increasing glycolysis 25, 26 . The body responds rapidly to infection by activating immune cells 4, 14 . Moreover, emerging evidence highlights that metabolic recycling of substrates, such as the utilization of phagocytosed bacteria by macrophages, is not merely a passive fuel-gathering process but actively shapes the immune response by suppressing pro-inflammatory pathways and enhancing anti-oxidant defenses.And failure to restore OXPHOS in the later stages of sepsis may lead to sustained pro-inflammatory states and myocardial injury 14, 23, 24 . Research on metabolic transition disorders in patients with SIMI should be given more attention 27 .Notably, while the prognostic value of LDH and albumin individually is well-established, our discovery of a U-shaped relationship between their ratio (LAR) and mortality reveals a more complex, non-linear interaction that had not been previously reported in SIMI.This U-shaped association delineates a bimodal risk profile: the elevated arm signifies severe metabolic stress and cellular injury consequent to failed metabolic reprogramming, while the depressed arm reflects profound inflammatory/nutritional depletion and impaired synthetic capacity. This pattern confirms that both extreme metabolic stress and severe inflammatory/nutritional depletion are detrimental. A key advancement of our study lies in the direct comparison of predictive performance. The Random Survival Forest model incorporating LAR achieved an AUC of 0.775, significantly outperforming the SOFA score alone (AUC = 0.698). This demonstrates that LAR provides prognostic information superior to conventional severity scores and supports its potential for improving risk stratification. The clinical relevance of LAR is underscored by its ability to identify distinct patient phenotypes.LDH is an essential enzyme involved in energy metabolism, and its prognostic value in sepsis patients has been well established 28 . Albumin reflects the inflammatory and nutritional status of sepsis patients 29 . The association between LAR and mortality in SIMI may be attributed to its reflection of the balance between metabolic and inflammatory stress in SIMI patients. Incorporating LAR into predictive models provides clinical value, as confirmed by DCA and calibration analyses. The U-shaped curve implies that LAR captures two distinct high-risk patient phenotypes, which could have important implications for personalized treatment strategies. It is important to acknowledge the limitations of this work. This study has several limitations. First, as a retrospective study, information bias may be present. Second, it is difficult to fully control for all potential confounders, which limits causal inference. Third, sample selection may be influenced by known or unknown factors, making the sample less representative and potentially affecting the study's external validity. Despite these limitations, our findings robustly establish LAR as a novel prognostic marker in SIMI.Future research should focus on the external validation of our findings in prospective, multi-center cohorts. 5. Conclusion In conclusion, LAR is an independent predictor of 28-day mortality in sepsis patients, showing a U-shaped association with mortality. Incorporating LAR into predictive models can improve risk stratification for mortality in sepsis-induced myocardial injury (SIMI) patients, thereby guiding treatment decisions. Declarations Author statements All authors have seen and approved the final version of the manuscript being submitted. They warrant that the article is the authors' original work, has not been published previously, and is not under consideration for publication elsewhere. Consent for publication All authors have read and approved the final manuscript and give their consent for its publication in .BMC Infectious Diseases Nixiang Jiang Xiangya School of Nursing, Central South University, Changsha 410013, China Email: [email protected] Jinwei Dai Department of Critical Care Medicine, Xiangya Hospital, Central South University, Changsha, Hunan, 410008, China National Clinical Research Center of Geriatric Disorders, Xiangya Hospital, Central South University, Changsha, China Email: [email protected] Yang Zhou (Corresponding Author) National Clinical Research Center of Geriatric Disorders, Xiangya Hospital, Central South University, Changsha, China Department of Clinical Nursing, Xiangya Hospital, Central South University, Changsha 410005, China Email: [email protected] Author contributions Yang Zhou: Study conception and design. Nixiang Jiang: Literature research, data extraction, and quality control of data and algorithms. Jinwei Dai: Statistical analysis and data interpretation. All authors: Manuscript writing, review, and approval. Funding This work was supported by the National Natural Science Foundation of China (82360448); the Natural Science Foundation of Hunan Province (2022JJ30503); and the Scientific Research Project of Health Commission of Hunan Province (2022030128723001). Availability of data and material Publicly available datasets were analyzed in this study. These data can be found at https://mimic.mit.edu/. Declaration of competing interest The authors declare that there are no conflicts of interest that could be construed as influencing the work reported in this paper. Ethics declarations Ethics approval and consent to participate This study utilized data from the MIMIC-IV database (version 3.1). The MIMIC-IV database is a publicly available, de-identified critical care database. Its creation was approved by the Institutional Review Boards (IRB) of the Massachusetts Institute of Technology (MIT) and Beth Israel Deaconess Medical Center (BIDMC). The requirement for individual patient consent was waived by these IRBs as the data is de-identified. All methods were carried out in accordance with the Declaration of Helsinki. Clinical trial number:Not applicable. Human Ethics and Consent to Participate declarations: Not applicable. References Singer, M. et al. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). Jama 315 (8), 801–810. 10.1001/jama.2016.0287 (2016). From NLM. Su, Z. D. et al. Melatonin alleviates lipopolysaccharide-induced myocardial injury by inhibiting inflammation and pyroptosis in cardiomyocytes. Ann. Transl Med. 9 (5), 413. 10.21037/atm-20-8196 (2021). From NLM. Walley, K. R. Sepsis-induced myocardial dysfunction. Curr. Opin. Crit. Care . 24 (4), 292–299. 10.1097/mcc.0000000000000507 (2018). From NLM. Hollenberg, S. M. & Singer, M. Pathophysiology of sepsis-induced cardiomyopathy. Nat. Rev. Cardiol. 18 (6), 424–434. 10.1038/s41569-020-00492-2 (2021). From NLM. Beesley, S. J. et al. Septic Cardiomyopathy. Crit. Care Med. 46 (4), 625–634. 10.1097/ccm.0000000000002851 (2018). From NLM. Kottmann, R. M. et al. Lactic acid is elevated in idiopathic pulmonary fibrosis and induces myofibroblast differentiation via pH-dependent activation of transforming growth factor-β. Am. J. Respir Crit. Care Med. 186 (8), 740–751. 10.1164/rccm.201201-0084OC (2012). From NLM. Garrabou, G. et al. The effects of sepsis on mitochondria. J. Infect. Dis. 205 (3), 392–400. 10.1093/infdis/jir764 (2012). From NLM. Liu, J., Zhou, G., Wang, X. & Liu, D. Metabolic reprogramming consequences of sepsis: adaptations and contradictions. Cell. Mol. Life Sci. 79 (8), 456. 10.1007/s00018-022-04490-0 (2022). From NLM. Preau, S. et al. Energetic dysfunction in sepsis: a narrative review. Ann. Intensive Care . 11 (1), 104. 10.1186/s13613-021-00893-7 (2021). From NLM. Yang, H. & Zhang, Z. Sepsis-induced myocardial dysfunction: the role of mitochondrial dysfunction. Inflamm. Res. 70 (4), 379–387. 10.1007/s00011-021-01447-0 (2021). From NLM. Feng, Y. et al. Lactate dehydrogenase A: A key player in carcinogenesis and potential target in cancer therapy. Cancer Med. 7 (12), 6124–6136. 10.1002/cam4.1820 (2018). From NLM. Vincent, J. L., De Backer, D. & Wiedermann, C. J. Fluid management in sepsis: The potential beneficial effects of albumin. J. Crit. Care . 35 , 161–167. 10.1016/j.jcrc.2016.04.019 (2016). From NLM. Liang, M. et al. The association between lactate dehydrogenase to serum albumin ratio and the 28-day mortality in patients with sepsis-associated acute kidney injury in intensive care: a retrospective cohort study. Ren. Fail. 45 (1), 2212080. 10.1080/0886022x.2023.2212080 (2023). From NLM. Li, X. et al. Inhibition of fatty acid oxidation enables heart regeneration in adult mice. Nature 622 (7983), 619–626. 10.1038/s41586-023-06585-5 (2023). From NLM. Chen, Y. et al. LDHA-mediated metabolic reprogramming promoted cardiomyocyte proliferation by alleviating ROS and inducing M2 macrophage polarization. Redox Biol. 56 , 102446. 10.1016/j.redox.2022.102446 (2022). From NLM. Dong, Y. et al. Aspirin is associated with improved outcomes in patients with sepsis-induced myocardial injury: An analysis of the MIMIC-IV database. J. Clin. Anesth. 99 , 111597. 10.1016/j.jclinane.2024.111597 (2024). From NLM. Lee, B. K. et al. Lactate dehydrogenase to albumin ratio as a prognostic factor in lower respiratory tract infection patients. Am. J. Emerg. Med. 52 , 54–58. 10.1016/j.ajem.2021.11.028 (2022). From NLM. Vallabhajosyula, S. et al. Role of Admission Troponin-T and Serial Troponin-T Testing in Predicting Outcomes in Severe Sepsis and Septic Shock. J. Am. Heart Assoc. 6 (9). 10.1161/jaha.117.005930 (2017). From NLM. Xie, R., Chen, Q., He, W. & Zeng, M. Association of Cardiac Troponin T Concentration on Admission with Prognosis in Critically Ill Patients without Myocardial Infarction: A Cohort Study. Int. J. Gen. Med. 14 , 2729–2739. 10.2147/ijgm.S318232 (2021). From NLM. Kalimouttou, A., Lerner, I., Cheurfa, C., Jannot, A. S. & Pirracchio, R. Machine-learning-derived sepsis bundle of care. Intensive Care Med. 49 (1), 26–36. 10.1007/s00134-022-06928-2 (2023). From NLM. Venugopalan, J., Chanani, N., Maher, K. & Wang, M. D. Novel Data Imputation for Multiple Types of Missing Data in Intensive Care Units. IEEE J. Biomed. Health Inf. 23 (3), 1243–1250. 10.1109/jbhi.2018.2883606 (2019). From NLM. Tao, T. et al. The top cited clinical research articles on sepsis: a bibliometric analysis. Crit Care 16 (3), R110. DOI: 10.1186/cc11401 From NLM. (2012). Yao, R. Q. et al. Publication Trends of Research on Sepsis and Host Immune Response during 1999–2019: A 20-year Bibliometric Analysis. Int. J. Biol. Sci. 16 (1), 27–37. 10.7150/ijbs.37496 (2020). From NLM. Bauerfeld, C., Talwar, H., Zhang, K., Liu, Y. & Samavati, L. MKP-1 Modulates Mitochondrial Transcription Factors, Oxidative Phosphorylation, and Glycolysis. Immunohorizons 4 (5), 245–258. 10.4049/immunohorizons.2000015 (2020). From NLM. Zuo, W., Sun, R., Ji, Z. & Ma, G. Macrophage-driven cardiac inflammation and healing: insights from homeostasis and myocardial infarction. Cell. Mol. Biol. Lett. 28 (1), 81. 10.1186/s11658-023-00491-4 (2023). From NLM. Gharavi, A. T., Hanjani, N. A., Movahed, E. & Doroudian, M. The role of macrophage subtypes and exosomes in immunomodulation. Cell. Mol. Biol. Lett. 27 (1), 83. 10.1186/s11658-022-00384-y (2022). From NLM. Yu, Q. et al. The role of histone deacetylases in cardiac energy metabolism in heart diseases. Metabolism 142 , 155532. 10.1016/j.metabol.2023.155532 (2023). From NLM. Duman, A. et al. Prognostic value of neglected biomarker in sepsis patients with the old and new criteria: predictive role of lactate dehydrogenase. Am. J. Emerg. Med. 34 (11), 2167–2171. 10.1016/j.ajem.2016.06.012 (2016). From NLM. Takegawa, R. et al. Serum albumin as a risk factor for death in patients with prolonged sepsis: An observational study. J. Crit. Care . 51 , 139–144. 10.1016/j.jcrc.2019.02.004 (2019). From NLM. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8189180","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":560855165,"identity":"dd5b1b5b-d3ec-484e-ac69-b85e67e0a52c","order_by":0,"name":"Nixiang Jiang","email":"","orcid":"","institution":"Central South University","correspondingAuthor":false,"prefix":"","firstName":"Nixiang","middleName":"","lastName":"Jiang","suffix":""},{"id":560855166,"identity":"85cfa252-41a4-4894-9e23-016fd2adfd7d","order_by":1,"name":"Jinwei Dai","email":"","orcid":"","institution":"Central South University","correspondingAuthor":false,"prefix":"","firstName":"Jinwei","middleName":"","lastName":"Dai","suffix":""},{"id":560855167,"identity":"ef3ca65c-ba45-49f2-9cfd-9c56d4e27441","order_by":2,"name":"Yang Zhou","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA/ElEQVRIiWNgGAWjYDACCSjNxsx+4MAHCNuAOC187D2JD2eQpEWO54CxMQ8xWuRnNx+T5qm5Y9cmkZAmbVNTl9jA3rxNgqHmDk4tjHOOpUnzHHuW3CaReEw659jhxAaeY2USDMee4dTCLJFjJs3DdjiZDWRLbsOBxAagiARjw2GcWtjAWv6BtZhJWzYAHSb/Br8WHpAW3rbDdmwg7zM2MANt4cGvRUIiLdlybt/hBDZQIPccO2zcxpNWbJFwDLcW+RnJB2+8+XbYXr4ZGJU/aupk+9kPb7zxoQa3FhBgAkZHYgPcdyAiAa8GYED/YGCwJ6BmFIyCUTAKRjIAANUrUJ1ZSKxNAAAAAElFTkSuQmCC","orcid":"","institution":"Central South University","correspondingAuthor":true,"prefix":"","firstName":"Yang","middleName":"","lastName":"Zhou","suffix":""}],"badges":[],"createdAt":"2025-11-24 04:38:25","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8189180/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8189180/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":98302890,"identity":"ef29bd62-eb2a-48e7-a28b-9773caced247","added_by":"auto","created_at":"2025-12-16 10:35:42","extension":"tif","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":127746,"visible":true,"origin":"","legend":"","description":"","filename":"Figure1.tif","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/8eb6966373d0f48d9a615f23.tif"},{"id":98435212,"identity":"a0fae271-bcd5-4f5e-8907-e10895226c8d","added_by":"auto","created_at":"2025-12-17 16:53:19","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1148543,"visible":true,"origin":"","legend":"","description":"","filename":"Manuscript.docx","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/a5fd2c37cb43f4805123c5d0.docx"},{"id":98302893,"identity":"381f95c1-812f-475f-b801-f19678bf0dda","added_by":"auto","created_at":"2025-12-16 10:35:42","extension":"tif","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1332964,"visible":true,"origin":"","legend":"","description":"","filename":"Figure5.tif","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/5f5a0ac96980d3103957c150.tif"},{"id":98435457,"identity":"ec80da0d-a09d-4731-ae42-0beb5964028f","added_by":"auto","created_at":"2025-12-17 16:53:51","extension":"tif","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":209352,"visible":true,"origin":"","legend":"","description":"","filename":"Figure6.tif","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/7c095685e9b5a0df90c11770.tif"},{"id":98302892,"identity":"0d63de90-2ebb-49dc-a240-5bf153f532d2","added_by":"auto","created_at":"2025-12-16 10:35:42","extension":"docx","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":20609,"visible":true,"origin":"","legend":"","description":"","filename":"Table.docx","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/55dcec7ccdbd2c590604d322.docx"},{"id":98302925,"identity":"291cd7dd-50d9-4b4d-8594-6bf3191ced0d","added_by":"auto","created_at":"2025-12-16 10:35:43","extension":"tif","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":57511176,"visible":true,"origin":"","legend":"","description":"","filename":"Figure2.tif","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/a109fdfc354f4840de5f938a.tif"},{"id":98435581,"identity":"d20509a3-5aef-411a-8c21-3a23876fbf03","added_by":"auto","created_at":"2025-12-17 16:54:05","extension":"tif","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":43205112,"visible":true,"origin":"","legend":"","description":"","filename":"Figure3.tif","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/2b39fb92607fae7f4f3d8d72.tif"},{"id":98302923,"identity":"158e4600-c552-4953-9a58-259009bd48e0","added_by":"auto","created_at":"2025-12-16 10:35:43","extension":"tif","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":37841040,"visible":true,"origin":"","legend":"","description":"","filename":"Figure4.tif","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/3dbb4ceae5a781710dc0ee1c.tif"},{"id":98435380,"identity":"ee7b7114-1c7e-4ac4-9028-fc9b35da331c","added_by":"auto","created_at":"2025-12-17 16:53:38","extension":"json","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5628,"visible":true,"origin":"","legend":"","description":"","filename":"850c67b5fc18431e98faeccbd3ab0b62.json","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/a05367a1930c9c537f81aa55.json"},{"id":98302905,"identity":"4f6f8be2-5416-490c-bcbb-fbcf0a1f9761","added_by":"auto","created_at":"2025-12-16 10:35:42","extension":"xml","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":124625,"visible":true,"origin":"","legend":"","description":"","filename":"850c67b5fc18431e98faeccbd3ab0b621enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/b9a4d80bd6520ff06d749198.xml"},{"id":98302899,"identity":"c10dfc04-bcf0-4d27-8dd0-b8fe4bd2890e","added_by":"auto","created_at":"2025-12-16 10:35:42","extension":"tif","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":127746,"visible":true,"origin":"","legend":"","description":"","filename":"Figure1.tif","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/547d261772ba84161b61bfd5.tif"},{"id":98302909,"identity":"8b71a966-087d-4715-8802-55a269e3b99e","added_by":"auto","created_at":"2025-12-16 10:35:42","extension":"tif","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1332964,"visible":true,"origin":"","legend":"","description":"","filename":"Figure5.tif","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/9dd720b3f12a263354002124.tif"},{"id":98434908,"identity":"96e2cb81-db8a-4719-8589-f0b74f35c7cb","added_by":"auto","created_at":"2025-12-17 16:52:44","extension":"tif","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":209352,"visible":true,"origin":"","legend":"","description":"","filename":"Figure6.tif","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/4486818b0b44546289be2116.tif"},{"id":98302924,"identity":"743ffc41-1a20-45bb-b3cc-0afbf3910e47","added_by":"auto","created_at":"2025-12-16 10:35:43","extension":"tif","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":57511176,"visible":true,"origin":"","legend":"","description":"","filename":"Figure2.tif","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/ab7caf1fb0913c7ec5279362.tif"},{"id":98302930,"identity":"3fb5f0e5-535d-40cf-b450-fc9415ab0504","added_by":"auto","created_at":"2025-12-16 10:35:43","extension":"tif","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":43205112,"visible":true,"origin":"","legend":"","description":"","filename":"Figure3.tif","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/ce7cbd1ea43e3ac5a6f7b888.tif"},{"id":98302927,"identity":"3b5a6c0f-52cf-4801-b898-14724d8c41bd","added_by":"auto","created_at":"2025-12-16 10:35:43","extension":"tif","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":37841040,"visible":true,"origin":"","legend":"","description":"","filename":"Figure4.tif","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/c119739f2a1fa6fa81b408b9.tif"},{"id":98435961,"identity":"e9893789-3df8-4dc2-a1d9-f78961f740a1","added_by":"auto","created_at":"2025-12-17 16:54:41","extension":"jpeg","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":127746,"visible":true,"origin":"","legend":"","description":"","filename":"Figure1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/757e74d412100292bdb94ce2.jpeg"},{"id":98435893,"identity":"fdc7aa30-8c7e-4a0d-a6af-e3d25350d80a","added_by":"auto","created_at":"2025-12-17 16:54:33","extension":"png","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":352035,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/8f6a01fd6ba0ce335dd94f7f.png"},{"id":98435232,"identity":"9a7ba90d-8139-43ff-808a-764ffa6558be","added_by":"auto","created_at":"2025-12-17 16:53:22","extension":"jpeg","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":452348,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/dc428a817f0d6dfadbef63e0.jpeg"},{"id":98435327,"identity":"2a56686e-c36c-4f5a-a430-6bc674473a16","added_by":"auto","created_at":"2025-12-17 16:53:29","extension":"jpeg","order_by":19,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":4585322,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/8793f8d63ca62894208ca84d.jpeg"},{"id":98435204,"identity":"d29bd100-115a-4f45-ac63-510716344b77","added_by":"auto","created_at":"2025-12-17 16:53:18","extension":"jpeg","order_by":20,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":111006,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage5.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/09de040f64b98d1045a5ff98.jpeg"},{"id":98435410,"identity":"3ad7f5fd-ff2f-4ec2-9ea1-d0481d387d2c","added_by":"auto","created_at":"2025-12-17 16:53:42","extension":"jpeg","order_by":21,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":209352,"visible":true,"origin":"","legend":"","description":"","filename":"Figure6.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/876d49574a6457af1e86648e.jpeg"},{"id":98302912,"identity":"c7517434-b786-4fcf-ac54-5d8c1ad6556b","added_by":"auto","created_at":"2025-12-16 10:35:42","extension":"png","order_by":22,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":25064,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigure1.png","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/1e44f45d36ac2033039229ef.png"},{"id":98302902,"identity":"5cb4b9f7-e038-4845-9eb6-516605e18fda","added_by":"auto","created_at":"2025-12-16 10:35:42","extension":"png","order_by":23,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":133760,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigure5.png","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/a6b8cd040d9de616f1d06bf3.png"},{"id":98435225,"identity":"7d48ad40-7ed7-44b0-9a20-5a7790c929e7","added_by":"auto","created_at":"2025-12-17 16:53:21","extension":"png","order_by":24,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":31551,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigure6.png","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/4ac6d10ae1e5e26ce2c8dd69.png"},{"id":98302916,"identity":"186a8686-01e0-4333-baa5-832928ce0265","added_by":"auto","created_at":"2025-12-16 10:35:42","extension":"png","order_by":25,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":346339,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigure2.png","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/32d24ec141b5043fc7f19ee6.png"},{"id":98436430,"identity":"47469b7b-85b0-4e4d-8ce0-65ce412ba163","added_by":"auto","created_at":"2025-12-17 16:55:40","extension":"png","order_by":26,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1127880,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigure3.png","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/cd3181244a95073792118bf4.png"},{"id":98435533,"identity":"ed3f4dae-bd38-443f-8c73-ca4274170075","added_by":"auto","created_at":"2025-12-17 16:54:01","extension":"png","order_by":27,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":345256,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigure4.png","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/4365007acdffed54bef6bd84.png"},{"id":98302914,"identity":"2ad17c48-34cc-4528-bbf7-d0dab7eefa08","added_by":"auto","created_at":"2025-12-16 10:35:42","extension":"png","order_by":28,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":25064,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/fd58c951432930a1725804d9.png"},{"id":98436809,"identity":"7fb82d57-6c57-4ac1-aab7-3d80059d181d","added_by":"auto","created_at":"2025-12-17 16:56:17","extension":"png","order_by":29,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":43110,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/ccc727dcfef5e3736c556467.png"},{"id":98302918,"identity":"dc042529-f8ff-444f-865d-e7097a6d81c0","added_by":"auto","created_at":"2025-12-16 10:35:43","extension":"png","order_by":30,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":71967,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/271c6db14190294e5763e058.png"},{"id":98436551,"identity":"213d453e-5538-4795-a71e-5ee6ba46e283","added_by":"auto","created_at":"2025-12-17 16:55:52","extension":"png","order_by":31,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":25045,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/c3fe3b7a45e36ca0336153a2.png"},{"id":98435694,"identity":"b8d33a6f-1b50-4617-a9a5-8cae0ef2a4ad","added_by":"auto","created_at":"2025-12-17 16:54:12","extension":"png","order_by":32,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":15317,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/2387772e10f29c956a65d88c.png"},{"id":98436755,"identity":"cf305cae-ed0e-419b-a084-54976fcd958f","added_by":"auto","created_at":"2025-12-17 16:56:13","extension":"png","order_by":33,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":31551,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/befaf1add5e2f2043fafdd11.png"},{"id":98302922,"identity":"910c380c-f2c9-4f2b-bd89-bcb5296b63ed","added_by":"auto","created_at":"2025-12-16 10:35:43","extension":"xml","order_by":34,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":122264,"visible":true,"origin":"","legend":"","description":"","filename":"850c67b5fc18431e98faeccbd3ab0b621structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/83b1290a3f15926445afb66d.xml"},{"id":98435567,"identity":"4729d1cc-77cd-4f7a-b4a3-2077441d6e37","added_by":"auto","created_at":"2025-12-17 16:54:04","extension":"html","order_by":35,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":137960,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/6c77f7a3563da564a4bcae89.html"},{"id":98302888,"identity":"401a9b39-3ca3-4164-96fa-6ebe75c950df","added_by":"auto","created_at":"2025-12-16 10:35:42","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":535153,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSelection of the Study Population from the MIMIC-IV Database.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFigure Caption: The diagram illustrates the selection process for ICU-diagnosed sepsis adult patients from the MIMIC-IV database.\u003c/p\u003e","description":"","filename":"image1.png","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/634bec28d27fdc582f41be9d.png"},{"id":98302889,"identity":"0600da23-c24d-46b6-9bee-fe9b08e076a7","added_by":"auto","created_at":"2025-12-16 10:35:42","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":352035,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eICU Mortality Across LAR Quartiles.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A): Comparison of hazard ratios (HR) for ICU mortality across LAR quartiles, displayed for both unadjusted and adjusted models.\u003cstrong\u003e \u003c/strong\u003e(B): Distribution of ICU mortality rates across LAR quartiles, showing the proportion of deaths (1) and survivors (0).\u003c/p\u003e","description":"","filename":"image2.png","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/0f57ecab35d02677d957408d.png"},{"id":98436740,"identity":"978663d2-e766-4688-a599-a0a359bcfba2","added_by":"auto","created_at":"2025-12-17 16:56:12","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":476801,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eRestricted Cubic Spline Analysis and ICU Mortality Under Different Models.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A)Restricted cubic spline analysis showing the association between LAR (log2-transformed) and ICU mortality under different models.(B)Adjusted restricted cubic spline analysis for the relationship between LAR (log2-transformed) and ICU mortality, demonstrating a non-linear \"U-shaped\" association. (C)Kaplan–Meier survival curves illustrating 28-day ICU survival probabilities across LAR quartiles, with significant differences observed (\u003cem\u003ep\u003c/em\u003e \u0026lt; 0.0001).\u003c/p\u003e","description":"","filename":"image3.png","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/757cdde315c7104068fe01ca.png"},{"id":98302895,"identity":"c840c37e-a68a-4a5a-ab39-d9f8eed9b32b","added_by":"auto","created_at":"2025-12-16 10:35:42","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":457903,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDeath Rates Across LAR Quartiles and Sub Group Analysis.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A)Stacked bar plot showing the death rates across LAR quartiles (Q1–Q4) stratified by subgroups of age (\u0026lt;65 and ≥65) and diabetes status (with or without diabetes). (B)Forest plot of hazard ratios (HR) for LAR quartiles (Q2–Q4, with Q1 as reference) across subgroups of age, sex, and diabetes status.\u003c/p\u003e","description":"","filename":"image4.png","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/5212b2f9f9fef6c9918e22aa.png"},{"id":98436602,"identity":"b943b204-c945-492f-8f36-ee7b884d3cef","added_by":"auto","created_at":"2025-12-17 16:55:59","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":258859,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe Feature Importance Rankings from the Boruta Algorithm.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eBox plot showing the feature importance rankings from the Boruta algorithm.\u003c/p\u003e","description":"","filename":"image5.png","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/267ce08ec813079017db47ae.png"},{"id":98436201,"identity":"a2b8d4ed-70ce-400f-9a05-a48bf6d2aa10","added_by":"auto","created_at":"2025-12-17 16:55:05","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":545433,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePerformance of the LAR and Laboratory Data and ROC Curves of the Machine Learning Algorithms.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) This panel shows the performance comparison of LAR and various laboratory data in predicting ICU mortality, with relevant statistical measures. (B) ROC curves of the machine learning algorithms. dt Rpart Survival Trees Survival Learner, \u0026nbsp;rsf Survival Random Forest SRC Learner, xgboost extreme gradient boosting survival learner, T days, AUC area under the curve.\u003c/p\u003e","description":"","filename":"image6.png","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/c25f6d9974e7e9386c06cbd7.png"},{"id":107704924,"identity":"f48c7d1d-b240-43b1-90fc-0ad978512831","added_by":"auto","created_at":"2026-04-24 09:04:10","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2937772,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8189180/v1/e12bee90-3a00-40c6-9799-c4e1ca8be0ed.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"The Lactate Dehydrogenase-to-Albumin Ratio (LAR) Predicts Mortality in Sepsis-Induced Myocardial Injury","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eSepsis has long been at the forefront of global health challenges, manifesting as life-threatening organ dysfunction caused by a dysregulated host response to infection\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e.SIMI is a serious, often reversible complication,\u003csup\u003e2\u0026ndash;4\u003c/sup\u003eyet it carries a poor prognosis\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e.With an incidence of 13\u0026ndash;65% and mortality exceeding 50% in severe cases, better prognostic tools are urgently needed.\u003c/p\u003e \u003cp\u003eLactate dehydrogenase converts pyruvate to lactate and is a common enzyme in energy metabolism\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e. Studies have observed that the energy metabolism of cardiomyocytes undergoes a significant shift from oxidative phosphorylation to glycolysis in the early stages of sepsis\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. Immune cells meet their energy demands by increasing glycolysis, thereby rapidly responding to infection \u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e. However, in the later stages of sepsis, failure to restore oxidative phosphorylation (OXPHOS) in a timely manner may lead to a persistent pro-inflammatory state and myocardial injury\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e.The pathophysiology of SIMI thus creates a clear rationale for biomarkers that integrate metabolic and inflammatory signals.Notably, recent research demonstrates that macrophages can directly link metabolic and immune states by recycling phagocytosed bacteria to fuel immunometabolic responses.\u003c/p\u003e \u003cp\u003eLactate dehydrogenase is involved in this metabolic process\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. Albumin is an important indicator reflecting systemic inflammation and nutritional status in critically ill patients \u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e. Previous studies have found that LAR is an important prognostic indicator in sepsis-associated acute kidney injury (AKI)\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. Some scholars have also observed the role of metabolic reprogramming in cardiac repair\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e.Therefore, we hypothesize that LAR, serving as a novel biomarker integrating metabolic stress and systemic inflammation, is an independent predictor of 28-day mortality specifically in patients with SIMI. While LAR has demonstrated prognostic value in other sepsis-related conditions such as acute kidney injury\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e, its association with outcomes in the SIMI population remains uninvestigated. To bridge this knowledge gap, we conducted a large-scale retrospective cohort study utilizing the MIMIC-IV database. We aim to comprehensively evaluate this association using both traditional statistical models and advanced machine learning algorithms, with the goal of validating LAR's utility in improving risk stratification for this high-risk population.\u003c/p\u003e"},{"header":"2. Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Data Source\u003c/h2\u003e \u003cp\u003eThe data used in this study were obtained from MIMIC-IV (3.1), a large database containing clinical information for ICU patients at Beth Israel Deaconess Medical Center from 2008 to 2019. After screening 31,910 ICU-diagnosed sepsis patients and applying exclusion criteria, 4,692 patients with sepsis-induced myocardial injury were included in the final analysis. The BIDMC Institutional Review Board approved a waiver of informed consent. The author (JW.D) obtained access to the database (certification number: 62317039). The establishment of this database was approved by the Massachusetts Institute of Technology and Beth Israel Deaconess Medical Center.\u003c/p\u003e \u003cp\u003e \u003cb\u003e2.2 Inclusion and exclusion criteria\u003c/b\u003e \u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003e \u003cb\u003eInclusion Criteria\u003c/b\u003e: Patients aged\u0026thinsp;\u0026ge;\u0026thinsp;18 years.Meeting the sepsis 3.0 criteria and SIMI criteria\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003eExclusion Criteria\u003c/b\u003e:\u003c/p\u003e \u003cp\u003e \u003col\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003ePatients with an ICU stay of less than 24 h.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eMissing LDH and ALBUMIN in the first laboratory test.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eSuffering from heart failure, myocardial infarction and other serious heart diseases.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003cspan\u003e \u003cli\u003e \u003cp\u003eFor patients with multiple ICU admissions, only data from the first hospitalization were included.\u003c/p\u003e \u003c/li\u003e \u003c/span\u003e \u003c/ol\u003e \u003c/p\u003e \u003cp\u003eData specification: For all laboratory tests, the most extreme value (i.e., the maximum for LDH, cTnT, lactate; the minimum for albumin) recorded within the first 24 hours of ICU stay was utilized in the analysis.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.3 Data collection\u003c/h2\u003e \u003cp\u003eData extraction was performed using the pgAdmin software. The collected patient characteristics included the following: basic demographic information such as age, gender, and weight. Information on comorbidities was extracted based on the International Classification of Diseases (ICD) coding system, including hypertension, type 2 diabetes, type 1 diabetes, heart failure, malignancy, chronic kidney disease (CKD), stroke, pneumonia, and septic shock. Vital signs (heart rate, systolic blood pressure (SBP), respiratory rate, arterial oxygen saturation (SaO2)), and laboratory tests [lactate dehydrogenase (LDH), troponin T (cTnT), red blood cell count (RBC), white blood cell count (WBC), platelet count (PLT), serum sodium, serum potassium, serum calcium, anion gap, pH, partial pressure of carbon dioxide (PaCO2), arterial oxygen partial pressure (PaO2), lactate, international normalized ratio (INR) of prothrombin time, total bilirubin, aspartate aminotransferase (AST), blood urea nitrogen, serum creatinine, and serum glucose] were included. Vital signs (heart rate, systolic blood pressure (SBP), respiratory rate, arterial oxygen saturation (SaO2)), and laboratory tests [lactate dehydrogenase (LDH), troponin T (cTnT), red blood cell count (RBC), white blood cell count (WBC), platelet count (PLT), serum sodium, serum potassium, serum calcium, anion gap, pH, partial pressure of carbon dioxide (PaCO2), arterial oxygen partial pressure (PaO2), lactate, international normalized ratio (INR) of prothrombin time, total bilirubin, aspartate aminotransferase (AST), blood urea nitrogen, serum creatinine, and serum glucose] were included. Interventions included mechanical ventilation and renal replacement therapy (RRT). SOFA score was used to evaluate disease severity. LAR was calculated as an index using the following formula: LAR = (admission lactate dehydrogenase (IU/L)) / (albumin (g/L)) \u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.4 Diagnosis of SIMI\u003c/h2\u003e \u003cp\u003eThe first cTnT result after ICU admission was extracted, and the worst value of the day was used for SIMI assessment. The 99th percentile of the upper reference limit for troponin T in this center is 0.01 ng/mL, and SIMI was defined as cTnT\u0026thinsp;\u0026gt;\u0026thinsp;0.01 ng/mL The first cTnT result after ICU admission was extracted, and the worst value of the day was used for SIMI assessment. The 99th percentile of the upper reference limit for troponin T in this center is 0.01 ng/mL, and SIMI was defined as cTnT\u0026thinsp;\u0026gt;\u0026thinsp;0.01 ng/mL \u003csup\u003e16, 18, 19\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e2.5 Outcome\u003c/h2\u003e \u003cp\u003eThe primary outcome was 28-day -ICU mortality.This was defined as the all-cause mortality occurring within a 28-day period starting from the time of ICU admission. Patients who were discharged from the hospital alive before day 28 were considered survivors (i.e., right-censored at their discharge date). The outcome was determined by extracting the dod (date of death) and discharge time stamps from the MIMIC-IV database. The survival time was calculated as the difference between ICU admission time and either the date of death or the date of hospital discharge, whichever occurred first within the 28-day window\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e2.6 Statistical analysis\u003c/h2\u003e \u003cp\u003eMissing data were handled using multiple imputation, while variables with a missing rate greater than 20% were excluded\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003eMulticollinearity of variables was assessed using the variance inflation factor (VIF), and variables with a VIF greater than 5 were excluded. Patients were divided into four groups\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e based on the quartiles of LAR levels. Continuous variables with a normal or approximately normal distribution were expressed as mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD and analyzed using analysis of variance (ANOVA). For non-normally distributed variables, Kruskal\u0026ndash;Wallis tests were used for analysis. Categorical variables were expressed as numbers and percentages and analyzed using the χ\u0026sup2; test or Fisher's exact test. Kaplan\u0026ndash;Meier survival curves were used to compare the 28-day survival rates among the four patient groups. We also used the Cox proportional hazards model to evaluate the hazard ratio (HR) and 95% confidence interval (95% CI) for event occurrence. Two models were constructed: Model I without covariate adjustments and Model II adjusted for age, weight, heart rate, respiratory rate, systolic blood pressure, gender, and SOFA score. Time-dependent ROC curves were used to evaluate LAR and other continuous variables. A p-value of \u0026lt;\u0026thinsp;0.05 was considered statistically significant. All statistical analyses were performed using RStudio (version 4.4.1).\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section3\"\u003e \u003ch2\u003e2.6.1 Restricted cubic splines\u003c/h2\u003e \u003cp\u003eWe controlled for covariates (age, weight, heart rate, SBP, respiratory rate, sex, and SOFA scores) and collected data on LAR and outcome variables. The potential non-linear relationship between changes in LAR and survival rates was examined using the \u003cb\u003ercssci\u003c/b\u003e R package, which automatically selects knots for the analysis.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section3\"\u003e \u003ch2\u003e2.6.2 Subgroup analysis\u003c/h2\u003e \u003cp\u003eSubgroup analyses were performed based on age, sex, and diabetes status. Multivariable analysis was adjusted for age, weight, heart rate, SBP, respiratory rate, sex, and SOFA scores. Cox regression analysis was conducted for each subgroup, and the results were visually presented using a forest plot.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e2.7 Establishment and validation of the prediction models\u003c/h2\u003e \u003cp\u003eThe Boruta algorithm was employed to identify the most informative predictors and reduce the risk of overfitting. As a wrapper method built around a random forest classifier, Boruta iteratively compares the importance of each real feature with that of randomly permuted \u0026ldquo;shadow features.\u0026rdquo; During each iteration, the algorithm computes the Z-score (mean decrease accuracy) of every variable and marks a feature as \u0026ldquo;confirmed\u0026rdquo; if its Z-score consistently exceeds the maximum Z-score among shadow features, or as \u0026ldquo;rejected\u0026rdquo; if it is consistently lower. The algorithm was repeated 500 times to ensure stability, and only the confirmed features were retained for downstream model construction. Meanwhile, the predictive importance of the lactate dehydrogenase-to-albumin ratio (LAR) was specifically evaluated to determine its independent contribution to the outcome.\u003c/p\u003e \u003cp\u003eThe dataset was randomly divided into a training set and a validation set at a ratio of 7:3. Machine learning survival models were then constructed using the mlr3proba framework in R, including RSF, extreme gradient boosting survival model (XGBoost), and the decision tree survival model (Rpart). RSF was implemented as a non-parametric ensemble approach based on bootstrap aggregation with the log-rank splitting rule and 1,000 trees to ensure robustness. XGBoost optimized the Cox partial likelihood loss function to model nonlinear associations and high-dimensional interactions. The Rpart survival tree provided interpretability by identifying hierarchical relationships among predictors. Hyperparameters were tuned through nested fivefold cross-validation within the training set to optimize predictive accuracy while avoiding overfitting.\u003c/p\u003e \u003cp\u003eModel performance was evaluated using time-dependent ROC curves, with AUC at 28 days serving as the primary metric for discrimination. Calibration curves were plotted to assess the agreement between predicted and observed survival probabilities. DCA was further performed to evaluate the net clinical benefit across a range of threshold probabilities, quantifying the practical value of LAR and the final model in clinical decision-making. Finally, feature importance was interpreted through SHapley Additive exPlanations (SHAP) values, which provided an intuitive visualization of each predictor\u0026rsquo;s contribution to the model output.\u003c/p\u003e \u003c/div\u003e"},{"header":"3. Results","content":"\u003cp\u003e\u003cstrong\u003e3.1 Baseline Characteristics\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA total of 4,692 patients were included, and they were divided into four groups based on LAR quartiles (\u003cstrong\u003eFigure 1\u003c/strong\u003e). Table 1 presents the baseline characteristics of the patients. Groups with higher LAR levels exhibited higher heart rates, respiratory rates, white blood cell counts, serum potassium, serum glucose, aspartate aminotransferase (AST), and SOFA scores, along with lower platelet counts and oxygen saturation levels. As LAR quartiles increased, the anion gap widened, and pH values decreased, indicating a higher prevalence of metabolic dysregulation in these patients.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 1. LAR quartile baseline table\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"118%\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 19.3878%;\"\u003e\n \u003cp\u003e\u003cstrong\u003echaracteristics\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eLAR Overall\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eLAR Quartile 1\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eLAR Quartile 2\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eLAR Quartile 3\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 17.3469%;\"\u003e\n \u003cp\u003e\u003cstrong\u003eLAR Quartile 4\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 19.3878%;\"\u003e\n \u003cp\u003en\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e4692\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e1173\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e1172\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e1174\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 17.3469%;\"\u003e\n \u003cp\u003e1173\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 19.3878%;\"\u003e\n \u003cp\u003eage (mean (SD))\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e68.32 (15.46)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e70.40 (14.40)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e70.28 (14.59)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e68.05 (15.76)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 17.3469%;\"\u003e\n \u003cp\u003e64.55 (16.27)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 19.3878%;\"\u003e\n \u003cp\u003eabps (mean (SD))\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e119.10 (30.13)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e122.12 (28.23)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e121.05 (28.47)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e120.30 (35.48)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 17.3469%;\"\u003e\n \u003cp\u003e112.92 (26.73)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 19.3878%;\"\u003e\n \u003cp\u003ehr (mean (SD))\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e93.09 (22.42)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e89.20 (21.69)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e91.58 (21.93)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e95.52 (21.82)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 17.3469%;\"\u003e\n \u003cp\u003e96.04 (23.51)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 19.3878%;\"\u003e\n \u003cp\u003err (mean (SD))\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e20.85 (6.66)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e20.03 (6.36)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e20.52 (5.74)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e21.36 (6.57)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 17.3469%;\"\u003e\n \u003cp\u003e21.48 (7.72)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 19.3878%;\"\u003e\n \u003cp\u003ewbc (mean (SD))\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e14.58 (11.29)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e12.12 (6.92)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e14.24 (9.45)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e15.27 (11.60)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 17.3469%;\"\u003e\n \u003cp\u003e16.69 (15.09)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 19.3878%;\"\u003e\n \u003cp\u003erbc (mean (SD))\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e3.54 (0.85)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e3.47 (0.79)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e3.51 (0.83)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e3.56 (0.85)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 17.3469%;\"\u003e\n \u003cp\u003e3.61 (0.90)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 19.3878%;\"\u003e\n \u003cp\u003eplatele (mean (SD))\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e202.20 (116.46)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e209.04 (110.86)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e209.03 (123.11)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e201.20 (117.83)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 17.3469%;\"\u003e\n \u003cp\u003e189.54 (112.69)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 19.3878%;\"\u003e\n \u003cp\u003esodium (mean (SD))\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e138.23 (6.45)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e137.79 (6.11)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e138.36 (6.35)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e138.65 (6.83)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 17.3469%;\"\u003e\n \u003cp\u003e138.10 (6.48)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 19.3878%;\"\u003e\n \u003cp\u003epotassium (mean (SD))\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e4.38 (0.89)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e4.31 (0.82)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e4.31 (0.83)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e4.31 (0.85)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 17.3469%;\"\u003e\n \u003cp\u003e4.57 (1.04)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 19.3878%;\"\u003e\n \u003cp\u003eanion_gap (median [IQR])\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e16.00 [13.00, 19.00]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e15.00 [13.00, 18.00]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e15.00 [13.00, 18.00]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e16.00 [13.00, 19.00]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 17.3469%;\"\u003e\n \u003cp\u003e17.00 [14.00, 21.00]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 19.3878%;\"\u003e\n \u003cp\u003eph (median [IQR])\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e7.35 [7.27, 7.41]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e7.36 [7.30, 7.42]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e7.36 [7.29, 7.42]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e7.35 [7.27, 7.41]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 17.3469%;\"\u003e\n \u003cp\u003e7.31 [7.22, 7.39]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 19.3878%;\"\u003e\n \u003cp\u003epco2 (median [IQR])\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e41.00 [34.00, 48.00]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e41.00 [35.00, 48.00]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e41.00 [34.00, 48.00]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e40.50 [34.00, 48.00]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 17.3469%;\"\u003e\n \u003cp\u003e40.00 [34.00, 48.00]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 19.3878%;\"\u003e\n \u003cp\u003elactate (median [IQR])\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e1.90 [1.30, 3.20]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e1.60 [1.10, 2.40]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e1.70 [1.30, 2.80]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e2.00 [1.40, 3.30]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 17.3469%;\"\u003e\n \u003cp\u003e2.70 [1.60, 4.90]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 19.3878%;\"\u003e\n \u003cp\u003espo2 (mean (SD))\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e96.25 (4.88)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e96.58 (4.25)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e96.54 (4.36)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e96.09 (4.72)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 17.3469%;\"\u003e\n \u003cp\u003e95.81 (5.97)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 19.3878%;\"\u003e\n \u003cp\u003einrpt (median [IQR])\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e1.30 [1.20, 1.70]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e1.30 [1.10, 1.60]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e1.30 [1.20, 1.60]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e1.30 [1.20, 1.70]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 17.3469%;\"\u003e\n \u003cp\u003e1.50 [1.20, 1.90]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 19.3878%;\"\u003e\n \u003cp\u003east (median [IQR])\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e59.00 [30.00, 162.00]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e30.00 [20.00, 49.00]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e47.00 [27.00, 84.00]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e73.00 [39.00, 155.00]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 17.3469%;\"\u003e\n \u003cp\u003e279.00 [90.00, 904.00]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 19.3878%;\"\u003e\n \u003cp\u003eglucose (mean (SD))\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e168.94 (96.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e160.45 (93.74)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e163.97 (90.14)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e169.41 (91.20)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 17.3469%;\"\u003e\n \u003cp\u003e181.93 (106.73)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 19.3878%;\"\u003e\n \u003cp\u003esofa (mean (SD))\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e7.55 (3.82)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e6.58 (3.46)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e7.02 (3.50)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e7.59 (3.70)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 17.3469%;\"\u003e\n \u003cp\u003e8.99 (4.13)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 19.3878%;\"\u003e\n \u003cp\u003egender = M (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e2807 (59.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e722 (61.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e682 (58.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e698 (59.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 17.3469%;\"\u003e\n \u003cp\u003e705 (60.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 19.3878%;\"\u003e\n \u003cp\u003et1dm = 1 (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e50 (1.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e13 (1.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e9 (0.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e12 (1.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 17.3469%;\"\u003e\n \u003cp\u003e16 (1.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 19.3878%;\"\u003e\n \u003cp\u003et2dm = 1 (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e1543 (32.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e455 (38.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e405 (34.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e360 (30.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 17.3469%;\"\u003e\n \u003cp\u003e323 (27.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 19.3878%;\"\u003e\n \u003cp\u003eckd = 1 (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e969 (20.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e268 (22.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e245 (20.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e228 (19.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 17.3469%;\"\u003e\n \u003cp\u003e228 (19.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 19.3878%;\"\u003e\n \u003cp\u003ehtn = 1 (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e2265 (48.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e603 (51.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e607 (51.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e576 (49.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 17.3469%;\"\u003e\n \u003cp\u003e479 (40.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 19.3878%;\"\u003e\n \u003cp\u003ehld = 1 (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e1604 (34.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e448 (38.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 15.3061%;\"\u003e\n \u003cp\u003e433 (36.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16.3265%;\"\u003e\n \u003cp\u003e384 (32.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 17.3469%;\"\u003e\n \u003cp\u003e339 (28.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.2 Clinical outcomes\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe 28-day ICU mortality rate among SIMI patients was higher in groups with elevated LAR levels (\u003cstrong\u003eFigure 2\u003c/strong\u003e). In the Cox regression analysis, using the Q1 group as the reference, the ICU mortality risk in the Q3 and Q4 groups significantly increased in both Model I and Model II. The hazard ratios (HR) for the Q3 and Q4 groups were 1.638 (95% CI: 1.404-1.911,\u003cem\u003e\u0026nbsp;p\u003c/em\u003e \u0026lt; 0.001) and 2.273 (95% CI: 1.953-2.644, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.001), respectively (\u003cstrong\u003eTable 2, Figure 2\u003c/strong\u003e). Kaplan\u0026ndash;Meier analysis demonstrated a clear and significant association between LAR quartiles and ICU survival, with patients in the fourth quartile (Q4) showing a significantly increased risk of mortality. The Log-rank test revealed a p-value \u0026lt; 0.001, indicating a significant difference between the survival curves of the different quartiles(\u003cstrong\u003eFigure 3\u003c/strong\u003e).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 2. 28-day ICU mortality\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"516\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 18.4109%;\"\u003e\n \u003cp\u003eLAR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 20.7364%;\"\u003e\n \u003cp\u003eUnadjusted HR (95% CI)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 20.7364%;\"\u003e\n \u003cp\u003eUnadjusted P value\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 20.155%;\"\u003e\n \u003cp\u003eAdjusted HR\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e(95% CI)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 19.9612%;\"\u003e\n \u003cp\u003eAdjusted P value\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 18.4109%;\"\u003e\n \u003cp\u003eQ1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 20.7364%;\"\u003e\n \u003cp\u003eReference\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 20.7364%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 20.155%;\"\u003e\n \u003cp\u003eReference\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 19.9612%;\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 18.4109%;\"\u003e\n \u003cp\u003eQ2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 20.7364%;\"\u003e\n \u003cp\u003e1.192 (1.013-1.403)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 20.7364%;\"\u003e\n \u003cp\u003e\u0026lt;0.05\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 20.155%;\"\u003e\n \u003cp\u003e1.150 (0.977-1.354)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 19.9612%;\"\u003e\n \u003cp\u003e0.0938\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 18.4109%;\"\u003e\n \u003cp\u003eQ3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 20.7364%;\"\u003e\n \u003cp\u003e1.708 (1.466-1.991)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 20.7364%;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 20.155%;\"\u003e\n \u003cp\u003e1.638 (1.404-1.911)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 19.9612%;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 18.4109%;\"\u003e\n \u003cp\u003eQ4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 20.7364%;\"\u003e\n \u003cp\u003e2.471 (2.135-2.860)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 20.7364%;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 20.155%;\"\u003e\n \u003cp\u003e2.273 (1.953-2.644)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 19.9612%;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.3 Restricted cubic spline\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe RCS analysis was conducted using the rcssci package, adjusting for the effects of age, weight, heart rate, systolic blood pressure, respiratory rate, sex, and SOFA score. The RCS analysis of 28-day ICU mortality (\u003cstrong\u003eFigure 4A\u003c/strong\u003e) revealed a U-shaped association between LAR (log2) and the risk of death. The turning point of the RCS curve was approximately at LAR（log2）\u0026nbsp;= 4.297 (\u003cstrong\u003eFigure 4B\u003c/strong\u003e). According to the Cox proportional hazards regression model, the turning point was located in Quartile 1, where the risk of death was minimal, consistent with the RCS results.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.4 Subgroup analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAs shown in \u003cstrong\u003eFigure 5\u003c/strong\u003e, the 28-day ICU mortality subgroup analysis is presented. We adjusted for age, sex, and comorbidities. In the subgroup analysis, it was found that in the subgroups of age \u0026gt;65, age \u0026lt;65, male, female, and diabetic status, groups with higher LAR levels consistently exhibited higher mortality rates. Regardless of whether confounding variables were adjusted, the high LAR group was significantly associated with higher mortality risk across all subgroups. This highlights the potential of LAR as an independent risk predictor. In the unadjusted model, the hazard ratio (HR) for LAR_groupQ4 was particularly significant in the age \u0026ge;65 group and the diabetes group, suggesting that elderly and diabetic patients are more sensitive to high LAR levels. Regarding sex, the risk associated with LAR_groupQ4 was slightly higher in males compared to females. After adjustment in the model, the significance of the risk associated with LAR_groupQ4 remained across all subgroups, particularly in the elderly and diabetic groups, further confirming the independent association between LAR and mortality risk. Following adjustment, the influence of sex on LAR was reduced, suggesting that the observed gender differences were likely driven by confounding variables (HR \u0026gt; 1,\u003cem\u003e\u0026nbsp;P\u003c/em\u003e \u0026lt; 0.001 in each subgroup).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.5 Boruta Algorithm\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFigure 5\u003c/strong\u003e illustrates the features selected by the Boruta algorithm. Variables in the red region were identified as important features, including LAR, age, SOFA score, lactate, and others. Figure S1 provides the scoring details of each variable in the Boruta analysis.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e3.6 Model Performance\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFigure 6\u003c/strong\u003e shows the ROC curves of various models, with AUC values indicating model performance. The AUC for the RSF model was 0.7745, for XGBoost was 0.7176, and for the decision tree model was 0.6863, indicating that the RSF model performed the best in predicting mortality risk. Higher LAR levels were associated with increased mortality risk (HR \u0026gt; 1, \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05). The machine learning models incorporating LAR demonstrated good predictive performance. DCA confirmed the net benefit of incorporating LAR into mortality prediction, particularly in the high-risk quartiles (Figure S2). Figure 6 demonstrates the performance of LAR compared to other continuous laboratory indicators, with LAR showing a higher AUC value than the others.\u003c/p\u003e"},{"header":"4. Discussion","content":"\u003cp\u003eThis retrospective cohort study of 4,692 ICU patients with SIMI confirms that the lactate dehydrogenase-to-albumin ratio (LAR) is a reliable predictor of 28-day mortality. This is supported by a strong, dose-response relationship, where the highest LAR quartile was associated with a greater than two-fold increase in mortality risk (adjusted HR: 2.273). After controlling for confounding factors, the results remained consistent in the subgroup analysis, demonstrating the robustness of our findings. To date, this study represents the first exploration of the relationship between LAR and adverse outcomes in patients with SIMI\u003csup\u003e22\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eWhen compared with existing literature, our findings both confirm and extend previous knowledge. Patients with sepsis-induced myocardial injury (SIMI) exhibit a high degree of complexity\u003csup\u003e23\u003c/sup\u003e，Current research has paid limited attention to the exploration of metabolic conditions in SIMI patients\u003csup\u003e22\u003c/sup\u003e. Recent studies have shown that metabolic reprogramming plays an important role in cardiac repair. Previous research has also observed a significant shift in myocardial energy metabolism from oxidative phosphorylation to glycolysis during the early stages of sepsis\u003csup\u003e14, 24\u003c/sup\u003e. Immune cells adapt to energy demands by increasing glycolysis\u003csup\u003e25, 26\u003c/sup\u003e. The body responds rapidly to infection by activating immune cells\u003csup\u003e4, 14\u003c/sup\u003e. Moreover, emerging evidence highlights that metabolic recycling of substrates, such as the utilization of phagocytosed bacteria by macrophages, is not merely a passive fuel-gathering process but actively shapes the immune response by suppressing pro-inflammatory pathways and enhancing anti-oxidant defenses.And failure to restore OXPHOS in the later stages of sepsis may lead to sustained pro-inflammatory states and myocardial injury\u003csup\u003e14, 23, 24\u003c/sup\u003e. Research on metabolic transition disorders in patients with SIMI should be given more attention\u003csup\u003e27\u003c/sup\u003e.Notably, while the prognostic value of LDH and albumin individually is well-established, our discovery of a U-shaped relationship between their ratio (LAR) and mortality reveals a more complex, non-linear interaction that had not been previously reported in SIMI.This U-shaped association delineates a bimodal risk profile: the elevated arm signifies severe metabolic stress and cellular injury consequent to failed metabolic reprogramming, while the depressed arm reflects profound inflammatory/nutritional depletion and impaired synthetic capacity. This pattern confirms that both extreme metabolic stress and severe inflammatory/nutritional depletion are detrimental.\u003c/p\u003e\n\u003cp\u003eA key advancement of our study lies in the direct comparison of predictive performance. The Random Survival Forest model incorporating LAR achieved an AUC of 0.775, significantly outperforming the SOFA score alone (AUC = 0.698). This demonstrates that LAR provides prognostic information superior to conventional severity scores and supports its potential for improving risk stratification.\u003c/p\u003e\n\u003cp\u003eThe clinical relevance of LAR is underscored by its ability to identify distinct patient phenotypes.LDH is an essential enzyme involved in energy metabolism, and its prognostic value in sepsis patients has been well established\u003csup\u003e28\u003c/sup\u003e. Albumin reflects the inflammatory and nutritional status of sepsis patients\u003csup\u003e29\u003c/sup\u003e. The association between LAR and mortality in SIMI may be attributed to its reflection of the balance between metabolic and inflammatory stress in SIMI patients. Incorporating LAR into predictive models provides clinical value, as confirmed by DCA and calibration analyses. The U-shaped curve implies that LAR captures two distinct high-risk patient phenotypes, which could have important implications for personalized treatment strategies.\u003c/p\u003e\n\u003cp\u003eIt is important to acknowledge the limitations of this work. This study has several limitations. First, as a retrospective study, information bias may be present. Second, it is difficult to fully control for all potential confounders, which limits causal inference. Third, sample selection may be influenced by known or unknown factors, making the sample less representative and potentially affecting the study\u0026apos;s external validity. Despite these limitations, our findings robustly establish LAR as a novel prognostic marker in SIMI.Future research should focus on the external validation of our findings in prospective, multi-center cohorts.\u0026nbsp;\u003c/p\u003e"},{"header":"5. Conclusion","content":"\u003cp\u003eIn conclusion, LAR is an independent predictor of 28-day mortality in sepsis patients, showing a U-shaped association with mortality. Incorporating LAR into predictive models can improve risk stratification for mortality in sepsis-induced myocardial injury (SIMI) patients, thereby guiding treatment decisions.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAuthor statements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll authors have seen and approved the final version of the manuscript being submitted. They warrant that the article is the authors\u0026apos; original work, has not been published previously, and is not under consideration for publication elsewhere.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll authors have read and approved the final manuscript and give their consent for its publication in .BMC Infectious Diseases\u003c/p\u003e\n\u003cp\u003eNixiang Jiang\u003c/p\u003e\n\u003cp\u003eXiangya School of Nursing, Central South University, Changsha 410013, China\u003c/p\u003e\n\u003cp\u003eEmail: [email protected]\u003c/p\u003e\n\u003cp\u003eJinwei Dai\u003c/p\u003e\n\u003cp\u003eDepartment of Critical Care Medicine, Xiangya Hospital, Central South University, Changsha, Hunan, 410008, China\u003c/p\u003e\n\u003cp\u003eNational Clinical Research Center of Geriatric Disorders, Xiangya Hospital, Central South University, Changsha, China\u003c/p\u003e\n\u003cp\u003eEmail: [email protected]\u003c/p\u003e\n\u003cp\u003eYang Zhou (Corresponding Author)\u003c/p\u003e\n\u003cp\u003eNational Clinical Research Center of Geriatric Disorders, Xiangya Hospital, Central South University, Changsha, China\u003c/p\u003e\n\u003cp\u003eDepartment of Clinical Nursing, Xiangya Hospital, Central South University, Changsha 410005, China\u003c/p\u003e\n\u003cp\u003eEmail: [email protected]\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eYang Zhou: Study conception and design.\u003c/p\u003e\n\u003cp\u003eNixiang Jiang: Literature research, data extraction, and quality control of data and algorithms.\u003c/p\u003e\n\u003cp\u003eJinwei Dai: Statistical analysis and data interpretation.\u003c/p\u003e\n\u003cp\u003eAll authors: Manuscript writing, review, and approval.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work was supported by the National Natural Science Foundation of China (82360448); the Natural Science Foundation of Hunan Province (2022JJ30503); and the Scientific Research Project of Health Commission of Hunan Province (2022030128723001).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and material\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003ePublicly available datasets were analyzed in this study. These data can be found at https://mimic.mit.edu/.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDeclaration of competing interest\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that there are no conflicts of interest that could be construed as influencing the work reported in this paper.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics declarations\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eEthics approval and consent to participate\u003c/p\u003e\n\u003cp\u003eThis study utilized data from the MIMIC-IV database (version 3.1). The MIMIC-IV database is a publicly available, de-identified critical care database. Its creation was approved by the Institutional Review Boards (IRB) of the Massachusetts Institute of Technology (MIT) and Beth Israel Deaconess Medical Center (BIDMC). The requirement for individual patient consent was waived by these IRBs as the data is de-identified. All methods were carried out in accordance with the Declaration of Helsinki. Clinical trial number:Not applicable. Human Ethics and Consent to Participate declarations: Not applicable.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eSinger, M. et al. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). \u003cem\u003eJama\u003c/em\u003e \u003cb\u003e315\u003c/b\u003e (8), 801\u0026ndash;810. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1001/jama.2016.0287\u003c/span\u003e\u003cspan address=\"10.1001/jama.2016.0287\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2016). From NLM.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSu, Z. D. et al. Melatonin alleviates lipopolysaccharide-induced myocardial injury by inhibiting inflammation and pyroptosis in cardiomyocytes. \u003cem\u003eAnn. Transl Med.\u003c/em\u003e \u003cb\u003e9\u003c/b\u003e (5), 413. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.21037/atm-20-8196\u003c/span\u003e\u003cspan address=\"10.21037/atm-20-8196\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021). From NLM.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWalley, K. R. Sepsis-induced myocardial dysfunction. \u003cem\u003eCurr. Opin. Crit. Care\u003c/em\u003e. \u003cb\u003e24\u003c/b\u003e (4), 292\u0026ndash;299. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1097/mcc.0000000000000507\u003c/span\u003e\u003cspan address=\"10.1097/mcc.0000000000000507\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2018). From NLM.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHollenberg, S. M. \u0026amp; Singer, M. Pathophysiology of sepsis-induced cardiomyopathy. \u003cem\u003eNat. Rev. Cardiol.\u003c/em\u003e \u003cb\u003e18\u003c/b\u003e (6), 424\u0026ndash;434. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41569-020-00492-2\u003c/span\u003e\u003cspan address=\"10.1038/s41569-020-00492-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021). From NLM.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBeesley, S. J. et al. Septic Cardiomyopathy. \u003cem\u003eCrit. Care Med.\u003c/em\u003e \u003cb\u003e46\u003c/b\u003e (4), 625\u0026ndash;634. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1097/ccm.0000000000002851\u003c/span\u003e\u003cspan address=\"10.1097/ccm.0000000000002851\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2018). From NLM.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKottmann, R. M. et al. Lactic acid is elevated in idiopathic pulmonary fibrosis and induces myofibroblast differentiation via pH-dependent activation of transforming growth factor-β. \u003cem\u003eAm. J. Respir Crit. Care Med.\u003c/em\u003e \u003cb\u003e186\u003c/b\u003e (8), 740\u0026ndash;751. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1164/rccm.201201-0084OC\u003c/span\u003e\u003cspan address=\"10.1164/rccm.201201-0084OC\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2012). From NLM.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGarrabou, G. et al. The effects of sepsis on mitochondria. \u003cem\u003eJ. Infect. Dis.\u003c/em\u003e \u003cb\u003e205\u003c/b\u003e (3), 392\u0026ndash;400. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/infdis/jir764\u003c/span\u003e\u003cspan address=\"10.1093/infdis/jir764\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2012). From NLM.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu, J., Zhou, G., Wang, X. \u0026amp; Liu, D. Metabolic reprogramming consequences of sepsis: adaptations and contradictions. \u003cem\u003eCell. Mol. Life Sci.\u003c/em\u003e \u003cb\u003e79\u003c/b\u003e (8), 456. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s00018-022-04490-0\u003c/span\u003e\u003cspan address=\"10.1007/s00018-022-04490-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2022). From NLM.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePreau, S. et al. Energetic dysfunction in sepsis: a narrative review. \u003cem\u003eAnn. Intensive Care\u003c/em\u003e. \u003cb\u003e11\u003c/b\u003e (1), 104. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s13613-021-00893-7\u003c/span\u003e\u003cspan address=\"10.1186/s13613-021-00893-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021). From NLM.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYang, H. \u0026amp; Zhang, Z. Sepsis-induced myocardial dysfunction: the role of mitochondrial dysfunction. \u003cem\u003eInflamm. Res.\u003c/em\u003e \u003cb\u003e70\u003c/b\u003e (4), 379\u0026ndash;387. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s00011-021-01447-0\u003c/span\u003e\u003cspan address=\"10.1007/s00011-021-01447-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021). From NLM.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFeng, Y. et al. Lactate dehydrogenase A: A key player in carcinogenesis and potential target in cancer therapy. \u003cem\u003eCancer Med.\u003c/em\u003e \u003cb\u003e7\u003c/b\u003e (12), 6124\u0026ndash;6136. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/cam4.1820\u003c/span\u003e\u003cspan address=\"10.1002/cam4.1820\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2018). From NLM.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVincent, J. L., De Backer, D. \u0026amp; Wiedermann, C. J. Fluid management in sepsis: The potential beneficial effects of albumin. \u003cem\u003eJ. Crit. Care\u003c/em\u003e. \u003cb\u003e35\u003c/b\u003e, 161\u0026ndash;167. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.jcrc.2016.04.019\u003c/span\u003e\u003cspan address=\"10.1016/j.jcrc.2016.04.019\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2016). From NLM.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiang, M. et al. The association between lactate dehydrogenase to serum albumin ratio and the 28-day mortality in patients with sepsis-associated acute kidney injury in intensive care: a retrospective cohort study. \u003cem\u003eRen. Fail.\u003c/em\u003e \u003cb\u003e45\u003c/b\u003e (1), 2212080. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1080/0886022x.2023.2212080\u003c/span\u003e\u003cspan address=\"10.1080/0886022x.2023.2212080\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023). From NLM.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi, X. et al. Inhibition of fatty acid oxidation enables heart regeneration in adult mice. \u003cem\u003eNature\u003c/em\u003e \u003cb\u003e622\u003c/b\u003e (7983), 619\u0026ndash;626. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41586-023-06585-5\u003c/span\u003e\u003cspan address=\"10.1038/s41586-023-06585-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023). From NLM.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen, Y. et al. LDHA-mediated metabolic reprogramming promoted cardiomyocyte proliferation by alleviating ROS and inducing M2 macrophage polarization. \u003cem\u003eRedox Biol.\u003c/em\u003e \u003cb\u003e56\u003c/b\u003e, 102446. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.redox.2022.102446\u003c/span\u003e\u003cspan address=\"10.1016/j.redox.2022.102446\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2022). From NLM.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDong, Y. et al. Aspirin is associated with improved outcomes in patients with sepsis-induced myocardial injury: An analysis of the MIMIC-IV database. \u003cem\u003eJ. Clin. Anesth.\u003c/em\u003e \u003cb\u003e99\u003c/b\u003e, 111597. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.jclinane.2024.111597\u003c/span\u003e\u003cspan address=\"10.1016/j.jclinane.2024.111597\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024). From NLM.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLee, B. K. et al. Lactate dehydrogenase to albumin ratio as a prognostic factor in lower respiratory tract infection patients. \u003cem\u003eAm. J. Emerg. Med.\u003c/em\u003e \u003cb\u003e52\u003c/b\u003e, 54\u0026ndash;58. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.ajem.2021.11.028\u003c/span\u003e\u003cspan address=\"10.1016/j.ajem.2021.11.028\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2022). From NLM.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVallabhajosyula, S. et al. Role of Admission Troponin-T and Serial Troponin-T Testing in Predicting Outcomes in Severe Sepsis and Septic Shock. \u003cem\u003eJ. Am. Heart Assoc.\u003c/em\u003e \u003cb\u003e6\u003c/b\u003e (9). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1161/jaha.117.005930\u003c/span\u003e\u003cspan address=\"10.1161/jaha.117.005930\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2017). From NLM.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXie, R., Chen, Q., He, W. \u0026amp; Zeng, M. Association of Cardiac Troponin T Concentration on Admission with Prognosis in Critically Ill Patients without Myocardial Infarction: A Cohort Study. \u003cem\u003eInt. J. Gen. Med.\u003c/em\u003e \u003cb\u003e14\u003c/b\u003e, 2729\u0026ndash;2739. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.2147/ijgm.S318232\u003c/span\u003e\u003cspan address=\"10.2147/ijgm.S318232\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021). From NLM.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKalimouttou, A., Lerner, I., Cheurfa, C., Jannot, A. S. \u0026amp; Pirracchio, R. Machine-learning-derived sepsis bundle of care. \u003cem\u003eIntensive Care Med.\u003c/em\u003e \u003cb\u003e49\u003c/b\u003e (1), 26\u0026ndash;36. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s00134-022-06928-2\u003c/span\u003e\u003cspan address=\"10.1007/s00134-022-06928-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023). From NLM.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVenugopalan, J., Chanani, N., Maher, K. \u0026amp; Wang, M. D. Novel Data Imputation for Multiple Types of Missing Data in Intensive Care Units. \u003cem\u003eIEEE J. Biomed. Health Inf.\u003c/em\u003e \u003cb\u003e23\u003c/b\u003e (3), 1243\u0026ndash;1250. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1109/jbhi.2018.2883606\u003c/span\u003e\u003cspan address=\"10.1109/jbhi.2018.2883606\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2019). From NLM.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTao, T. et al. The top cited clinical research articles on sepsis: a bibliometric analysis. \u003cem\u003eCrit Care 16\u003c/em\u003e (3), R110. DOI: 10.1186/cc11401 From NLM. (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYao, R. Q. et al. Publication Trends of Research on Sepsis and Host Immune Response during 1999\u0026ndash;2019: A 20-year Bibliometric Analysis. \u003cem\u003eInt. J. Biol. Sci.\u003c/em\u003e \u003cb\u003e16\u003c/b\u003e (1), 27\u0026ndash;37. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.7150/ijbs.37496\u003c/span\u003e\u003cspan address=\"10.7150/ijbs.37496\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020). From NLM.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBauerfeld, C., Talwar, H., Zhang, K., Liu, Y. \u0026amp; Samavati, L. MKP-1 Modulates Mitochondrial Transcription Factors, Oxidative Phosphorylation, and Glycolysis. \u003cem\u003eImmunohorizons\u003c/em\u003e \u003cb\u003e4\u003c/b\u003e (5), 245\u0026ndash;258. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.4049/immunohorizons.2000015\u003c/span\u003e\u003cspan address=\"10.4049/immunohorizons.2000015\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020). From NLM.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZuo, W., Sun, R., Ji, Z. \u0026amp; Ma, G. Macrophage-driven cardiac inflammation and healing: insights from homeostasis and myocardial infarction. \u003cem\u003eCell. Mol. Biol. Lett.\u003c/em\u003e \u003cb\u003e28\u003c/b\u003e (1), 81. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s11658-023-00491-4\u003c/span\u003e\u003cspan address=\"10.1186/s11658-023-00491-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023). From NLM.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGharavi, A. T., Hanjani, N. A., Movahed, E. \u0026amp; Doroudian, M. The role of macrophage subtypes and exosomes in immunomodulation. \u003cem\u003eCell. Mol. Biol. Lett.\u003c/em\u003e \u003cb\u003e27\u003c/b\u003e (1), 83. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s11658-022-00384-y\u003c/span\u003e\u003cspan address=\"10.1186/s11658-022-00384-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2022). From NLM.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYu, Q. et al. The role of histone deacetylases in cardiac energy metabolism in heart diseases. \u003cem\u003eMetabolism\u003c/em\u003e \u003cb\u003e142\u003c/b\u003e, 155532. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.metabol.2023.155532\u003c/span\u003e\u003cspan address=\"10.1016/j.metabol.2023.155532\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023). From NLM.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDuman, A. et al. Prognostic value of neglected biomarker in sepsis patients with the old and new criteria: predictive role of lactate dehydrogenase. \u003cem\u003eAm. J. Emerg. Med.\u003c/em\u003e \u003cb\u003e34\u003c/b\u003e (11), 2167\u0026ndash;2171. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.ajem.2016.06.012\u003c/span\u003e\u003cspan address=\"10.1016/j.ajem.2016.06.012\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2016). From NLM.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTakegawa, R. et al. Serum albumin as a risk factor for death in patients with prolonged sepsis: An observational study. \u003cem\u003eJ. Crit. Care\u003c/em\u003e. \u003cb\u003e51\u003c/b\u003e, 139\u0026ndash;144. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.jcrc.2019.02.004\u003c/span\u003e\u003cspan address=\"10.1016/j.jcrc.2019.02.004\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2019). From NLM.\u003c/span\u003e\u003c/li\u003e \u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Lactate Dehydrogenase-to-Albumin Ratio, Sepsis, Machine Learning, Mortality Prediction, Boruta","lastPublishedDoi":"10.21203/rs.3.rs-8189180/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8189180/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eBackground\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSepsis-induced myocardial injury (SIMI), a fatal complication seen in 13%-65% of septic patients with mortality exceeding 50% in severe cases, is characterized by profound metabolic and inflammatory dysregulation.While the Sequential Organ Failure Assessment (SOFA) score is commonly used to assess disease severity in these patients, there remains a need for biomarkers that more directly reflect the metabolic and inflammatory components of SIMI.The lactate dehydrogenase-to-albumin ratio (LAR), which integrates markers of cellular damage and systemic inflammation, has shown promise in sepsis but its prognostic utility specifically in SIMI remains unexplored.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMethod \u003c/strong\u003eThis retrospective cohort study analyzed data from the MIMIC-IV database involving 4,692 ICU patients with sepsis-induced myocardial injury (SIMI). The primary endpoint was 28-day all-cause mortality. The association was assessed using Cox regression, with restricted cubic splines examining nonlinearity. Subgroup analyses were performed by age, sex, and diabetes. Machine learning models (random forest, XGBoost, decision tree) were developed to validate LAR's predictive value, with performance evaluated by receiver operating characteristic (ROC) curves and decision curve analysis.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults\u003c/strong\u003e\u003cbr\u003e\n The Random Survival Forest (RSF) model incorporating LAR achieved an AUC of 0.775 (95% CI: 0.752–0.803) for predicting 28-day ICU mortality, outperforming the SOFA score alone (AUC = 0.698, \u003cem\u003eP \u003c/em\u003e\u0026lt; 0.001).The prognostic value of LAR remained consistent across key subgroups, including age, sex, and diabetic status (all interaction \u003cem\u003eP\u003c/em\u003e\u0026gt; 0.05), underscoring its robustness as a risk stratifier. Furthermore, decision curve analysis confirmed the clinical utility of the model, demonstrating a superior net benefit across a wide range of risk thresholds.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusion\u003c/strong\u003e\u003cbr\u003e\n The lactate dehydrogenase-to-albumin ratio (LAR) is an independent predictor of 28-day mortality in sepsis-induced myocardial injury. Incorporating LAR into clinical risk stratification could improve early identification of high-risk patients.\u003c/p\u003e","manuscriptTitle":"The Lactate Dehydrogenase-to-Albumin Ratio (LAR) Predicts Mortality in Sepsis-Induced Myocardial Injury","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-12-16 10:35:33","doi":"10.21203/rs.3.rs-8189180/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"93d5b845-8814-4921-96c3-0c7868d62a29","owner":[],"postedDate":"December 16th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":59711825,"name":"Health sciences/Biomarkers"},{"id":59711826,"name":"Health sciences/Cardiology"},{"id":59711827,"name":"Health sciences/Diseases"},{"id":59711828,"name":"Health sciences/Medical research"},{"id":59711829,"name":"Health sciences/Risk factors"}],"tags":[],"updatedAt":"2026-04-15T16:41:20+00:00","versionOfRecord":[],"versionCreatedAt":"2025-12-16 10:35:33","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8189180","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8189180","identity":"rs-8189180","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}