Prognostic Value of Lactate-to-Albumin Ratio in Patients with Asthma in the Intensive Care Unit: Development and External Validation of a Predictive Model

Prognostic Value of Lactate-to-Albumin Ratio in Patients with Asthma in the Intensive Care Unit: Development and External Validation of a Predictive Model | 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 Prognostic Value of Lactate-to-Albumin Ratio in Patients with Asthma in the Intensive Care Unit: Development and External Validation of a Predictive Model Chenxi Wang, Qin Chen, Yajing Li, Li Zhang This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9082854/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 7 You are reading this latest preprint version Abstract Background The lactate-to-albumin ratio (LAR), a composite biomarker integrating metabolic stress and systemic inflammation, has demonstrated significant prognostic value across various critical illnesses. However, its clinical significance in patients with life-threatening asthma remains poorly defined. This study aimed to elucidate the association between LAR and mortality risk in asthma patients admitted to the ICU and to develop and validate a LAR-based individualized predictive model. Methods We retrospectively analyzed 1,038 adult asthma patients from the MIMIC-IV (v3.1) database (2008–2022) and 467 patients from the eICU-CRD for external validation. The lactate-to-albumin ratio (LAR) was calculated using admission laboratory values, with patients stratified by the median LAR (0.5385). Multivariable Cox regression, restricted cubic spline (RCS) analysis, and Kaplan-Meier curves were employed to assess mortality risks. Subgroup analyses were performed to ensure robustness, adjusting for demographics, comorbidities, and clinical severity scores. Results In the discovery cohort, 28-day and 60-day mortality rates were 17.5% and 20.9%, respectively, with significantly higher rates observed in the high LAR group ( P < 0.001). After multivariable adjustment, each 1-unit increase in LAR was independently associated with 28-day (HR = 1.22, P = 0.002) and 60-day (HR = 1.16, P < 0.001) mortality. These findings were highly consistent in the eICU-CRD validation cohort (OR = 1.454, P = 0.012). Restricted cubic spline (RCS) analysis confirmed a significant linear dose-response relationship between LAR and 28-day mortality. A gender interaction was identified for 28-day mortality ( P interaction = 0.044), showing a stronger association in females. The final predictive model demonstrated excellent stability and performance, with an AUC of 0.809 in the external validation set, closely matching the discovery set (AUC 0.815). Conclusion Elevated LAR at ICU admission is a powerful independent predictor of both short-term and long-term mortality in asthma patients admitted to the ICU, characterized by a robust linear dose-response relationship. As a robust and accessible composite biomarker, LAR effectively integrates metabolic stress with systemic immune dysregulation, such as cytokine amplification and capillary leak. Our cross-center validation confirms that LAR-based risk stratification provides a stable and reliable tool for identifying high-risk individuals, potentially informing the optimal timing for immunomodulatory interventions (e.g., precise anti-inflammatory or biologic therapies) to improve clinical outcomes. Health sciences/Biomarkers Health sciences/Diseases Biological sciences/Immunology Health sciences/Medical research Health sciences/Risk factors lactate-to-albumin ratio (LAR) asthma mortality critically ill patients dose-response relationship prognostic model external validation Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 1 Introduction Asthma is a chronic respiratory condition characterized by airway inflammation, hyperresponsiveness, and reversible obstruction, affecting millions worldwide and posing a significant public health challenge. Driven by complex immunological cascades such as Th2-mediated inflammation, eosinophilic infiltration, and cytokine storms (e.g., IL-4, IL-5, and IL-13), asthma often involves IgE-dependent mast cell activation and allergic responses, exacerbating systemic immune imbalances in severe cases. Globally, the prevalence of asthma has been estimated at approximately 262 million people in 2019, with around 455,000 deaths attributed to the disease annually [ 1 ]. Projections indicate that by 2045, the burden may increase, particularly in low- and middle-income countries where access to care is limited [ 2 ]. In critically ill populations, asthma exacerbations can lead to severe respiratory failure requiring intensive care unit (ICU) admission, with short-term mortality rates ranging from 1–5% overall, but escalating to 8–25% among those needing mechanical ventilation [ 3 ]. Factors such as comorbidities, socioeconomic disparities, and delayed intervention contribute to these outcomes, highlighting the need for early risk stratification to improve survival [ 4 , 5 ]. The lactate-to-albumin ratio (LAR) has emerged as a promising biomarker integrating markers of metabolic stress (lactate) and nutritional and inflammatory reserves(albumin), offering insights into systemic derangements in critically ill patients [ 6 ]. Elevated lactate reflects tissue hypoperfusion and anaerobic metabolism, while hypoalbuminemia indicates inflammation, malnutrition, or liver dysfunction—both of which are independently associated with poor prognosis in various conditions [ 7 ]. Studies have demonstrated LAR's utility as a prognostic marker in sepsis, septic shock, and other ICU scenarios, where higher ratios correlate with increased mortality [ 8 , 9 ]. For instance, in critically ill patients with sepsis, LAR outperforms individual lactate or albumin levels in predicting 28-day mortality [ 10 ]. Similarly, in cardiac arrest and acute pancreatitis cohorts, LAR has shown strong predictive value for short- and long-term outcomes [ 11 , 12 ]. Despite these advancements, a notable gap exists in the literature regarding LAR's role in ICU patients with asthma. While LAR has been validated in broad critical illness populations [ 13 , 14 ], its association with short-term mortality in acute asthma exacerbations requiring intensive care remains underexplored[ 15 , 16 ]. Asthma pathophysiology involves unique elements like bronchospasm and dynamic hyperinflation, which may interact differently with metabolic and inflammatory markers compared to other respiratory failures [ 17 ]. Moreover, existing studies on asthma mortality in the ICU often focus on clinical scores like SOFA or APACHE II, and the precise dose-response relationship between LAR and asthma-specific outcomes remains to be fully characterized.[ 18 , 19 ]. Non-linear associations between biomarkers and mortality have been observed in ICU settings, such as U-shaped or L-shaped patterns in platelet counts or globulin levels with sepsis mortality[ 20 ], characterizing the dose-response relationship is therefore essential for accurate risk assessment in clinical practice. [ 21 , 22 ]. To address this knowledge gap, our study evaluates the association between LAR and short-term (28-day and 60-day) all-cause mortality in asthma patients admitted to the ICU, while characterizing the dose-response relationship. By leveraging a discovery cohort of 1,038 patients from the MIMIC-IV (v3.1) database and an external validation cohort of 467 patients from the eICU-CRD, this research utilizes a dual-center approach to provide asthma-specific insights into LAR's prognostic value, potentially guiding early interventions and personalized management strategies targeted at modulating immunological pathways to mitigate mortality risks. 2 Methods 2.1 Data source This is a retrospective cohort study utilizing two large-scale clinical databases: the Medical Information Mart for Intensive Care IV (MIMIC-IV) and the eICU Collaborative Research Database (eICU-CRD) The discovery cohort was derived from MIMIC-IV version 3.1, which encompasses comprehensive clinical records for 194,458 admissions to the Intensive Care Units (ICUs) of Beth Israel Deaconess Medical Center (Boston, MA, USA) between 2008 and 2022. The database provides detailed longitudinal information, including patients’ hospital stays, vital signs, laboratory measurements, medication treatments, and clinical outcomes. To evaluate the generalizability of our findings, we utilized the eICU-CRD (version 2.0) as an external validation cohort. The eICU-CRD is a multi-center database comprising de-identified data from over 200,000 ICU admissions across 208 hospitals in the United States between 2014 and 2015. It contains high-granularity clinical data similar to MIMIC-IV, facilitating robust cross-center validation. To ensure patient privacy, all personal identifiers in both databases were de-identified and anonymized. Therefore, the Institutional Review Boards (IRBs) of the involved institutions (including Beth Israel Deaconess Medical Center and the Massachusetts Institute of Technology) approved the data use and waived the requirement for written informed consent. All methods were performed in accordance with the relevant guidelines and regulations (e.g., the Declaration of Helsinki). The first author successfully completed the Collaborative Institutional Training Initiative (CITI) program and obtained access to both datasets (Access No. 65378269). Both the MIMIC-IV and eICU-CRD datasets are accessible via the PhysioNet platform (https://physionet.org/). 2.2 Patient selection A standardized screening protocol was applied to both the discovery and validation cohorts to ensure methodological consistency (Figure 1). For the MIMIC-IV discovery cohort, 4,416 adult asthma patients were initially identified. A total of 3,378 patients were excluded based on the following hierarchy: (1) missing LAR data at admission (n = 3,223); (2) severe liver disease (n = 116); (3) ICU length of stay < 24 hours (n = 32); and (4) pregnancy (n = 7). Consequently, 1,038 patients were enrolled and stratified into low-LAR (n = 521) and high-LAR (n = 517) groups. Similarly, in the eICU-CRD validation cohort, 2,200 asthma cases were initially screened. Following the same exclusion criteria, 1,733 patients were excluded, including those lacking admission LAR data (n = 1,654), severe liver disease (n = 56), ICU stay < 24 hours (n = 20), and pregnancy (n = 3). This resulted in a final validation population of 467 patients, with 195 assigned to the low-LAR group and 272 to the high-LAR group based on the median LAR value Variable Extraction and Data Processing Relevant clinical variables were extracted from both the MIMIC-IV (v3.1) and eICU-CRD (v2.0) databases using Structured Query Language (SQL). To ensure methodological consistency and minimize measurement bias, only the first recorded values within the initial 24 hours of the first ICU admission were utilized for analysis. Patient Identification and Selection Criteria Asthma patients were identified based on the International Classification of Diseases (ICD) codes. Specifically, we included adult patients (≥18 years) with the following diagnosis codes: ICD-9: 493.x (including 493.00, 493.01, 493.10, 493.11, 493.20, 493.21, 493.90, 493.91) and ICD-10: J45.x (including J45.0–J45.5, J45.8, J45.9) or J46. To ensure the prognostic integrity of the lactate-to-albumin ratio (LAR), patients with severe liver disease were strictly excluded, as impaired hepatic synthetic function can significantly confound albumin levels and lactate metabolism. Severe liver disease was defined using the following codes: ICD-9: 570, 571.x, 572.x; ICD-10: K70.3, K72.x, K74.x, K76.6–K76.7, R18, and I85.x. Additional exclusions included pregnancy and an ICU length of stay < 24 hours. Data Elements and Predictor Calculation The primary predictor, LAR, was calculated by dividing the admission arterial lactate concentration (mmol/L) by the serum albumin concentration (g/dL). The following covariates were also extracted: Demographics and Vitals: Age, gender, race, heart rate, systolic blood pressure (SBP), diastolic blood pressure (DBP), mean arterial pressure (MAP), respiratory rate, body temperature, and peripheral oxygen saturation (SpO2). Laboratory Parameters: PCO2, PO2, blood urea nitrogen (BUN), creatinine, hemoglobin, and white blood cell (WBC) count. Comorbidities and Severity Scores: Smoking status, congestive heart failure, renal disease, diabetes, coronary heart disease (CHD), Alzheimer's disease (AD), atrial fibrillation (AF), and aortic regurgitation (AR). Illness severity was assessed using the Sequential Organ Failure Assessment (SOFA) and Simplified Acute Physiology Score II (SAPS II). Clinical Interventions: The use of mechanical ventilation and continuous renal replacement therapy (CRRT) during the ICU stay. Cross-Center Validation Strategy The same extraction logic and variable definitions were applied to the eICU-CRD validation cohort. To ensure robust external validation, variables from the multi-center eICU network were meticulously mapped to the MIMIC-IV schema. In instances of multiple measurements within the first 24 hours, only the baseline (first available) value was used to represent the patient’s clinical status at the time of acute decompensation. 2.3 Statistical analysis All statistical analyses were performed using R Statistical Software (Version 4.2.2, The R Foundation) and the Free Statistics Analysis Platform (Version 1.9, Beijing, China). A two-sided P < 0.05 was considered statistically significant. Continuous variables were expressed as mean ± standard deviation (SD) for normally distributed data, while skewed distributions were presented as median (interquartile range [IQR]). Categorical variables were displayed as frequencies and percentages (%). Differences between groups were assessed using the independent samples t-test or Mann-Whitney U-test for continuous variables, and the chi-square test or Fisher’s exact test for categorical data, as appropriate. Survival curves for 28-day and 60-day all-cause mortality were estimated using the Kaplan–Meier method and compared via the log-rank test. To evaluate the independent association between the LAR and mortality, both univariate and multivariate Cox proportional hazards models were utilized to compute hazard ratios (HRs) and 95% confidence intervals (CIs). Three stepwise adjustment models were constructed: Model I adjusted for age and sex. Model II further adjusted for comorbidities, including diabetes mellitus, CHD, AR, AD, and AF. Model III (fully adjusted model) included additional adjustments for smoking status, mechanical ventilation, CRRT, SAPS II, and SOFA score. To explore potential non-linear dose-response relationships between LAR and mortality, we employed a restricted cubic spline (RCS) model with four knots. The median LAR value (0.5385 was selected )as the reference point (HR = 1.0). Subgroup and interaction analyses were conducted to examine the consistency of LAR’s predictive value across various clinical strata, with interactions assessed using the likelihood ratio test. The predictive performance of the final model was validated in the eICU-CRD cohort using the Area Under the Receiver Operating Characteristic Curve (AUC). Regarding missing data (less than 5% of the dataset), list-wise deletion was employed as this missingness proportion has minimal impact on the stability of the model estimates. 3 Results 3.1 Characteristics of participants A total of 1,038 patients from the MIMIC-IV database and 467 patients from the eICU-CRD database were finally enrolled in this study (Figure 1). The baseline characteristics of the discovery cohort (MIMIC-IV) are detailed in Table 1, while those of the validation cohort (eICU-CRD) are presented in Supplementary Table S1. In the MIMIC-IV cohort, the overall 28-day all-cause mortality rate among the enrolled asthma patients was 17.5%. Based on the median LAR value (0.5385), patients were stratified into a low-LAR group (n = 517) and a high-LAR group (n = 521). No significant statistical differences were observed between the two groups regarding age ( P = 0.571), sex ( P = 0.393), or racial distribution ( P = 0.354), indicating well-balanced baseline demographics. However, patients in the high-LAR group exhibited a significantly higher burden of comorbidities, including mild liver disease (15.0% vs. 7.9%, P < 0.001), malignant cancer (14.8% vs. 8.1%, P < 0.001), and acute kidney injury (87.3% vs. 82.4%, P < 0.05). Correspondingly, severity-of-illness scores, including SAPS II and SOFA, were significantly higher in the high-LAR group (both P < 0.001). Hemodynamic and Laboratory Findings Regarding hemodynamic status in the MIMIC-IV cohort, the high-LAR group presented with lower systolic blood pressure (120.5 ± 26.6 vs. 126.6 ± 26.0 mmHg, P < 0.001) and significantly elevated heart and respiratory rates (both P < 0.001), suggesting increased physiological stress and potential tissue hypoperfusion. Laboratory analysis revealed that the high-LAR group had markedly higher lactate levels (3.4 ± 2.2 vs. 1.2 ± 0.3 mmol/L, P < 0.001) and significantly lower albumin levels (2.8 ± 0.7 vs. 3.2 ± 0.6 g/dL, P < 0.001). The requirement for continuous renal replacement therapy (CRRT) was also significantly higher in the high-LAR group (14.6% vs. 6.0%, P < 0.001). Ultimately, mortality rates for both 28-day (23.8% vs. 11.2%) and 60-day (28.2% vs. 13.5%) periods were significantly higher in the high-LAR group (both P < 0.001). External Validation in the eICU-CRD Cohort To ensure the generalizability of our findings, we analyzed an external validation cohort of 467 patients from the eICU-CRD (Supplementary Table S1). Consistent with the discovery cohort, patients in the high-LAR group (n = 272) were characterized by higher SOFA scores (6.8 ± 3.9 vs. 6.1 ± 3.5, P = 0.042) and significantly elevated lactate levels compared to the low-LAR group (n = 195). Notably, the ICU mortality rate was significantly higher in the high-LAR group than in the low-LAR group (11.4% vs. 4.6%, P = 0.01), further confirming the robust prognostic value of the LAR across different clinical environments. Table 1 Baseline characteristic. Variables Total (n = 1038) Low LAR (n = 517) High LAR (n = 521) P- value Demographic variables Gender, n (%) 0.393 Female 631 (60.8) 321 (62.1) 310 (59.5) Male 407 (39.2) 196 (37.9) 211 (40.5) Age (yr) 61.0 ± 17.8 60.6 ± 18.2 61.3 ± 17.4 0.571 Race, n (%) 0.354 Black 165 (15.9) 85 (16.4) 80 (15.4) Other 294 (28.3) 136 (26.3) 158 (30.3) White 579 (55.8) 296 (57.3) 283 (54.3) Smoking, n (%) 0.836 No 881 (84.9) 440 (85.1) 441 (84.6) Yes 157 (15.1) 77 (14.9) 80 (15.4) Vital signs HR (bpm) 94.6 ± 22.0 90.6 ± 20.0 98.5 ± 23.2 < 0.001 SBP (mmHg) 123.5 ± 26.4 126.6 ± 26.0 120.5 ± 26.6 < 0.001 DBP (mmHg) 69.6 ± 20.3 70.4 ± 20.3 68.9 ± 20.4 0.243 RR (breaths/min) 20.9 ± 6.6 20.2 ± 6.1 21.6 ± 6.9 < 0.001 T (°C) 36.8 ± 1.0 36.8 ± 0.9 36.7 ± 1.0 0.261 SpO 2 (%) 96.6 ± 4.1 96.6 ± 4.1 96.5 ± 4.2 0.821 Laboratory test results PCO 2 (mmHg) 45.0 ± 14.0 46.4 ± 14.4 43.7 ± 13.5 0.002 PO 2 (mmHg) 130.5 ± 110.2 128.9 ± 106.1 132.2 ± 114.1 0.632 Bun (mg/dL) 27.9 ± 24.8 27.7 ± 26.3 28.2 ± 23.2 0.735 Creatinine (mg/dL) 1.6 ± 2.1 1.6 ± 2.2 1.6 ± 2.0 0.897 WBC (K/uL) 13.5 ± 9.1 12.1 ± 6.8 14.9 ± 10.7 < 0.001 Hemoglobin (g/dL) 10.8 ± 2.3 10.8 ± 2.2 10.7 ± 2.4 0.857 Lactate (mmol/L) 2.3 ± 2.0 1.2 ± 0.3 3.4 ± 2.2 < 0.001 Albumin (g/dL) 3.0 ± 0.7 3.2 ± 0.6 2.8 ± 0.7 < 0.001 Comorbidities AKI, n (%) 0.027 No 157 (15.1) 91 (17.6) 66 (12.7) Yes 881 (84.9) 426 (82.4) 455 (87.3) Myocardial infarct, n (%) 173 (16.7) 75 (14.5) 98 (18.8) 0.063 Congestive heart failure, n (%) 0.812 No 683 (65.8) 342 (66.2) 341 (65.5) Yes 355 (34.2) 175 (33.8) 180 (34.5) Chronic pulmonary disease, n (%) 0.624 No 4 (0.4) 1 (0.2) 3 (0.6) Yes 1,034 (99.6) 516 (99.8) 518 (99.4) Mild liver disease, n (%) < 0.001 No 919 (88.5) 476 (92.1) 443 (85) Yes 119 (11.5) 41 (7.9) 78 (15.0) Diabetes, n (%) 0.277 No 713 (68.7) 347 (67.1) 366 (70.2) Yes 325 (31.3) 170 (32.9) 155 (29.8) Renal disease, n (%) 0.091 No 829 (79.9) 402 (77.8) 427 (82) Yes 209 (20.1) 115 (22.2) 94 (18.0) Cancer, n (%) < 0.001 No 919 (88.5) 475 (91.9) 444 (85.2) Yes 119 (11.5) 42 (8.1) 77 (14.8) CHD, n (%) 0.667 No 913 (88.0) 457 (88.4) 456 (87.5) Yes 125 (12.0) 60 (11.6) 65 (12.5) AR, n (%) 0.287 No 1,030 (99.2) 515 (99.6) 515 (98.8) Yes 8 (0.8) 2 (0.4) 6 (1.2) AD, n (%) 0.051 No 849 (81.8) 435 (84.1) 414 (79.5) Yes 189 (18.2) 82 (15.9) 107 (20.5) AF, n (%) 0.734 No 752 (72.4) 377 (72.9) 375 (72.0) Yes 286 (27.6) 140 (27.1) 146 (28.0) Hypertension, n (%) 0.421 No 384 (37.0) 185 (35.8) 199 (38.2) Yes 654 (63.0) 332 (64.2) 322 (61.8) Treatment received Ventilation, n (%) 0.647 No 102 (9.8) 53 (10.3) 49 (9.4) Yes 936 (90.2) 464 (89.7) 472 (90.6) CRRT, n (%) < 0.001 No 931 (89.7) 486 (94.0) 445 (85.4) Yes 107 (10.3) 31 (6.0) 76 (14.6) Severity scale and prognosis SOFA 2.0 ± 2.3 1.6 ± 2.0 2.4 ± 2.5 < 0.001 SAPSII 39.5 ± 15.1 35.8 ± 13.3 43.2 ± 15.8 < 0.001 Prognosis 28‑day non‑survivors, n (%) < 0.001 No 856 (82.5) 459 (88.8) 397 (76.2) Yes 182 (17.5) 58 (11.2) 124 (23.8) 60‑day non‑survivors, n (%) < 0.001 No 821 (79.1) 447 (86.5) 374 (71.8) Yes 217 (20.9) 70 (13.5) 147 (28.2) HR, Heart Rate; SBP, Systolic Blood Pressure; DBP, Diastolic Blood Pressure; RR, Respiratory Rate; T, Temperature; WBC, White Blood Cell; AKI, Acute Kidney Injury; CHD, Coronary Heart Disease; AR, Aortic Regurgitation; AD: Alzheimer's disease; AF, Atrial Fibrillation. 3.2 Independent Prognostic Value of LAR for Short-term Mortality To systematically evaluate the independent prognostic significance of LAR for short-term mortality, we conducted a series of univariate and multivariate Cox proportional hazards analyses. The following sections detail the association between LAR levels and clinical outcomes in the discovery and validation cohorts. 3.2.1 Association Between LAR and Mortality in ICU Patients with Asthma To evaluate the independent association between the LAR and clinical outcomes, three Cox proportional hazard models were constructed (Table 2). When analyzed as a continuous variable, every 1-unit increase in LAR was associated with a significant 40% increase in 28-day mortality risk in Model I (adjusted for age and sex; HR = 1.40, 95% CI: 1.28–1.53, P < 0.001). This association remained stable in Model II after further adjustment for comorbidities (HR = 1.41, 95% CI: 1.29–1.55, P < 0.001). In the fully adjusted model (Model III), which accounted for all potential confounders, a significant positive correlation between LAR levels and 28-day mortality persisted (HR = 1.22, 95% CI: 1.08–1.37, P = 0.002). Similarly, elevated LAR levels were consistently associated with an increased risk of 60-day all-cause mortality across Model I (HR = 1.29), Model II (HR = 1.31), and Model III (HR = 1.16), with all P -values < 0.05. To verify the robustness of these findings, the LAR was converted into a categorical variable based on the median value (0.5385). Compared to the low-LAR group (reference), the high-LAR group exhibited significantly higher risks for both 28-day mortality (Model III HR = 1.65, P < 0.003) and 60-day mortality (Model III HR = 1.53, P < 0.001). These results were further validated in the external eICU-CRD cohort (OR = 1.454, P = 0.012), demonstrating the high consistency of the LAR's prognostic value. Finally, to eliminate the influence of baseline confounding factors, a Cox regression analysis was performed in the propensity score-matched (PSM) cohort (all covariates SMD < 0.1). Consistent with the primary analysis, elevated LAR remained significantly associated with increased mortality in the matched cohort ( P < 0.05). These findings collectively confirm that the LAR is a robust and independent prognostic factor for short-term mortality in ICU patients with asthma, independent of baseline clinical characteristics Table 2 Cox proportional HRs for 28-day and 60-day all-cause mortality. Model 1 Model 2 Model 3 HR-(95%CI) P -value HR (95%CI) P -value HR (95%CI) P -value 28-day all-cause mortality LAR 1.40 (1.28~1.53) <0.001 1.41 (1.29~1.55) <0.001 1.22 (1.08~1.37) 0.002 LAR（low LAR） 1(Ref) 1(Ref) 1(Ref) LAR（high LAR） 2.32 (1.7~3.17) <0.001 2.33 (1.7~3.18) <0.001 1.65 (1.19~2.29) 0.003 60-day all-cause mortality LAR 1.29 (1.16~1.44) <0.001 1.31 (1.17~1.46) <0.001 1.16 (1.02~1.32) <0.001 LAR（low LAR） 1(Ref) 1(Ref) 1(Ref) LAR（high LAR） 1.78 (1.3~2.43) <0.001 1.77 (1.29~2.42) <0.001 1.53 (1.1~2.11) <0.001 Results are shown as HRs with 95% CIs. Model I adjust for age and gender. Model II adjust for age, gender, diabetes mellitus, CHD, AR, AD, and AF. Model III adjust for age, gender, diabetes mellitus, CHD, AR, AD, AF, smoking status, mechanical ventilation, CRRT, SAPS II, and SOFA score. 3.2.2 Discriminatory Performance and Confounding Control To ensure the reliability of the adjusted analyses, the discriminatory performance of the propensity score model and the effectiveness of covariate balancing were evaluated (Figure 2). The propensity score model exhibited strong discriminatory power, with an Area Under the Curve (AUC) of 0.815 (Figure 2a). At the optimal propensity score cutoff of 0.1, the model achieved a sensitivity of 75.6% and a specificity of 73.4%, indicating that the model effectively distinguished between the high-LAR and low-LAR groups based on baseline clinical features. To assess the balance of covariates, Standardized Mean Differences (SMDs) were calculated across the original and adjusted cohorts (Figure 2b). In the unmatched cohort (red), significant imbalances (SMD > 0.1) were observed in several critical parameters, including SAPS II and SOFA scores. After applying 1:1 nearest-neighbor matching (cyan) and various weighting methods—including Inverse Probability of Treatment Weighting (IPTW), Standardized Mortality/Morbidity Ratio Weighting (SMRW), Propensity Score Adjustment (PA), and Overlap Weighting (OW)—all covariates achieved an SMD < 0.1, confirming excellent equilibration of baseline features. Consistent with the primary analysis, the association between elevated LAR and increased mortality risk demonstrated notable stability across various sensitivity analyses. As detailed in Supplementary Table S2, the positive trend between LAR and mortality remained consistent across all propensity score-based adjustment models, including PSM, IPTW, and weighting methods 3.2.3 Visualization of the Dose-Response Relationship To further characterize the relationship between the LAR and 28-day mortality, a restricted cubic spline model was employed (Figure 3). The analysis demonstrated a monotonic, linear increase in the risk of mortality as LAR levels rose. The dose-response relationship between the LAR and 28-day mortality was visualized using Restricted Cubic Spline (RCS) analysis (Figure 4). The association was appropriately characterized by a robust linear dose-response pattern ( P non-linearity = 0.141). Using the median LAR value of 0.5385 as the reference point (HR = 1.0), the hazard ratio for death increased steadily and monotonically across the entire distribution of the LAR. This linear progression reinforces the findings from our multivariable Cox regression, indicating that every incremental increase in the LAR confers a progressively higher risk of mortality without evidence of a plateau effect or a specific safety threshold. In contrast to then on-linear associations observed in some previous studies of other critical conditions, our RCS analysis demonstrates that the association between the LAR and mortality risk in asthma patients admitted to the ICU is characterized by a sustained linear progression. This finding implies that clinicians should not merely focus on whether the LAR exceeds a specific 'danger zone' or cutoff. Instead, any incremental elevation in the LAR warrants clinical vigilance, as even a minor increase in the ratio is indicative of a progressive deterioration in prognosis, reflecting the continuous nature of metabolic and inflammatory derangement in asthma patients in critical care settings. 3.3 Association between LAR and all-cause mortality in ICU patients with asthma In the propensity score-matched cohort with well-balanced baseline covariates, we proceeded to investigate the relationship between LAR and all-cause mortality at 28 days and 60 days among critically ill patients with asthma. A total of 182 and 217 patients died within 28 and 60 days, respectively. Kaplan–Meier survival curves (Figure 4) revealed that the 28-day and 60-day cumulative survival rates in the high-LAR group were significantly lower than those in the low-LAR group (log-rank P < 0.001). The substantial separation between the survival curves from the early stages of ICU admission suggests that elevated LAR is not only a marker of acute metabolic derangement but also a persistent predictor of poor clinical outcomes throughout the follow-up period. These findings, consistent across both short-term (28-day) and longer-term (60-day) intervals, reinforce the robust prognostic utility of the LAR in risk-stratifying asthma patients in critical care settings. 3.4 Subgroup analysis To further evaluate the stability of the association between the LAR and 28-day and 60-day all-cause mortality in ICU patients with asthma, subgroup analyses were performed based on smoking status, gender, and various comorbidities, including congestive heart failure, mild liver disease, renal disease, diabetes, coronary heart disease, gastroesophageal reflux, atrial fibrillation, and aortic regurgitation (Figure 5). Subgroup analysis results demonstrated that the positive association between the LAR and 28-day all-cause mortality remained consistent across most subgroups (Figure 5(a)). Notably, a significant interaction was observed within the gender subgroup ( P interaction = 0.044): in female patients, elevated LAR was significantly associated with a higher risk of mortality (HR = 1.29, 95% CI: 1.13–1.46, P < 0.001), whereas this association did not reach statistical significance in male patients (HR = 0.96, 95% CI: 0.71–1.30, P = 0.779). Except for gender, no significant interactions were identified in other subgroups, such as smoking status, diabetes, or renal disease (all P interaction > 0.05), indicating that the predictive value of the LAR for 28-day mortality risk is robust across diverse clinical settings. For 60-day all-cause mortality, the subgroup analysis further confirmed the independent predictive role of the LAR (Figure 5(b)). The results showed that no significant interactions were observed across all included subgroups, including smoking status, gender, and various comorbidities (all P interaction > 0.05). This suggests that the significant positive correlation between the LAR and 60-day all-cause mortality is highly consistent across different demographic characteristics and clinical comorbidity statuses, remaining unmodulated by the aforementioned confounding factors. In summary, the subgroup analysis indicates that the LAR is a reliable predictor of both short-term (28-day) and mid-to-long-term (60-day) all-cause mortality risk in ICU patients with asthma. Although gender exhibited a specific interaction effect in the 28-day analysis, the overall trend demonstrates that patients with higher LAR levels face an increased risk of mortality across various clinical subgroups. 3.5 External Validation of the Predictive Model in the eICU Cohort To rigorous evaluate the generalizability and clinical utility of the LAR-based predictive model, external validation was performed using a multi-center cohort from the eICU-CRD database (n = 467). 3.5.1 Construction and Performance of the Predictive Model Based on the independent predictors identified in the discovery cohort, a clinical nomogram was constructed to facilitate individualized prediction of ICU mortality risk, incorporating the LAR, age, gender, and SOFA score (Figure 6). The independent prognostic value of these variables was further confirmed in the eICU validation cohort through multivariable Logistic regression analysis (Supplementary Table S3). In this external set, the LAR remained a significant predictor of mortality (OR = 1.454, 95% CI: 1.087–1.943, P = 0.012), alongside the SOFA score and age, ensuring the stability of the model's components across different clinical settings. The model demonstrated robust discriminative ability in the external validation cohort, achieving an Area Under the ROC Curve (AUC) of 0.809 (95% CI: 0.753–0.865) (Figure 7a). This performance was highly consistent with the results obtained in the primary analysis, indicating that the LAR-based predictive model maintains stable and reliable performance across different clinical environments and patient populations. 3.5.2 Discrimination and Calibration Performance The model demonstrated robust discriminative ability in the validation cohort, achieving an Area Under the ROC Curve (AUC) of 0.809 (95% CI: 0.753–0.865) (Figure 7a). Calibration analysis indicated excellent agreement between the predicted and observed probabilities of ICU mortality (Figure 7b). This was statistically substantiated by the Hosmer–Lemeshow test ( c2= 11.373, P = 0.181, Table 3), which indicated no significant lack of fit. Furthermore, the detailed decile-based validation data, illustrating the precise alignment between mean predicted and observed outcomes, are summarized in Table 4. Table 3. Hosmer–Lemeshow Goodness-of-Fit Test for the Clinical Predictive Model in the eICU Cohort. Statistical Index Value Chi-square (c 2 ) 11.373 Degrees of freedom (df) 8 P -value 0.181 Note : A P -value > 0.05 indicates no significant difference between the predicted and observed mortality, reflecting the high calibration of the model. Table 4. Decile-based calibration data of the predictive model in the eICU validation cohort. Group n Observed Events Expected Events Mean Predicted Prob. Mean Observed Prob. 1 47 0 0.4 0.009 0 2 47 0 1.07 0.023 0 3 47 0 1.83 0.039 0 4 46 2 2.65 0.058 0.043 5 47 2 3.52 0.075 0.043 6 47 2 4.41 0.094 0.043 7 46 4 5.89 0.128 0.087 8 47 10 7.39 0.157 0.213 9 47 8 10.37 0.221 0.17 10 46 12 15.49 0.337 0.261 Notes: 1. Decile-based grouping: Patients in the eICU-CRD validation cohort (n = 467) were divided into ten equal-sized groups (deciles) based on their predicted probability of mortality, ranging from the lowest risk (Group 1) to the highest risk (Group 10). 2. Observed vs. Expected Events: "Observed Events" refers to the actual number of deaths occurring in each decile, while "Expected Events" represents the number of deaths predicted by the LAR-based model. 3. Mean Predicted vs. Observed Probabilities: These columns display the average predicted risk versus the actual mortality rate within each group. The close alignment between these values across all deciles reflects the model’s excellent calibration performance. 4. Statistical Significance: A Hosmer-Lemeshow goodness-of-fit test was performed on this data, yielding a p- value of 0.181 ( P > 0.05), indicating no significant difference between the predicted and observed mortality and confirming the model's reliability in the external validation set. 3.5.3 Survival Analysis in the Validation Cohort Decision Curve Analysis (DCA) was employed to assess the clinical usefulness of the model. As shown in Figure 7c, the LAR-based model provided a significant net benefit across a wide range of threshold probabilities (1% to 45%). The Clinical Impact Curve (Figure 7d) further confirmed that the model's predicted high-risk cases closely align with actual events. Finally, Kaplan–Meier survival analysis in the eICU cohort confirmed that patients in the high-LAR group had significantly lower cumulative survival rates than those in the low-LAR group (log-rank P = 0.0066, Figure 8). 4 Discussion This study demonstrates a significant association between elevated LAR and increased short-term all-cause mortality in critically ill patients with asthma, utilizing a large retrospective cohort from the MIMIC-IV v3.1 database. To further verify the universality of this finding, we obtained highly consistent results in the external validation cohort of eICU-CRD, which included 467 patients (OR = 1.454, P = 0.012). The research results show that a higher LAR level at the time of ICU admission is independently associated with an increased risk of death. After adjusting for demographic characteristics, comorbidities, and clinical severity scores, the hazard ratios (HRs/ORs) remained significant. Kaplan-Meier survival analysis further corroborated these results, showing markedly lower survival probabilities in the high LAR group. These observations align with the growing body of evidence highlighting LAR as a composite biomarker that integrates metabolic stress and inflammatory/nutritional status, offering superior prognostic utility compared to lactate or albumin alone in various critical care settings [23,24]. In this study, the linear dose-response relationship between LAR and 28-day mortality, identified through RCS analysis, provides a straightforward and reliable clinical indicator. Unlike biomarkers that exhibit complex non-linear or U-shaped trajectories, the risk of mortality in our asthma-specific cohort increases proportionally with rising LAR levels. This consistent linear association explains why the LAR-based model maintained high predictive stability (AUC 0.815 in MIMIC vs. 0.809 in eICU-CRD), despite the significant difference in clinical severity between the two cohorts (median SOFA score of 6 vs. 2). This demonstrates the exceptional robustness and generalizability of the LAR as a prognostic tool across diverse clinical environments[19]. The linear pattern ensures that even incremental elevations in LAR can be interpreted by clinicians as a direct signal of escalating risk[20]. By capturing the continuous progression of metabolic and inflammatory derangements[25],this linear model overcomes the potential 'prediction blind spots' often encountered in traditional scoring systems like SOFA[21], which may not fully reflect the acute metabolic crises specific to asthma patients in critical care settings." [25]. Mechanistically, elevated LAR in asthmatic ICU patients likely reflects intertwined processes of tissue hypoperfusion, anaerobic metabolism, and systemic inflammation exacerbated by bronchospasm and dynamic hyperinflation[22].. Lactate accumulation signals hypoxic stress from severe airway obstruction, while hypoalbuminemia indicates capillary leak, malnutrition, or hepatic involvement amid inflammatory cascades [6]. In our cohort, the high LAR group exhibited a higher incidence of comorbidities and hemodynamic instability. The data from the external validation set further supported this point. Patients in the high LAR group exhibited more significant characteristics of organ dysfunction, and the prominent performance of the LAR weight (following the SOFA score in the nomogram and ranking second) proved its core role in identifying asthma-related metabolic crises. [7,26]. Notably, while mechanical ventilation rates did not differ between LAR groups, CRRT utilization was higher in the high LAR cohort, suggesting LAR's potential to flag renal hypoperfusion early in asthma crises [11]. Subgroup analyses reinforced LAR's robustness as a predictor across most strata, with no significant interactions except for gender in 28-day mortality. Females showed a stronger association between elevated LAR and mortality risk, potentially attributable to hormonal influences on asthma severity, such as estrogen-driven airway inflammation or progesterone effects on ventilation [27]. This gender disparity aligns with broader evidence of higher asthma hospitalization and mortality rates in women, particularly post-menopause, where comorbidities amplify risks [28,29]. No such interaction was observed for 60-day outcomes, implying LAR's gender-specific prognostic value may be more acute. These findings highlight the need for gender-stratified approaches in asthma management, as supported by studies showing differential biomarker responses and treatment outcomes between sexes [30,31]. To translate research into clinical tools, the LAR-based nomogram developed in this study demonstrated excellent calibration ( P = 0.181) and discrimination (AUC = 0.809) in the external validation set. Decision curve analysis (DCA) further confirmed that the model has significant clinical net benefits within the risk threshold range of 1%–45%Our results extend the role of LAR in non-asthmatic critical illnesses, such as trauma and COVID-19. [13,32-34]. For instance, in surgical ICU patients, elevated LAR correlates with hypovolemia and poor survival, mirroring our observations of hemodynamic derangements [35]. In community-acquired pneumonia, LAR determines ICU admission and mortality risks, paralleling its utility in respiratory failure[36],thereby highlighting the robust prognostic consistency of LAR across different respiratory phenotypes thereby highlighting the robust prognostic consistency of LAR across different respiratory phenotypes consistent with its reliable performance observed across various cohorts [37]. Integrating LAR with existing scores like SOFA enhanced predictive accuracy in similar cohorts, suggesting potential for composite tools in asthma prognostication [12]. Despite these strengths, limitations warrant consideration. Although this study enhanced extrapolability through dual-center cross-database validation, as a retrospective analysis, unmeasured confounding factors may still introduce bias. In addition, LAR is calculated based on values from the initial 24 hours, and continuous measurement may further improve its prognostic value . [38]. Future prospective studies should validate these findings in diverse populations, explore LAR's response to interventions like corticosteroids or biologics, and assess its integration into clinical decision-making algorithms [39]. In conclusion, LAR has become a simple and easily accessible biomarker. It is independent of traditional scoring systems and shows robustness across various subgroups, including gender effects. The consistent performance across multiple centers demonstrates its excellent ability in identifying high-risk patients with asthma patients in critical care settings. Further research is crucial to clarify the underlying mechanisms and optimize its clinical application. 5 Conclusion In summary, this study, leveraging dual-center databases (1,038 patients in the MIMIC-IV discovery cohort and 467 patients in the eICU-CRD validation cohort), indicates that an elevated LAR is a robust independent predictor of both short-term and long-term mortality in patients with asthma patients in critical care settings. Our analysis reveals a consistent linear dose-response relationship, suggesting that mortality risk increases proportionally with escalating LAR levels. As an easily accessible composite biomarker, LAR effectively integrates dual pathophysiological signals reflecting metabolic stress (lactate) and systemic inflammatory/nutritional status (albumin). It not only captures the systemic immune dysregulation at the core of life-threatening asthma exacerbations—including neutrophil activation and cytokine amplification—but also demonstrates exceptional predictive stability across different clinical settings. These findings highlight the significant potential of LAR for early risk stratification and for optimizing the timing of personalized interventions. This study provides strong evidence-based support for incorporating LAR into clinical decision-making to advance precision medicine for asthma patients admitted to the ICU. Declarations 6 Data Availability Statement The data that support the findings of this study are available from the PhysioNet repository (MIMIC-IV and eICU-CRD databases). However, restrictions apply to the availability of these data, which were used under license for the current study, and so are not publicly available. Data are however available from the corresponding author upon reasonable request and with permission of PhysioNet, provided the requester completes the required CITI training and signs the data use agreement. 7 Conflict of Interest The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. 8 Ethics statement The establishment of this database was sanctioned by the Massachusetts Institute of Technology (Cambridge, MA) and Beth Israel Deaconess Medical Center (Boston, MA), and consent was procured for the original data collection. Thus, the statement of ethical approval and the requirement for informed consent were waived for this manuscript. 9 Author Contributions Chenxi Wang was primarily responsible for the conception, design, data collection, data analysis, and drafting of the manuscript. Qin Chen and Yajing Li performed the data processing, literature search, and data code review. Chenxi Wang and Li Zhang were responsible for the statistical analysis and the interpretation of the results. Li Zhang provided overall supervision, revised the manuscript for important intellectual content, and double-checked the statistical analysis results. All authors made substantial contributions to this study and approved the final version of the manuscript for publication. 10 Funding Tianjin Key Medical Discipline Construction Project, No. TJYXZDXK-3-032C. 11 Acknowledgments We thank Dr. Jie Liu from the People’s Liberation Army General Hospital, Beijing, China, for his valuable assistance in revising this manuscript. The authors would like to thank the team at the Laboratory for Computational Physiology (LCP) at the Massachusetts Institute of Technology (MIT) for maintaining the availability of the MIMIC-IV v3.1 databases. References Global Asthma Network. The Global Asthma Report 2022. Int. J. Tuberc Lung Dis. 26 , 1–104 (2022). Ong, K. L. et al. Global, regional, and national burden of diabetes from 1990 to 2021, with projections of prevalence to 2050: a systematic analysis for the Global Burden of Disease Study 2021. Lancet 402 , 203–234 (2023). Louie, S. et al. The Critically Ill Asthmatic—from ICU to Discharge. Clin. Rev. Allergy Immunol. 43 , 30–44 (2012). Manyeruke, F. et al. Asthma in the intensive care unit: A review of patient characteristics and outcomes. Afr. J. Thorac. Crit. Care Med. 29 , 48–51 (2023). Cao, Y. et al. Global trends in the incidence and mortality of asthma from 1990 to 2019: An age-period-cohort analysis using the global burden of disease study 2019. Front. Public. Health . 10 , 1036674 (2022). Gharipour, A. et al. Lactate/albumin ratio: An early prognostic marker in critically ill patients. Am. J. Emerg. Med. 38 , 2088–2095 (2020). Acharya, C. P. et al. Prognostic significance of lactate/albumin ratio in respiratory failure and sepsis. Ann. Med. 57 , 2482024 (2025). Huang, Y. et al. Lactate to albumin ratio as a prognostic marker for all-cause mortality in patients with venous thromboembolism: a retrospective cohort study. Front. Cardiovasc. Med. 12 , 1609295 (2025). Mu, Y. et al. Lactate to albumin ratio predicts one year mortality in ICU patients receiving invasive mechanical ventilation. Sci. Rep. (2026). Wang, J. et al. Lactate-to-albumin ratio as a potential prognostic predictor in patients with cirrhosis and sepsis: a retrospective cohort study. BMC Infect. Dis. 25 , 223 (2025). Droś, J. et al. Association of the lactate/albumin ratio with mortality and hypovolemia in critically Ill patients: a retrospective cohort study. J. Clin. Med. 14 , 6321 (2025). Gou, J. et al. Prognostic value of the lactate-to-albumin ratio in critically ill chronic heart failure patients with sepsis: insights from a retrospective cohort study. Front. Med. (Lausanne) . 12 , 1593524 (2025). Yang, X. et al. Lactate-to-albumin ratio as an independent predictor of mortality in critically ill patients with gastrointestinal bleeding: a retrospective cohort study. BMC Gastroenterol. 25 , 653 (2025). Ge, Y. et al. Predicting 30-day in-hospital mortality in ICU asthma patients: a retrospective machine learning study with external validation. BMC Pulm Med. 25 , 387 (2025). Zhuang, R. et al. Relationship between dietary protein, serum albumin, and mortality in asthmatic populations: a cohort study. Front. Immunol. 15 , 1396740. 10.3389/fimmu.2024.1396740 (2024). Ke, J. et al. Associations of complete blood cell count-derived inflammatory biomarkers with asthma and mortality in adults: a population-based study. Front. Immunol. 14 , 1205687. 10.3389/fimmu.2023.1205687 (2023). Gayen, S. et al. Critical Care Management of Severe Asthma Exacerbations. J. Clin. Med. 13 , 859 (2024). Younan, R. et al. Severe asthma exacerbation: Changes in patient characteristics, management, and outcomes from 1997 to 2016 in 40 ICUs in the greater Paris area. J. Intensive Med. 4 , 209–215 (2023). Liu, Q. et al. Association between lactate-to-albumin ratio and 28-days all-cause mortality in patients with acute pancreatitis: A retrospective analysis of the MIMIC-IV database. Front. Immunol. 13 , 1076121. 10.3389/fimmu.2022.1076121 (2022). Wang, J. et al. Association between serum creatinine to albumin ratio and short- and long-term all-cause mortality in patients with acute pancreatitis admitted to the intensive care unit: a retrospective analysis based on the MIMIC-IV database. Front. Immunol. 15 , 1373371. 10.3389/fimmu.2024.1373371 (2024). Ling, T. et al. Non-linear relationship between platelet count and 30-day in-hospital mortality in ICU patients with acute myocardial infarction: a multicenter retrospective cohort study. Sci. Rep. 15 , 21821 (2025). She, X. et al. Nonlinear Association Between Calculated Globulin Levels and 28-Day Mortality in Patients with Sepsis: A Retrospective Cohort Study. Risk Manag Healthc. Policy . 18 , 2743–2757 (2025). Song, Z. et al. Elevated lactate-to-albumin ratio predicts short- and long-term mortality in trauma and surgical intensive care patients: a retrospective MIMIC-IV cohort study. Sci. Rep. 15 , 35754 (2025). Han, K. et al. Lactate-to-albumin ratio as an independent predictor of 28-day ICU mortality in patients with hypertension: a retrospective cohort study. Sci. Rep. 15 , 33018 (2025). Umemura, Y. et al. Non-linear and Interaction Analyses of Biomarkers for Organ Dysfunctions as Predictive Markers for Sepsis: A Nationwide Retrospective Study. J. Pers. Med. 12 , 44 (2022). Erdoğan, M. & Findikli, H. A. Prognostic value of the lactate/albumin ratio for predicting mortality in patients with pneumosepsis in intensive care units. Medicine 101 , e28748 (2022). Zein, J. G. & Erzurum, S. C. Asthma is Different in Women. Curr. Allergy Asthma Rep. 15 , 1–10 (2015). Woods, S. E. et al. Young Adults Admitted for Asthma: Does Gender Influence Outcomes? J. Womens Health . 12 , 481 (2003). Sunyer, J. et al. Sex differences in mortality of people who visited emergency rooms for asthma and chronic obstructive pulmonary disease. Am. J. Respir Crit. Care Med. 158 , 851–856 (1998). Lin, R. Y. & Lee, G. B. The gender disparity in adult asthma hospitalizations dynamically relates to age. J. Asthma . 45 , 931–935 (2008). Trawick, D. R. et al. Influence of gender on rates of hospitalization, hospital course, and hypercapnea in high-risk patients admitted for asthma: a 10-year retrospective study at Yale-New Haven Hospital. Chest 119 , 115–119 (2001). Suzuki, Y. et al. Predictive accuracy of lactate albumin ratio for mortality in intensive care units: a nationwide cohort study. BMJ Open. 14 , e088926 (2024). Hancı, P. et al. Lactate to albumin ratio as a determinant of intensive care unit admission and mortality in hospitalized patients with community-acquired pneumonia. BMC Pulm Med. 25 , 224 (2025). Kokkoris, S. et al. Lactate to Albumin Ratio and Mortality in Patients with Severe Coronavirus Disease-2019 Admitted to an Intensive Care Unit. J. Clin. Med. 13 , 7106 (2024). Shi, X. et al. Clinical significance of the lactate-to-albumin ratio on prognosis in critically ill patients with acute kidney injury. Ren. Fail. 46 , 2350238 (2024). Cetin, E. U. et al. Lactate-albumin ratio improves combined predictive value of qSOFA and MEWS for 30-day mortality in ICU patients with sepsis: A retrospective cohort study. Medicine 104 , e43097 (2025). Zhao, K. et al. Association Between Lactate-to-Albumin Ratio and 28-Day All-Cause Mortality in Critical Care Patients with COPD: Can Both Arterial and Peripheral Venous Lactate Serve as Predictors? Int. J. Chron. Obstruct Pulmon Dis. 20 , 1419–1434 (2025). Liu, J. et al. Association between lactate/albumin ratio and prognosis in critically ill patients with acute kidney injury undergoing continuous renal replacement therapy. Ren. Fail. 46 , 2374451 (2024). Zhang, R. et al. Lactate to albumin ratio as a novel predictor of short-term prognosis for liver cirrhosis in ICU. Sci. Rep. 15 , 35754 (2025). Additional Declarations No competing interests reported. Supplementary Files SupplementaryMaterial.pdf Cite Share Download PDF Status: Under Review Version 1 posted Reviews received at journal 28 Apr, 2026 Reviewers agreed at journal 26 Apr, 2026 Reviewers invited by journal 23 Apr, 2026 Editor assigned by journal 26 Mar, 2026 Editor invited by journal 20 Mar, 2026 Submission checks completed at journal 18 Mar, 2026 First submitted to journal 13 Mar, 2026 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-9082854","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":633672425,"identity":"5e1b6f1d-5f94-4db5-8d41-7081a05d40be","order_by":0,"name":"Chenxi Wang","email":"","orcid":"","institution":"Tianjin Chest Hospital","correspondingAuthor":false,"prefix":"","firstName":"Chenxi","middleName":"","lastName":"Wang","suffix":""},{"id":633672426,"identity":"6b3737bf-d33b-4a03-af91-e18c0ff24f8b","order_by":1,"name":"Qin Chen","email":"","orcid":"","institution":"Tianjin Chest Hospital","correspondingAuthor":false,"prefix":"","firstName":"Qin","middleName":"","lastName":"Chen","suffix":""},{"id":633672427,"identity":"38fee546-d758-46b6-bfd3-e342daa91827","order_by":2,"name":"Yajing Li","email":"","orcid":"","institution":"Tianjin Chest Hospital","correspondingAuthor":false,"prefix":"","firstName":"Yajing","middleName":"","lastName":"Li","suffix":""},{"id":633672428,"identity":"b670ae1b-c8de-46bf-addc-9015da3173ae","order_by":3,"name":"Li Zhang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAsUlEQVRIiWNgGAWjYJACZhDBz8x8+AFpWiTb2dIMSNNicJ5HQYIo5fL9xx9/Lqiwyzc+zMNgwFBjE01QC2PDgQTjGWeSLbcd5j3wgOFYWm4DQUcB9STztjEbmB3mSzBgbDhMWAsbUM9h3n/1BsbNPAYSRGnhAepp5m04bGDATKwWCaAe5hnHjhtIHAYGcgIxfoGEWE21AX//4cMPPtTYENaCChJIUz4KRsEoGAWjABcAAAVuNxGIzBMqAAAAAElFTkSuQmCC","orcid":"","institution":"Tianjin Chest Hospital","correspondingAuthor":true,"prefix":"","firstName":"Li","middleName":"","lastName":"Zhang","suffix":""}],"badges":[],"createdAt":"2026-03-10 10:54:17","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9082854/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9082854/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":108412453,"identity":"caf16e4a-4ef1-41a7-b328-ba247e08699e","added_by":"auto","created_at":"2026-05-04 10:26:07","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":771874,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFlowchart of patient selection and study design. \u003c/strong\u003eThe screening process followed a consistent hierarchy across both the discovery\u003cstrong\u003e (MIMIC-IV) \u003c/strong\u003eand validation \u003cstrong\u003e(eICU-CRD) c\u003c/strong\u003eohorts. After identifying initial asthma admissions, patients were excluded based on age, pregnancy, ICU stay duration, missing LAR data, and severe liver disease.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAbbreviations: MIMIC-IV, \u003c/strong\u003eMedical Information Mart for Intensive Care IV\u003cstrong\u003e; eICU-CRD, e\u003c/strong\u003eICU Collaborative Research Database; \u003cstrong\u003eICU, \u003c/strong\u003eintensive care unit\u003cstrong\u003e; LAR, \u003c/strong\u003elactate-to-albumin ratio\u003cstrong\u003e.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"image1.png","url":"https://assets-eu.researchsquare.com/files/rs-9082854/v1/58a27aa803c9a6498fe29aeb.png"},{"id":108412431,"identity":"99b87197-f89f-4927-ac24-5b5030921ffa","added_by":"auto","created_at":"2026-05-04 10:26:01","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":215992,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePerformance and covariate balance of the propensity score-based models.\u003c/strong\u003e \u003cstrong\u003e(a)\u003c/strong\u003e Receiver operating characteristic (ROC) curve for the propensity score model, showing an AUC of 0.815. \u003cstrong\u003e(b)\u003c/strong\u003e Standardized mean differences (SMDs) of baseline covariates before (Unmatched, red) and after (Matched/Weighted, cyan) various propensity score-based adjustments. The vertical dashed lines indicate the 0.1 threshold for adequate balance. Abbreviations: IPTW, inverse probability of treatment weighting; SMRW, standardized mortality/morbidity ratio weighting; PA, propensity score adjustment; OW, overlap weighting.\u003c/p\u003e","description":"","filename":"image2.png","url":"https://assets-eu.researchsquare.com/files/rs-9082854/v1/7fc647e3ba8901f6876ec290.png"},{"id":108412430,"identity":"019f8d8c-3a70-49a9-97dc-2581565730b4","added_by":"auto","created_at":"2026-05-04 10:26:01","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":56332,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDose-response relationship between the lactate-to-albumin ratio (LAR) and 28-day mortality.\u003c/strong\u003e The restricted cubic spline (RCS) illustrates a linear association between the LAR and the risk of 28-day all-cause mortality in ICU patients with asthma (\u003cem\u003e\u003cstrong\u003eP \u003c/strong\u003e\u003c/em\u003efor non-linearity = 0.141). The solid line represents the estimated hazard ratios, and the shaded area indicates the 95% confidence intervals. The median LAR of 0.5385 is used as the reference point (HR = 1.0), denoted by the horizontal and vertical dashed lines.\u003c/p\u003e","description":"","filename":"image3.png","url":"https://assets-eu.researchsquare.com/files/rs-9082854/v1/84d33c3c79d510871f0947c9.png"},{"id":108412428,"identity":"e6edc1a0-e2fc-4855-b30d-b7e9b528fc5d","added_by":"auto","created_at":"2026-05-04 10:26:01","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":210759,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eKaplan–Meier curves and cumulative incidence of mortality stratified by LAR groups. (a) 28-day all-cause mortality. (b) 60-day all-cause mortality.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"image4.png","url":"https://assets-eu.researchsquare.com/files/rs-9082854/v1/8fabef525875a8c8e60f3c8c.png"},{"id":108412436,"identity":"8f9e24ae-d1c2-45b5-8219-2fc6fa3c4393","added_by":"auto","created_at":"2026-05-04 10:26:02","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":513938,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSubgroup analysis of the correlation between LAR and mortality. (a) 28-day all-cause mortality. (b) 60-day all-cause mortality.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"image5.png","url":"https://assets-eu.researchsquare.com/files/rs-9082854/v1/098c1ecc0b729f059210bd44.png"},{"id":108412434,"identity":"1dbee988-77ad-4a8c-bc95-b87b0e646be7","added_by":"auto","created_at":"2026-05-04 10:26:02","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":51082,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFor predicting the risk of ICU mortality in patients with asthma.\u003c/strong\u003e The nomogram integrates the lactate-to-albumin ratio (LAR), age, gender, and SOFA score to provide an individualized mortality risk score.\u003c/p\u003e","description":"","filename":"image6.png","url":"https://assets-eu.researchsquare.com/files/rs-9082854/v1/f27e247f83de4c8f8b2042b2.png"},{"id":108412449,"identity":"ae7b317a-01c5-48e6-a2ca-c3dad22bb2cd","added_by":"auto","created_at":"2026-05-04 10:26:06","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":279219,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eEvaluation of the predictive model in the external validation cohort (eICU-CRD). \u003c/strong\u003e(a) Receiver operating characteristic (ROC) curve; (b) Calibration plot; (c) Decision curve analysis (DCA); (d) Clinical impact curve.\u003c/p\u003e","description":"","filename":"image7.png","url":"https://assets-eu.researchsquare.com/files/rs-9082854/v1/1e1cad07f38d63c06915b685.png"},{"id":108412437,"identity":"095097c9-ed11-4c68-b67e-d284eebdd558","added_by":"auto","created_at":"2026-05-04 10:26:03","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":69172,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eKaplan–Meier survival curves for ICU mortality in the eICU validation cohort. \u003c/strong\u003ePatients were stratified by the median LAR value (0.5385). Log-rank test \u003cem\u003e\u003cstrong\u003eP\u003c/strong\u003e\u003c/em\u003e\u003cstrong\u003e \u003c/strong\u003e= 0.0066.\u003c/p\u003e","description":"","filename":"image8.png","url":"https://assets-eu.researchsquare.com/files/rs-9082854/v1/95f586880adefd61084ea59e.png"},{"id":108412478,"identity":"70c18ac3-ef6d-4c0c-bd8e-27a627351e2c","added_by":"auto","created_at":"2026-05-04 10:26:19","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2453584,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9082854/v1/07fbffd6-f810-41b4-9667-a3ec0c1c30d3.pdf"},{"id":108412448,"identity":"16a2dcf5-1241-4242-a084-474c68032da3","added_by":"auto","created_at":"2026-05-04 10:26:06","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":64808,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryMaterial.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9082854/v1/779cffceb23cc31dd0ad8f8b.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Prognostic Value of Lactate-to-Albumin Ratio in Patients with Asthma in the Intensive Care Unit: Development and External Validation of a Predictive Model","fulltext":[{"header":"1 Introduction","content":"\u003cp\u003eAsthma is a chronic respiratory condition characterized by airway inflammation, hyperresponsiveness, and reversible obstruction, affecting millions worldwide and posing a significant public health challenge. Driven by complex immunological cascades such as Th2-mediated inflammation, eosinophilic infiltration, and cytokine storms (e.g., IL-4, IL-5, and IL-13), asthma often involves IgE-dependent mast cell activation and allergic responses, exacerbating systemic immune imbalances in severe cases. Globally, the prevalence of asthma has been estimated at approximately 262\u0026nbsp;million people in 2019, with around 455,000 deaths attributed to the disease annually [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. Projections indicate that by 2045, the burden may increase, particularly in low- and middle-income countries where access to care is limited [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. In critically ill populations, asthma exacerbations can lead to severe respiratory failure requiring intensive care unit (ICU) admission, with short-term mortality rates ranging from 1\u0026ndash;5% overall, but escalating to 8\u0026ndash;25% among those needing mechanical ventilation [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. Factors such as comorbidities, socioeconomic disparities, and delayed intervention contribute to these outcomes, highlighting the need for early risk stratification to improve survival [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eThe lactate-to-albumin ratio (LAR) has emerged as a promising biomarker integrating markers of metabolic stress (lactate) and nutritional and inflammatory reserves(albumin), offering insights into systemic derangements in critically ill patients [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. Elevated lactate reflects tissue hypoperfusion and anaerobic metabolism, while hypoalbuminemia indicates inflammation, malnutrition, or liver dysfunction\u0026mdash;both of which are independently associated with poor prognosis in various conditions [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. Studies have demonstrated LAR's utility as a prognostic marker in sepsis, septic shock, and other ICU scenarios, where higher ratios correlate with increased mortality [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. For instance, in critically ill patients with sepsis, LAR outperforms individual lactate or albumin levels in predicting 28-day mortality [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. Similarly, in cardiac arrest and acute pancreatitis cohorts, LAR has shown strong predictive value for short- and long-term outcomes [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eDespite these advancements, a notable gap exists in the literature regarding LAR's role in ICU patients with asthma. While LAR has been validated in broad critical illness populations [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e], its association with short-term mortality in acute asthma exacerbations requiring intensive care remains underexplored[\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. Asthma pathophysiology involves unique elements like bronchospasm and dynamic hyperinflation, which may interact differently with metabolic and inflammatory markers compared to other respiratory failures [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. Moreover, existing studies on asthma mortality in the ICU often focus on clinical scores like SOFA or APACHE II, and the precise dose-response relationship between LAR and asthma-specific outcomes remains to be fully characterized.[\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. Non-linear associations between biomarkers and mortality have been observed in ICU settings, such as U-shaped or L-shaped patterns in platelet counts or globulin levels with sepsis mortality[\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e], characterizing the dose-response relationship is therefore essential for accurate risk assessment in clinical practice. [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eTo address this knowledge gap, our study evaluates the association between LAR and short-term (28-day and 60-day) all-cause mortality in asthma patients admitted to the ICU, while characterizing the dose-response relationship. By leveraging a discovery cohort of 1,038 patients from the MIMIC-IV (v3.1) database and an external validation cohort of 467 patients from the eICU-CRD, this research utilizes a dual-center approach to provide asthma-specific insights into LAR's prognostic value, potentially guiding early interventions and personalized management strategies targeted at modulating immunological pathways to mitigate mortality risks.\u003c/p\u003e"},{"header":"2 Methods","content":"\u003ch2\u003e2.1 Data source\u003c/h2\u003e\n\u003cp\u003eThis is a retrospective cohort study utilizing two large-scale clinical databases: the Medical Information Mart for Intensive Care IV (MIMIC-IV) and the eICU Collaborative Research Database (eICU-CRD)\u003c/p\u003e\n\u003cp\u003eThe discovery cohort was derived from MIMIC-IV version 3.1, which encompasses comprehensive clinical records for 194,458 admissions to the Intensive Care Units (ICUs) of Beth Israel Deaconess Medical Center (Boston, MA, USA) between 2008 and 2022. The database provides detailed longitudinal information, including patients’ hospital stays, vital signs, laboratory measurements, medication treatments, and clinical outcomes.\u003c/p\u003e\n\u003cp\u003eTo evaluate the generalizability of our findings, we utilized the eICU-CRD (version 2.0) as an external validation cohort. The eICU-CRD is a multi-center database comprising de-identified data from over 200,000 ICU admissions across 208 hospitals in the United States between 2014 and 2015. It contains high-granularity clinical data similar to MIMIC-IV, facilitating robust cross-center validation.\u003c/p\u003e\n\u003cp\u003eTo ensure patient privacy, all personal identifiers in both databases were de-identified and anonymized. Therefore, the Institutional Review Boards (IRBs) of the involved institutions (including Beth Israel Deaconess Medical Center and the Massachusetts Institute of Technology) approved the data use and waived the requirement for written informed consent. All methods were performed in accordance with the relevant guidelines and regulations (e.g., the Declaration of Helsinki). The first author successfully completed the Collaborative Institutional Training Initiative (CITI) program and obtained access to both datasets (Access No. 65378269). Both the MIMIC-IV and eICU-CRD datasets are accessible via the PhysioNet platform (https://physionet.org/).\u003c/p\u003e\n\u003ch2\u003e2.2 Patient selection\u003c/h2\u003e\n\u003cp\u003eA standardized screening protocol was applied to both the discovery and validation cohorts to ensure methodological consistency (Figure 1). For the MIMIC-IV discovery cohort, 4,416 adult asthma patients were initially identified. A total of 3,378 patients were excluded based on the following hierarchy: (1) missing LAR data at admission (n = 3,223); (2) severe liver disease (n = 116); (3) ICU length of stay \u0026lt; 24 hours (n = 32); and (4) pregnancy (n = 7). Consequently, 1,038 patients were enrolled and stratified into low-LAR (n = 521) and high-LAR (n = 517) groups. Similarly, in the eICU-CRD validation cohort, 2,200 asthma cases were initially screened. Following the same exclusion criteria, 1,733 patients were excluded, including those lacking admission LAR data (n = 1,654), severe liver disease (n = 56), ICU stay \u0026lt; 24 hours (n = 20), and pregnancy (n = 3). This resulted in a final validation population of 467 patients, with 195 assigned to the low-LAR group and 272 to the high-LAR group based on the median LAR value\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eVariable Extraction and Data Processing\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eRelevant clinical variables were extracted from both the MIMIC-IV (v3.1) and eICU-CRD (v2.0) databases using Structured Query Language (SQL). To ensure methodological consistency and minimize measurement bias, only the first recorded values within the initial 24 hours of the first ICU admission were utilized for analysis.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePatient Identification and Selection Criteria\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAsthma patients were identified based on the International Classification of Diseases (ICD) codes. Specifically, we included adult patients (≥18 years) with the following diagnosis codes: ICD-9: 493.x (including 493.00, 493.01, 493.10, 493.11, 493.20, 493.21, 493.90, 493.91) and ICD-10: J45.x (including J45.0–J45.5, J45.8, J45.9) or J46.\u003c/p\u003e\n\u003cp\u003eTo ensure the prognostic integrity of the lactate-to-albumin ratio (LAR), patients with severe liver disease were strictly excluded, as impaired hepatic synthetic function can significantly confound albumin levels and lactate metabolism. Severe liver disease was defined using the following codes: ICD-9: 570, 571.x, 572.x; ICD-10: K70.3, K72.x, K74.x, K76.6–K76.7, R18, and I85.x. Additional exclusions included pregnancy and an ICU length of stay \u0026lt; 24 hours.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Elements and Predictor Calculation\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe primary predictor, LAR, was calculated by dividing the admission arterial lactate concentration (mmol/L) by the serum albumin concentration (g/dL). The following covariates were also extracted:\u003c/p\u003e\n\u003cp\u003eDemographics and Vitals: Age, gender, race, heart rate, systolic blood pressure (SBP), diastolic blood pressure (DBP), mean arterial pressure (MAP), respiratory rate, body temperature, and peripheral oxygen saturation (SpO2).\u003c/p\u003e\n\u003cp\u003eLaboratory Parameters: \u0026nbsp;PCO2, PO2, blood urea nitrogen (BUN), creatinine, hemoglobin, and white blood cell (WBC) count.\u003c/p\u003e\n\u003cp\u003eComorbidities and Severity Scores: Smoking status, congestive heart failure, renal disease, diabetes, coronary heart disease (CHD), Alzheimer's disease (AD), atrial fibrillation (AF), and aortic regurgitation (AR). Illness severity was assessed using the Sequential Organ Failure Assessment (SOFA) and Simplified Acute Physiology Score II (SAPS II).\u003c/p\u003e\n\u003cp\u003eClinical Interventions: The use of mechanical ventilation and continuous renal replacement therapy (CRRT) during the ICU stay.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCross-Center Validation Strategy\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe same extraction logic and variable definitions were applied to the eICU-CRD validation cohort. To ensure robust external validation, variables from the multi-center eICU network were meticulously mapped to the MIMIC-IV schema. In instances of multiple measurements within the first 24 hours, only the baseline (first available) value was used to represent the patient’s clinical status at the time of acute decompensation.\u003c/p\u003e\n\u003ch2\u003e2.3 Statistical analysis\u003c/h2\u003e\n\u003cp\u003eAll statistical analyses were performed using R Statistical Software (Version 4.2.2, The R Foundation) and the Free Statistics Analysis Platform (Version 1.9, Beijing, China). A two-sided \u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u003c/strong\u003e \u0026lt; 0.05 was considered statistically significant.\u003c/p\u003e\n\u003cp\u003eContinuous variables were expressed as mean ± standard deviation (SD) for normally distributed data, while skewed distributions were presented as median (interquartile range [IQR]). Categorical variables were displayed as frequencies and percentages (%). Differences between groups were assessed using the independent samples t-test or Mann-Whitney U-test for continuous variables, and the chi-square test or Fisher’s exact test for categorical data, as appropriate.\u003c/p\u003e\n\u003cp\u003eSurvival curves for 28-day and 60-day all-cause mortality were estimated using the Kaplan–Meier method and compared via the log-rank test. To evaluate the independent association between the LAR and mortality, both univariate and multivariate Cox proportional hazards models were utilized to compute hazard ratios (HRs) and 95% confidence intervals (CIs). Three stepwise adjustment models were constructed: Model I adjusted for age and sex. Model II further adjusted for comorbidities, including diabetes mellitus, CHD, AR, AD, and AF. Model III (fully adjusted model) included additional adjustments for smoking status, mechanical ventilation, CRRT, SAPS II, and SOFA score.\u003c/p\u003e\n\u003cp\u003eTo explore potential non-linear dose-response relationships between LAR and mortality, we employed a restricted cubic spline (RCS) model with four knots. The median LAR value (0.5385 was selected )as the reference point (HR = 1.0). Subgroup and interaction analyses were conducted to examine the consistency of LAR’s predictive value across various clinical strata, with interactions assessed using the likelihood ratio test.\u003c/p\u003e\n\u003cp\u003eThe predictive performance of the final model was validated in the eICU-CRD cohort using the Area Under the Receiver Operating Characteristic Curve (AUC). Regarding missing data (less than 5% of the dataset), list-wise deletion was employed as this missingness proportion has minimal impact on the stability of the model estimates.\u003c/p\u003e"},{"header":"3 Results","content":"\u003ch2\u003e3.1 Characteristics of participants\u003c/h2\u003e\n\u003cp\u003eA total of 1,038 patients from the MIMIC-IV database and 467 patients from the eICU-CRD database were finally enrolled in this study (Figure 1). The baseline characteristics of the discovery cohort (MIMIC-IV) are detailed in Table 1, while those of the validation cohort (eICU-CRD) are presented in Supplementary Table S1.\u003c/p\u003e\n\u003cp\u003eIn the MIMIC-IV cohort, the overall 28-day all-cause mortality rate among the enrolled asthma patients was 17.5%. Based on the median LAR value (0.5385), patients were stratified into a low-LAR group (n = 517) and a high-LAR group (n = 521). No significant statistical differences were observed between the two groups regarding age (\u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u003c/strong\u003e= 0.571), sex (\u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u003c/strong\u003e= 0.393), or racial distribution (\u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u003c/strong\u003e = 0.354), indicating well-balanced baseline demographics. However, patients in the high-LAR group exhibited a significantly higher burden of comorbidities, including mild liver disease (15.0% vs. 7.9%, \u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u0026nbsp;\u003c/strong\u003e\u0026lt; 0.001), malignant cancer (14.8% vs. 8.1%, \u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u0026nbsp;\u003c/strong\u003e\u0026lt; 0.001), and acute kidney injury (87.3% vs. 82.4%, \u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u0026nbsp;\u003c/strong\u003e\u0026lt; 0.05). Correspondingly, severity-of-illness scores, including SAPS II and SOFA, were significantly higher in the high-LAR group (both \u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u0026nbsp;\u003c/strong\u003e\u0026lt; 0.001).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eHemodynamic and Laboratory Findings\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eRegarding hemodynamic status in the MIMIC-IV cohort, the high-LAR group presented with lower systolic blood pressure (120.5 \u0026plusmn; \u0026nbsp; 26.6 vs. 126.6 \u0026plusmn; \u0026nbsp; 26.0 mmHg, \u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u0026nbsp;\u003c/strong\u003e\u0026lt; 0.001) and significantly elevated heart and respiratory rates (both \u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u0026nbsp;\u003c/strong\u003e\u0026lt; 0.001), suggesting increased physiological stress and potential tissue hypoperfusion. Laboratory analysis revealed that the high-LAR group had markedly higher lactate levels (3.4 \u0026plusmn; \u0026nbsp; \u0026nbsp;2.2 vs. 1.2 \u0026plusmn; \u0026nbsp; 0.3 mmol/L, \u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u0026nbsp;\u003c/strong\u003e\u0026lt; 0.001) and significantly lower albumin levels (2.8 \u0026plusmn; \u0026nbsp; 0.7 vs. 3.2 \u0026plusmn; \u0026nbsp; 0.6 g/dL,\u003cem\u003e\u0026nbsp;\u003cstrong\u003eP\u003c/strong\u003e\u003c/em\u003e\u0026lt; 0.001). The requirement for continuous renal replacement therapy (CRRT) was also significantly higher in the high-LAR group (14.6% vs. 6.0%,\u003cem\u003e\u0026nbsp;\u003cstrong\u003eP\u003c/strong\u003e\u003c/em\u003e\u0026lt; 0.001). Ultimately, mortality rates for both 28-day (23.8% vs. 11.2%) and 60-day (28.2% vs. 13.5%) periods were significantly higher in the high-LAR group (both \u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u0026nbsp;\u003c/strong\u003e\u0026lt; 0.001).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eExternal Validation in the eICU-CRD Cohort\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo ensure the generalizability of our findings, we analyzed an external validation cohort of 467 patients from the eICU-CRD (Supplementary Table S1). Consistent with the discovery cohort, patients in the high-LAR group (n = 272) were characterized by higher SOFA scores (6.8 \u0026plusmn; \u0026nbsp; 3.9 vs. 6.1 \u0026plusmn; \u0026nbsp; 3.5, \u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u0026nbsp;\u003c/strong\u003e= 0.042) and significantly elevated lactate levels compared to the low-LAR group (n = 195). Notably, the ICU mortality rate was significantly higher in the high-LAR group than in the low-LAR group (11.4% vs. 4.6%, \u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u003c/strong\u003e = 0.01), further confirming the robust prognostic value of the LAR across different clinical environments.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 1 Baseline characteristic.\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"100%\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eVariables\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eTotal (n = 1038)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eLow LAR\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003e(n\u0026nbsp;\u003c/strong\u003e=\u003cstrong\u003e\u0026nbsp;517)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eHigh LAR\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003e(n\u0026nbsp;\u003c/strong\u003e=\u003cstrong\u003e\u0026nbsp;521)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003eP-\u003c/em\u003e\u003c/strong\u003e\u003cstrong\u003evalue\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eDemographic variables\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eGender, n (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.393\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eFemale\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e631 (60.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e321 (62.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e310 (59.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eMale\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e407 (39.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e196 (37.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e211 (40.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eAge (yr)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e61.0 \u0026plusmn; 17.8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e60.6 \u0026plusmn; 18.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e61.3 \u0026plusmn; 17.4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.571\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eRace, n (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.354\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eBlack\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e165 (15.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e85 (16.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e80 (15.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eOther\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e294 (28.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e136 (26.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e158 (30.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eWhite\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e579 (55.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e296 (57.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e283 (54.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"top\"\u003e\n \u003cp\u003eSmoking, n (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.836\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e881 (84.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e440 (85.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e441 (84.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e157 (15.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e77 (14.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e80 (15.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eVital signs\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eHR (bpm)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e94.6 \u0026plusmn; 22.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e90.6 \u0026plusmn; 20.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e98.5 \u0026plusmn; 23.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026lt; 0.001\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eSBP (mmHg)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e123.5 \u0026plusmn; 26.4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e126.6 \u0026plusmn; 26.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e120.5 \u0026plusmn; 26.6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026lt; 0.001\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eDBP (mmHg)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e69.6 \u0026plusmn; 20.3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e70.4 \u0026plusmn; 20.3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e68.9 \u0026plusmn; 20.4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.243\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eRR (breaths/min)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e20.9 \u0026plusmn; 6.6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e20.2 \u0026plusmn; 6.1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e21.6 \u0026plusmn; 6.9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026lt; 0.001\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eT (\u0026deg;C)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e36.8 \u0026plusmn; 1.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e36.8 \u0026plusmn; 0.9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e36.7 \u0026plusmn; 1.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.261\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eSpO\u003csub\u003e2\u003c/sub\u003e (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e96.6 \u0026plusmn; 4.1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e96.6 \u0026plusmn; 4.1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e96.5 \u0026plusmn; 4.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.821\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eLaboratory test results\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003ePCO\u003csub\u003e2\u003c/sub\u003e (mmHg)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e45.0 \u0026plusmn; 14.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e46.4 \u0026plusmn; 14.4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e43.7 \u0026plusmn; 13.5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.002\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003ePO\u003csub\u003e2\u003c/sub\u003e (mmHg)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e130.5 \u0026plusmn; 110.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e128.9 \u0026plusmn; 106.1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e132.2 \u0026plusmn; 114.1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.632\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eBun (mg/dL)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e27.9 \u0026plusmn; 24.8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e27.7 \u0026plusmn; 26.3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e28.2 \u0026plusmn; 23.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.735\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eCreatinine (mg/dL)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.6 \u0026plusmn; 2.1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.6 \u0026plusmn; 2.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.6 \u0026plusmn; 2.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.897\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eWBC (K/uL)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e13.5 \u0026plusmn; 9.1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e12.1 \u0026plusmn; 6.8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e14.9 \u0026plusmn; 10.7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026lt; 0.001\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eHemoglobin (g/dL)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e10.8 \u0026plusmn; 2.3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e10.8 \u0026plusmn; 2.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e10.7 \u0026plusmn; 2.4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.857\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eLactate (mmol/L)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2.3 \u0026plusmn; 2.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.2 \u0026plusmn; 0.3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e3.4 \u0026plusmn; 2.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026lt; 0.001\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eAlbumin (g/dL)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e3.0 \u0026plusmn; 0.7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e3.2 \u0026plusmn; 0.6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2.8 \u0026plusmn; 0.7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026lt; 0.001\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eComorbidities\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eAKI, n (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003e0.027\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e157 (15.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e91 (17.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e66 (12.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e881 (84.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e426 (82.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e455 (87.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eMyocardial infarct, n (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e173 (16.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e75 (14.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e98 (18.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.063\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"3\" valign=\"top\"\u003e\n \u003cp\u003eCongestive heart failure, n (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.812\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e683 (65.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e342 (66.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e341 (65.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e355 (34.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e175 (33.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e180 (34.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"3\" valign=\"top\"\u003e\n \u003cp\u003eChronic pulmonary disease, n (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.624\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e4 (0.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1 (0.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e3 (0.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1,034 (99.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e516 (99.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e518 (99.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"top\"\u003e\n \u003cp\u003eMild liver disease, n (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026lt; 0.001\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e919 (88.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e476 (92.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e443 (85)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e119 (11.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e41 (7.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e78 (15.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"top\"\u003e\n \u003cp\u003eDiabetes, n (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.277\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e713 (68.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e347 (67.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e366 (70.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e325 (31.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e170 (32.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e155 (29.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"top\"\u003e\n \u003cp\u003eRenal disease, n (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.091\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e829 (79.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e402 (77.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e427 (82)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e209 (20.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e115 (22.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e94 (18.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"top\"\u003e\n \u003cp\u003eCancer, n (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026lt; 0.001\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e919 (88.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e475 (91.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e444 (85.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e119 (11.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e42 (8.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e77 (14.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eCHD, n (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.667\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e913 (88.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e457 (88.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e456 (87.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e125 (12.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e60 (11.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e65 (12.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eAR, n (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.287\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1,030 (99.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e515 (99.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e515 (98.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e8 (0.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2 (0.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e6 (1.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eAD, n (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.051\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e849 (81.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e435 (84.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e414 (79.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e189 (18.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e82 (15.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e107 (20.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eAF, n (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.734\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e752 (72.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e377 (72.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e375 (72.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e286 (27.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e140 (27.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e146 (28.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"top\"\u003e\n \u003cp\u003eHypertension, n (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.421\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e384 (37.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e185 (35.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e199 (38.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e654 (63.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e332 (64.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e322 (61.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eTreatment received\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"top\"\u003e\n \u003cp\u003eVentilation, n (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.647\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e102 (9.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e53 (10.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e49 (9.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e936 (90.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e464 (89.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e472 (90.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eCRRT, n (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026lt; 0.001\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e931 (89.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e486 (94.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e445 (85.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e107 (10.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e31 (6.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e76 (14.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eSeverity scale and prognosis\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eSOFA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2.0 \u0026plusmn; 2.3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.6 \u0026plusmn; 2.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2.4 \u0026plusmn; 2.5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026lt; 0.001\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eSAPSII\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e39.5 \u0026plusmn; 15.1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e35.8 \u0026plusmn; 13.3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e43.2 \u0026plusmn; 15.8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026lt; 0.001\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003ePrognosis\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"top\"\u003e\n \u003cp\u003e28‑day non‑survivors, n (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026lt; 0.001\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e856 (82.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e459 (88.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e397 (76.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e182 (17.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e58 (11.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e124 (23.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd colspan=\"2\" valign=\"top\"\u003e\n \u003cp\u003e60‑day non‑survivors, n (%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026lt; 0.001\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e821 (79.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e447 (86.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e374 (71.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e217 (20.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e70 (13.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e147 (28.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eHR, Heart Rate; SBP, Systolic Blood Pressure; DBP, Diastolic Blood Pressure; RR, Respiratory Rate; T, Temperature; WBC, White Blood Cell; AKI, Acute Kidney Injury; CHD, Coronary Heart Disease; AR, Aortic Regurgitation; AD: Alzheimer\u0026apos;s disease; AF, Atrial Fibrillation.\u003c/p\u003e\n\u003ch2\u003e3.2 \u0026nbsp; \u0026nbsp; Independent Prognostic Value of LAR for Short-term Mortality\u003c/h2\u003e\n\u003cp\u003eTo systematically evaluate the independent prognostic significance of LAR for short-term mortality, we conducted a series of univariate and multivariate Cox proportional hazards analyses. The following sections detail the association between LAR levels and clinical outcomes in the discovery and validation cohorts.\u003c/p\u003e\n\u003ch3\u003e3.2.1 \u0026nbsp;Association Between LAR and Mortality in ICU Patients with Asthma\u003c/h3\u003e\n\u003cp\u003eTo evaluate the independent association between the LAR and clinical outcomes, three Cox proportional hazard models were constructed (Table 2).\u003c/p\u003e\n\u003cp\u003eWhen analyzed as a continuous variable, every 1-unit increase in LAR was associated with a significant 40% increase in 28-day mortality risk in Model I (adjusted for age and sex; HR = 1.40, 95% CI: 1.28\u0026ndash;1.53, \u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u0026nbsp;\u003c/strong\u003e\u0026lt; 0.001). This association remained stable in Model II after further adjustment for comorbidities (HR = 1.41, 95% CI: 1.29\u0026ndash;1.55, \u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u0026nbsp;\u003c/strong\u003e\u0026lt; 0.001). In the fully adjusted model (Model III), which accounted for all potential confounders, a significant positive correlation between LAR levels and 28-day mortality persisted (HR = 1.22, 95% CI: 1.08\u0026ndash;1.37, \u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u0026nbsp;\u003c/strong\u003e= 0.002). Similarly, elevated LAR levels were consistently associated with an increased risk of 60-day all-cause mortality across Model I (HR = 1.29), Model II (HR = 1.31), and Model III (HR = 1.16), with all \u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u003c/strong\u003e-values \u0026lt; 0.05.\u003c/p\u003e\n\u003cp\u003eTo verify the robustness of these findings, the LAR was converted into a categorical variable based on the median value (0.5385). Compared to the low-LAR group (reference), the high-LAR group exhibited significantly higher risks for both 28-day mortality (Model III HR = 1.65, \u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u0026nbsp;\u003c/strong\u003e\u0026lt; 0.003) and 60-day mortality (Model III HR = 1.53, \u003cstrong\u003e\u003cem\u003eP\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e\u0026lt; 0.001). These results were further validated in the external \u003cstrong\u003eeICU-CRD\u003c/strong\u003e cohort (OR = 1.454, \u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u0026nbsp;\u003c/strong\u003e= 0.012), demonstrating the high consistency of the LAR\u0026apos;s prognostic value.\u003c/p\u003e\n\u003cp\u003eFinally, to eliminate the influence of baseline confounding factors, a Cox regression analysis was performed in the propensity score-matched (PSM) cohort (all covariates SMD \u0026lt; 0.1). Consistent with the primary analysis, elevated LAR remained significantly associated with increased mortality in the matched cohort (\u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u0026nbsp;\u003c/strong\u003e\u0026lt; 0.05). These findings collectively confirm that the LAR is a robust and independent prognostic factor for short-term mortality in ICU patients with asthma, independent of baseline clinical characteristics\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 2 Cox proportional HRs for\u0026nbsp;\u003c/strong\u003e\u003cstrong\u003e28-day\u003c/strong\u003e\u003cstrong\u003e\u0026nbsp;and 60-day all-cause mortality.\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"100%\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003eModel 1\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003eModel 2\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd colspan=\"2\"\u003e\n \u003cp\u003e\u003cstrong\u003eModel 3\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eHR-(95%CI)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u003c/strong\u003e\u003cstrong\u003e-value\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eHR (95%CI)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u003c/strong\u003e\u003cstrong\u003e-value\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eHR (95%CI)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u003c/strong\u003e\u003cstrong\u003e-value\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003e28-day all-cause mortality\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eLAR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.40 (1.28~1.53)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.41 (1.29~1.55)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.22 (1.08~1.37)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.002\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eLAR（low LAR）\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1(Ref)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1(Ref)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1(Ref)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eLAR（high LAR）\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2.32 (1.7~3.17)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2.33 (1.7~3.18)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.65 (1.19~2.29)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.003\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003e60-day all-cause mortality\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eLAR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.29 (1.16~1.44)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.31 (1.17~1.46)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.16 (1.02~1.32)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eLAR（low LAR）\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1(Ref)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1(Ref)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1(Ref)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eLAR（high LAR）\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.78 (1.3~2.43)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.77 (1.29~2.42)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.53 (1.1~2.11)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\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\u003eResults are shown as HRs with 95% CIs.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eModel I\u003c/strong\u003e adjust for age and gender.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eModel II\u003c/strong\u003e adjust for age, gender, diabetes mellitus, CHD, AR, AD, and AF.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eModel III\u003c/strong\u003e adjust for age, gender, diabetes mellitus, CHD, AR, AD, AF, smoking status, mechanical ventilation, CRRT, SAPS II, and SOFA score.\u003c/p\u003e\n\u003ch3\u003e3.2.2 \u0026nbsp;Discriminatory Performance and Confounding Control\u003c/h3\u003e\n\u003cp\u003eTo ensure the reliability of the adjusted analyses, the discriminatory performance of the propensity score model and the effectiveness of covariate balancing were evaluated (Figure 2).\u003c/p\u003e\n\u003cp\u003eThe propensity score model exhibited strong discriminatory power, with an Area Under the Curve (AUC) of 0.815 (Figure 2a). At the optimal propensity score cutoff of 0.1, the model achieved a sensitivity of 75.6% and a specificity of 73.4%, indicating that the model effectively distinguished between the high-LAR and low-LAR groups based on baseline clinical features.\u003c/p\u003e\n\u003cp\u003eTo assess the balance of covariates, Standardized Mean Differences (SMDs) were calculated across the original and adjusted cohorts (Figure 2b). In the unmatched cohort (red), significant imbalances (SMD \u0026gt; 0.1) were observed in several critical parameters, including SAPS II and SOFA scores. After applying 1:1 nearest-neighbor matching (cyan) and various weighting methods\u0026mdash;including Inverse Probability of Treatment Weighting (IPTW), Standardized Mortality/Morbidity Ratio Weighting (SMRW), Propensity Score Adjustment (PA), and Overlap Weighting (OW)\u0026mdash;all covariates achieved an SMD \u0026lt; 0.1, confirming excellent equilibration of baseline features.\u003c/p\u003e\n\u003cp\u003eConsistent with the primary analysis, the association between elevated LAR and increased mortality risk demonstrated notable stability across various sensitivity analyses. As detailed in Supplementary Table S2, the positive trend between LAR and mortality remained consistent across all propensity score-based adjustment models, including PSM, IPTW, and weighting methods\u003c/p\u003e\n\u003ch3\u003e3.2.3 \u0026nbsp;Visualization of the Dose-Response Relationship\u003c/h3\u003e\n\u003cp\u003eTo further characterize the relationship between the LAR and 28-day mortality, a restricted cubic spline model was employed (Figure 3). The analysis demonstrated a monotonic, linear increase in the risk of mortality as LAR levels rose.\u003c/p\u003e\n\u003cp\u003eThe dose-response relationship between the LAR and 28-day mortality was visualized using Restricted Cubic Spline (RCS) analysis (Figure 4). The association was appropriately characterized by a robust linear dose-response pattern (\u003cstrong\u003e\u0026nbsp;\u003cem\u003eP\u003c/em\u003e\u003c/strong\u003e\u003csub\u003enon-linearity\u003c/sub\u003e = 0.141). Using the median LAR value of 0.5385 as the reference point (HR = 1.0), the hazard ratio for death increased steadily and monotonically across the entire distribution of the LAR. This linear progression reinforces the findings from our multivariable Cox regression, indicating that every incremental increase in the LAR confers a progressively higher risk of mortality without evidence of a plateau effect or a specific safety threshold.\u003c/p\u003e\n\u003cp\u003eIn contrast to then on-linear associations observed in some previous studies of other critical conditions, our RCS analysis demonstrates that the association between the LAR and mortality risk in asthma patients admitted to the ICU is characterized by a sustained linear progression. This finding implies that clinicians should not merely focus on whether the LAR exceeds a specific \u0026apos;danger zone\u0026apos; or cutoff. Instead, any incremental elevation in the LAR warrants clinical vigilance, as even a minor increase in the ratio is indicative of a progressive deterioration in prognosis, reflecting the continuous nature of metabolic and inflammatory derangement in asthma patients in critical care settings.\u003c/p\u003e\n\u003ch2\u003e3.3 \u0026nbsp; \u0026nbsp; Association between LAR and all-cause mortality in ICU patients with asthma\u003c/h2\u003e\n\u003cp\u003eIn the propensity score-matched cohort with well-balanced baseline covariates, we proceeded to investigate the relationship between LAR and all-cause mortality at 28 days and 60 days among critically ill patients with asthma.\u003c/p\u003e\n\u003cp\u003eA total of 182 and 217 patients died within 28 and 60 days, respectively. Kaplan\u0026ndash;Meier survival curves (Figure 4) revealed that the 28-day and 60-day cumulative survival rates in the high-LAR group were significantly lower than those in the low-LAR group (log-rank \u003cstrong\u003e\u003cem\u003eP\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e\u0026lt; 0.001).\u003c/p\u003e\n\u003cp\u003eThe substantial separation between the survival curves from the early stages of ICU admission suggests that elevated LAR is not only a marker of acute metabolic derangement but also a persistent predictor of poor clinical outcomes throughout the follow-up period. These findings, consistent across both short-term (28-day) and longer-term (60-day) intervals, reinforce the robust prognostic utility of the LAR in risk-stratifying asthma patients in critical care settings.\u003c/p\u003e\n\u003ch2\u003e3.4 \u0026nbsp; \u0026nbsp; Subgroup analysis\u003c/h2\u003e\n\u003cp\u003eTo further evaluate the stability of the association between the LAR and 28-day and 60-day all-cause mortality in ICU patients with asthma, subgroup analyses were performed based on smoking status, gender, and various comorbidities, including congestive heart failure, mild liver disease, renal disease, diabetes, coronary heart disease, gastroesophageal reflux, atrial fibrillation, and aortic regurgitation (Figure 5).\u003c/p\u003e\n\u003cp\u003eSubgroup analysis results demonstrated that the positive association between the LAR and 28-day all-cause mortality remained consistent across most subgroups (Figure 5(a)). Notably, a significant interaction was observed within the gender subgroup (\u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u003c/strong\u003e \u003csub\u003einteraction\u003c/sub\u003e= 0.044): in female patients, elevated LAR was significantly associated with a higher risk of mortality (HR = 1.29, 95% CI: 1.13\u0026ndash;1.46, \u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u0026nbsp;\u003c/strong\u003e\u0026lt; 0.001), whereas this association did not reach statistical significance in male patients (HR = 0.96, 95% CI: 0.71\u0026ndash;1.30, \u003cstrong\u003e\u003cem\u003eP\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e= 0.779). Except for gender, no significant interactions were identified in other subgroups, such as smoking status, diabetes, or renal disease (all \u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u003c/strong\u003e \u003csub\u003einteraction\u003c/sub\u003e \u0026gt; 0.05), indicating that the predictive value of the LAR for 28-day mortality risk is robust across diverse clinical settings.\u003c/p\u003e\n\u003cp\u003eFor 60-day all-cause mortality, the subgroup analysis further confirmed the independent predictive role of the LAR (Figure 5(b)). The results showed that no significant interactions were observed across all included subgroups, including smoking status, gender, and various comorbidities (all \u003cem\u003eP\u003c/em\u003e \u003csub\u003einteraction\u003c/sub\u003e\u0026gt; 0.05). This suggests that the significant positive correlation between the LAR and 60-day all-cause mortality is highly consistent across different demographic characteristics and clinical comorbidity statuses, remaining unmodulated by the aforementioned confounding factors.\u003c/p\u003e\n\u003cp\u003eIn summary, the subgroup analysis indicates that the LAR is a reliable predictor of both short-term (28-day) and mid-to-long-term (60-day) all-cause mortality risk in ICU patients with asthma. Although gender exhibited a specific interaction effect in the 28-day analysis, the overall trend demonstrates that patients with higher LAR levels face an increased risk of mortality across various clinical subgroups.\u003c/p\u003e\n\u003ch2\u003e3.5 \u0026nbsp; \u0026nbsp; External Validation of the Predictive Model in the eICU Cohort\u003c/h2\u003e\n\u003cp\u003eTo rigorous evaluate the generalizability and clinical utility of the LAR-based predictive model, external validation was performed using a multi-center cohort from the eICU-CRD database (n = 467).\u003c/p\u003e\n\u003ch3\u003e3.5.1 \u0026nbsp;Construction and Performance of the Predictive Model\u003c/h3\u003e\n\u003cp\u003eBased on the independent predictors identified in the discovery cohort, a clinical nomogram was constructed to facilitate individualized prediction of ICU mortality risk, incorporating the LAR, age, gender, and SOFA score (Figure 6). The independent prognostic value of these variables was further confirmed in the eICU validation cohort through multivariable Logistic regression analysis (Supplementary Table S3). In this external set, the LAR remained a significant predictor of mortality (OR = 1.454, 95% CI: 1.087\u0026ndash;1.943, \u003cstrong\u003e\u003cem\u003eP\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e= 0.012), alongside the SOFA score and age, ensuring the stability of the model\u0026apos;s components across different clinical settings.\u003c/p\u003e\n\u003cp\u003eThe model demonstrated robust discriminative ability in the external validation cohort, achieving an Area Under the ROC Curve (AUC) of 0.809 (95% CI: 0.753\u0026ndash;0.865) (Figure 7a). This performance was highly consistent with the results obtained in the primary analysis, indicating that the LAR-based predictive model maintains stable and reliable performance across different clinical environments and patient populations.\u003c/p\u003e\n\u003ch3\u003e3.5.2 \u0026nbsp;Discrimination and Calibration Performance\u003c/h3\u003e\n\u003cp\u003eThe model demonstrated robust discriminative ability in the validation cohort, achieving an Area Under the ROC Curve (AUC) of 0.809 (95% CI: 0.753\u0026ndash;0.865) (Figure 7a). Calibration analysis indicated excellent agreement between the predicted and observed probabilities of ICU mortality (Figure 7b). This was statistically substantiated by the Hosmer\u0026ndash;Lemeshow test ( \u0026nbsp;c2= 11.373, \u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u0026nbsp;\u003c/strong\u003e= 0.181, Table 3), which indicated no significant lack of fit. Furthermore, the detailed decile-based validation data, illustrating the precise alignment between mean predicted and observed outcomes, are summarized in Table 4.\u003c/p\u003e\n\u003cp\u003eTable 3. Hosmer\u0026ndash;Lemeshow Goodness-of-Fit Test for the Clinical Predictive Model in the eICU Cohort.\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"100%\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 77.8555%;\"\u003e\n \u003cp\u003eStatistical Index\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22.1445%;\"\u003e\n \u003cp\u003eValue\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 77.8555%;\"\u003e\n \u003cp\u003eChi-square (c\u003csup\u003e2\u003c/sup\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22.1445%;\"\u003e\n \u003cp\u003e11.373\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 77.8555%;\"\u003e\n \u003cp\u003eDegrees of freedom (df)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22.1445%;\"\u003e\n \u003cp\u003e8\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 77.8555%;\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u003c/strong\u003e-value\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 22.1445%;\"\u003e\n \u003cp\u003e0.181\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cstrong\u003eNote\u003c/strong\u003e: A \u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u003c/strong\u003e-value \u0026gt; 0.05 indicates no significant difference between the predicted and observed mortality, reflecting the high calibration of the model.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 4. Decile-based calibration data of the predictive model in the eICU validation cohort.\u003c/strong\u003e\u003c/p\u003e\n \u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"100%\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eGroup\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003en\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eObserved Events\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eExpected Events\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eMean Predicted Prob.\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eMean Observed Prob.\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e47\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.009\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e47\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e1.07\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.023\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e47\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e1.83\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.039\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e46\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e2.65\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.058\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.043\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e47\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e3.52\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.075\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.043\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e47\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e4.41\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.094\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.043\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e46\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e5.89\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.128\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.087\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e47\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e7.39\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.157\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.213\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e47\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e10.37\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.221\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.17\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e46\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e12\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e15.49\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.337\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.261\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003e\u003cstrong\u003eNotes:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e1. Decile-based grouping: Patients in the eICU-CRD validation cohort (n = 467) were divided into ten equal-sized groups (deciles) based on their predicted probability of mortality, ranging from the lowest risk (Group 1) to the highest risk (Group 10).\u003c/p\u003e\n\u003cp\u003e2. Observed vs. Expected Events: \u0026quot;Observed Events\u0026quot; refers to the actual number of deaths occurring in each decile, while \u0026quot;Expected Events\u0026quot; represents the number of deaths predicted by the LAR-based model.\u003c/p\u003e\n\u003cp\u003e3. Mean Predicted vs. Observed Probabilities: These columns display the average predicted risk versus the actual mortality rate within each group. The close alignment between these values across all deciles reflects the model\u0026rsquo;s excellent calibration performance.\u003c/p\u003e\n\u003cp\u003e4. Statistical Significance: A Hosmer-Lemeshow goodness-of-fit test was performed on this data, yielding a\u003cstrong\u003e\u003cem\u003e\u0026nbsp;p-\u003c/em\u003e\u003c/strong\u003evalue of 0.181 (\u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u003c/strong\u003e \u0026gt; 0.05), indicating no significant difference between the predicted and observed mortality and confirming the model\u0026apos;s reliability in the external validation set.\u003c/p\u003e\n\u003ch3\u003e3.5.3 \u0026nbsp;Survival Analysis in the Validation Cohort\u003c/h3\u003e\n\u003cp\u003eDecision Curve Analysis (DCA) was employed to assess the clinical usefulness of the model. As shown in Figure 7c, the LAR-based model provided a significant net benefit across a wide range of threshold probabilities (1% to 45%). The Clinical Impact Curve (Figure 7d) further confirmed that the model\u0026apos;s predicted high-risk cases closely align with actual events. Finally, Kaplan\u0026ndash;Meier survival analysis in the eICU cohort confirmed that patients in the high-LAR group had significantly lower cumulative survival rates than those in the low-LAR group (log-rank \u003cstrong\u003e\u003cem\u003eP\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e= 0.0066, Figure 8).\u003c/p\u003e"},{"header":"4 Discussion","content":"\u003cp\u003eThis study demonstrates a significant association between elevated LAR and increased short-term all-cause mortality in critically ill patients with asthma, utilizing a large retrospective cohort from the MIMIC-IV v3.1 database. To further verify the universality of this finding, we obtained highly consistent results in the external validation cohort of eICU-CRD, which included 467 patients (OR = 1.454, \u003cstrong\u003e\u003cem\u003eP\u003c/em\u003e\u0026nbsp;\u003c/strong\u003e= 0.012). The research results show that a higher LAR level at the time of ICU admission is independently associated with an increased risk of death. After adjusting for demographic characteristics, comorbidities, and clinical severity scores, the hazard ratios (HRs/ORs) remained significant. Kaplan-Meier survival analysis further corroborated these results, showing markedly lower survival probabilities in the high LAR group. These observations align with the growing body of evidence highlighting LAR as a composite biomarker that integrates metabolic stress and inflammatory/nutritional status, offering superior prognostic utility compared to lactate or albumin alone in various critical care settings [23,24].\u003c/p\u003e\n\u003cp\u003eIn this study, the linear dose-response relationship between LAR and 28-day mortality, identified through RCS analysis, provides a straightforward and reliable clinical indicator. Unlike biomarkers that exhibit complex non-linear or U-shaped trajectories, the risk of mortality in our asthma-specific cohort increases proportionally with rising LAR levels. This consistent linear association explains why the LAR-based model maintained high predictive stability (AUC 0.815 in MIMIC vs. 0.809 in eICU-CRD), despite the significant difference in clinical severity between the two cohorts (median SOFA score of 6 vs. 2). This demonstrates the exceptional robustness and generalizability of the LAR as a prognostic tool across diverse clinical environments[19]. The linear pattern ensures that even incremental elevations in LAR can be interpreted by clinicians as a direct signal of escalating risk[20]. By capturing the continuous progression of metabolic and inflammatory derangements[25],this linear model overcomes the potential 'prediction blind spots' often encountered in traditional scoring systems like SOFA[21], which may not fully reflect the acute metabolic crises specific to asthma patients in critical care settings.\" [25].\u003c/p\u003e\n\u003cp\u003eMechanistically, elevated LAR in asthmatic ICU patients likely reflects intertwined processes of tissue hypoperfusion, anaerobic metabolism, and systemic inflammation exacerbated by bronchospasm and dynamic hyperinflation[22].. Lactate accumulation signals hypoxic stress from severe airway obstruction, while hypoalbuminemia indicates capillary leak, malnutrition, or hepatic involvement amid inflammatory cascades [6]. In our cohort, the high LAR group exhibited a higher incidence of comorbidities and hemodynamic instability. The data from the external validation set further supported this point. Patients in the high LAR group exhibited more significant characteristics of organ dysfunction, and the prominent performance of the LAR weight (following the SOFA score in the nomogram and ranking second) proved its core role in identifying asthma-related metabolic crises. \u0026nbsp;[7,26]. Notably, while mechanical ventilation rates did not differ between LAR groups, CRRT utilization was higher in the high LAR cohort, suggesting LAR's potential to flag renal hypoperfusion early in asthma crises \u0026nbsp;[11].\u003c/p\u003e\n\u003cp\u003eSubgroup analyses reinforced LAR's robustness as a predictor across most strata, with no significant interactions except for gender in 28-day mortality. Females showed a stronger association between elevated LAR and mortality risk, potentially attributable to hormonal influences on asthma severity, such as estrogen-driven airway inflammation or progesterone effects on ventilation [27]. This gender disparity aligns with broader evidence of higher asthma hospitalization and mortality rates in women, particularly post-menopause, where comorbidities amplify risks [28,29]. No such interaction was observed for 60-day outcomes, implying LAR's gender-specific prognostic value may be more acute. These findings highlight the need for gender-stratified approaches in asthma management, as supported by studies showing differential biomarker responses and treatment outcomes between sexes [30,31].\u003c/p\u003e\n\u003cp\u003eTo translate research into clinical tools, the LAR-based nomogram developed in this study demonstrated excellent calibration (\u003cstrong\u003e\u003cem\u003eP\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e= 0.181) and discrimination (AUC = 0.809) in the external validation set. Decision curve analysis (DCA) further confirmed that the model has significant clinical net benefits within the risk threshold range of 1%–45%Our results extend the role of LAR in non-asthmatic critical illnesses, such as trauma and COVID-19. [13,32-34]. For instance, in surgical ICU patients, elevated LAR correlates with hypovolemia and poor survival, mirroring our observations of hemodynamic derangements [35]. In community-acquired pneumonia, LAR determines ICU admission and mortality risks, paralleling its utility in respiratory failure[36],thereby highlighting the robust prognostic consistency of LAR across different respiratory phenotypes thereby highlighting the robust prognostic consistency of LAR across different respiratory phenotypes consistent with its reliable performance observed across various cohorts \u0026nbsp;[37]. Integrating LAR with existing scores like SOFA enhanced predictive accuracy in similar cohorts, suggesting potential for composite tools in asthma prognostication [12].\u003c/p\u003e\n\u003cp\u003eDespite these strengths, limitations warrant consideration. Although this study enhanced extrapolability through dual-center cross-database validation, as a retrospective analysis, unmeasured confounding factors may still introduce bias. In addition, LAR is calculated based on values from the initial 24 hours, and continuous measurement may further improve its prognostic value\u003cs\u003e.\u003c/s\u003e[38]. Future prospective studies should validate these findings in diverse populations, explore LAR's response to interventions like corticosteroids or biologics, and assess its integration into clinical decision-making algorithms [39].\u003c/p\u003e\n\u003cp\u003eIn conclusion, LAR has become a simple and easily accessible biomarker. It is independent of traditional scoring systems and shows robustness across various subgroups, including gender effects. The consistent performance across multiple centers demonstrates its excellent ability in identifying high-risk patients with asthma patients in critical care settings. \u0026nbsp; Further research is crucial to clarify the underlying mechanisms and optimize its clinical application.\u003c/p\u003e"},{"header":"5 Conclusion","content":"\u003cp\u003eIn summary, this study, leveraging dual-center databases (1,038 patients in the MIMIC-IV discovery cohort and 467 patients in the eICU-CRD validation cohort), indicates that an elevated LAR is a robust independent predictor of both short-term and long-term mortality in patients with asthma patients in critical care settings. Our analysis reveals a consistent linear dose-response relationship, suggesting that mortality risk increases proportionally with escalating LAR levels.\u003c/p\u003e\n\u003cp\u003eAs an easily accessible composite biomarker, LAR effectively integrates dual pathophysiological signals reflecting metabolic stress (lactate) and systemic inflammatory/nutritional status (albumin). It not only captures the systemic immune dysregulation at the core of life-threatening asthma exacerbations—including neutrophil activation and cytokine amplification—but also demonstrates exceptional predictive stability across different clinical settings. These findings highlight the significant potential of LAR for early risk stratification and for optimizing the timing of personalized interventions. This study provides strong evidence-based support for incorporating LAR into clinical decision-making to advance precision medicine for asthma patients admitted to the ICU.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003e6 Data Availability Statement\u003c/h2\u003e\n\u003cp\u003eThe data that support the findings of this study are available from the PhysioNet repository (MIMIC-IV and eICU-CRD databases). However, restrictions apply to the availability of these data, which were used under license for the current study, and so are not publicly available. Data are however available from the corresponding author upon reasonable request and with permission of PhysioNet, provided the requester completes the required CITI training and signs the data use agreement.\u003c/p\u003e\n\u003ch2\u003e7 Conflict of Interest\u003c/h2\u003e\n\u003cp\u003eThe authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.\u003c/p\u003e\n\u003ch2\u003e8 Ethics statement\u003c/h2\u003e\n\u003cp\u003eThe establishment of this database was sanctioned by the Massachusetts Institute of Technology (Cambridge, MA) and Beth Israel Deaconess Medical Center (Boston, MA), and consent was procured for the original data collection. Thus, the statement of ethical approval and the requirement for informed consent were waived for this manuscript.\u003c/p\u003e\n\u003ch2\u003e9 Author Contributions\u003c/h2\u003e\n\u003cp\u003eChenxi Wang was primarily responsible for the conception, design, data collection, data analysis, and drafting of the manuscript. Qin Chen and Yajing Li performed the data processing, literature search, and data code review. Chenxi Wang and Li Zhang were responsible for the statistical analysis and the interpretation of the results. Li Zhang provided overall supervision, revised the manuscript for important intellectual content, and double-checked the statistical analysis results. All authors made substantial contributions to this study and approved the final version of the manuscript for publication.\u003c/p\u003e\n\u003ch2\u003e10 Funding\u003c/h2\u003e\n\u003cp\u003eTianjin Key Medical Discipline Construction Project, No. TJYXZDXK-3-032C.\u003c/p\u003e\n\u003ch2\u003e11 Acknowledgments\u003c/h2\u003e\n\u003cp\u003eWe thank Dr. Jie Liu from the People’s Liberation Army General Hospital, Beijing, China, for his valuable assistance in revising this manuscript. The authors would like to thank the team at the Laboratory for Computational Physiology (LCP) at the Massachusetts Institute of Technology (MIT) for maintaining the availability of the MIMIC-IV v3.1 databases.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eGlobal Asthma Network. The Global Asthma Report 2022. \u003cem\u003eInt. J. Tuberc Lung Dis.\u003c/em\u003e \u003cb\u003e26\u003c/b\u003e, 1\u0026ndash;104 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOng, K. L. et al. Global, regional, and national burden of diabetes from 1990 to 2021, with projections of prevalence to 2050: a systematic analysis for the Global Burden of Disease Study 2021. \u003cem\u003eLancet\u003c/em\u003e \u003cb\u003e402\u003c/b\u003e, 203\u0026ndash;234 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLouie, S. et al. The Critically Ill Asthmatic\u0026mdash;from ICU to Discharge. \u003cem\u003eClin. Rev. Allergy Immunol.\u003c/em\u003e \u003cb\u003e43\u003c/b\u003e, 30\u0026ndash;44 (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eManyeruke, F. et al. Asthma in the intensive care unit: A review of patient characteristics and outcomes. \u003cem\u003eAfr. J. Thorac. Crit. Care Med.\u003c/em\u003e \u003cb\u003e29\u003c/b\u003e, 48\u0026ndash;51 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCao, Y. et al. Global trends in the incidence and mortality of asthma from 1990 to 2019: An age-period-cohort analysis using the global burden of disease study 2019. \u003cem\u003eFront. Public. Health\u003c/em\u003e. \u003cb\u003e10\u003c/b\u003e, 1036674 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGharipour, A. et al. Lactate/albumin ratio: An early prognostic marker in critically ill patients. \u003cem\u003eAm. J. Emerg. Med.\u003c/em\u003e \u003cb\u003e38\u003c/b\u003e, 2088\u0026ndash;2095 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAcharya, C. P. et al. Prognostic significance of lactate/albumin ratio in respiratory failure and sepsis. \u003cem\u003eAnn. Med.\u003c/em\u003e \u003cb\u003e57\u003c/b\u003e, 2482024 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuang, Y. et al. Lactate to albumin ratio as a prognostic marker for all-cause mortality in patients with venous thromboembolism: a retrospective cohort study. \u003cem\u003eFront. Cardiovasc. Med.\u003c/em\u003e \u003cb\u003e12\u003c/b\u003e, 1609295 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMu, Y. et al. Lactate to albumin ratio predicts one year mortality in ICU patients receiving invasive mechanical ventilation. \u003cem\u003eSci. Rep.\u003c/em\u003e (2026).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang, J. et al. Lactate-to-albumin ratio as a potential prognostic predictor in patients with cirrhosis and sepsis: a retrospective cohort study. \u003cem\u003eBMC Infect. Dis.\u003c/em\u003e \u003cb\u003e25\u003c/b\u003e, 223 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDroś, J. et al. Association of the lactate/albumin ratio with mortality and hypovolemia in critically Ill patients: a retrospective cohort study. \u003cem\u003eJ. Clin. Med.\u003c/em\u003e \u003cb\u003e14\u003c/b\u003e, 6321 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGou, J. et al. Prognostic value of the lactate-to-albumin ratio in critically ill chronic heart failure patients with sepsis: insights from a retrospective cohort study. \u003cem\u003eFront. Med. (Lausanne)\u003c/em\u003e. \u003cb\u003e12\u003c/b\u003e, 1593524 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYang, X. et al. Lactate-to-albumin ratio as an independent predictor of mortality in critically ill patients with gastrointestinal bleeding: a retrospective cohort study. \u003cem\u003eBMC Gastroenterol.\u003c/em\u003e \u003cb\u003e25\u003c/b\u003e, 653 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGe, Y. et al. Predicting 30-day in-hospital mortality in ICU asthma patients: a retrospective machine learning study with external validation. \u003cem\u003eBMC Pulm Med.\u003c/em\u003e \u003cb\u003e25\u003c/b\u003e, 387 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhuang, R. et al. Relationship between dietary protein, serum albumin, and mortality in asthmatic populations: a cohort study. \u003cem\u003eFront. Immunol.\u003c/em\u003e \u003cb\u003e15\u003c/b\u003e, 1396740. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fimmu.2024.1396740\u003c/span\u003e\u003cspan address=\"10.3389/fimmu.2024.1396740\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKe, J. et al. Associations of complete blood cell count-derived inflammatory biomarkers with asthma and mortality in adults: a population-based study. \u003cem\u003eFront. Immunol.\u003c/em\u003e \u003cb\u003e14\u003c/b\u003e, 1205687. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fimmu.2023.1205687\u003c/span\u003e\u003cspan address=\"10.3389/fimmu.2023.1205687\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGayen, S. et al. Critical Care Management of Severe Asthma Exacerbations. \u003cem\u003eJ. Clin. Med.\u003c/em\u003e \u003cb\u003e13\u003c/b\u003e, 859 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYounan, R. et al. Severe asthma exacerbation: Changes in patient characteristics, management, and outcomes from 1997 to 2016 in 40 ICUs in the greater Paris area. \u003cem\u003eJ. Intensive Med.\u003c/em\u003e \u003cb\u003e4\u003c/b\u003e, 209\u0026ndash;215 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu, Q. et al. Association between lactate-to-albumin ratio and 28-days all-cause mortality in patients with acute pancreatitis: A retrospective analysis of the MIMIC-IV database. \u003cem\u003eFront. Immunol.\u003c/em\u003e \u003cb\u003e13\u003c/b\u003e, 1076121. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fimmu.2022.1076121\u003c/span\u003e\u003cspan address=\"10.3389/fimmu.2022.1076121\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang, J. et al. Association between serum creatinine to albumin ratio and short- and long-term all-cause mortality in patients with acute pancreatitis admitted to the intensive care unit: a retrospective analysis based on the MIMIC-IV database. \u003cem\u003eFront. Immunol.\u003c/em\u003e \u003cb\u003e15\u003c/b\u003e, 1373371. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fimmu.2024.1373371\u003c/span\u003e\u003cspan address=\"10.3389/fimmu.2024.1373371\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLing, T. et al. Non-linear relationship between platelet count and 30-day in-hospital mortality in ICU patients with acute myocardial infarction: a multicenter retrospective cohort study. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cb\u003e15\u003c/b\u003e, 21821 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShe, X. et al. Nonlinear Association Between Calculated Globulin Levels and 28-Day Mortality in Patients with Sepsis: A Retrospective Cohort Study. \u003cem\u003eRisk Manag Healthc. Policy\u003c/em\u003e. \u003cb\u003e18\u003c/b\u003e, 2743\u0026ndash;2757 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSong, Z. et al. Elevated lactate-to-albumin ratio predicts short- and long-term mortality in trauma and surgical intensive care patients: a retrospective MIMIC-IV cohort study. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cb\u003e15\u003c/b\u003e, 35754 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHan, K. et al. Lactate-to-albumin ratio as an independent predictor of 28-day ICU mortality in patients with hypertension: a retrospective cohort study. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cb\u003e15\u003c/b\u003e, 33018 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eUmemura, Y. et al. Non-linear and Interaction Analyses of Biomarkers for Organ Dysfunctions as Predictive Markers for Sepsis: A Nationwide Retrospective Study. \u003cem\u003eJ. Pers. Med.\u003c/em\u003e \u003cb\u003e12\u003c/b\u003e, 44 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eErdoğan, M. \u0026amp; Findikli, H. A. Prognostic value of the lactate/albumin ratio for predicting mortality in patients with pneumosepsis in intensive care units. \u003cem\u003eMedicine\u003c/em\u003e \u003cb\u003e101\u003c/b\u003e, e28748 (2022).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZein, J. G. \u0026amp; Erzurum, S. C. Asthma is Different in Women. \u003cem\u003eCurr. Allergy Asthma Rep.\u003c/em\u003e \u003cb\u003e15\u003c/b\u003e, 1\u0026ndash;10 (2015).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWoods, S. E. et al. Young Adults Admitted for Asthma: Does Gender Influence Outcomes? \u003cem\u003eJ. Womens Health\u003c/em\u003e. \u003cb\u003e12\u003c/b\u003e, 481 (2003).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSunyer, J. et al. Sex differences in mortality of people who visited emergency rooms for asthma and chronic obstructive pulmonary disease. \u003cem\u003eAm. J. Respir Crit. Care Med.\u003c/em\u003e \u003cb\u003e158\u003c/b\u003e, 851\u0026ndash;856 (1998).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLin, R. Y. \u0026amp; Lee, G. B. The gender disparity in adult asthma hospitalizations dynamically relates to age. \u003cem\u003eJ. Asthma\u003c/em\u003e. \u003cb\u003e45\u003c/b\u003e, 931\u0026ndash;935 (2008).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTrawick, D. R. et al. Influence of gender on rates of hospitalization, hospital course, and hypercapnea in high-risk patients admitted for asthma: a 10-year retrospective study at Yale-New Haven Hospital. \u003cem\u003eChest\u003c/em\u003e \u003cb\u003e119\u003c/b\u003e, 115\u0026ndash;119 (2001).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSuzuki, Y. et al. Predictive accuracy of lactate albumin ratio for mortality in intensive care units: a nationwide cohort study. \u003cem\u003eBMJ Open.\u003c/em\u003e \u003cb\u003e14\u003c/b\u003e, e088926 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHancı, P. et al. Lactate to albumin ratio as a determinant of intensive care unit admission and mortality in hospitalized patients with community-acquired pneumonia. \u003cem\u003eBMC Pulm Med.\u003c/em\u003e \u003cb\u003e25\u003c/b\u003e, 224 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKokkoris, S. et al. Lactate to Albumin Ratio and Mortality in Patients with Severe Coronavirus Disease-2019 Admitted to an Intensive Care Unit. \u003cem\u003eJ. Clin. Med.\u003c/em\u003e \u003cb\u003e13\u003c/b\u003e, 7106 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShi, X. et al. Clinical significance of the lactate-to-albumin ratio on prognosis in critically ill patients with acute kidney injury. \u003cem\u003eRen. Fail.\u003c/em\u003e \u003cb\u003e46\u003c/b\u003e, 2350238 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCetin, E. U. et al. Lactate-albumin ratio improves combined predictive value of qSOFA and MEWS for 30-day mortality in ICU patients with sepsis: A retrospective cohort study. \u003cem\u003eMedicine\u003c/em\u003e \u003cb\u003e104\u003c/b\u003e, e43097 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhao, K. et al. Association Between Lactate-to-Albumin Ratio and 28-Day All-Cause Mortality in Critical Care Patients with COPD: Can Both Arterial and Peripheral Venous Lactate Serve as Predictors? \u003cem\u003eInt. J. Chron. Obstruct Pulmon Dis.\u003c/em\u003e \u003cb\u003e20\u003c/b\u003e, 1419\u0026ndash;1434 (2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu, J. et al. Association between lactate/albumin ratio and prognosis in critically ill patients with acute kidney injury undergoing continuous renal replacement therapy. \u003cem\u003eRen. Fail.\u003c/em\u003e \u003cb\u003e46\u003c/b\u003e, 2374451 (2024).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang, R. et al. Lactate to albumin ratio as a novel predictor of short-term prognosis for liver cirrhosis in ICU. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cb\u003e15\u003c/b\u003e, 35754 (2025).\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"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":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"lactate-to-albumin ratio (LAR), asthma, mortality, critically ill patients, dose-response relationship, prognostic model, external validation","lastPublishedDoi":"10.21203/rs.3.rs-9082854/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9082854/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e \u003cp\u003eThe lactate-to-albumin ratio (LAR), a composite biomarker integrating metabolic stress and systemic inflammation, has demonstrated significant prognostic value across various critical illnesses. However, its clinical significance in patients with life-threatening asthma remains poorly defined. This study aimed to elucidate the association between LAR and mortality risk in asthma patients admitted to the ICU and to develop and validate a LAR-based individualized predictive model.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eWe retrospectively analyzed 1,038 adult asthma patients from the MIMIC-IV (v3.1) database (2008\u0026ndash;2022) and 467 patients from the eICU-CRD for external validation. The lactate-to-albumin ratio (LAR) was calculated using admission laboratory values, with patients stratified by the median LAR (0.5385). Multivariable Cox regression, restricted cubic spline (RCS) analysis, and Kaplan-Meier curves were employed to assess mortality risks. Subgroup analyses were performed to ensure robustness, adjusting for demographics, comorbidities, and clinical severity scores.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eIn the discovery cohort, 28-day and 60-day mortality rates were 17.5% and 20.9%, respectively, with significantly higher rates observed in the high LAR group ( \u003cb\u003eP\u003c/b\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001). After multivariable adjustment, each 1-unit increase in LAR was independently associated with 28-day (HR\u0026thinsp;=\u0026thinsp;1.22, \u003cb\u003eP\u003c/b\u003e\u0026thinsp;=\u0026thinsp;0.002) and 60-day (HR\u0026thinsp;=\u0026thinsp;1.16, \u003cb\u003eP\u003c/b\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001) mortality. These findings were highly consistent in the eICU-CRD validation cohort (OR\u0026thinsp;=\u0026thinsp;1.454, \u003cb\u003eP\u003c/b\u003e\u0026thinsp;=\u0026thinsp;0.012). Restricted cubic spline (RCS) analysis confirmed a significant linear dose-response relationship between LAR and 28-day mortality. A gender interaction was identified for 28-day mortality ( \u003cb\u003eP\u003c/b\u003e\u003csub\u003einteraction\u003c/sub\u003e = 0.044), showing a stronger association in females. The final predictive model demonstrated excellent stability and performance, with an AUC of 0.809 in the external validation set, closely matching the discovery set (AUC 0.815).\u003c/p\u003e\u003ch2\u003eConclusion\u003c/h2\u003e \u003cp\u003eElevated LAR at ICU admission is a powerful independent predictor of both short-term and long-term mortality in asthma patients admitted to the ICU, characterized by a robust linear dose-response relationship. As a robust and accessible composite biomarker, LAR effectively integrates metabolic stress with systemic immune dysregulation, such as cytokine amplification and capillary leak. Our cross-center validation confirms that LAR-based risk stratification provides a stable and reliable tool for identifying high-risk individuals, potentially informing the optimal timing for immunomodulatory interventions (e.g., precise anti-inflammatory or biologic therapies) to improve clinical outcomes.\u003c/p\u003e","manuscriptTitle":"Prognostic Value of Lactate-to-Albumin Ratio in Patients with Asthma in the Intensive Care Unit: Development and External Validation of a Predictive Model","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-05-04 10:25:07","doi":"10.21203/rs.3.rs-9082854/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"editorInvitedReview","content":"","date":"2026-04-28T06:58:34+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"306258994698146757882094953403553859656","date":"2026-04-26T23:53:01+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-04-23T09:58:02+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-03-26T14:07:17+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2026-03-20T07:12:14+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-03-18T17:59:12+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2026-03-13T04:23:26+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"9c164ee8-a5c9-49ee-ac2a-8c76518b9a24","owner":[],"postedDate":"May 4th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":67441894,"name":"Health sciences/Biomarkers"},{"id":67441895,"name":"Health sciences/Diseases"},{"id":67441896,"name":"Biological sciences/Immunology"},{"id":67441897,"name":"Health sciences/Medical research"},{"id":67441898,"name":"Health sciences/Risk factors"}],"tags":[],"updatedAt":"2026-05-04T10:25:07+00:00","versionOfRecord":[],"versionCreatedAt":"2026-05-04 10:25:07","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9082854","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9082854","identity":"rs-9082854","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}