Abstract
Background
Serum albumin (ALB) has traditionally been regarded as a marker of nutritional status. However, recent studies suggest its changes are closely linked to inflammation, metabolic dysregulation, and disease severity, limiting its role as a sole indicator of nutritional status. Yet, clinical practice continues to rely on ALB to monitor nutritional interventions, with a paucity of high-quality evidence on its dynamic associations with clinical outcomes. This study aimed to investigate the comprehensive associations of ALB dynamics with inflammation, nutritional status, and clinical outcomes in hospitalized patients, providing evidence to optimize clinical management.
Methods
This secondary analysis utilized data from a prospective observational cohort study conducted in 34 tertiary hospitals across China between June and September 2014. A total of 2959 patients hospitalized for 7–30 days with complete data were included. Standardized protocols were used to collect demographics, nutritional indices (Nutritional Risk Screening 2002, Subjective Global Assessment), hematology, biochemistry results, and clinical outcomes (complications, length of stay, costs). Subgroup analyses were performed based on inflammatory status changes, nutritional therapy administration, department type, baseline nutritional status, and advanced age. Receiver operating characteristic curves identified cutoff values for infection-related complications. Correlation analyses and multivariable linear regression models determined independent predictors of ALB changes.
Results
Among 2959 patients, 1894 (64.0%) experienced a decrease in ALB during hospitalization, which significantly impacted primary outcomes, including prolonged length of stay, increased hospitalization costs, and higher complication rates. Significant ALB decline was also strongly associated with worsened nutritional status, weight loss at discharge, exacerbated gastrointestinal symptoms, functional impairments, and edema (P < 0.001 for all). Compared to binary categorization (increase vs. decrease), the magnitude of ALB change demonstrated a stronger correlation with infection-related complications across all subgroups. Subgroup-specific cutoff values stratified patients into high- and low-risk groups, with significant differences in infection-related complication rates (P < 0.05), aiding early identification and intervention. Independent predictors of ALB decline included advanced age, surgical status, lower baseline handgrip strength and its change during hospitalization, higher baseline ALB and globulin levels, baseline Prognostic Nutritional Index, baseline inflammatory status and its exacerbation, elevated platelet-to-lymphocyte ratio, and intensive care unit admission.
Conclusions
Dynamic changes in ALB serve as a critical indicator of inflammation–nutrition interplay, with its reduction effectively predicting infection-related complications, clinical outcomes, and nutritional deterioration. This is particularly valuable in older adults, inflammatory-variable, surgical, and non-malnourished patients. The conventional view of ALB as a pure nutritional marker requires revision. Joint monitoring with inflammatory biomarkers and multidisciplinary interventions targeting high-risk populations are recommended.
Supplementary Information
The online version contains supplementary material available at 10.1186/s40001-025-02925-5.
Keywords: Serum albumin, Malnutrition, Complications, Inflammatory status, Clinical outcomes
Background
Serum albumin (ALB) has long been regarded as a critical biomarker for assessing patients’ nutritional status. Clinicians traditionally view its concentration as directly reflecting total circulating plasma proteins, closely correlating with protein-energy malnutrition, and thus widely utilized to guide nutritional interventions [1]. However, emerging evidence has unveiled additional roles beyond nutrition, including stabilization of cell growth and DNA replication, maintenance of sex hormone homeostasis, and modulation of systemic inflammation [2].
Accumulating research suggests ALB functions more as a negative acute phase reactant, with its concentration influenced by multidimensional factors, such as inflammatory status, disease severity, and fluid balance, rather than merely nutritional intake. In acute or chronic illnesses, inflammation-mediated physiological alterations can induce hepatic protein synthesis reprogramming and increased vascular permeability, leading to rapid declines in ALB levels. This process demonstrates weak correlation with nutritional intake, implying that ALB fluctuations may predominantly reflect inflammatory burdens and systemic physiological adaptations [3]. The American Society for Parenteral and Enteral Nutrition (ASPEN) guidelines accordingly emphasize that visceral proteins, including ALB, should be interpreted as inflammation-related “nutritional risk” markers rather than pure malnutrition indicators [4]. Consequently, relying solely on ALB concentrations to monitor nutritional therapy efficacy appears overly simplistic, necessitating comprehensive consideration of underlying pathophysiological mechanisms.
Despite evolving conceptual frameworks, clinical practice continues to depend heavily on ALB monitoring for nutritional assessment. However, high-quality evidence supporting this practice remains insufficient, particularly in hospitalized patients. The synergistic interplay between malnutrition and inflammation often exacerbates adverse clinical outcomes [5–7]. While nutritional support improves prognoses in selected high-risk populations, precise identification of intervention-responsive patients and effective utilization of ALB-based biomarkers for personalized strategies remain unresolved challenges. Furthermore, existing studies predominantly focus on prognostic values of baseline visceral protein levels [8], while dynamic changes over short intervals have received limited attention. The interplay between ALB variations, inflammatory/nutritional status transitions, and clinical outcomes remains poorly defined. F. Boesiger et al. [9] first investigated 7-day ALB changes in hospitalized patients, but the narrow time window (relative to 19-day half-life [10]) compromised reliability. The present study addresses this gap by analyzing patients with 7–30-day admissions to systematically evaluate the integrated impacts of in-hospital ALB dynamics on nutritional status, inflammatory responses, and clinical outcomes, providing novel insights for optimizing inpatient management.
This investigation represents the first large-scale prospective cohort analysis examining associations between ALB variations and inflammatory markers, nutritional indices, and clinical endpoints in hospitalized patients. Findings aim to elucidate the potential of dynamic ALB monitoring in predicting outcomes and guiding individualized interventions, advancing beyond conventional nutritional assessment paradigms.
Materials and methods
Patients
These patients were screened from a cohort of 7122 subjects between June and September 2014. They participated in a large-scale prospective observational study covering 34 Grade-A Class-III hospitals, which represent the highest level of general hospitals, in 18 cities across China [11]. All patients had complete medical record tracking and underwent a Subjective Global Assessment (SGA) upon admission. This study included ward patients who were conscious and had signed written informed consent forms. Patients with a hospital stay of less than 7 days or more than 30 days, those who were temporarily discharged or died during hospitalization, and those with incomplete data (including SGA assessment results, serum test indicators, the mode of nutritional supply during hospitalization, hospitalization costs and the number of days, as well as the occurrence of complications) were excluded from the data set. The patient screening process is detailed in Fig. 1. This study was approved by the Ethics Committee of Beijing Hospital (No. 2014BJYYEC-022-02), adhered to the Declaration of Helsinki, and was registered at the Chinese Clinical Trial Registry (Registration number: ChiCTR-EPC-14005253, Registration Center: Beijing Hospital, Date of Registration: 2014-09-09).
Fig. 1.
Flowchart of patient selection
Data collection
A standardized research protocol was adopted for this study. The data collected included: (1) demographic parameters: gender, age, and marital status; (2) medical history, weight loss, and food intake; and (3) anthropometric parameters: height, weight, calf circumference, and grip strength measured using standard methods. All researchers underwent relevant standardized training. Calf circumference was measured when the patient was seated with the knees bent at 90° and the feet flat on the ground. The thickest part of the calf was measured using a non-elastic tape measure. Grip strength was measured using an electronic grip strength meter (EH101, Xiangshan, Guangdong, China). Patients were seated with shoulders adducted and elbows bent at 90°. Measurements were taken three times for both the left and right hands, and the maximum value was recorded, with measurements accurate to 0.1 kg. (4) Laboratory parameters: complete blood count and blood biochemistry. The measurement of ALB after admission was conducted in the morning of the second day after admission. The albumin measurement at discharge usually occurred on the day of discharge or at the last time in the hospital, and all measurements were taken under fasting conditions. Based on the diagnostic criteria for systemic inflammatory response syndrome, patients with a white blood cell (WBC) count less than 4 × 10^9/L or greater than 12 × 10^9/L were considered to be in an inflammatory state. If a patient’s WBC count is within the normal range upon admission but meets the criteria for an “inflammatory state” at discharge, they are categorized as inflammation-worsened. Conversely, they are categorized as inflammation-alleviated. If the patient’s WBC count remains within the normal range or continues to meet the criteria for an “inflammatory state” at both admission and discharge, they are classified as persistent status. The platelet-to-lymphocyte ratio (PLR) was calculated from the patient’s blood cell analysis results, and the prognostic nutritional index (PNI) was calculated using the formula: PNI = 10 × ALB (g/dl) + 5 × lymphocyte count (× 10^9/L). All patients completed the Nutritional Risk Screening 2002 (NRS 2002) scale [12] within 24 h of admission, and malnutrition was diagnosed by a trained clinician using the SGA scale.
Adverse clinical outcomes
The primary outcomes of this study included the occurrence of various complications within 30 days after admission, length of hospital stay, and hospitalization costs. Complications were defined as any deviation from the ideal treatment process, such as infections, anastomotic leaks, anemia, electrolyte disturbances, myocardial infarction, etc., excluding the untreated primary disease. Secondary outcomes included changes in nutritional status, weight loss, gastrointestinal symptoms, abnormalities in functional activities, the degree of muscle and subcutaneous fat depletion, and changes in edema.
Statistical analysis
Data were analyzed using IBM SPSS Statistics (V27.0.1) and R software (V4.4.2). Continuous variables were expressed as mean ± standard deviation (Mean ± SD), while categorical variables were described by frequency (percentage). Different statistical methods were applied based on data types for inter-group comparisons. Continuous variables were compared using independent-samples t tests, one-way ANOVA, or non-parametric tests (Mann–Whitney U/Kruskal–Wallis H test). Categorical variables were compared using Chi-square tests or Fisher’s exact test. Associations between ALB changes and indicators such as complications/nutritional status were evaluated using Pearson or Spearman correlation coefficients. Multiple linear regression models (stepwise selection) were constructed to identify independent predictors, with included variables comprising baseline ALB, PLR, PNI, age, admitting department, etc. Model collinearity was assessed via variance inflation factors (VIF). Complication risk stratification was determined by identifying optimal cutoff values based on receiver operating characteristic curves. All tests were two-tailed, with statistical significance set at P < 0.05.
Results
Patient characteristics
As shown in Table 1, this study included 2959 patients aged 18–97 years with a mean age of 58.80 ± 15.47 years. Among all patients, 1065 showed stable or increased ALB levels at discharge, while 1894 experienced decreases. Significant differences were observed in baseline grip strength and its changes between the two groups, with the ALB-decreasing group demonstrating higher baseline grip strength (27.41 vs. 25.84, P < 0.001) and greater grip strength variation (1.28 vs. 0.32, P < 0.001). Pre-admission weight loss, dietary intake, and baseline ALB status also differed significantly, indicating that ALB changes during hospitalization were associated with pre-admission nutritional status and in-hospital interventions. In addition, significant differences were found in inflammatory status changes, nutritional intervention modalities, disease types, treatment departments, and comorbidity presence between groups (all P < 0.001). Figure 2 illustrates the magnitude of changes in serum albumin in hospitalized patients with different categories of disease.
Table 1.
Clinical characteristics of 2959 study patients
| All patients (n = 2959) | ALB decreased at discharge (n = 1894) | ALB stable/increased at discharge (n = 1065) | P | |||
|---|---|---|---|---|---|---|
| Sex | 0.653 | |||||
| Male | 1726 (58.3%) | 1099 (58.0%) | 627 (58.9%) | |||
| Female | 1233 (41.7%) | 795 (42.0%) | 438 (41.1%) | |||
| Age | 0.098 | |||||
| 58.80 (15.47) | 58.45 (15.25) | 59.43 (15.85) | ||||
| Marital status | 0.823 | |||||
| Married | 2763 (93.4%) | 1770 (93.5%) | 993 (93.2%) | |||
| Single or divorced | 196 (6.6%) | 124 (6.5%) | 72 (6.8%) | |||
| Payment Method | 0.034 | |||||
| Self-paid | 569 (19.2%) | 359 (19.0%) | 210 (19.7%) | |||
| Medical insurance | 2000 (67.6%) | 1284 (67.8%) | 716 (67.2%) | |||
| Commercial insurance | 51 (1.7%) | 36 (1.9%) | 15 (1.4%) | |||
| Other | 200 (6.8%) | 140 (7.4%) | 60 (5.6%) | |||
| Government healthcare | 139 (4.7%) | 75 (4.0%) | 64 (6.0%) | |||
| Baseline calf circumference | 0.559 | |||||
| 32.85 (4.54) | 32.89 (4.41) | 32.78 (4.75) | ||||
| Baseline grip strength | < 0.001 | |||||
| 26.84 (10.68) | 27.41 (10.74) | 25.84 (10.50) | ||||
| Baseline BMI | 0.169 | |||||
| 23.05 (3.42) | 23.12 (3.41) | 22.94 (3.44) | ||||
| Calf circumference change | 0.324 | |||||
| − 0.30 (1.46) | − 0.32 (1.39) | − 0.27 (1.56) | ||||
| Grip strength change | < 0.001 | |||||
| − 0.94 (5.85) | − 1.28 (6.08) | − 0.32 (5.36) | ||||
| BMI change | 0.982 | |||||
| − 0.01 (0.10) | − 0.01 (0.10) | − 0.01 (0.12) | ||||
| Baseline PNI | < 0.001 | |||||
| 67.94 (49.12) | 70.78 (50.07) | 62.89 (46.99) | ||||
| Baseline PLR | < 0.001 | |||||
| 158.04 (210.96) | 147.38 (189.00) | 177.09 (244.30) | ||||
| PLR change | < 0.001 | |||||
| 16.94 (183.04) | 30.22 (175.72) | − 6.86 (193.29) | ||||
| Recent 3-month weight loss | < 0.001 | |||||
| Yes | 1004 (33.9%) | 593 (31.3%) | 411 (38.6%) | |||
| No | 1955 (66.1%) | 1301 (68.7%) | 654 (61.4%) | |||
| Recent 1-week reduction in diet | 0.002 | |||||
| Yes | 1184 (40.0%) | 719 (38.0%) | 465 (43.7%) | |||
| No | 1775 (60.0%) | 1175 (62.0%) | 600 (56.3%) | |||
| Baseline NRS 2002 score | 0.155 | |||||
| 0 | 503 (17.0%) | 345 (18.2%) | 158 (14.8%) | |||
| 1 | 604 (20.4%) | 386 (20.4%) | 218 (20.5%) | |||
| 2 | 641 (21.7%) | 395 (20.9%) | 246 (23.1%) | |||
| 3 | 480 (16.2%) | 305 (16.1%) | 175 (16.4%) | |||
| 4 | 416 (14.1%) | 265 (14.0%) | 151 (14.2%) | |||
| 5 | 240 (8.1%) | 150 (7.9%) | 90 (8.5%) | |||
| 6 | 64 (2.2%) | 41 (2.2%) | 23 (2.2%) | |||
| 7 | 11 (0.4%) | 7 (0.4%) | 4 (0.4%) |
| All patients (n = 2959) | ALB decreased at discharge (n = 1894) | ALB stable/increased at discharge (n = 1065) | P | Change in ALB M (SD) | P | |
|---|---|---|---|---|---|---|
| Change in inflammatory | < 0.001 | < 0.001 | ||||
| Inflammation-alleviated | 324 (21.1%) | 180 (9.5%) | 144 (13.5%) | − 0.87 (5.92) | ||
| Inflammation-worsened | 372 (12.6%) | 265 (14.0%) | 107 (10.0%) | − 3.57 (6.10) | ||
| Persistent inflammatory | 254 (8.6%) | 139 (7.3%) | 115 (10.8%) | − 1.26 (5.40) | ||
| Persistent non-inflammatory | 2009 (67.9%) | 1310 (69.2%) | 699 (65.6%) | − 2.64 (5.68) | ||
| Nutrition intervention | 0.018 | < 0.001 | ||||
| No | 1708 (57.7%) | 1062 (56.1%) | 646 (60.7%) | − 2.12 (4.98) | ||
| Parenteral nutrition | 672 (22.7%) | 433 (22.9%) | 239 (22.4%) | − 2.57 (6.61) | ||
| Enteral nutrition | 157 (5.3%) | 102 (5.4%) | 55 (5.2%) | − 2.49 (7.16) | ||
| Both | 422 (14.3%) | 297 (15.7%) | 125 (11.7%) | − 3.51 (6.69) | ||
| Disease Types | < 0.001 | < 0.001 | ||||
| Endocrine diseases | 40 (1.4%) | 26 (1.4%) | 14 (1.3%) | − 1.85 (4.62) | ||
| Neurological diseases | 388 (13.1%) | 253 (13.4%) | 135 (12.7%) | − 2.33 (6.09) | ||
| Musculoskeletal diseases | 171 (5.8%) | 136 (7.2%) | 35 (3.3%) | − 5.05 (5.41) | ||
| Gastrointestinal diseases | 720 (24.3%) | 420 (22.2%) | 300 (28.2%) | − 1.52 (6.06) | ||
| Thyroid and breast diseases | 76 (2.6%) | 45 (2.4%) | 31 (2.9%) | − 1.89 (3.75) | ||
| Benign respiratory diseases | 172 (5.8%) | 89 (4.7%) | 83 (7.8%) | − 0.83 (4.27) | ||
| Cardiovascular diseases | 57 (1.9%) | 34 (1.8%) | 23 (2.2%) | − 1.71 (4.22) | ||
| Malignant tumors | 1190 (40.2%) | 811 (42.8%) | 379 (35.6%) | − 3.16 (5.75) | ||
| Other | 145 (4.9%) | 80 (4.2%) | 65 (6.1%) | − 0.94 (5.80) | ||
| Inpatient department | < 0.001 | < 0.001 | ||||
| Orthopedics department | 148 (5.0%) | 124 (6.5%) | 24 (2.3%) | − 5.76 (5.03) | ||
| Respiratory department | 135 (4.6%) | 58 (3.1%) | 77 (7.2%) | − 0.01 (4.01) | ||
| Geriatrics department | 206 (7.0%) | 110 (5.8%) | 96 (9.0%) | − 0.58 (4.64) | ||
| Cardiology department | 50 (1.7%) | 28 (1.5%) | 22 (2.1%) | − 1.38 (3.99) | ||
| General surgery department | 1324 (44.7%) | 884 (46.7%) | 440 (41.3%) | − 2.82 (6.17) | ||
| Neurology department | 400 (13.5%) | 261 (13.8%) | 139 (13.1%) | − 2.35 (6.15) | ||
| Gastroenterology department | 268 (9.1%) | 136 (7.2%) | 132 (12.4%) | − 0.26 (4.82) | ||
| Cardiothoracic surgery department | 242 (8.2%) | 187 (9.9%) | 55 (5.2%) | − 4.87 (5.27) | ||
| Medical oncology department | 186 (6.3%) | 106 (5.6%) | 80 (7.5%) | − 1.43 (4.27) | ||
| Baseline hypoproteinemia | < 0.001 | < 0.001 | ||||
| No | 2400 (81.8%) | 1706 (90.1%) | 694 (65.2%) | − 3.55 (5.22) | ||
| Mild | 380 (12.8%) | 149 (7.9%) | 231 (21.7%) | 1.22 (5.06) | ||
| Moderate | 134 (4.5%) | 39 (2.1%) | 95 (8.9%) | 3.47 (5.82) | ||
| Severe | 45 (1.5%) | 0 (0.0%) | 45 (4.2%) | 8.34 (5.93) | ||
| Baseline malnutrition | < 0.001 | < 0.001 | ||||
| Yes | 896 (30.3%) | 533 (28.1%) | 363 (34.1%) | − 1.61 (6.09) | ||
| No | 2063 (69.7%) | 1361 (71.9%) | 702 (65.9%) | − 2.80 (5.61) | ||
| Baseline comorbidity | 0.039 | 0.014 | ||||
| Yes | 1850 (62.5%) | 1158 (61.1%) | 692 (65.0%) | − 2.24 (5.62) | ||
| No | 1109 (37.5%) | 736 (38.9%) | 373 (35.0%) | − 2.79 (6.06) |
Fig. 2.
Magnitude of changes in serum ALB in hospitalized patients with different categories of disease
Impact of different nutritional modalities on ALB and primary outcomes across baseline ALB levels
As shown in Table 2, regardless of admission hypoalbuminemia status (< 35 g/L), different nutritional support modalities significantly affected albumin changes. These effects persisted after adjusting for age, gender, and disease type. In hypoalbuminemic patients, the enteral nutrition group demonstrated the greatest ALB improvement, with a mean change of 6.67 ± 8.31 g/L.
Table 2.
Effects of different nutritional modalities on ALB and primary outcomes stratified by baseline ALB levels
| No (n = 1708) | PN (n = 672) | EN (n = 157) | Both (n = 422) | P | Pa | |
|---|---|---|---|---|---|---|
| A) All patients | ||||||
| Baseline ALB (g/L) | 40.07 (5.69) | 39.04 (6.83) | 40.99 (6.62) | 39.92 (6.38) | < 0.001 | |
| ALB at discharge (g/L) | 37.95 (5.66) | 36.50 (5.99) | 38.49 (5.82) | 36.35 (5.54) | < 0.001 | |
| Change in ALB (g/L) | − 2.12 (4.98) | − 2.57 (6.61) | − 2.49 (7.16) | − 3.51 (6.69) | < 0.001 | < 0.001 |
| 0.018 | < 0.001 | |||||
| ALB stable/increased at discharge | 646 (37.8%) | 239 (35.6%) | 55 (35.0%) | 125 (29.6%) | ||
| ALB decreased at discharge | 1062 (62.2%) | 433 (64.4%) | 102 (65.0%) | 297 (70.4%) | ||
| B) Subgroup analysis: baseline ALB < 35 g/L | ||||||
| Baseline ALB (g/L) | 31.01 (3.50) | 30.44 (3.90) | 30.37 (3.34) | 30.40 (3.67) | 0.311 | |
| ALB at discharge (g/L) | 32.56 (4.66) | 33.07 (5.91) | 37.04 (7.49) | 33.63 (5.66) | 0.002 | |
| Change in ALB (g/L) | 1.55 (4.80) | 2.63 (5.93) | 6.67 (8.31) | 3.23 (6.46) | < 0.001 | 0.003 |
| 0.508 | 0.480 | |||||
| ALB stable/increased at discharge | 181 (65.1%) | 119 (67.2%) | 17 (81.0%) | 54 (65.1%) | ||
| ALB decreased at discharge | 97 (34.9%) | 58 (32.8%) | 4 (19.0%) | 29 (34.9%) | ||
| C) Subgroup analysis: baseline ALB ≥ 35 g/L | ||||||
| Baseline ALB (g/L) | 41.83 (4.16) | 42.12 (4.68) | 42.63 (5.36) | 42.25 (4.44) | 0.090 | |
| ALB at discharge (g/L) | 38.99 (5.20) | 37.69 (5.48) | 38.72 (5.49) | 37.09 (5.22) | < 0.001 | |
| Change in ALB (g/L) | − 2.83 (4.69) | − 4.43 (5.80) | − 3.91 (5.83) | − 5.16 (5.63) | < 0.001 | < 0.001 |
| < 0.001 | < 0.001 | |||||
| ALB stable/increased at discharge | 465 (32.5%) | 120 (24.2%) | 38 (27.9%) | 71 (20.9%) | ||
| ALB decreased at discharge | 965 (67.5%) | 375 (75.8%) | 98 (72.1%) | 268 (79.1%) | ||
a: After adjusting for age, gender, and disease type
Effects of ALB levels on primary and secondary outcomes
As shown in Table 3, ALB changes significantly impacted primary outcomes (infection-related complications, non-infectious complications, and hospitalization costs) (all P < 0.05). Significant ALB decreases were also associated with in-hospital nutritional status changes, discharge weight loss, gastrointestinal symptom exacerbation, functional activity impairments, and edema development (all P < 0.05).
Table 3.
Effects of ALB levels on primary and secondary outcomes
| All patients (n = 2959) | ALB decreased at discharge (n = 1894) | ALB stable/increased at discharge (n = 1065) | P | X2 | Change in ALB M (SD) | P | |
|---|---|---|---|---|---|---|---|
| Primary Outcome | |||||||
| Infection-related complications | 0.009 | 6.854 | < 0.001 | ||||
| Yes | 81 (2.7%) | 63 (3.3%) | 18 (1.7%) | − 6.51 (8.06) | |||
| No | 2878 (97.3%) | 1831 (96.7%) | 1047 (98.3%) | − 2.33 (5.67) | |||
| Non-infection-related complications | 0.918 | 0.011 | 0.028 | ||||
| Yes | 65 (2.2%) | 42 (2.2%) | 23 (2.2%) | − 4.62 (7.92) | |||
| No | 2894 (97.8%) | 1852 (97.8%) | 1042 (97.8%) | − 2.39 (5.72) | |||
| Inpatient duration | 0.12 | ||||||
| 14.49 (5.98) | 14.62 (5.89) | 14.27 (6.13) | |||||
| hospitalization expenses | 0.013 | ||||||
| 39,430.7 (63,625.8) | 41,615.2 (30,506.1) | 35,542.9 (97,872.5) | |||||
| Secondary Outcome | |||||||
| Changes in nutritional status | < 0.001 | 20.826 | < 0.001 | ||||
| Improvement | 229 (7.7%) | 135 (7.1%) | 94 (8.8%) | − 1.62 (5.77) | |||
| Worsening | 328 (11.1%) | 246 (13.0%) | 82 (7.7%) | − 4.11 (6.27) | |||
| Status quo | 2402 (81.2%) | 1513 (79.9%) | 889 (83.5%) | − 2.29 (5.68) | |||
| Weight loss status | 0.001 | 13.694 | < 0.001 | ||||
| Improvement | 114 (3.9%) | 107 (5.6%) | 72 (6.8%) | − 1.56 (5.82) | |||
| Worsening | 382 (12.9%) | 276 (14.6%) | 106 (10.0%) | − 3.38 (6.11) | |||
| Status quo | 2398 (81.0%) | 1511 (79.8%) | 887 (83.3%) | − 2.36 (5.72) | |||
| Gastrointestinal symptoms | 0.002 | 12.976 | < 0.001 | ||||
| Improvement | 372 (12.6%) | 214 (11.3%) | 158 (14.8%) | − 1.62 (5.86) | |||
| Worsening | 265 (9.0%) | 189 (10.0%) | 76 (7.1%) | − 3.57 (5.96) | |||
| Status quo | 2322 (78.57%) | 1491 (78.7%) | 831 (78.0%) | − 2.44 (5.74) | |||
| Mobility and physical functional status | < 0.001 | 23.591 | < 0.001 | ||||
| Improvement | 287 (9.7%) | 166 (8.8%) | 121 (11.4%) | − 1.44 (4.01) | |||
| Worsening | 518 (17.5%) | 377 (19.9%) | 141 (13.2%) | − 3.44 (5.96) | |||
| Status quo | 2154 (72.8%) | 1351 (71.3%) | 803 (75.4%) | − 2.33 (5.819) | |||
| Muscle or subcutaneous fat depletion | 0.754 | 0.564 | 0.773 | ||||
| Improvement | 801 (27.1%) | 504 (26.6%) | 297 (27.9%) | − 2.34 (6.27) | |||
| Worsening | 317 (10.7%) | 204 (10.8%) | 113 (10.6%) | − 2.34 (5.06) | |||
| Status quo | 1841 (62.2%) | 1186 (62.6%) | 655 (61.5%) | − 2.50 (5.69) | |||
| Edema status | 0.040 | 6.456 | 0.14 | ||||
| Improvement | 54 (1.8%) | 26 (1.4%) | 28 (2.6%) | − 0.91 (5.59) | |||
| Worsening | 55 (1.9%) | 33 (1.7%) | 22 (2.1%) | − 2.27 (5.25) | |||
| Status quo | 1455 (96.3%) | 1835 (96.9%) | 1015 (95.3%) | − 2.47 (5.80) | |||
Based on Table 1 findings, we further stratified patients by inflammation severity, nutritional delivery methods, hospital departments, and baseline nutritional status. As shown in Table 4, all patients were categorized into three inflammation groups: inflammation-improved (n = 324), inflammation-worsened (n = 372), and persistent status (n = 2263, including both persistently asymptomatic and symptomatic patients). Results indicated that in persistent status patients, ALB decreases were significantly associated with complications, higher hospitalization costs, worsened malnutrition, progressive weight loss, exacerbated gastrointestinal symptoms, and reduced functional activity. However, these associations were attenuated in inflammation-improved/worsened groups, despite their ALB changes remaining significantly correlated with infection-related complications.
Table 4.
Impact of ALB levels on primary and secondary outcomes in patients with different inflammatory statuses
| Inflammation-alleviated (n = 324) | Inflammation-worsened (n = 372) | Persistent status (n = 2263) | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Primary Outcome | ALB stable/increased at discharge (n = 144) | ALB decreased at discharge (n = 180) | P | Change in ALB M (SD) | P | ALB stable/increased at discharge (n = 107) | ALB decreased at discharge (n = 265) | P | Change in ALB M (SD) | P | ALB stable/increased at discharge (n = 814) | ALB decreased at discharge (n = 1449) | P | Change in ALB M (SD) | P |
| Infection-related complications | 0.307 | 0.019 | 0.569 | < 0.001 | 0.029 | < 0.001 | |||||||||
| Yes | 2 (1.4%) | 6 (3.3%) | − 5.71 (9.39) | 3 (2.8%) | 12 (4.5%) | − 9.13 (10.25) | 13 (1.6%) | 45 (3.1%) | − 5.95 (7.24) | ||||||
| No | 142 (98.6%) | 174 (96.7%) | − 0.74 (5.77) | 104 (97.2%) | 253 (95.5%) | − 3.34 (5.77) | 801 (98.4%) | 1404 (96.9%) | − 2.39 (5.59) | ||||||
| Non–infection-related complications | 0.248 | 0.139 | 0.720 | 0.598 | 0.429 | < 0.001 | |||||||||
| Yes | 5 (3.5%) | 2 (1.1%) | 2.41 (10.74) | 3 (2.8%) | 6 (2.3%) | − 4.63 (8.05) | 15 (1.8%) | 34 (2.3%) | − 5.62 (5.61) | ||||||
| No | 139 (96.5%) | 178 (98.9%) | − 0.94 (5.78) | 104 (97.2%) | 259 (71.3%) | − 3.54 (6.06) | 799 (98.2%) | 1415 (97.7%) | − 2.41 (5.61) | ||||||
| Inpatient duration | 0.809 | 0.319 | 0.137 | ||||||||||||
| 14.69 (5.95) | 14.53 (5.85) | 14.19 (5.02) | 14.83 (5.81) | 14.20 (6.30) | 14.60 (5.92) | ||||||||||
| hospitalization expenses | 0.143 | < 0.001 | 0.093 | ||||||||||||
| 36,053.2 (26,797.9) | 40,533.6 (27,360.7) | 36,159.2 (22,909.9) | 47,935.0 (31,370.2) | 35,373.5 (110,945.1) | 40,601.1 (30,599.3) | ||||||||||
| Secondary Outcome | |||||||||||||||
| Changes in nutritional status | 0.046 | 0.077 | < 0.001 | < 0.001 | 0.004 | < 0.001 | |||||||||
| Improvement | 21 (14.6%) | 12 (6.7%) | 0.89 (4.15) | 6 (5.6%) | 22 (8.3%) | − 3.37 (5.69) | 67 (8.2%) | 101 (7.0%) | − 1.83 (5.93) | ||||||
| Worsening | 13 (9.0%) | 13 (7.2%) | 0.45 (7.68) | 4 (3.7%) | 54 (20.4%) | − 6.98 (5.36) | 65 (8.0%) | 179 (12.4%) | − 3.92 (5.99) | ||||||
| Status quo | 110 (76.4%) | 155 (86.1%) | − 1.21 (5.87) | 97 (90.7%) | 189 (71.3%) | − 2.90 (6.07) | 682 (83.8%) | 1169 (80.7%) | − 2.35 (5.57) | ||||||
| Weight loss status | 0.824 | 0.414 | 0.182 | 0.151 | 0.001 | < 0.001 | |||||||||
| Improvement | 10 (6.9%) | 10 (5.6%) | − 0.09 (4.38) | 10 (9.3%) | 18 (6.8%) | − 3.16 (6.42) | 52 (6.4%) | 79 (5.5%) | − 1.47 (5.83) | ||||||
| Worsening | 20 (13.9%) | 23 (12.8%) | − 0.01 (7.42) | 9 (8.4%) | 40 (15.1%) | − 5.15 (6.67) | 77 (9.5%) | 213 (14.7%) | − 3.58 (5.62) | ||||||
| Status quo | 114 (79.2%) | 147 (81.7%) | − 1.08 (5.74) | 88 (82.2%) | 207 (78.1%) | − 3.35 (5.96) | 685 (84.2%) | 1157 (79.8%) | − 2.38 (5.64) | ||||||
| Gastrointestinal symptoms | 0.138 | 0.006 | 0.213 | 0.259 | 0.010 | < 0.001 | |||||||||
| Improvement | 22 (15.3%) | 18 (10.0%) | − 0.62 (6.12) | 14 (13.1%) | 29 (10.9%) | − 2.83 (5.34) | 122 (15.0%) | 167 (11.5%) | − 1.58 (5.89) | ||||||
| Worsening | 8 (5.6%) | 5 (2.8%) | 4.15 (8.04) | 9 (8.4%) | 40 (15.1%) | − 4.80 (5.02) | 59 (7.2%) | 144 (9.9%) | − 3.78 (5.68) | ||||||
| Status quo | 114 (79.2%) | 157 (87.2%) | − 1.14 (5.68) | 84 (78.5%) | 196 (74.0%) | − 3.47 (6.37) | 633 (77.8%) | 1138 (78.57%) | − 2.48 (5.60) | ||||||
| Mobility and physical functional status | 0.128 | 0.641 | 0.009 | 0.043 | < 0.001 | < 0.001 | |||||||||
| Improvement | 25 (17.4%) | 18 (10.0%) | − 0.23 (5.38) | 7 (6.5%) | 23 (8.7%) | − 2.61 (3.32) | 89 (10.9%) | 125 (8.6%) | − 1.52 (5.09) | ||||||
| Worsening | 23 (16.0%) | 36 (20.0%) | − 0.57 (6.96) | 16 (15.0%) | 77 (29.1%) | − 4.92 (5.81) | 102 (12.5%) | 264 (18.2%) | − 3.53 (5.68) | ||||||
| Status quo | 96 (66.7%) | 126 (70.0%) | − 1.07 (5.73) | 84 (78.5%) | 165 (62.3%) | − 3.18 (6.40) | 623 (76.5%) | 1060 (73.2%) | − 2.38 (5.70) | ||||||
| Muscle or subcutaneous fat depletion | 0.307 | 0.637 | 0.958 | 0.858 | 0.998 | 0.980 | |||||||||
| Improvement | 48 (33.3%) | 46 (25.6%) | − 0.43 (7.19) | 21 (19.6%) | 53 (20.0%) | − 3.91 (7.21) | 228 (28.0%) | 405 (28.0%) | − 2.44 (5.94) | ||||||
| Worsening | 18 (12.5%) | 26 (14.4%) | − 0.71 (4.35) | 8 (7.5%) | 22 (8.3%) | − 3.31 (4.17) | 87 (10.7%) | 156 (10.8%) | − 2.52 (5.23) | ||||||
| Status quo | 78 (54.2%) | 108 (60.0%) | − 1.13 (5.53) | 78 (72.9%) | 190 (71.7%) | − 3.51 (5.97) | 499 (61.3%) | 888 (61.3%) | − 2.49 (5.61) | ||||||
| Edema status | 0.793 | 0.964 | 0.040 | 0.056 | 0.128 | 0.491 | |||||||||
| Improvement | 4 (2.8%) | 5 (2.8%) | − 0.54 (4.98) | 5 (4.7%) | 2 (0.8%) | 1.34 (7.46) | 19 (2.3%) | 19 (1.3%) | − 1.41 (5.39) | ||||||
| Worsening | 3 (2.1%) | 6 (3.3%) | − 1.29 (3.04) | 2 (1.9%) | 5 (1.5%) | − 1.05 (3.87) | 17 (2.1%) | 23 (1.6%) | − 2.67 (5.80) | ||||||
| Status quo | 137 (95.1%) | 169 (93.9%) | − 0.86 (6.01) | 100 (93.5%) | 259 (97.7%) | − 3.71 (6.07) | 778 (95.6%) | 1407 (97.1%) | − 2.50 (5.67) | ||||||
As shown in Supplementary Table S1, compared to non-intervention patients, those receiving nutritional support during hospitalization showed stronger associations between discharge ALB decreases and adverse clinical outcomes/nutritional deterioration. Among treated patients, both infection-related and non-infection-related complications demonstrated significant differences in ALB changes, with mean values of (− 6.96 ± 8.33 vs. − 2.70 ± 6.59, P < 0.001) and (− 5.55 ± 7.98 vs. − 2.78 ± 6.65, P = 0.007), respectively. Supplementary Table S2 reveals that in both surgical and medical patients, ALB changes were significantly associated with multiple clinical outcomes, including infection-related complications. However, in surgical patients, ALB increases showed no significant correlation with weight loss improvement. Supplementary Table S3 demonstrates that regardless of admission malnutrition diagnosis, ALB changes were associated with infectious complications (P < 0.001). In non-malnourished patients, ALB changes significantly correlated with non-infectious complications, hospitalization duration/costs, and nutritional status changes, which was not observed in malnourished patients. Supplementary Table S4 shows that in older adults patients (≥ 60 years), ALB changes exhibited stronger associations with both infection-related and non-infection-related complications than simple “increase/decrease” patterns. In addition, we performed a sensitivity analysis on exogenous ALB administration (Supplementary Table S5). ICU patients represent a subgroup highly likely to receive exogenous ALB. Comparing outcomes between non-ICU patients and ICU patients revealed that ALB changes remained significantly associated with outcomes in both groups (P < 0.05). This suggests that exogenous ALB infusion may supplement or maintain ALB levels but cannot counteract the overall impact of the underlying disease state on ALB.
Cutoff values for risk stratification of infection-related complications
As shown in Table 5, we conducted cutoff value analysis for ALB changes across different inflammatory statuses, nutritional intervention statuses, hospital departments, baseline nutritional statuses, and age groups. This approach aimed to establish a more precise risk stratification indicator for infection-related complications than simple “increase/decrease” categorization, particularly targeting patient populations, where ALB changes demonstrated significant associations with infectious complications as previously described. For example, in patients with worsened inflammatory status, when ALB decreased by ≥ 9.35 g/L, the infection complication rates reached 12.50% vs. 2.78% (P = 0.030) between high-risk and low-risk groups. In patients receiving nutritional therapy, a ≥ 10.80 g/L decrease in ALB correlated with 13.04% infection complication rate (P < 0.001). For surgical and medical patients, cutoff values of ≥ 11.0 g/L and ≥ 7.10 g/L, respectively, resulted in 10.06% (P < 0.001) and 6.30% (P = 0.044) infection rates. Malnourished patients showed 12.77% infection rate when ALB decreased by ≥ 9.00 g/L. In older adults and non-older adult patients, cutoff values of ≥ 9.35 g/L and ≥ 5.80 g/L, respectively, correlated with 11.04% (P < 0.001) and 4.83% (P < 0.001) infection rates. Figure 3 shows the distribution of patients in subgroups with different magnitudes of albumin change according to the presence or absence of infection-related complications.
Table 5.
Cutoff values for risk stratification of infection-related complications in different patient subgroups
| Subgroups | Cutoff Values | Number of patients | Infection-related complications | P | X2 |
|---|---|---|---|---|---|
| Inflammation-alleviated | 0.042 | 9.744 | |||
| High risk | ALB ≤ − 8.10 | 29 (8.95%) | 3 (10.34%) | ||
| Low risk | ALB > − 8.10 | 295 (91.05%) | 5 (1.69%) | ||
| Inflammation-worsened | 0.030 | 10.450 | |||
| High risk | ALB ≤ − 9.35 | 48 (12.90%) | 6 (12.50%) | ||
| Low risk | ALB > − 9.35 | 324 (87.10%) | 9 (2.78%) | ||
| Persistent status | < 0.001 | 26.860 | |||
| High risk | ALB ≤ − 6.86 | 489 (21.61%) | 28 (5.73%) | ||
| Low risk | ALB > − 6.86 | 1774 (78.39%) | 30 (1.69%) | ||
| Without nutritional intervention | < 0.001 | 33.100 | |||
| High risk | ALB ≤ − 7.49 | 234 (15.49%) | 14 (5.98%) | ||
| Low risk | ALB > − 7.49 | 1474 (84.51%) | 15 (1.02%) | ||
| With nutritional intervention | < 0.001 | 27.010 | |||
| High risk | ALB ≤ − 10.80 | 138 (11.03%) | 18 (13.04%) | ||
| Low risk | ALB > − 10.80 | 1113 (88.97%) | 34 (3.05%) | ||
| Surgical department | < 0.001 | 39.112 | |||
| High risk | ALB ≤ − 11.00 | 179 (10.44%) | 18 (10.06%) | ||
| Low risk | ALB > − 11.00 | 1535 (89.56%) | 33 (2.15%) | ||
| Medical department | 0.044 | 9.646 | |||
| High risk | ALB ≤ − 7.10 | 127 (10.20%) | 8 (6.30%) | ||
| Low risk | ALB > − 7.10 | 1118 (89.80%) | 22 (1.97%) | ||
| Baseline malnutrition | < 0.001 | 23.006 | |||
| High risk | ALB ≤ − 9.00 | 94 (10.49%) | 12 (12.77%) | ||
| Low risk | ALB > − 9.00 | 802 (89.51%) | 21 (2.62%) | ||
| Baseline non-malnutrition | < 0.001 | 26.153 | |||
| High risk | ALB ≤ − 6.10 | 529 (25.64%) | 27 (5.10%) | ||
| Low risk | ALB > − 6.10 | 1534 (74.36%) | 21 (1.37%) | ||
| Older adults patients | < 0.001 | 35.135 | |||
| High risk | ALB ≤ − 9.35 | 154 (10.29%) | 17 (11.04%) | ||
| Low risk | ALB > − 9.35 | 1342 (89.71%) | 35 (2.61%) | ||
| Non-older adults patients | < 0.001 | 23.048 | |||
| High risk | ALB ≤ − 5.80 | 414 (28.30%) | 20 (4.83%) | ||
| Low risk | ALB > − 5.80 | 1049 (71.70%) | 9 (0.86%) |
Fig. 3.
Distribution of ALB changes across patient subgroups (a stratification by inflammatory status; b stratification by nutritional therapy; c stratification by hospital department; d stratification by baseline nutritional status; e stratification by age group). (+) Infection-related complications occurred; (−) infection-related complications did not occur
Correlation analysis between ALB changes and other clinical indicators
As shown in Table 6, this study further analyzed the correlations between ALB changes and relevant clinical factors. Results demonstrated significant associations (P < 0.05) between ALB changes and multiple clinical characteristics, including age, hospital department, recent weight loss, reduced food intake, disease type, baseline handgrip strength, handgrip strength changes, calf circumference changes, baseline ALB, baseline globulin, presence of comorbidities, PLR changes, PNI, complication incidence, nutritional support modalities, malnutrition improvement degree, and hospitalization costs. These findings suggest that ALB changes, as a robust clinical predictor, may correlate with various nutritional indices, inflammatory statuses, and clinical outcomes during hospitalization. Scatter plots illustrating the relationships between relevant factors and ALB changes are presented in Fig. 4.
Table 6.
Correlation analysis between ALB changes and other clinical indicators
| Clinical characteristics | r | P |
|---|---|---|
| Sex | − 0.021 | 0.258 |
| Age | 0.045 | 0.015 |
| Inpatient department | − 0.208 | < 0.001 |
| Recent 3-month weight loss | 0.091 | < 0.001 |
| Recent 1-week reduction in diet | 0.080 | < 0.001 |
| Disease Types | − 0.070 | < 0.001 |
| Baseline calf circumference | − 0.036 | 0.062 |
| Baseline grip strength | − 0.094 | < 0.001 |
| Baseline BMI | − 0.030 | 0.102 |
| Grip strength change | 0.139 | < 0.001 |
| Calf circumference change | 0.052 | 0.007 |
| BMI change | 0.019 | 0.290 |
| Baseline nutritional status | 0.098 | < 0.001 |
| Baseline ALB | − 0.537 | < 0.001 |
| Baseline globulin | 0.049 | 0.008 |
| Discharge globulin | 0.062 | 0.001 |
| Change in globulin | 0.013 | 0.500 |
| Baseline Comorbidity | 0.057 | 0.002 |
| Change in nutritional status | 0.048 | 0.009 |
| Degree of improvement in weight loss | 0.033 | 0.075 |
| Degree of improvement in gastrointestinal symptoms | − 0.015 | 0.426 |
| Degree of improvement in subcutaneous fat and muscle wasting | − 0.012 | 0.521 |
| Degree of improvement in functional abnormalities | 0.020 | 0.286 |
| Degree of improvement in edema | − 0.036 | 0.047 |
| Baseline inflammatory status | 0.122 | < 0.001 |
| Inflammatory status at discharge | − 0.003 | 0.859 |
| Change in inflammatory status | − 0.059 | 0.001 |
| Baseline PLR | 0.066 | < 0.001 |
| PLR change | − 0.367 | < 0.001 |
| Baseline PNI | − 0.295 | < 0.001 |
| Infection-related complications | − 0.086 | < 0.001 |
| Non-infection-related complications | − 0.040 | 0.029 |
| ICU treatment | − 0.087 | < 0.001 |
| Total hospital stay (days) | − 0.052 | 0.005 |
| Total healthcare expenses (CNY) | − 0.250 | < 0.001 |
| Nutritional support method | − 0.091 | < 0.001 |
Fig. 4.
Scatter plots of relevant factors versus ALB changes (a grip strength change; b baseline ALB; c PLR change; d baseline PNI; e total healthcare expenses (CNY); f calf circumference change; g baseline grip strength; h age)
Multiple linear regression analysis for predicting ALB changes
As shown in Table 7, we further identified independent predictors of ALB changes through multiple linear regression analysis. Advanced age, surgical patients, lower baseline handgrip strength and its changes, higher baseline ALB and globulin levels, baseline PNI, baseline inflammatory status and its exacerbation, and ICU treatment experience were significant predictors of ALB decrease. Notably, baseline ALB showed the strongest correlation with ALB changes (β = − 0.671, P < 0.001), while baseline PNI was a positive predictor (β = 0.335, P < 0.001). In addition, patients with ICU treatment experience exhibited a greater reduction in ALB (β = − 0.071, P < 0.001). The model demonstrated an adjusted R2 of 0.421, Durbin–Watson statistic of 1.990, and all VIF < 2, indicating no multicollinearity issues. Figure 5 presents the P–P plot, which indicates that the test residuals conform to a normal distribution. Figure 6 shows a scatter plot of albumin change magnitude vs. standardized predicted values from regression.
Table 7.
Multiple linear regression analysis for predicting ALB changes
| R | R2 | Adjusted R2 | Standard estimation error | F change | P (F change) | Durbin–Watson statistic |
|---|---|---|---|---|---|---|
| 0.652a | 0.425 | 0.421 | 4.431 | 88.891 | < 0.001 | 1.99 |
| Clinical characteristics | Beta | t | P | 95% CI for Beta | VIF | |
|---|---|---|---|---|---|---|
| Age | − 0.057 | − 3.290 | 0.001 | − 0.034 | − 0.009 | 1.309 |
| Inpatient department | − 0.080 | − 4.534 | < 0.001 | − 1.342 | − 0.531 | 1.354 |
| Baseline grip strength | 0.038 | 2.150 | 0.032 | 0.002 | 0.041 | 1.401 |
| Grip strength change | 0.055 | 3.295 | 0.001 | 0.021 | 0.085 | 1.230 |
| Baseline ALB | − 0.671 | − 37.603 | < 0.001 | − 0.663 | − 0.597 | 1.399 |
| Baseline globulin | − 0.038 | − 2.413 | 0.016 | − 0.073 | − 0.008 | 1.108 |
| Change in inflammatory status | 0.088 | 4.285 | < 0.001 | 0.263 | 0.708 | 1.872 |
| Baseline inflammatory status | 0.080 | 3.882 | < 0.001 | 0.580 | 1.763 | 1.866 |
| PNI | 0.335 | 17.109 | < 0.001 | 0.072 | 0.091 | 1.682 |
| ICU treatment | − 0.071 | − 4.184 | < 0.001 | − 1.885 | − 0.682 | 1.250 |
Fig. 5.
P–P plot of this model
Fig. 6.
Scatter plot of albumin change magnitude versus standardized predicted values from regression
Discussion
This study reveals that dynamic changes in ALB concentrations among hospitalized patients not only reflect nutritional status but also serve as a comprehensive indicator of inflammatory status changes, metabolic imbalances, and clinical outcome deterioration. As the first study systematically exploring the dynamic associations between ALB changes and multidimensional clinical indicators, we demonstrate the unique value of dynamic ALB monitoring in risk stratification and personalized intervention. These findings further challenge the conventional view of ALB as a mere nutritional biomarker.
Notably, 64.0% of patients in this study exhibited decreased ALB levels at discharge. We initially investigated whether nutritional support influenced ALB changes during hospitalization. After adjusting for age, gender, and disease type, nutritional support modalities still significantly affected ALB changes across all patients. Subsequent subgroup analysis based on baseline ALB levels (≥ 35 g/L vs. < 35 g/L) showed that nutritional support significantly impacted ALB changes regardless of admission hypoalbuminemia status. This aligns with findings by Boesiger et al. [9] that nutritional support did not significantly correlate with ALB increase in this study, suggesting ALB elevation may be influenced by multiple factors, such as disease severity rather than solely reflecting nutritional responses. A previous randomized controlled trial (RCT) also found no significant outcome differences between ALB and crystalloid infusion groups in severe sepsis patients [13]. These conclusions concur with a recent ASPEN statement emphasizing that visceral proteins are not proven sensitive markers of adequate energy/protein intake and should not guide treatment decisions [4]. Remarkably, varying degrees of ALB decrease during hospitalization significantly correlated with infection-related complications irrespective of nutritional therapy. In patients receiving nutritional support, ALB changes significantly associated with nutritional status improvement. Beyond infection-related complications, ALB reduction also indicated non-infectious complications and higher costs. For patients without nutritional support, ALB changes correlated with hospital stay duration and weight loss but not other nutritional improvements.
ALB, functioning as an “inflammatory biomarker,” demonstrates levels influenced by the body’s inflammatory status. During inflammatory states, the release of inflammatory mediators and vasoactive substances increases vascular permeability, promoting plasma protein extravasation into interstitial spaces and subsequently reducing ALB levels. In addition, impaired hepatic synthesis of visceral proteins and enhanced protein catabolism during inflammation contribute to this phenomenon [14, 15]. This study involving 2959 patients demonstrated a significant correlation between ALB changes and inflammatory status variations during hospitalization (P < 0.001), aligning with findings by Arik Sheinenzon et al. [16]. Considering the impact of inflammatory dynamics on ALB levels, we categorized patients into three groups based on inflammatory trajectory: alleviated, exacerbated, and persistent inflammation groups. Results showed that only in the persistent inflammation group did ALB reduction significantly correlate with clinical outcomes, such as infection-related complications and malnutrition progression. In contrast, both alleviated and exacerbated inflammation groups required assessing ALB change magnitude to establish such correlations. This discrepancy may stem from passive ALB concentration elevation during inflammation alleviation due to interstitial edema resolution and hemoconcentration, which does not reflect nutritional improvement [17]. Conversely, during inflammation exacerbation, ALB reduction often occurs, and its elevation may indicate nutritional recovery. These findings suggest that simple “increase/decrease” patterns of ALB may not reflect nutritional status and clinical outcomes under varying inflammatory conditions, echoing previous research perspectives [8, 18, 19]. Therefore, we introduced ALB change magnitude to evaluate clinical outcomes and nutritional status variations.
Regardless of admission malnutrition diagnosis, ALB changes correlated with infectious complications, though in malnourished patients, simple “increase/decrease” patterns did not associate with clinical outcomes. For nutritional status changes, dynamic ALB variations significantly correlated with nutritional status in non-malnourished patients but not in malnourished counterparts. This disparity may originate from baseline metabolic reserve differences: non-malnourished patients with higher baseline ALB may exhibit more sensitive responses to nutritional interventions, while malnourished patients often suffer from chronic diseases or prolonged nutritional deficits impairing ALB synthesis. In such cases, further ALB reduction may be driven more by inflammation or pathological states rather than mere nutritional deficiency [20]. Furthermore, malnourished patients frequently exhibit chronic inflammation, where inflammatory mediators suppress ALB synthesis and accelerate catabolism, maintaining low ALB levels. ALB concentrations independently correlate with nutritional and inflammatory factors yet interact mutually, as inflammatory changes amplify nutritional impacts on body composition to varying degrees [18, 21]. Under these circumstances, inpatient nutritional interventions may struggle to rapidly reverse ALB levels, making changes more reflective of inflammatory burden than nutritional status. Therefore, dynamic ALB changes serve as sensitive indicators of nutritional intervention efficacy in non-malnourished patients, while in malnourished patients, they may better reflect inflammation and metabolic disorders, necessitating comprehensive assessment with other indicators. Future studies should develop a Nutrition–Inflammation Composite Score integrating ALB kinetics and inflammation status to differentiate nutritionally driven from inflammation-driven hypoalbuminemia.
Across the entire study cohort, we observed that declining ALB levels were significantly associated with both infection-related and non-infection-related complications, as well as increased hospitalization costs. Previous investigations by Sacks et al. [22] utilizing SGA demonstrated that lower baseline ALB correlated with infections, pressure ulcers, and nutrition-related readmissions. Similarly, Derek et al. [22] and Schwartz et al. [23] independently reported associations between baseline hypoalbuminemia and infectious complications following gastrointestinal and laryngeal surgeries, respectively. However, these prior studies primarily relied on static baseline measurements for prognostic predictions. Our current analysis extends these findings by establishing the dynamic relationship between changes in ALB concentrations and clinical outcomes, particularly in reflecting nutritional status fluctuations. Among the patients stratified by inflammatory status, the patients in the persistent inflammation subgroup showed a significant correlation between decreased ALB and the occurrence of infection complications. While inflammatory responses themselves contribute to complications and cost escalation, these findings reinforce ALB’s dual role as both a nutritional biomarker and inflammation-sensitive parameter [4]. Supporting this dual role, Hatice et al. [24] and Etienne et al. [25] previously reported prolonged hospital stays in pediatric ICU and orthopedic patients with baseline hypoalbuminemia compared to normoalbuminemic counterparts. This study employed dynamic evaluation based on changes in ALB levels. Although the decline in ALB was unrelated to the length of hospital stay in all patients included in this study, increases in ALB were significantly associated with shorter hospital stays in patients without prior nutritional intervention, surgical patients, non-baseline malnourished patients, and non-older adult’s patients.
The findings of this study indicate significant correlations between reduced ALB levels and the occurrence of infectious complications, as well as improvements in malnutrition status, among different departments. These discoveries suggest that ALB, as a “nutritional risk” biomarker, may better reflect clinical disease progression and the effectiveness of nutritional interventions in surgical and internal medicine patients. According to research by Derek et al. [22], patients with lower ALB levels are more prone to postoperative infections, and improvements in ALB are significantly correlated with the success of anti-infective treatments. Furthermore, the persistent stress induced by surgery may lead to a hypermetabolic state, making ALB more susceptible to changes, thereby potentially serving as an important indicator for clinically assessing infection risks and nutritional recovery [26]. Internal medicine patients also demonstrated significant associations between reduced ALB and infectious complications, as well as deterioration in malnutrition status. Among these, the incidence of ALB reduction during hospitalization in neurology patients was 65.2%. Previous literature reports that 66.1% of older adults stroke patients suffer from malnutrition during hospitalization. It is noteworthy that underlying diseases in older adult’s stroke patients may predispose them to malnutrition even before illness onset, and the incidence of malnutrition significantly increases when combined with dysphagia [27]. Therefore, reasonable nutritional diagnosis, treatment, and support can improve nutritional status, reduce the risk of primary diseases and complications, and ultimately improve clinical outcomes. However, due to the heterogeneity of internal medicine diseases, differences in patients’ underlying conditions, and variations in nutritional intervention measures, clinical outcomes vary across different diseases. Future research needs to further refine the classification of internal medicine patients, explore the clinical significance of ALB changes under specific disease states, and evaluate the effectiveness of different nutritional intervention strategies, with the aim of providing more precise guidance for nutritional interventions and infection management.
Correlation analysis between changes in ALB and clinical indicators revealed significant associations with baseline ALB, disease type, and changes in nutritional status. PLR reflects the intensity of systemic inflammatory responses; high PLR indicates inflammatory activation or immune imbalance, which can exacerbate ALB breakdown and vascular leakage, leading to decreased concentration [28]. PNI assesses nutritional reserves and immune status while predicting disease prognosis; low PNI suggests malnutrition and immune deficiencies, predisposing to infections [29]. Correlation analysis showed a significant negative correlation between baseline PNI and ALB reduction, suggesting a strong association between low ALB levels and clinical outcomes. Changes in ALB can serve as an important indicator for assessing nutritional recovery and infection risks. Both baseline PLR and changes in PLR were significantly associated with dynamic changes in ALB. In addition, patients’ inflammatory states derived from leukocyte counts were correlated with dynamic ALB changes at both baseline and variation levels. This further demonstrates that ALB changes are related to dynamic inflammatory changes and influenced by baseline inflammatory states. These findings indicate that changes in PNI and PLR may be involved in the pathological process of inflammation–nutrition imbalance, thereby affecting infection risks and clinical outcomes. Furthermore, results showed a significant positive correlation between ALB trends and baseline/discharge globulin levels. Serum globulins are mainly related to immune function and inflammatory responses, with immunoglobulins more affected by infections and immune status, while transferrin correlates with nutrition. Changes in globulin levels may be associated with immune system function, nutritional intake, and disease states. The ratio of ALB to globulin is an important indicator of liver function, further reflecting changes in hepatic synthetic function and immune status [30]. Notably, this study found a positive correlation between age and ALB variation in correlation analysis, whereas age emerged as a negative factor in linear regression analysis. Although linear regression adjusted results by introducing other variables, age remains a research focus regarding its relationship with ALB levels and nutritional status changes [31, 32]. Previous studies reported that ALB levels in healthy older adults (even over 90 years) typically remain within normal ranges (> 38 g/L), with age not directly causing hypoalbuminemia [33]. However, Ikuko et al. [34] found that ALB concentrations may slightly decrease annually at rates of 0.08–0.17 g/L, with greater declines in males. Despite focusing on hospitalized patients across multiple disciplines, Table S4 further demonstrates significant associations between ALB levels and nutritional status changes in older adult’s patients. We propose that age’s impact on ALB may manifest indirectly through inflammation, nutrition, or disease states rather than direct causation. This study identified a positive correlation between ALB reduction and declines in grip strength and calf circumference, suggesting a close association between low ALB levels and sarcopenia. Although some studies found no significant reduction in ALB in pediatric sarcopenia patients [4], Kübra et al.’s [35] cross-sectional study identified low ALB as an independent factor significantly associated with sarcopenia. This discrepancy may arise, because dynamic ALB changes more sensitively reflect visceral protein reserve depletion, while skeletal muscle mass is influenced by multiple factors. In certain disease states, selective protein loss may mobilize and decompose skeletal muscle proteins, impacting structural and functional integrity. Notably, this study found a negative correlation between ALB elevation and baseline grip strength/calf circumference. Potential explanations include individual variations and nutritional intervention effects. Therefore, combined monitoring of ALB and muscle indicators may provide a more comprehensive assessment of body protein metabolism. No significant association was observed between ALB reduction and body mass index (BMI) changes. Jessica et al.’s [36] meta-analysis similarly concluded that even in severely malnourished healthy patients, ALB and prealbumin levels remain normal until extreme starvation states (BMI < 12 or prolonged fasting exceeding 6 weeks). By dynamically monitoring ALB changes, clinicians can early identify high-cost-risk patients and implement preventive measures. In addition, ALB changes were significantly associated with functional status deterioration, suggesting its utility as a sensitive indicator for rehabilitation interventions [32].
This study revealed the relationship between changes in ALB levels and clinical outcomes by analyzing ALB variations in patients classified into different subgroups. Specifically, we found that changes in ALB levels more accurately reflect patients’ nutritional status and infection risk than simply relying on the “increase or decrease” of ALB. In inflammatory environments, ALB, as an acute-phase response protein, may not accurately reflect the patient’s true nutritional status [16]. Xin et al. [37] also concluded that hypoalbuminemia was not directly related to prognosis in all patients by studying the ALB trajectories of 1950 patients in the intensive care unit. Monitoring dynamic changes in ALB levels throughout treatment is critical, as timely correction of low ALB correlates with improved clinical outcomes. For all patients, tracking ALB trends minimizes misinterpretation from single-point measurements, offering a more accurate reflection of disease progression, nutritional recovery, and complication risks. This study established ALB cutoff values to stratify high-risk patients: a ≥ 6.10 g/L ALB decline in non-malnourished patients predicts infection-related complications, while malnourished individuals require combined nutritional/inflammatory markers for risk assessment. Based on the cutoff value, we propose a stratified management strategy for hospitalized patients based on dynamic changes in ALB. It is recommended that the low-risk group receive routine nutritional support, with a focus on preventing iatrogenic malnutrition and strengthening anti-inflammatory treatment when necessary. For high-risk groups, multidisciplinary intervention should be promptly implemented, combined with immunonutrition, and prioritized to control inflammation and correct microcirculatory disorders to address protein metabolism disorders. This approach avoids the socio-economic loss caused by blindly administering nutritional supplements such as ALB infusions and provides an important reference for clinicians to develop targeted treatment strategies [38].
This study is the first to quantify the clinical significance of the dynamic changes of ALB, breaking through the traditional limitation of only focusing on the baseline level. A multivariate prediction model was established to clarify the independent effects of baseline ALB, PNI and inflammation status on ALB trajectory, so as to provide a basis for individualized monitoring. Among them, baseline ALB and PLR change were both independent predictors of ALB change. These findings could aid clinical assessment of nutritional status, predict worsening dehydration, and set goals for nutritional improvement that go beyond inflammatory management. Prospective studies are needed to explore the ALB optimization scheme, and the clinical scene-specific ALB threshold can improve the operability of the results.
Nevertheless, this study has limitations. First, as a secondary analysis of a large prospective observational cohort, residual confounding is possible. Crucially, data on exogenous ALB administration during hospitalization were unavailable. This may obscure whether ALB changes reflect underlying pathophysiology due to two potential biases. On one hand, the true association between spontaneous ALB decline and adverse outcomes may be underestimated throughout the cohort, because ALB infusion artificially mitigates declines in high-risk patients. On the other hand, confounding by indication: patients receiving ALB typically have greater illness severity, lower baseline ALB, or higher complication risk (e.g., critically ill, severe hypoalbuminemia, cirrhosis, large-volume fluid resuscitation). If ALB is infused before an outcome occurs, it may confound the observed relationship between ALB changes and outcomes (e.g., infused patients may show less decline but still develop complications due to underlying severity). Although precise ALB usage data were lacking, ALB infusion is typically reserved for specific indications (e.g., severe hypoalbuminemia in critical illness, large-volume paracentesis in cirrhosis, and certain shock states) and such patients are likely concentrated in settings like ICUs. Sensitivity analyses in both non-ICU patients and ICU patients showed significant associations between ALB changes and outcomes, suggesting exogenous ALB may supplement/maintain levels but cannot fully counteract the overall impact of disease state. Nevertheless, prospectively collecting detailed ALB infusion data are crucial to definitively address this confounder. Second, only admission and discharge serum ALB levels were available, and the exact timing of complications during hospitalization was undetermined, making it difficult to establish the temporal sequence between ALB changes and complications. Third, inflammation status was defined using WBC count, whereas C-reactive protein is more commonly used clinically, potentially introducing bias. Fourth, the study did not examine how different nutritional interventions affect ALB changes, nutritional improvement, or clinical outcomes across patient subgroups; future multicenter RCTs should address this. Finally, the molecular mechanisms underlying ALB changes were not explored and require validation via isotope labeling experiments.
Conclusion
By integrating multidimensional clinical data, this study revealed the pivotal role of dynamic changes in ALB in the inflammation–nutrition network. Changes in ALB not only reflect the patient’s nutritional status but are also closely related to clinical outcomes, such as the occurrence of complications, hospitalization costs, and length of stay. Its dynamic changes are key indicators reflecting the interaction between inflammation and nutrition. Especially, for older adult’s patients and those with severe inflammatory states, changes in ALB levels have important value as a clinical predictive marker. Future research needs to further verify individualized intervention strategies guided by changes in ALB levels.
Supplementary Information
Author contributions
Yonghao Li: methodology, formal analysis, investigation, data curation, writing (original draft),writing (review and editing) and visualization. Liru Chen: conceptualization, writing (review and editing). Xin Yang: methodology, writing (review and editing), Hongyuan Cui: conceptualization, investigation. Zijian Li: formal analysis, data curation. Wei Chen: investigation. Hanping Shi: writing (review and editing) and supervision. Mingwei Zhu: conceptualization, methodology, formal analysis, investigation, writing (review and editing), and supervision.
Funding
This study was supported by the National Key Research and Development Program of the 13th 5-Year Plan (2022YFC2010101), the Medical and Health Science and Technology Innovation Project of the Chinese Academy of Medical Sciences (2021-12 M-C&T-B-094), and the Food Science and Technology Fund of the Chinese Society of Food Science and Technology (2020–14).
Data availability
No datasets were generated or analysed during the current study.
Declarations
Ethics approval and consent to participate
All patients signed informed consent forms and agreed to participate in the study.
Conflict of interest
The authors declare no competing interests.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Wu JM, Ho TW, Lai IR, Chen CN, Lin MT. Parenteral glutamine supplementation improves serum albumin values in surgical cancer patients. Clin Nutr. 2021;40(2):645–50. [DOI] [PubMed] [Google Scholar]
- 2.Gupta D, Lis CG. Pretreatment serum albumin as a predictor of cancer survival: a systematic review of the epidemiological literature. Nutr J. 2010;9:69. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Bretschera C, Boesiger F, Kaegi-Braun N, Hersberger L, Lobo DN, Evans DC, Tribolet P, Gomes F, Hoess C, Pavlicek V, Bilz S, Sigrist S, Brändle M, Henzen C, Thomann R, Rutishauser J, Aujesky D, Rodondi N, Donzé J, Stanga Z, Mueller B, Schuetz P. Admission serum albumin concentrations and response to nutritional therapy in hospitalised patients at malnutrition risk: Secondary analysis of a randomised clinical trial. EClinicalMedicine. 2022;45: 101301. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Evans DC, Corkins MR, Malone A, Miller S, Mogensen KM, Guenter P, Jensen GL. ASPEN Malnutrition Committee. The use of visceral proteins as nutrition markers: an ASPEN Position Paper. Nutr Clin Pract. 2021;36(1):22–8. [DOI] [PubMed] [Google Scholar]
- 5.Schuetz P, Seres D, Lobo DN, Gomes F, Kaegi-Braun N, Stanga Z. Management of disease-related malnutrition for patients being treated in hospital. Lancet. 2021;398(10314):1927–38. [DOI] [PubMed] [Google Scholar]
- 6.Qi XY, Tan F, Liu MX, Xu K, Gao P, Yao ZD, Zhang N, Yang H, Zhang CH, Xing JD, Cui M, Su XQ. Serum and peritoneal biomarkers for the early prediction of symptomatic anastomotic leakage in patients following laparoscopic low anterior resection: a single-center prospective cohort study. Cancer Rep (Hoboken). 2023;6(4): e1781. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Wu H, Wang Y, Deng M, Zhai Z, Xue D, Luo F, Li H. Preoperative inflammatory markers and tumor markers in predicting lymphatic metastasis and postoperative complications in colorectal cancer: a retrospective study. BMC Surg. 2025;25(1):71. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Loftus TJ, Brown MP, Slish JH, Rosenthal MD. Serum levels of prealbumin and albumin for preoperative risk stratification. Nutr Clin Pract. 2019;34(3):340–8. [DOI] [PubMed] [Google Scholar]
- 9.Boesiger F, Poggioli A, Netzhammer C, Bretscher C, Kaegi-Braun N, Tribolet P, Wunderle C, Kutz A, Lobo DN, Stanga Z, Mueller B, Schuetz P. Changes in serum albumin concentrations over 7 days in medical inpatients with and without nutritional support. A secondary post-hoc analysis of a randomized clinical trial. Eur J Clin Nutr. 2023;77(10):989–97. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Cho SY, Han J, Cha SH, Yoon SI. Structural basis of serum albumin recognition by SL335, an antibody Fab extending the serum half-life of protein therapeutics. Biochem Biophys Res Commun. 2020;526(4):941–6. [DOI] [PubMed] [Google Scholar]
- 11.Ren SS, Zhang KW, Chen BW, Yang C, Xiao R, Li PG, Zhu MW. Machine Learning-Based Prediction of Complications and Prolonged Hospitalization with the GLIM Criteria Combinations Containing Calf Circumference in Elderly Asian Patients. Nutrients. 2023;15(19):4146. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Kondrup J, Allison SP, Elia M, Vellas B, Plauth M. Educational and Clinical Practice Committee, European Society of Parenteral and Enteral Nutrition (ESPEN). ESPEN guidelines for nutrition screening 2002. Clin Nutr. 2003;22(4):415–21. [DOI] [PubMed] [Google Scholar]
- 13.Caironi P, Tognoni G, Masson S, Fumagalli R, Pesenti A, Romero M, Fanizza C, Caspani L, Faenza S, Grasselli G, Iapichino G, Antonelli M, Parrini V, Fiore G, Latini R, Gattinoni L. ALBIOS Study Investigators. Albumin replacement in patients with severe sepsis or septic shock. N Engl J Med. 2014;370(15):1412–21. [DOI] [PubMed] [Google Scholar]
- 14.Rozga J, Piątek T, Małkowski P. Human albumin: old, new, and emerging applications. Ann Transplant. 2013;18:205–17. [DOI] [PubMed] [Google Scholar]
- 15.Gabay C, Kushner I. Acute-phase proteins and other systemic responses to inflammation. N Engl J Med. 1999;340(6):448–54. [DOI] [PubMed] [Google Scholar]
- 16.Sheinenzon A, Shehadeh M, Michelis R, Shaoul E, Ronen O. Serum albumin levels and inflammation. Int J Biol Macromol. 2021;184:857–62. [DOI] [PubMed] [Google Scholar]
- 17.Gonzales GB, Njunge JM, Gichuki BM, Wen B, Ngari M, Potani I, Thitiri J, Laukens D, Voskuijl W, Bandsma R, Vanmassenhove J, Berkley JA. The role of albumin and the extracellular matrix on the pathophysiology of oedema formation in severe malnutrition. EBioMedicine. 2022;79: 103991. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Don BR, Kaysen G. Serum albumin: relationship to inflammation and nutrition. Semin Dial. 2004;17(6):432–7. [DOI] [PubMed] [Google Scholar]
- 19.Sharma K, Mogensen KM, Robinson MK. Pathophysiology of Critical Illness and Role of Nutrition. Nutr Clin Pract. 2019;34(1):12–22. [DOI] [PubMed] [Google Scholar]
- 20.Muscaritoli M, Imbimbo G, Jager-Wittenaar H, Cederholm T, Rothenberg E, di Girolamo FG, Amabile MI, Sealy M, Schneider S, Barazzoni R, Biolo G, Molfino A. Disease-related malnutrition with inflammation and cachexia. Clin Nutr. 2023;42(8):1475–9. [DOI] [PubMed] [Google Scholar]
- 21.Eckart A, Struja T, Kutz A, Baumgartner A, Baumgartner T, Zurfluh S, Neeser O, Huber A, Stanga Z, Mueller B, Schuetz P. Relationship of nutritional status, inflammation, and serum albumin levels during acute illness: a prospective study. Am J Med. 2020;133(6):713-722.e7. [DOI] [PubMed] [Google Scholar]
- 22.Sacks GS, Dearman K, Replogle WH, Cora VL, Meeks M, Canada T. Use of subjective global assessment to identify nutrition-associated complications and death in geriatric long-term care facility residents. J Am Coll Nutr. 2000;19(5):570–7. [DOI] [PubMed] [Google Scholar]
- 23.Hennessey DB, Burke JP, Ni-Dhonochu T, Shields C, Winter DC, Mealy K. Preoperative hypoalbuminemia is an independent risk factor for the development of surgical site infection following gastrointestinal surgery: a multi-institutional study. Ann Surg. 2010;252(2):325–9. [DOI] [PubMed] [Google Scholar]
- 24.Ari HF, Turanli EE, Yavuz S, Guvenc K, Avci A, Keskin A, Aslan N, Yildizdas D. Association between serum albumin levels at admission and clinical outcomes in pediatric intensive care units: a multi-center study. BMC Pediatr. 2024;24(1):844. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Flamant EM, Goltz DE, Burnett RA, Wickman JR, Belay ES, Saltzman EB, Nicholson GP, Garrigues GE, Lassiter T, Anakwenze OA, Klifto CS. Malnutrition in elective shoulder arthroplasty: a multi-institutional retrospective study of preoperative albumin and adverse outcomes. J Shoulder Elbow Surg. 2021;30(11):2491–7. [DOI] [PubMed] [Google Scholar]
- 26.Wang R, He M, Yin W, Liao X, Wang B, Jin X, Ma Y, Yue J, Bai L, Liu D, Zhu T, Huang Z, Kang Y. The Prognostic Nutritional Index is associated with mortality of COVID-19 patients in Wuhan, China. J Clin Lab Anal. 2020;34(10): e23566. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Hu S, Gao M, He Y, Xie X. Comparison of the performance of different nutritional indicators for predicting poststroke depression in older adults with ischemic stroke. Int J Nurs Sci. 2024;11(3):349–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Xu N, Zhang JX, Zhang JJ, Huang Z, Mao LC, Zhang ZY, Jin WD. The prognostic value of the neutrophil-to-lymphocyte ratio (NLR) and platelet-to-lymphocyte ratio (PLR) in colorectal cancer and colorectal anastomotic leakage patients: a retrospective study. BMC Surg. 2025;25(1):57. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Zheng J, Zheng L, Wang X, Mao X, Wang Q, Yang Y, Mo D. The Clinical Value of the Combined Detection of Systemic Immune-Inflammation Index (SII), Systemic Inflammation Response Index (SIRI), and Prognostic Nutritional Index (PNI) in Early Diagnosis of Gastric Cancer. J Inflamm Res. 2025;18:813–26. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Li H, Dai J, Lan T, Liu H, Wang J, Cai B, Xu L, Yuan K, Wang G, Wu H. Combination of albumin-globulin score and skeletal muscle index predicts long-term outcomes of intrahepatic cholangiocarcinoma patients after curative resection. Clin Nutr. 2021;40(6):3891–900. [DOI] [PubMed] [Google Scholar]
- 31.Agarwal E, Miller M, Yaxley A, Isenring E. Malnutrition in the elderly: a narrative review. Maturitas. 2013;76(4):296–302. [DOI] [PubMed] [Google Scholar]
- 32.Cabrerizo S, Cuadras D, Gomez-Busto F, Artaza-Artabe I, Marín-Ciancas F, Malafarina V. Serum albumin and health in older people: review and meta analysis. Maturitas. 2015;81(1):17–27. [DOI] [PubMed] [Google Scholar]
- 33.Campion EW, deLabry LO, Glynn RJ. The effect of age on serum albumin in healthy males: report from the Normative Aging Study. J Gerontol. 1988;43(1):M18-20. [DOI] [PubMed] [Google Scholar]
- 34.Gom I, Fukushima H, Shiraki M, Miwa Y, Ando T, Takai K, Moriwaki H. Relationship between serum albumin level and aging in community-dwelling self-supported elderly population. J Nutr Sci Vitaminol (Tokyo). 2007;53(1):37–42. [DOI] [PubMed] [Google Scholar]
- 35.Erdoğan K, Kara M, Şener FE, Durmuş ME, Durmuşoğlu BNÇ, Abdulsalam AJ, Sezer S, Kara Ö, Kaymak B, Özçakar L. Serum albumin as a biomarker of (nutritional status in) sarcopenia. J Bone Miner Metab. 2024;2024:1–6. [DOI] [PubMed] [Google Scholar]
- 36.Bouillanne O, Hay P, Liabaud B, Duché C, Cynober L, Aussel C. Evidence that albumin is not a suitable marker of body composition-related nutritional status in elderly patients. Nutrition. 2011;27(2):165–9. [DOI] [PubMed] [Google Scholar]
- 37.Tie X, Zhao Y, Sun T, Zhou R, Li J, Su J, Yin W. Associations between serum albumin level trajectories and clinical outcomes in sepsis patients in ICU: insights from longitudinal group trajectory modeling. Front Nutr. 2024;11:1433544. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Yu YT, Liu J, Hu B, Wang RL, Yang XH, Shang XL, Wang G, Wang CS, Li BL, Gong Y, Zhang S, Li X, Wang L, Shao M, Meng M, Zhu F, Shang Y, Xu QH, Wu ZX, Chen DC. Expert consensus on the use of human serum albumin in critically ill patients. Chin Med J (Engl). 2021;134(14):1639–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
No datasets were generated or analysed during the current study.