A Preliminary Predictive Panel for Pre-frailty Based on Serum Proteomic Biomarkers: A Two-Phase Cross-Sectional Study

A Preliminary Predictive Panel for Pre-frailty Based on Serum Proteomic Biomarkers: A Two-Phase Cross-Sectional Study | 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 Research Article A Preliminary Predictive Panel for Pre-frailty Based on Serum Proteomic Biomarkers: A Two-Phase Cross-Sectional Study Yu Ye, Jinwei Liu, Zhen Zhang, Shuaixuan Xu, Chenghao Chang, Mengyu Cao, and 5 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9388693/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 6 You are reading this latest preprint version Abstract Background Frailty is associated with anomalies in protein metabolism; however, the underlying serum proteomic signatures and their driving mechanisms remain unclear. This study aims to uncover key protein profiles during the transition from robust health to frailty and evaluate their potential for early identification through serum proteomic analysis. Methods Well-matched older adults were enrolled and categorized into robust, prefrail, and frail groups. In a cross-sectional design, untargeted proteomic screening was first performed in a discovery cohort of 30 participants, followed by targeted validation of candidate biomarkers using parallel reaction monitoring (PRM) in an independent validation cohort of 99 individuals. Multidimensional clinical parameters and differentially expressed proteins were integrated within machine learning pipelines to refine the search for characteristic features of prefrailty. Results In the discovery phase, 166 proteins were found to be differentially expressed across frailty statuses, with 15 significantly frailty-associated proteins (e.g., CP, VWF, UBE2O, and ACTN3) subsequently confirmed by PRM validation. These proteins were functionally enriched in pathways related to inflammation/immunity, coagulation, muscle structure, and protein metabolism. A Random Forest model, further assembled from gait speed, skeletal muscle mass, UBE2O, Timed Up and Go test, ACTN3, and Mini-Mental State Examination score, exhibited the most robust performance for early frailty identification among multiple algorithms compared (AUC: 0.896, 95% CI: 0.792–0.971). Conclusions This study reveals a panel of serum protein biomarkers closely linked to frailty progression. Through machine learning algorithms integrating clinical indices with circulating UBE2O and ACTN3, we evaluated the discriminatory capacity of these features for frailty, shedding light on the heterogeneity and protein alterations among community-dwelling older adults. Frailty Pre-frailty Older Adults Proteomics Machine Learning Figures Figure 1 Figure 3 Figure 4 Highlights 1.Serum proteomic analysis reveals a distinct panel of biomarkers associated with frailty progression. 2.Novel protein markers such as UBE2O and ACTN3 demonstrate predictive value for early‑stage frailty. 3. The integration of clinical indices and protein markers within a machine learning model establishes an interpretable framework for the classification of pre-frailty. Introduction Frailty is a geriatric syndrome characterized by decreased physiological reserve, increased vulnerability, and reduced resistance to stressors due to multi-system dysregulation 1 . Focusing on the global proportion of frailty prevalence, approximately 18% of older adults meet the criteria for frailty, while the proportion of prefrailty, the transitional phase from robustness to frailty, reaches 49% 2 . The transition of older adults to frailty poses a significant threat to the aging process and is undoubtedly one of the most serious public health challenges of the 21st century 3 , 4 . Notably, pre‑frailty is a dynamic and reversible state, making its effective identification and timely intervention crucial for improving health outcomes in the elderly 5 . The pathological changes underlying early frailty are closely linked to dysregulated protein expression, with core mechanisms such as chronic inflammation, energy metabolism disturbances, cellular senescence, and skeletal muscle homeostasis imbalance all exhibiting corresponding protein‑level alterations 6 – 8 . For instance, elevated levels of classic inflammatory markers such as C‑reactive protein (CRP) and interleukin‑6 (IL‑6) contribute directly to muscle functional decline through inflammatory catabolism in skeletal muscle 9 . A proteomic study of older women identified 32 pro‑inflammatory proteins overexpressed in frail individuals, among which eight core proteins were significantly correlated with frailty indices and functionally implicated in multi‑system changes including renal homeostasis, skeletal muscle regulation, and immune response 10 。Furthermore, proteins involved in lipid metabolism and energy balance, such as fatty acid‑binding proteins and leptin, also directly influence muscle development and function in frail individuals 11 . These findings collectively demonstrate that the development of frailty is accompanied by extensive and dynamic alterations in the proteome 12 . Despite the recognized role of proteins in the pathogenesis of frailty, existing studies have largely focused on individual biomarkers or isolated pathways, leaving the network‑level interactions among proteins in pre‑frailty insufficiently explored. In recent years, the integration of clinical rapid assessments with machine‑learning algorithms have shown promise in translating high‑throughput omics data into clinically actionable insights 13 , 14 . In this study, mass spectrometry-based proteomics was employed to screen for key proteins associated with frailty. By incorporating differentially expressed proteins into algorithmic analyses based on clinical features of frailty, a preliminary interpretable classification framework for prefrailty was established, providing a reference for the precise identification of prefrailty. Methods Study design Between December 2022 and June 2023, participants were consecutively recruited from community settings in Beijing, China, with the aim of identifying serum characteristics associated with frailty in older adults. The overall workflow is illustrated in the accompanying Fig. 1. During the discovery phase, 10 individuals each from the robust, prefrail, and frail groups were selected for clinical data collection and serum sampling. Untargeted proteomics screening was then performed to identify potential biomarkers. In the validation phase, targeted proteomic validation was carried out using 99 independent participants, comprising 20 robust, 59 prefrail, and 20 frail older adults. Based on data from this validation cohort, feature selection and comparison of multiple machine‑learning algorithms were employed to determine the optimal combination for identifying early frailty. All eligible participants provided written informed consent. The study was conducted in accordance with the principles of the Declaration of Helsinki and received formal approval from the Ethics Committee of the Chinese PLA General Hospital (Approval No. S2019‑140‑03). Assessment of Frailty Frailty was assessed using the Fried Physical Frailty Phenotype, developed by Fried et al. in 2001 15 . This tool evaluates five core components: unintentional weight loss, reduced grip strength, slowed walking speed, low physical activity, and self‑reported exhaustion. Based on the assessment results, participants were categorized as follows: those meeting three or more criteria were classified as frail, those meeting one to two criteria as prefrail, and those meeting none as robust. Clinical Assessments Clinical data included gender, age, years of education, body mass index (BMI), smoking history, alcohol consumption, insomnia status, sedentary behavior, Mini-Mental State Examination (MMSE) score 16 , Fried frailty score, number of chronic diseases 17 , pain severity measured with the Visual Analogue Scale (VAS) 18 , nutritional status evaluated by the Mini-Nutritional Assessment-Short Form (MNA-SF) 19 , depressive symptoms measured by the Geriatric Depression Scale-15 (GDS-15) 20 , and activities of daily living assessed by the Modified Barthel Index (MBI) 21 . Functional measurements included customary gait speed, grip strength, Timed Up and Go Test (TUG) time, and the Short Physical Performance Battery (SPPB) 22 . Body composition was measured via direct segmental multi‑frequency bioimpedance analysis (DSM‑BIA) using the InBody 770 device (Korea). Parameters obtained included total body protein, body fat mass (BFM), skeletal muscle mass (SLM). Physiological reserve indicators comprised serum levels of vitamin D (Vit D), cystatin C (CysC), insulin‑like growth factor‑1 (IGF‑1), CRP, creatine kinase (CK), creatinine (Cr), and triglycerides (TG) 23 . Serum Sample Preparation Blood samples were collected from participants under fasting conditions using vacuum tubes, with a volume of 5 mL per sample. Within one hour after collection, the blood was centrifuged at 3,000 rpm for 10 minutes to separate the serum. The supernatant was carefully aliquoted and stored at − 80°C until analysis. For subsequent testing, frozen serum samples were thawed to room temperature prior to processing. If turbidity was observed, the samples were re‑centrifuged at 3,500 rpm for 10 minutes, and the clarified supernatant was used for further assays. Untargeted proteomic Untargeted proteomic profiling was performed using data-independent acquisition (DIA) for quantitative analysis 24 . After protein extraction with lysis buffer (8 M urea containing protease inhibitors), high‑abundance proteins were removed using an Agilent MARS Hu‑14 immunoaffinity column to enrich low‑abundance proteins from the flow‑through fraction. Protein concentration was determined by the BCA assay. Subsequent sample processing included reduction (10 mM DTT, 56°C), alkylation (55 mM IAA, in the dark), and digestion via the filter‑aided sample preparation (FASP) method. The resulting peptides were desalted using a C18 column prior to analysis. Peptide separation was carried out on an EASY‑nLC 1200 nano‑LC system equipped with a C18 analytical column, using mobile phase A (0.1% formic acid in water) and mobile phase B (0.1% formic acid in 80% acetonitrile). The gradient elution program was as follows: 7–12% B, 12–30% B, 30–45% B, 45–95% B, and maintained until 60 min. Mass spectrometry analysis was performed on a Q Exactive HF mass spectrometer equipped with a Nanospray Flex ion source. The spray voltage was set to 2.2 kV and the ion transfer temperature to 320°C. DIA acquisition parameters included a full‑scan range of 350–1150 m/z at a resolution of 120,000, followed by 30 MS/MS windows acquired at a resolution of 30,000 with a normalized collision energy of 33%. DIA data were processed using Spectronaut software against the UniProt human database. Search settings included trypsin digestion with up to two missed cleavages, fixed modification of carbamidomethylation on cysteine, and variable modifications of methionine oxidation and N‑terminal acetylation. Both peptide and protein identifications were filtered at a false discovery rate (FDR) < 1%. Label‑free quantification was performed based on MS/MS spectra. Targeted Proteomic Targeted proteomic quantification was performed using mass spectrometry-based parallel reaction monitoring (PRM). This method enables both quantitative and qualitative analysis of specific proteins by monitoring the ion signals of predetermined target peptides, with at least two unique peptides selected per protein 25 . Following tryptic digestion, peptides were separated by liquid chromatography and subsequently analyzed by mass spectrometry. PRM acquisition comprises two main steps: Precursor Ion Selection—specific precursor ions corresponding to the target peptides were isolated based on their predefined mass-to-charge (m/z) ratios; and Fragment Ion Monitoring—after fragmentation, the resulting product ions were recorded. Quantification was performed based on the signal intensity of these fragment ions, allowing relative or absolute protein abundance to be determined. Analyses were carried out on a Q-Exactive high-resolution mass spectrometer, which offers precise mass accuracy and high sensitivity to ensure reliable detection of low-abundance proteins. Peak identification was performed using chromatographic and spectral data generated from the instrument. The signal intensity (peak area) of each peptide was extracted for quantification. Where applicable, absolute quantification was achieved by constructing a calibration curve using synthetic peptide standards of known concentration. Statistical Analysis Data were processed and analyzed using IBM SPSS Statistics 27.0. All statistical tests were two‑sided, with P < 0.05 considered statistically significant. Missing values were imputed five times via Markov Chain Monte Carlo (MCMC) multiple imputation. Outlier treatment was performed in conjunction with the actual pathophysiological status of the participants. When outliers were consistent with pathological upregulation, they were considered valid data and retained for analysis. Normality of continuous variables, including age, years of education, BMI, MMSE score, Fried score, number of chronic diseases, VAS score, MNA‑SF score, GDS‑15 score, MBI step count, grip strength, TUG time, SPPB score, body composition measures, and routine hematological parameters, was assessed using the Shapiro–Wilk test across the three groups. For variables following a normal distribution, one‑way analysis of variance (ANOVA) was used for group comparisons; otherwise, the Kruskal–Wallis rank‑sum test was applied. Continuous variables with a normal distribution were expressed as mean ± standard deviation (Mean ± SD), while those with a non‑normal distribution were presented as median (interquartile range) [M (P25–P75)]. Categorical variables, including gender, smoking history, alcohol consumption, insomnia status, and sedentary behavior, were described using constituent ratios and analyzed using the chi‑squared test. Spearman’s rank correlation analysis was performed to evaluate the correlations between differentially expressed proteins and clinical characteristics. Bioinformatics Analysis Quantitative peptide data obtained from proteomic profiling were processed using DIA‑NN software against the UniProt human proteome database. Peptide intensities were aggregated into protein‑level relative expression values and represented as log2‑transformed label‑free quantities. Following quantile normalization and missing‑value filtering, differentially expressed proteins among the three study groups were initially identified by Kruskal‑Wallis test (FDR 0.58 and FDR < 0.05 were considered statistically significant differentially expressed proteins. To further explore their dynamic expression patterns, these significant proteins were subjected to Mfuzz clustering. Pathway enrichment analysis of differentially expressed proteins was performed using the Gene Ontology (GO) database ( http://www.geneontology.org/ ) and the Kyoto Encyclopedia of Genes and Genomes (KEGG) database ( https://www.kegg.jp/kegg/pathway.html ). For GO analysis, enriched terms were identified by comparing the frequency of differentially expressed proteins assigned to each term against the background proteome, with significantly over‑represented terms ( P < 0.05) reported to highlight biological functions, cellular components, and molecular processes associated with frailty‑related proteomic changes. For KEGG analysis, enriched pathways ( P < 0.05) were visualized in bubble plots, where the enrichment score (rich factor), P ‑value, and number of mapped proteins were integrated to identify key biological pathways involved in frailty progression. Protein-protein interaction (PPI) network analysis was performed using the STRING database ( https://string-db.org/ ). Genomic associations between proteins were calculated based on a scoring framework that integrates different types of association benchmarks derived from commonly used reference datasets, with each prediction assigned a single confidence score. Modular analysis was subsequently conducted to infer biological processes. Feature Selection The dependent variable was defined as prefrailty, with a Fried phenotype score of 1–2 serving as the outcome of interest. Robust and frail groups were included as controls to help exclude false‑positive proteins potentially arising from non‑specific aging in the older adult population. A total of 26 candidate parameters were considered as independent variables, among which those showing statistically significant differences across the three groups in the validation cohort were selected ( P < 0.05). These included 11 clinical parameters, namely sedentary behavior, MMSE, MBI, gait speed, grip strength, TUG, SPPB, protein, SLM, CysC, and IGF-1, and 15 differentially expressed proteins, including CP, F8, HPX, VWF, RPLP0, ATP5F1B, LPL, F5, MYH9, HSPG2, SVEP1, BPIFB1, UBE2O, ACTN3, and TREML1. First, the least absolute shrinkage and selection operator (LASSO) regression algorithm was applied for feature standardization by adding an L1 norm to the objective function to penalize the regression coefficients. This resulted in the shrinkage of regression coefficients of redundant variables to zero, thereby achieving feature selection, reducing model complexity, and addressing multicollinearity 26 . Subsequently, Kendall's rank correlation coefficient (τ) was used to detect monotonic correlations among features to ensure model parsimony and stability 27 . Pairwise Kendall correlation coefficients were calculated for all features to generate a symmetric matrix. A threshold of |τ| ≥ 0.8 was set, where feature pairs with absolute correlation values above this threshold were considered redundant. In such cases, the feature with a stronger correlation to the target variable was retained, and the other was removed. Machine Learning Model development and comparison were performed in the Python 3.7 programming environment using eight machine learning algorithms based on the selected important features. These included ensemble learning algorithms suitable for high-dimensional data: random forest (RF), gradient boosting (GB), and extreme gradient boosting (XGBoost) 28 – 30 ; linear algorithms with good interpretability: logistic regression (LR) and ridge regression (RR) 31 , 32 ; a probabilistic algorithm: naive Bayes (NB) 33 ; an instance‑based learning method: K‑nearest neighbors (KNN) 34 ; and a single‑tree algorithm: decision tree (DT) 35 . During the modeling process, five rounds of cross‑validation and grid search for parameter optimization were repeated to prevent overfitting. Hyperparameter tuning for all algorithms was performed on the training set using nested cross‑validation. The final results were summarized across five rounds of validation, and the average performance of each algorithm was calculated. Model performance was evaluated primarily using the receiver operating characteristic (ROC) curve and the area under the curve (AUC) along with its 95% confidence interval (CI), supplemented by accuracy, precision, recall, and F1 score for comprehensive assessment of the classification performance of the eight machine learning algorithms. Model Interpretability Shapley Additive Explanations (SHAP) were applied to interpret the optimal model. Rooted in cooperative game theory, SHAP assigns each feature an importance value (Shapley value) by averaging its marginal contribution across all possible feature subsets. This approach provides a mathematically grounded, transparent explanation of how each feature influences model predictions, thereby enhancing the interpretability and credibility of the model 36 。 Results Clinical Characteristics in Discovery Cohort In the discovery phase, the mean age of the participants was 77.00 ± 3.17 years, and 15 participants were male (50%). No statistically significant differences were observed in general demographic characteristics among the three groups (all P > 0.05). In terms of clinical functional performance, Fried score ( P < 0.001), VAS score ( P = 0.010), GDS-15 score ( P = 0.013), and TUG time increased with the severity of frailty ( P = 0.009). MNA-SF score ( P = 0.024), MBI ( P = 0.009), gait speed ( P = 0.003), grip strength ( P = 0.013), and SPPB score ( P = 0.010) were all decreased in both the prefrail and frail groups. Circulating levels of CK ( P = 0.038) also decreased progressively from the robust to the frail state. No significant inter‑group differences were observed for number of chronic diseases, protein, BFM, SLM, Vit D, CysC, IGF-1, CRP, Cr, or TG (all P > 0.05, Table 1 ). Table 1 Clinical characteristics of participants in the discovery cohort Characteristic Robust (n = 10) Prefrail (n = 10) Frail (n = 10) F/Z/χ 2 P Value Demographic characteristics Gender (male/female, n) 5/5 5/5 5/5 0.000 1.000 Age(years) 75.60 ± 1.17 78.10 ± 4.58 77.30 ± 2.54 1.697 0.202 Education(years)* 12.00(9.00–16.00) 9.00(8.25-16.00) 11.32(9.00–12.00) 0.476 0.788 BMI(kg/m2) 23.86 ± 1.01 24.07 ± 3.46 24.16 ± 4.80 0.020 0.980 Smoking history(yes/no) 4/6 2/8 4/6 1.200 0.549 Alcohol consumption(yes/no) 6/4 2/8 3/7 3.732 0.155 Insomnia(yes/no) 3/7 7/3 8/2 5.833 0.054 Sedentary behavior(yes/no) 2/8 5/5 7/3 5.089 0.079 MMSE* 27.50(25.00–29.00) 26.50(25.00–29.00) 27.50(25.00-28.25) 0.327 0.849 Clinical characteristics Fried* 0.00(0.00–0.00) 1.00(1.00-1.25) 3.00(3.00-3.25) 27.878 < 0.001 Number of chronic diseases* 2.00(1.00-2.50) 2.22(1.00-4.02) 3.50(1.75-4.00) 2.421 0.298 VAS* 0.00(0.00–0.00) 0.00(0.00–5.00) 4.00(0.28–5.25) 9.280 0.010 MNA-SF* 14.00(14.00–14.00) 12.50(10.00–14.00) 13.00(10.75-14.00) 7.441 0.024 GDS-15* 1.00(0.00–2.00) 1.00(1.00–2.00) 3.20(1.75–6.25) 8.730 0.013 MBI* 100.00(99.50–100.00) 99.00(94.25–100.00) 93.50(86.75–98.50) 9.368 0.009 Gait speed 1.45 ± 0.56 1.21 ± 0.37 0.76 ± 0.25 7.066 0.003 Grip strength 31.93 ± 9.46 29.75 ± 6.42 22.03 ± 5.35 5.093 0.013 TUG* 8.49(7.07–10.40) 8.96(8.10-12.51) 11.82(10.71–13.24) 9.363 0.009 SPPB* 12.00(10.75-12.00) 11.00(9.75-11.00) 9.00(7.75–10.25) 9.206 0.010 Protein 9.16 ± 1.58 8.75 ± 1.64 8.21 ± 1.58 0.887 0.424 BFM 17.75 ± 2.06 21.00 ± 7.30 22.07 ± 9.93 0.973 0.391 SLM 44.18 ± 7.47 42.47 ± 7.94 40.25 ± 7.78 0.648 0.531 Vit D (ng/ml) 22.55 ± 6.52 23.19 ± 13.19 17.68 ± 4.21 1.163 0.328 CysC 0.09 ± 0.01 0.11 ± 0.02 0.11 ± 0.03 2.940 0.070 IGF-1(ng/mL)* 13.99(10.90-22.02) 29.41(18.96–33.26) 21.02(15.37–27.80) 5.772 0.056 CRP (mg/L)* 0.92(0.13–1.66) 0.49(0.35–0.63) 0.63(0.16–2.80) 0.321 0.852 CK (U/L)* 89.91(64.75-110.88) 87.80(67.10-138.48) 57.95(53.30-84.37) 6.521 0.038 Cr (umol/L) 71.99 ± 12.11 70.15 ± 15.64 70.01 ± 13.41 0.064 0.938 TG (mmol/L) 1.10(0.97–1.86) 1.27(0.86–1.85) 1.18(1.07–1.71) 0.156 0.925 Notes : *, the Kruskal‑Wallis test was used for non‑normally distributed data. Abbreviations: BMI, body mass index; MMSE, Mini-Mental State Examination; VAS, Visual AnalogueScale; MNA-SF, Mini-Nutritional Assessment-Short Form; GDS-15, Geriatric Depression Scale (15 items); MBI, Modified Barthel Index; TUG, Timed Up and Go Test; SPPB, Short Physical Performance Battery; BFM, body fat mass; SLM, skeletal muscle mass; Vit D, vitamin D; CysC, cystatin C; IGF 1, insulin like growth factor 1; CRP, C-reactive protein; CK, creatine kinase; Cr, creatinine; TG, triglycerides. Proteins in the Discovery Cohort In the discovery cohort, a total of 1,145 raw proteins were identified at quantifiable levels. Comparative analysis revealed 166 proteins that were differentially expressed among the robust, prefrail, and frail groups (Figure 2A). To elucidate the biological functions associated with frailty progression, enrichment analyses were performed using the GO and KEGG databases. Significantly enriched GO terms included vesicle‑mediated transport, regulation of response to stimulus, regulation of immune system processes, platelet activation, regulation of cell communication, proteolysis, muscle structure development, and muscle cell differentiation (Figure 2B). KEGG pathway analysis revealed that the transition from robust to frail states involved complement and coagulation cascades, ferroptosis, focal adhesion, riboflavin metabolism, platelet activation, cholesterol metabolism, glutathione metabolism, glycerolipid metabolism, and ubiquitin‑mediated proteolysis (Figure 2C). PPI was constructed to visualize functional connectivity among the differentially expressed proteins (Figure 2D). Clustering analysis highlighted 57 key proteins involved in biological processes such as regulation of adaptive immune response, regulation of cytokine production, protein polyubiquitination, regulation of skeletal muscle tissue development, lipid metabolism, and proteolysis. Based on the above bioinformatic analysis of differentially expressed proteins, candidate proteins were prioritized according to their correlation with clinical phenotypes (Figure 2E), expression trends (Figure 2F), and sex specificity (Figure 2G). This led to the selection of 30 candidate proteins as potential frailty biomarkers for subsequent PRM quantitative validation in the verification cohort (all P < 0.05; Table 2). Table 2 Candidate proteins associated with frailty Accession No. Protein Name F P Value E9PAV3 NACA Nascent-Polypeptide-Associated Complex Alpha Polypeptide 3.592 0.043 P00450 CP Ceruloplasmin 3.919 0.030 P00451 F8 Coagulation Factor VIII 4.579 0.031 P02787 TF Transcription Factor 3.923 0.023 P02790 HPX Hemopexin 3.861 0.024 P04275 VWF Von Willebrand Factor 4.771 0.017 P05156 CFI Complement Factor I 3.749 0.039 P05388 RPLP0 60S Acidic Ribosomal Protein P0 3.448 0.048 P06576 ATP5F1B ATP Synthase Subunit Beta， Mitochondrial 6.353 0.003 P06681 C2 Complement component 2 4.247 0.043 P06858 LPL Lipoprotein Lipase 4.594 0.019 P07585 DCN Decorin 17.270 0.000 P08311 CTSG Cathepsin G 3.752 0.027 P12259 F5 Coagulation Factor V 4.579 0.042 P20742 PZP Pregnancy Zone Protein 3.927 0.024 P28289 TMOD1 Tropomodulin-1 3.358 0.048 P35443 THBS4 Thrombospondin 4 4.743 0.025 P35579 MYH9 Myosin-9 6.566 0.006 P51884 LUM Lumican 3.758 0.029 P55010 EIF5 Eukaryotic Translation Initiation Factor 5 6.952 0.003 P62256 UBE2H Ubiquitin-conjugating Enzyme E2 H 5.162 0.010 P98160 HSPG2 Basement Membrane-Specific Heparan Sulfate Proteoglycan Core Protein 4.872 0.022 Q08043 ACTN3 Alpha-Actinin-3 4.855 0.013 Q08495 DEMA Dynamin-Related Protein 1 / Dermatological Evaluation of Mucocutaneous Adverse Events 3.919 0.027 Q4LDE5 SVEP1 Sushi，Von Willebrand Factor Type A，EGF and Pentraxin Domain-containing Protein 1 3.878 0.023 Q86YW5 TREML1 Trem-like Transcript 1 Protein 3.542 0.035 Q8TDL5 BPIFB1 BPI Fold-containing Family B Member 1 6.241 0.006 Q8WWA0 ITLN1 Intelectin 1 3.424 0.041 Q9BY43 CHMP4A Charged Multivesicular Body Protein 4A 4.253 0.015 Q9C0C9 UBE2O (E3-independent) E2 Ubiquitin-conjugating Enzyme 4.273 0.021 Clinical Characteristics in Validation Cohort In the validation phase, the mean age of participants was 75.32 ± 5.09 years, with 27 males (27.3%). The prevalence of sedentary behavior was significantly higher in the pre‑frail and frail groups (47.5% and 95.0%, respectively; P = 0.001) compared to the robust group (45.0%). MMSE scores decreased progressively with worsening frailty status ( P = 0.004). No significant differences were observed among the three groups in gender, age, years of education, BMI, smoking history, alcohol consumption use, or insomnia (all P > 0.05). Frailty status differed significantly among groups ( P < 0.001), with Fried scores increasing across the robust, pre‑frail, and frail categories. Compared to the robust group, the frail group showed significantly lower scores in the MBI, gait speed, handgrip strength, and SPPB (all P < 0.05). TUG time was significantly longer in both the pre‑frail and frail groups than in the robust group ( P < 0.001). Total body protein and SLM were notably lower in the pre‑frail group compared to the robust group ( P = 0.017 and P = 0.009, respectively). CysC and IGF‑1 levels were significantly altered in the frail group compared with the robust group. ( P = 0.040 and P = 0.004, respectively). In contrast, no significant inter‑group differences were detected for number of chronic diseases, VAS, MNA‑SF, GDS‑15, BFM, Vit D, CRP, CK, Cr, or TG levels (all P > 0.05; Table 3). Table 3 Comparison of clinical characteristics in the validation cohort Characteristic Robust(n=20) Prefrail(n=59) Frail(n=20) F/Z/χ 2 P Value Demographic characteristics Gender (male/female, n) 7/13 12/47 8/12 3.666 0.160 Age(years) 75.55±4.20 74.71±4.97 76.90±6.06 1.419 0.247 Education(years)* 11.28(9.00-16.00) 12.00(9.00-14.00) 12.00(9.00-16.00) 0.452 0.798 BMI(kg/m2) 25.70±3.65 23.91±3.02 23.69±3.48 2.616 0.078 Smoking history(yes/no) 1/19 8/51 2/18 1.139 0.566 Alcohol consumption(yes/no) 6/14 13/46 2/18 2.452 0.293 Insomnia(yes/no) 10/10 32/27 11/9 0.129 0.937 Sedentary behavior(yes/no) 9/11 28/31 19/1 15.106 0.001 MMSE* 28.00(27.00-29.00) 28.00(26.30-29.00) 26.00(25.00-27.00) 10.941 0.004 Clinical characteristics Fried* 0.00(0.00-0.00) 1.00(1.00-2.00) 3.00(3.00-4.00) 61.308 <0.001 Number of chronic diseases* 2.50(2.00-4.00) 2.00(1.00-4.00) 3.00(2.00-5.50) 4.454 0.110 VAS* 0.00(0.00-1.00) 0.00(0.00-5.00) 3.00(0.00-5.50) 5.557 0.060 MNA-SF* 14.00(13.00-14.00) 13.00(12.00-14.00) 13.00(12.00-14.00) 4.895 0.090 GDS-15* 1.00(0.00-2.00) 2.00(1.00-3.50) 2.00(1.00-4.50) 5.494 0.060 MBI* 100.00(98.00-100.00) 100.00(96.00-100.00) 88.00(83.5-94.00) 30.948 <0.001 Gait speed 1.13(0.90-1.40) 1.09(0.97-1.30) 0.69(0.47-0.81) 23.608 <0.001 Grip strength 25.45(23.60-35.35) 24.10(21.95-29.55) 21.70(18.80-25.10) 9.785 0.010 TUG* 8.74(8.15-9.64) 9.00(7.80-11.51) 15.42(13.44-25.72) 24.813 <0.001 SPPB* 11.00(9.00-11.50) 11.00(10.00-12.00) 7.00(3.50-8.50) 28.535 <0.001 Protein 8.60(7.80-10.55) 7.60(7.25-8.50) 8.40(7.30-9.25) 8.205 0.017 BFM 22.08±6.52 20.59±5.86 20.23±5.67 0.589 0.557 SLM 41.70(38.05-51.15) 37.40(35.00-41.35) 41.15(35.75-45.85) 9.403 0.009 Vit D (ng/ml) 17.93(11.26-27.93) 17.28(13.45-24.64) 16.77(11.04-23.03) 0.983 0.612 CysC 0.10(0.09-0.12) 0.10(0.09-0.11) 0.12(0.10-0.14) 6.441 0.040 IGF-1(ng/mL)* 20.83(12.92-34.12) 18.97(13.62-27.64) 38.75(21.71-49.42) 11.007 0.004 CRP (mg/L)* 0.52(0.30-1.33) 0.54(0.30-1.22) 0.73(0.20-1.22) 0.072 0.965 CK (U/L)* 85.15(72.25-114.30) 75.80(58.10-100.00) 62.65(44.05-110.90) 4.887 0.087 Cr (umol/L) 65.53(48.20-81.25) 65.00(53.65-77.55) 70.65(59.85-91.00) 4.767 0.092 TG (mmol/L)* 1.55(1.22-2.17) 1.23(0.90-1.63) 1.34(1.12-1.96) 4.859 0.088 Notes: *, the Kruskal‑Wallis test was used for non‑normally distributed data . Abbreviations: BMI,body mass index;MMSE,Mini-Mental State Examination;VAS,Visual Analogue Scale;MNA-SF,Mini-Nutritional Assessment-Short Form;GDS-15,Geriatric Depression Scale (15 items);MBI,Modified Barthel Index;TUG,Timed Up and Go Test;SPPB,Short Physical Performance Battery; BFM,body fat mass;SLM,skeletal muscle mass;Vit D,vitamin D;CysC,cystatin C;IGF 1,insulin like growth factor 1;CRP,C-reactive protein;CK,creatine kinase;Cr,creatinine;TG,triglycerides Targeted Proteins in the Validation Cohort Targeted proteomic analysis via PRM was performed in the validation cohort to quantify the expression levels of 30 candidate proteins previously identified in the discovery phase (Figure 3A). Among these, 15 proteins exhibited statistically significant differential expression across the robust, pre‑frail, and frail groups ( P < 0.05; Figure 3B), including ceruloplasmin (CP), coagulation factor VIII (F8), hemopexin (HPX), von Willebrand factor (VWF), 60S acidic ribosomal protein P0 (RPLP0), ATP synthase subunit beta mitochondrial (ATP5F1B), lipoprotein lipase (LPL), coagulation factor V (F5), myosin‑9 (MYH9), basement membrane‑specific heparan sulfate proteoglycan core protein (HSPG2), SVEP1 (sushi, von Willebrand factor type A, EGF and pentraxin domain‑containing protein 1), BPI fold‑containing family B member 1 (BPIFB1), ubiquitin‑conjugating enzyme E2O (UBE2O), alpha‑actinin‑3 (ACTN3), and trem‑like transcript 1 protein (TREML1). Feature Selection A total of 26 independent variables showing statistical significance in the validation phase were subjected to LASSO regression analysis (Figure 4A). Feature selection was performed using LASSO with cross‑validation, and seven features with non‑zero coefficients corresponding to the optimal lambda on the left side of the curve were selected: MMSE, gait speed, TUG, protein, SLM, UBE2O, and ACTN3 (Figure 4B). Based on the LASSO results, Kendall correlation analysis (Figure 4C) revealed high collinearity between SLM and total protein content (τ = 0.95). The redundant feature with relatively lower nonlinear association, “protein” was removed. The remaining six features, namely MMSE, gait speed, TUG, SLM, UBE2O, and ACTN3, were subsequently used to construct the machine learning model. Performance Comparison of Machine Learning Based on the six selected important features, the study systematically evaluated the classification performance of eight widely used machine learning algorithms, including RF, GB, XGBoost, LR, RR, NB, KNN, and DT (Figure 4D). The results showed that ensemble learning methods exhibited the most prominent overall performance. The RF algorithm achieved the highest AUC (0.896, 95% CI: 0.792–0.971), indicating excellent discriminative ability and stability. GB ranked second (AUC = 0.875, 95% CI: 0.775–0.983), followed by XGBoost (AUC = 0.792, 95% CI: 0.707–0.969), which demonstrated relatively narrow CIs and good generalizability. Among traditional models, LR (AUC = 0.844, 95% CI: 0.475–0.948) and RR (AUC = 0.833, 95% CI: 0.464–0.948) showed moderate performance, yet their wide CIs suggested sensitivity to data distribution. The probabilistic model NB (AUC = 0.760, 95% CI: 0.448–0.929) and distance‑sensitive KNN (AUC = 0.745, 95% CI: 0.531–0.955) performed similarly, with NB displaying greater variability. In contrast, DT yielded the poorest results (AUC = 0.625, 95% CI: 0.599–0.964), reflecting its instability and limited applicability in this classification task (Table 4). SHAP Analysis The best‑performing RF algorithm was further interpreted using SHAP analysis. Figure 4E illustrates the mean absolute SHAP values for each feature, ranked in descending order of importance: gait speed, SLM, UBE2O, TUG, ACTN3, and MMSE. Figure 4F further reveals the contribution magnitude and direction of each feature to the classification results in the RF model. Gait speed exhibited the strongest predictive influence, with lower speed (blue dots) associated with higher risk of pre‑frailty (positive SHAP values). SLM and the protein markers UBE2O and ACTN3 showed more complex, non‑linear relationships, where both high and low levels could contribute to frailty risk. TUG time, while less influential, was positively associated with pre‑frailty risk, whereas MMSE score only contributed substantially under conditions of pronounced cognitive decline. Table 4 Comparison of predictive performance among the eight model Algorithm AUC(95%CI) Accuracy Precision Recall F1 Score RF 0.896(0.792 – 0.971) 0.750 0.733 0.917 0.815 GB 0.875(0.775 – 0.983) 0.700 0.714 0.833 0.769 XGBoost 0.792(0.707 – 0.969) 0.600 0.625 0.833 0.714 LR 0.844(0.475 – 0.948) 0.707 0.694 0.834 0.732 RR 0.833(0.464 – 0.948) 0.658 0.641 0.825 0.723 NB 0.760(0.448 – 0.929) 0.713 0.696 0.880 0.778 KNN 0.745(0.531 – 0.955) 0.650 0.667 0.833 0.741 DT 0.625(0.599 – 0.964) 0.650 0.692 0.750 0.720 Discussion To our knowledge, this is the first attempt to integrate high‑throughput proteomic data with clinical phenotypes to systematically delineate the molecular trajectory of proteins during the transition from robustness to frailty, with prefrailty examined as a distinct status. The results revealed that among 161 significantly differentially expressed proteins detected in older adults across different health statuses, 15 core proteins associated with frailty were validated by targeted proteomics, including CP, F8, HPX, VWF, RPLP0, ATP5F1B, LPL, F5, MYH9, HSPG2, SVEP1, BPIFB1, UBE2O, ACTN3, and TREML1. Through machine learning combined with SHAP‑based interpretability analysis, UBE2O and ACTN3, together with clinical features including gait speed, SLM, TUG time, and MMSE score, were further identified as the most promising indicators for early frailty detection. Our findings provide new insights into the mechanisms underlying frailty development and offer a reference for precise clinical identification. Integrating data from two independent cohorts, the results demonstrated that different frailty statuses were accompanied by widespread proteomic alterations. A total of 15 proteins involved in intercellular signaling (SVEP1, HPX, RPLP0), protein metabolism (ATP5F1B, CP, LPL), immune regulation (VWF, F5, F8, BPIFB1, TREML1), maintenance of skeletal muscle structure and function (MYH9, ACTN3, HSPG2), and proteolysis (UBE2O) maintained statistically significant differences. This trend of proteomic changes detected during the transition from robustness to frailty suggests that the onset of frailty typically emerges gradually under the influence of multisystem dysregulation. The study further combined LASSO regression with Kendall correlation analysis for feature selection. Using eight machine learning algorithms with complementary advantages, namely RF, GB, XGBoost, LR, RR, NB, KNN, and DT, coupled with SHAP‑based interpretability analysis, the classification mechanism of the optimal model was systematically elucidated. Ultimately, an RF‑based feature set comprising gait speed, SLM, UBE2O, TUG, ACTN3, and MMSE score was established. These findings provide a multidimensional perspective on the characteristic patterns of early frailty and offer an interpretable machine learning decision framework to facilitate the clinical translation of protein biomarkers. Clinical characteristics have long been strong discriminators of frailty and represent the most intuitive manifestations of frailty in older adults and populations with various chronic diseases 37 。Our findings align with related research on frailty risk factor prediction, where functional indicators (e.g., gait speed, grip strength, balance, mobility), socio-psychological factors (e.g., cognition, loneliness, economic status), and physiological parameters (e.g., BMI, waist circumference, muscle mass, visceral fat area) are the most frequently included risk factors in frailty prediction models and dominate clinical frailty assessment 37,38 . In a large Chinese cohort study with an average follow-up of five years, a prediction model based on nine risk factors was developed and validated. This model incorporated cognitive function, lifestyle behaviors, and body composition as predictive variables, providing significant insights for the early screening and risk stratification of frailty in the elderly 39 . Parameters related to functional status, psychosocial factors, and basic physiological indicators have become core variables in frailty discrimination models due to their operability and well‑defined mechanisms. However, with the advancement of precision medicine, integrating dynamic molecular marker data with clinical observational indicators has emerged as a new direction for personalized early risk assessment of frailty. In this study, ACTN3 and UBE2O, which were highly expressed in serum, were identified as two key protein molecules for identifying early frailty through machine learning algorithms. ACTN3, which was upregulated in the peripheral circulation of prefrail older adults, is a key structural protein of type II muscle fibers and is predominantly expressed within skeletal muscle cells; the R577X polymorphism in its associated gene is known to influence fast-twitch muscle fiber function, regulating muscle strength, power, and athletic potential 40 . Numerous studies on muscle degenerative diseases report that elevated blood levels of ACTN3 may result from stress-induced activation of the X allele in damaged skeletal muscle cells, leading to its release into the peripheral circulation via extracellular vesicle transport pathways 41 . Furthermore, the compensatory increase in ACTN3 can disrupt the balance between muscle protein synthesis and degradation, negatively regulating muscle mass. Particularly in females, aberrant ACTN3 expression may further impair muscle growth and function by inhibiting downstream effectors of the mTOR signaling pathway, a key regulator of muscle hypertrophy 42 . While ACTN3's effect on protein synthesis and degradation is indirect, UBE2O, functioning as a ubiquitin-conjugating enzyme, can directly influence normal skeletal muscle development by ubiquitinating and degrading target proteins (e.g., the AMPKα2 subunit), thereby blunting the mTOR signaling pathway 43 . Although current literature typically discusses the functions of ACTN3 and UBE2O independently, our initial PPI analysis captured an indirect interaction between them. Both proteins may converge on the AMPK/mTOR signaling pathway, where the ubiquitin-proteasome system selectively degrades proteins, disrupting metabolic homeostasis between protein synthesis and breakdown. Studies in animal models of obesity or metabolic syndrome have precisely defined that abnormal increases in UBE2O accelerate protein ubiquitination and hydrolysis, leading to the disruption of skeletal muscle protein metabolic homeostasis 44 . In clinical research, the elevated expression of various ubiquitination-related proteins, including UBE2O and UBE2T, has been implicated in the pathogenesis of aging and cancer and is closely associated with poor prognosis 45 . ACTN3 and UBE2O elucidate the complex pathological regulatory network of prefrailty from two perspectives: skeletal muscle structural development and protein ubiquitin-mediated degradation. Their interaction may synergistically exacerbate the onset and progression of frailty. This negative impact could represent a potential pathological basis for early frailty and an important risk factor for the prefrail state. However, it is crucial to acknowledge the potential influence of sample size limitations. First, although FDR correction was applied to mitigate false positives, only 50% (15/30) of the candidate proteins were replicated in the validation cohort, suggesting possible false discoveries. Second, while power analysis indicated 90% statistical power for detecting medium-effect-size differences with the current sample size, proteins with important biological significance might still have been missed. Limitations This study has several limitations. The cross‑sectional design precludes the observation of dynamic progression from robustness to frailty, and the causal relationship between proteomic alterations and early frailty remains unclear. Future longitudinal studies with extended follow‑up or multi‑time‑point sampling may help elucidate the temporal sequence linking protein expression and functional decline. The modest sample size may limit the accuracy and generalizability of the findings; although key results reached statistical significance, subtle changes in secondary indicators may have been underpowered. Furthermore, the single‑center cohort may not fully represent the heterogeneity of older adults across different geographic and cultural contexts. Future investigations could expand sample size, adopt a multicenter design, and conduct stratified analyses accounting for varying levels of frailty, comorbidities, and medication use. Computational constraints also restricted the breadth of cohort coverage, potentially introducing data bias, and limited the depth of algorithmic comparison. Additional external validation is needed to further assess the generalizability of the machine‑learning models. Conclusion This study revealed widespread proteomic alterations across different frailty statuses and identified 15 proteins potentially associated with prefrailty, which were involved in key pathways including cell signaling, protein metabolism, immune regulation, skeletal muscle function maintenance, and proteolysis. Based on the random forest algorithm, a preliminary combination of clinical indicators and protein markers was established for frailty identification. Among these, ACTN3 and UBE2O demonstrated discriminatory potential as key molecules in early frailty, providing a scientific basis for the clinical translation of omics findings and the development of multidimensional tools for frailty assessment. Future studies could leverage the predictive performance of these markers to expand frailty phenotyping and facilitate their translation into clinical applications. Declarations Ethics approval and informed consent The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Ethics Committee of the Chinese PLA General Hospital (Approval No. S2019‑140‑03). All participants provided written informed consent for participation and data collection. Sources of funding This study was supported by the National Key Research and Development Program of China (Grant No. 2018YFC2002004). Author contributions YY contributed to study design, data collection, data analysis, manuscript drafting, and manuscript editing. JWL and ZZ contributed to study conceptualization, data collection, and analysis. SHC contributed to data visualization. SXX contributed to data curation. MYC, YYZ, FW, and NHZ contributed to study conceptualization. GBW contributed to data analysis, visualization, and conceptualization. NP contributed to study design, conceptualization, supervision, manuscript review, and funding acquisition. Conflicts of interest disclosure The authors declare no competing interests. References Orkaby AR, Schwartz AW, Callahan KE, Frailty. Ann Intern Med. 2026;179(2):ITC17–32. 10.7326/ANNALS-25-04412 . Cai S, Li J, Fang Y, et al. Frailty and pre-frailty prevalence in community-dwelling elderly with multimorbidity: A systematic review and meta-analysis. Arch Gerontol Geriatr. 2025;132:105782. 10.1016/j.archger.2025.105782 . Lapi F, Marconi E, Aprile PL, et al. Frailty, comorbidity, and multimorbidity and their relation with medications adherence in primary care older adults. Eur Geriatr Med. 2024. 10.1007/s41999-024-01098-4 . Sato R, Vatic M, Peixoto da Fonseca GW, Anker SD, von Haehling S. Biological basis and treatment of frailty and sarcopenia. Cardiovasc Res. 2024;120(9):982–98. 10.1093/cvr/cvae073 . Serra-Prat M, Fortuny Borsot A, Burdoy E, Lavado Cuevas À, Muñoz Ortiz L, Cabré M. Frailty reversal and its main determinants: a population-based observational and longitudinal study. Fam Med Community Health. 2025;13(2). 10.1136/fmch-2024-003250 . Wilkinson DJ, Piasecki M, Atherton PJ. The age-related loss of skeletal muscle mass and function: Measurement and physiology of muscle fibre atrophy and muscle fibre loss in humans. Ageing Res Rev. 2018;47:123–32. 10.1016/j.arr.2018.07.005 . Hodgins ML, Maxwell SP, Howlett SE, Rockwood K. Blood biomarkers of frailty and cognition: A scoping review. Neurobiol Aging. 2026;161:14–30. 10.1016/j.neurobiolaging.2026.01.003 . Sciancalepore F, Salemme S, Valletta M, et al. Frailty and the brain: A narrative review of functional and pathological correlates. Ageing Res Rev. 2025;113:102945. 10.1016/j.arr.2025.102945 . Liu G, Jiang S, Xie W, et al. Biomarkers for sarcopenia, muscle mass, muscle strength, and physical performance: an umbrella review. J Transl Med Jun. 2025;12(1):650. 10.1186/s12967-025-06575-3 . Mitchell A, Malmgren L, Bartosch P, McGuigan FE, Akesson KE. Pro-Inflammatory Proteins Associated with Frailty and Its Progression-A Longitudinal Study in Community-Dwelling Women. J Bone Min Res. 2023;38(8):1076–91. 10.1002/jbmr.4861 . Sathyan S, Ayers E, Gao T, Milman S, Barzilai N, Verghese J. Plasma proteomic profile of frailty. Aging Cell Sep. 2020;19(9):e13193. 10.1111/acel.13193 . Ubaida-Mohien C, Lyashkov A, Gonzalez-Freire M, et al. Discovery proteomics in aging human skeletal muscle finds change in spliceosome, immunity, proteostasis and mitochondria. Elife. 2019;8. 10.7554/eLife.49874 . Pedro KM, Alvi MA, Hejrati N, Quddusi AI, Singh A, Fehlings MG. Machine learning-based cluster analysis identifies four unique phenotypes of patients with degenerative cervical myelopathy with distinct clinical profiles and long-term functional and neurological outcomes. EBioMedicine. 2024;106:105226. 10.1016/j.ebiom.2024.105226 . Niu L, Thiele M, Geyer PE, et al. Noninvasive proteomic biomarkers for alcohol-related liver disease. Nat Med. 2022;28(6):1277–87. 10.1038/s41591-022-01850-y . Fried LP, Tangen CM, Walston J, et al. Frailty in older adults: evidence for a phenotype. Journals Gerontol Ser Biol Sci Med Sci. 2001;56(3):M146–56. Jia X, Wang Z, Huang F, et al. A comparison of the Mini-Mental State Examination (MMSE) with the Montreal Cognitive Assessment (MoCA) for mild cognitive impairment screening in Chinese middle-aged and older population: a cross-sectional study. BMC Psychiatry. 2021;21(1):485. 10.1186/s12888-021-03495-6 . Che Y, Xin H, Gu Y, Ma X, Xiang Z, He C. Associated factors of frailty among community-dwelling older adults with multimorbidity from a health ecological perspective: a cross-sectional study. BMC Geriatr Mar. 2025;14(1):172. 10.1186/s12877-025-05777-0 . Anbarasan D, Merchant RA. Association between chronic pain severity, falls, frailty and perceived health in older adults at risk of falls. Eur J Med Res Oct. 2025;31(1):1047. 10.1186/s40001-025-03304-w . Ishiwari M, Kono Y, Togashi Y, et al. Mini nutritional assessment short-form as a risk factor for mortality in patients with respiratory disease undergoing urgent hospitalization. BMC Geriatr Feb. 2025;19(1):113. 10.1186/s12877-025-05767-2 . Park S-H, Kwak M-J. Performance of the Geriatric Depression Scale-15 with Older Adults Aged over 65 Years: An Updated Review 2000–2019. Clin Gerontol. 2020;44(2):83–96. 10.1080/07317115.2020.1839992 . Sainsbury A, Seebass G, Bansal A, Young JB. Reliability of the Barthel Index when used with older people. Age Ageing May. 2005;34(3):228–32. 10.1093/ageing/afi063 . Houston DK, Marsh AP, Neiberg RH, et al. Vitamin D Supplementation and Muscle Power, Strength and Physical Performance in Older Adults: A Randomized Controlled Trial. Am J Clin Nutr Jun. 2023;117(6):1086–95. 10.1016/j.ajcnut.2023.04.021 . Horn DB, Linetzky B, Davies MJ, et al. Cardiometabolic Parameter Change by Weight Regain on Tirzepatide Withdrawal in Adults With Obesity: A Post Hoc Analysis of the SURMOUNT-4 Trial. JAMA Intern Med. 2026;186(2):157–67. 10.1001/jamainternmed.2025.6112 . Li K, Teo GC, Yang KL, Yu F, Nesvizhskii AI. diaTracer enables spectrum-centric analysis of diaPASEF proteomics data. Nat Commun. 2025;16(1):95. 10.1038/s41467-024-55448-8 . Chambers AG, Sweet SMM, Chain D, Kim YJ. Global Extraction from Parallel Reaction Monitoring to Quantify Background Peptides for Improved Normalization and Quality Control in Targeted Proteomics. Anal Chem. 2021;93(40):13434–40. 10.1021/acs.analchem.1c01375 . Deng M, Sun S, Zhao R, et al. The pyroptosis-related gene signature predicts prognosis and indicates immune activity in hepatocellular carcinoma. Mol Med. 2022;28(1):16. 10.1186/s10020-022-00445-0 . Takefuji Y. Reevaluating feature importance in machine learning: concerns regarding SHAP interpretations in the context of the EU artificial intelligence act. Water Res. 2025;280:123514. 10.1016/j.watres.2025.123514 . Owens CD, Pinto CB, Mukli P, et al. Neurovascular coupling, functional connectivity, and cerebrovascular endothelial extracellular vesicles as biomarkers of mild cognitive impairment. Alzheimer's Dement J Alzheimer's Assoc. 2024;20(8):5590–606. 10.1002/alz.14072 . Yi Y, Tae M, Shin S, Choi S-I. Predicting acute kidney injury in trauma using an extreme gradient boosting model. Clin Kidney J. 2025;18(4):sfaf002. 10.1093/ckj/sfaf002 . Liang D, Wang L, Zhong P, et al. Perspective: Global Burden of Iodine Deficiency: Insights and Projections to 2050 Using XGBoost and SHAP. Adv Nutr. 2025;16(3):100384. 10.1016/j.advnut.2025.100384 . Dey D, Haque MS, Islam MM, et al. The proper application of logistic regression model in complex survey data: a systematic review. BMC Med Res Methodol. 2025;25(1):15. 10.1186/s12874-024-02454-5 . Akhtar N, Alharthi MF. Enhancing accuracy in modelling highly multicollinear data using alternative shrinkage parameters for ridge regression methods. Sci Rep. 2025;15(1):10774. 10.1038/s41598-025-94857-7 . Sugahara S, Ueno M. Exact Learning Augmented Naive Bayes Classifier. Entropy (Basel). 2021;23(12). 10.3390/e23121703 . Wang J, Geng X. Large Margin Weighted k-Nearest Neighbors Label Distribution Learning for Classification. IEEE Trans Neural Netw Learn Syst. 2024;35(11):16720–32. 10.1109/TNNLS.2023.3297261 . Hou R, Wang X, Zhang W, et al. Decision Tree Extraction for Clinical Decision Support System With If-Else Pseudocode and PlanSelect Strategy. IEEE J Biomed Health Inf. 2025;PPdoi. 10.1109/JBHI.2025.3529682 . Cao S, Hu Y. Creating machine learning models that interpretably link systemic inflammatory index, sex steroid hormones, and dietary antioxidants to identify gout using the SHAP (SHapley Additive exPlanations) method. Front Immunol. 2024;15:1367340. 10.3389/fimmu.2024.1367340 . Perdiguero GG, Spina JC, Martínez J, et al. Enhancing ACLF prediction by integrating sarcopenia assessment and frailty in liver transplant candidates on the waiting list. JHEP Rep Mar. 2024;6(3):100985. 10.1016/j.jhepr.2023.100985 . Shah SJ, Oreper S, Jeon SY, Boscardin WJ, Fang MC, Covinsky KE. Social Frailty Index: Development and validation of an index of social attributes predictive of mortality in older adults. Proc Natl Acad Sci U S A Feb. 2023;14(7):e2209414120. 10.1073/pnas.2209414120 . Dong Y, Wang Q, Zhang K, et al. Development and validation of a prediction model of frailty risk in community-dwelling older adults: From a national longitudinal survey. Public Health Jan. 2025;27:240:63–70. 10.1016/j.puhe.2024.12.055 . El Ouali EM, Barthelemy B, Del Coso J, et al. A Systematic Review and Meta-analysis of the Association Between ACTN3 R577X Genotypes and Performance in Endurance Versus Power Athletes and Non-athletes. Sports Med Open Apr. 2024;12(1):37. 10.1186/s40798-024-00711-x . Giza S, Mojica-Santiago JA, Parafati M, et al. Microphysiological system for studying contractile differences in young, active, and old, sedentary adult derived skeletal muscle cells. Aging Cell Jul. 2022;21(7):e13650. 10.1111/acel.13650 . Seto JT, Roeszler KN, Meehan LR, et al. ACTN3 genotype influences skeletal muscle mass regulation and response to dexamethasone. Sci Adv Jul. 2021;7(27). 10.1126/sciadv.abg0088 . Huang H, Zhu W, Huang B, et al. Structural insights into the biochemical mechanism of the E2/E3 hybrid enzyme UBE2O. Structure. 2025;33(2). 10.1016/j.str.2024.12.002 . Vila IK, Park MK, Setijono SR, et al. A muscle-specific UBE2O/AMPKα2 axis promotes insulin resistance and metabolic syndrome in obesity. JCI Insight. 2019;4(13). 10.1172/jci.insight.128269 . de Carvalho LGA, Komoto TT, Moreno DA, et al. USP15-USP7 Axis and UBE2T Differential Expression May Predict Pathogenesis and Poor Prognosis in De Novo Myelodysplastic Neoplasm. Int J Mol Sci Jun. 2023;13(12). 10.3390/ijms241210058 . Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Reviewers agreed at journal 11 May, 2026 Reviewers invited by journal 07 May, 2026 Editor assigned by journal 07 May, 2026 Editor invited by journal 27 Apr, 2026 Submission checks completed at journal 24 Apr, 2026 First submitted to journal 24 Apr, 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-9388693","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":641475547,"identity":"385b20bb-6f65-43dc-b0c4-3c8e4740709a","order_by":0,"name":"Yu Ye","email":"","orcid":"","institution":"Chinese PLA General Hospital","correspondingAuthor":false,"prefix":"","firstName":"Yu","middleName":"","lastName":"Ye","suffix":""},{"id":641475549,"identity":"2d6baf4a-a9ac-4d76-9efc-a3c44443a700","order_by":1,"name":"Jinwei Liu","email":"","orcid":"","institution":"Chinese PLA General Hospital","correspondingAuthor":false,"prefix":"","firstName":"Jinwei","middleName":"","lastName":"Liu","suffix":""},{"id":641475550,"identity":"a3e39a20-8dff-4533-bedd-d091f0204ff9","order_by":2,"name":"Zhen Zhang","email":"","orcid":"","institution":"Chinese PLA General Hospital","correspondingAuthor":false,"prefix":"","firstName":"Zhen","middleName":"","lastName":"Zhang","suffix":""},{"id":641475552,"identity":"46681e89-b93e-45b8-9aa2-1040a2de76bc","order_by":3,"name":"Shuaixuan Xu","email":"","orcid":"","institution":"Chinese PLA General Hospital","correspondingAuthor":false,"prefix":"","firstName":"Shuaixuan","middleName":"","lastName":"Xu","suffix":""},{"id":641475554,"identity":"4d9b2b09-6fa2-432d-b807-0cff3689cbb3","order_by":4,"name":"Chenghao Chang","email":"","orcid":"","institution":"Chinese PLA General Hospital","correspondingAuthor":false,"prefix":"","firstName":"Chenghao","middleName":"","lastName":"Chang","suffix":""},{"id":641475556,"identity":"f5848c36-688e-4e37-ac59-89fc8b155b98","order_by":5,"name":"Mengyu Cao","email":"","orcid":"","institution":"Chinese PLA General Hospital","correspondingAuthor":false,"prefix":"","firstName":"Mengyu","middleName":"","lastName":"Cao","suffix":""},{"id":641475558,"identity":"c0158eab-1bbd-4bac-a0e1-610d163aea91","order_by":6,"name":"Yongyi Zhang","email":"","orcid":"","institution":"Medical School of Chinese PLA","correspondingAuthor":false,"prefix":"","firstName":"Yongyi","middleName":"","lastName":"Zhang","suffix":""},{"id":641475560,"identity":"05c3a558-76e1-464b-90cc-3bc7f488f903","order_by":7,"name":"Fang Wang","email":"","orcid":"","institution":"Chinese PLA General Hospital","correspondingAuthor":false,"prefix":"","firstName":"Fang","middleName":"","lastName":"Wang","suffix":""},{"id":641475562,"identity":"8ecf2a2e-8e74-484b-bbc5-73d4417b0b88","order_by":8,"name":"Nihui Zhang","email":"","orcid":"","institution":"Chinese PLA General Hospital","correspondingAuthor":false,"prefix":"","firstName":"Nihui","middleName":"","lastName":"Zhang","suffix":""},{"id":641475567,"identity":"f2829787-b96b-42d7-9b6b-67a53ae4fed7","order_by":9,"name":"Guibin Wang","email":"","orcid":"","institution":"Beijing Institute of Lifeomics","correspondingAuthor":false,"prefix":"","firstName":"Guibin","middleName":"","lastName":"Wang","suffix":""},{"id":641475569,"identity":"7e58d9e7-f6c8-4a99-bf6e-b82b24db8119","order_by":10,"name":"Nan Peng","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA2UlEQVRIiWNgGAWjYBAC+wYIbcfGzHzgwIcfRGgxOAChk/nZ2RIPzuwhQQvjzH4e48McbMRoOd57+HVBzR1mg8M8Hw4z8DDI84sdwK/FvudcmvWMY8/4DA7zbjhcYMFgOHN2AgFbJHLMjHnYDjODtczgYUgwuE1Ii/wboJZ/hxk3HOZ5cJiHjRgtEjzGj3nbDjPObOZhIFILT44ZM2/f4WR+ZjYDYCBLEOEX9jPGn3m+HbZj4z/8+MOHHzby/NIEtAABmwQSRwKnMmTA/IEoZaNgFIyCUTByAQC0OkTHToXnFAAAAABJRU5ErkJggg==","orcid":"","institution":"Chinese PLA General Hospital","correspondingAuthor":true,"prefix":"","firstName":"Nan","middleName":"","lastName":"Peng","suffix":""}],"badges":[],"createdAt":"2026-04-11 14:23:57","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9388693/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9388693/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":109462716,"identity":"beaa9a6d-99bf-4bec-b10d-c62a52bf2e9d","added_by":"auto","created_at":"2026-05-18 11:22:52","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":274049,"visible":true,"origin":"","legend":"\u003cp\u003eSee image above for figure legend.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-9388693/v1/b355379fa835e63af21de389.png"},{"id":109759208,"identity":"55afd2fe-5fb5-4ccf-a018-3d4535a97e4d","added_by":"auto","created_at":"2026-05-22 07:26:07","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":333245,"visible":true,"origin":"","legend":"\u003cp\u003eSee image above for figure legend.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-9388693/v1/490c1438763ca509e023535a.png"},{"id":109462717,"identity":"9792295a-d5b4-41b4-b7a7-55e7afd5f14f","added_by":"auto","created_at":"2026-05-18 11:22:52","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":404638,"visible":true,"origin":"","legend":"\u003cp\u003eSee image above for figure legend.\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-9388693/v1/89ae56a156a83abf2e978aa1.png"},{"id":109760465,"identity":"625e2d97-8e34-45cf-836f-e71d5aca0300","added_by":"auto","created_at":"2026-05-22 07:28:43","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1495997,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9388693/v1/5f452955-bd29-4481-a263-63f02c4ece73.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"A Preliminary Predictive Panel for Pre-frailty Based on Serum Proteomic Biomarkers: A Two-Phase Cross-Sectional Study","fulltext":[{"header":"Highlights","content":"\u003cp\u003e1.Serum proteomic analysis reveals a distinct panel of biomarkers associated with frailty progression.\u003c/p\u003e\u003cp\u003e2.Novel protein markers such as UBE2O and ACTN3 demonstrate predictive value for early‑stage frailty.\u003c/p\u003e\u003cp\u003e3. The integration of clinical indices and protein markers within a machine learning model establishes an interpretable framework for the classification of pre-frailty.\u003c/p\u003e"},{"header":"Introduction","content":"\u003cp\u003eFrailty is a geriatric syndrome characterized by decreased physiological reserve, increased vulnerability, and reduced resistance to stressors due to multi-system dysregulation\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e. Focusing on the global proportion of frailty prevalence, approximately 18% of older adults meet the criteria for frailty, while the proportion of prefrailty, the transitional phase from robustness to frailty, reaches 49%\u003csup\u003e2\u003c/sup\u003e. The transition of older adults to frailty poses a significant threat to the aging process and is undoubtedly one of the most serious public health challenges of the 21st century \u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e. Notably, pre‑frailty is a dynamic and reversible state, making its effective identification and timely intervention crucial for improving health outcomes in the elderly\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe pathological changes underlying early frailty are closely linked to dysregulated protein expression, with core mechanisms such as chronic inflammation, energy metabolism disturbances, cellular senescence, and skeletal muscle homeostasis imbalance all exhibiting corresponding protein‑level alterations\u003csup\u003e\u003cspan additionalcitationids=\"CR7\" citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. For instance, elevated levels of classic inflammatory markers such as C‑reactive protein (CRP) and interleukin‑6 (IL‑6) contribute directly to muscle functional decline through inflammatory catabolism in skeletal muscle\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e. A proteomic study of older women identified 32 pro‑inflammatory proteins overexpressed in frail individuals, among which eight core proteins were significantly correlated with frailty indices and functionally implicated in multi‑system changes including renal homeostasis, skeletal muscle regulation, and immune response\u003csup\u003e\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e。Furthermore, proteins involved in lipid metabolism and energy balance, such as fatty acid‑binding proteins and leptin, also directly influence muscle development and function in frail individuals\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. These findings collectively demonstrate that the development of frailty is accompanied by extensive and dynamic alterations in the proteome \u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eDespite the recognized role of proteins in the pathogenesis of frailty, existing studies have largely focused on individual biomarkers or isolated pathways, leaving the network‑level interactions among proteins in pre‑frailty insufficiently explored. In recent years, the integration of clinical rapid assessments with machine‑learning algorithms have shown promise in translating high‑throughput omics data into clinically actionable insights\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e,\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eIn this study, mass spectrometry-based proteomics was employed to screen for key proteins associated with frailty. By incorporating differentially expressed proteins into algorithmic analyses based on clinical features of frailty, a preliminary interpretable classification framework for prefrailty was established, providing a reference for the precise identification of prefrailty.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eStudy design\u003c/h2\u003e \u003cp\u003e Between December 2022 and June 2023, participants were consecutively recruited from community settings in Beijing, China, with the aim of identifying serum characteristics associated with frailty in older adults. The overall workflow is illustrated in the accompanying Fig.\u0026nbsp;1. During the discovery phase, 10 individuals each from the robust, prefrail, and frail groups were selected for clinical data collection and serum sampling. Untargeted proteomics screening was then performed to identify potential biomarkers. In the validation phase, targeted proteomic validation was carried out using 99 independent participants, comprising 20 robust, 59 prefrail, and 20 frail older adults. Based on data from this validation cohort, feature selection and comparison of multiple machine‑learning algorithms were employed to determine the optimal combination for identifying early frailty. All eligible participants provided written informed consent. The study was conducted in accordance with the principles of the Declaration of Helsinki and received formal approval from the Ethics Committee of the Chinese PLA General Hospital (Approval No. S2019‑140‑03).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eAssessment of Frailty\u003c/h3\u003e\n\u003cp\u003eFrailty was assessed using the Fried Physical Frailty Phenotype, developed by Fried et al. in 2001 \u003csup\u003e15\u003c/sup\u003e. This tool evaluates five core components: unintentional weight loss, reduced grip strength, slowed walking speed, low physical activity, and self‑reported exhaustion. Based on the assessment results, participants were categorized as follows: those meeting three or more criteria were classified as frail, those meeting one to two criteria as prefrail, and those meeting none as robust.\u003c/p\u003e\n\u003ch3\u003eClinical Assessments\u003c/h3\u003e\n\u003cp\u003eClinical data included gender, age, years of education, body mass index (BMI), smoking history, alcohol consumption, insomnia status, sedentary behavior, Mini-Mental State Examination (MMSE) score\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e, Fried frailty score, number of chronic diseases\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e, pain severity measured with the Visual Analogue Scale (VAS) \u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e, nutritional status evaluated by the Mini-Nutritional Assessment-Short Form (MNA-SF) \u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e, depressive symptoms measured by the Geriatric Depression Scale-15 (GDS-15) \u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e, and activities of daily living assessed by the Modified Barthel Index (MBI) \u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e. Functional measurements included customary gait speed, grip strength, Timed Up and Go Test (TUG) time, and the Short Physical Performance Battery (SPPB) \u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eBody composition was measured via direct segmental multi‑frequency bioimpedance analysis (DSM‑BIA) using the InBody 770 device (Korea). Parameters obtained included total body protein, body fat mass (BFM), skeletal muscle mass (SLM).\u003c/p\u003e \u003cp\u003ePhysiological reserve indicators comprised serum levels of vitamin D (Vit D), cystatin C (CysC), insulin‑like growth factor‑1 (IGF‑1), CRP, creatine kinase (CK), creatinine (Cr), and triglycerides (TG)\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\n\u003ch3\u003eSerum Sample Preparation\u003c/h3\u003e\n\u003cp\u003eBlood samples were collected from participants under fasting conditions using vacuum tubes, with a volume of 5 mL per sample. Within one hour after collection, the blood was centrifuged at 3,000 rpm for 10 minutes to separate the serum. The supernatant was carefully aliquoted and stored at \u0026minus;\u0026thinsp;80\u0026deg;C until analysis. For subsequent testing, frozen serum samples were thawed to room temperature prior to processing. If turbidity was observed, the samples were re‑centrifuged at 3,500 rpm for 10 minutes, and the clarified supernatant was used for further assays.\u003c/p\u003e\n\u003ch3\u003eUntargeted proteomic\u003c/h3\u003e\n\u003cp\u003eUntargeted proteomic profiling was performed using data-independent acquisition (DIA) for quantitative analysis\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u003c/sup\u003e. After protein extraction with lysis buffer (8 M urea containing protease inhibitors), high‑abundance proteins were removed using an Agilent MARS Hu‑14 immunoaffinity column to enrich low‑abundance proteins from the flow‑through fraction. Protein concentration was determined by the BCA assay. Subsequent sample processing included reduction (10 mM DTT, 56\u0026deg;C), alkylation (55 mM IAA, in the dark), and digestion via the filter‑aided sample preparation (FASP) method. The resulting peptides were desalted using a C18 column prior to analysis.\u003c/p\u003e \u003cp\u003ePeptide separation was carried out on an EASY‑nLC 1200 nano‑LC system equipped with a C18 analytical column, using mobile phase A (0.1% formic acid in water) and mobile phase B (0.1% formic acid in 80% acetonitrile). The gradient elution program was as follows: 7\u0026ndash;12% B, 12\u0026ndash;30% B, 30\u0026ndash;45% B, 45\u0026ndash;95% B, and maintained until 60 min.\u003c/p\u003e \u003cp\u003eMass spectrometry analysis was performed on a Q Exactive HF mass spectrometer equipped with a Nanospray Flex ion source. The spray voltage was set to 2.2 kV and the ion transfer temperature to 320\u0026deg;C. DIA acquisition parameters included a full‑scan range of 350\u0026ndash;1150 m/z at a resolution of 120,000, followed by 30 MS/MS windows acquired at a resolution of 30,000 with a normalized collision energy of 33%.\u003c/p\u003e \u003cp\u003eDIA data were processed using Spectronaut software against the UniProt human database. Search settings included trypsin digestion with up to two missed cleavages, fixed modification of carbamidomethylation on cysteine, and variable modifications of methionine oxidation and N‑terminal acetylation. Both peptide and protein identifications were filtered at a false discovery rate (FDR)\u0026thinsp;\u0026lt;\u0026thinsp;1%. Label‑free quantification was performed based on MS/MS spectra.\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eTargeted Proteomic\u003c/h2\u003e \u003cp\u003eTargeted proteomic quantification was performed using mass spectrometry-based parallel reaction monitoring (PRM). This method enables both quantitative and qualitative analysis of specific proteins by monitoring the ion signals of predetermined target peptides, with at least two unique peptides selected per protein\u003csup\u003e\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eFollowing tryptic digestion, peptides were separated by liquid chromatography and subsequently analyzed by mass spectrometry. PRM acquisition comprises two main steps: Precursor Ion Selection\u0026mdash;specific precursor ions corresponding to the target peptides were isolated based on their predefined mass-to-charge (m/z) ratios; and Fragment Ion Monitoring\u0026mdash;after fragmentation, the resulting product ions were recorded. Quantification was performed based on the signal intensity of these fragment ions, allowing relative or absolute protein abundance to be determined.\u003c/p\u003e \u003cp\u003eAnalyses were carried out on a Q-Exactive high-resolution mass spectrometer, which offers precise mass accuracy and high sensitivity to ensure reliable detection of low-abundance proteins. Peak identification was performed using chromatographic and spectral data generated from the instrument. The signal intensity (peak area) of each peptide was extracted for quantification. Where applicable, absolute quantification was achieved by constructing a calibration curve using synthetic peptide standards of known concentration.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003eStatistical Analysis\u003c/h2\u003e \u003cp\u003eData were processed and analyzed using IBM SPSS Statistics 27.0. All statistical tests were two‑sided, with \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05 considered statistically significant. Missing values were imputed five times via Markov Chain Monte Carlo (MCMC) multiple imputation. Outlier treatment was performed in conjunction with the actual pathophysiological status of the participants. When outliers were consistent with pathological upregulation, they were considered valid data and retained for analysis.\u003c/p\u003e \u003cp\u003eNormality of continuous variables, including age, years of education, BMI, MMSE score, Fried score, number of chronic diseases, VAS score, MNA‑SF score, GDS‑15 score, MBI step count, grip strength, TUG time, SPPB score, body composition measures, and routine hematological parameters, was assessed using the Shapiro\u0026ndash;Wilk test across the three groups. For variables following a normal distribution, one‑way analysis of variance (ANOVA) was used for group comparisons; otherwise, the Kruskal\u0026ndash;Wallis rank‑sum test was applied. Continuous variables with a normal distribution were expressed as mean\u0026thinsp;\u0026plusmn;\u0026thinsp;standard deviation (Mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD), while those with a non‑normal distribution were presented as median (interquartile range) [M (P25\u0026ndash;P75)]. Categorical variables, including gender, smoking history, alcohol consumption, insomnia status, and sedentary behavior, were described using constituent ratios and analyzed using the chi‑squared test. Spearman\u0026rsquo;s rank correlation analysis was performed to evaluate the correlations between differentially expressed proteins and clinical characteristics.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eBioinformatics Analysis\u003c/h3\u003e\n\u003cp\u003eQuantitative peptide data obtained from proteomic profiling were processed using DIA‑NN software against the UniProt human proteome database. Peptide intensities were aggregated into protein‑level relative expression values and represented as log2‑transformed label‑free quantities. Following quantile normalization and missing‑value filtering, differentially expressed proteins among the three study groups were initially identified by Kruskal‑Wallis test (FDR\u0026thinsp;\u0026lt;\u0026thinsp;0.05), followed by pairwise comparisons using Dunn\u0026rsquo;s test with Benjamini‑Hochberg correction. Proteins meeting both |log2 FC| \u0026gt; 0.58 and FDR\u0026thinsp;\u0026lt;\u0026thinsp;0.05 were considered statistically significant differentially expressed proteins. To further explore their dynamic expression patterns, these significant proteins were subjected to Mfuzz clustering.\u003c/p\u003e \u003cp\u003ePathway enrichment analysis of differentially expressed proteins was performed using the Gene Ontology (GO) database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.geneontology.org/\u003c/span\u003e\u003cspan address=\"http://www.geneontology.org/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) and the Kyoto Encyclopedia of Genes and Genomes (KEGG) database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.kegg.jp/kegg/pathway.html\u003c/span\u003e\u003cspan address=\"https://www.kegg.jp/kegg/pathway.html\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). For GO analysis, enriched terms were identified by comparing the frequency of differentially expressed proteins assigned to each term against the background proteome, with significantly over‑represented terms (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05) reported to highlight biological functions, cellular components, and molecular processes associated with frailty‑related proteomic changes. For KEGG analysis, enriched pathways (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05) were visualized in bubble plots, where the enrichment score (rich factor), \u003cem\u003eP\u003c/em\u003e‑value, and number of mapped proteins were integrated to identify key biological pathways involved in frailty progression.\u003c/p\u003e \u003cp\u003eProtein-protein interaction (PPI) network analysis was performed using the STRING database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://string-db.org/\u003c/span\u003e\u003cspan address=\"https://string-db.org/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). Genomic associations between proteins were calculated based on a scoring framework that integrates different types of association benchmarks derived from commonly used reference datasets, with each prediction assigned a single confidence score. Modular analysis was subsequently conducted to infer biological processes.\u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eFeature Selection\u003c/h2\u003e \u003cp\u003eThe dependent variable was defined as prefrailty, with a Fried phenotype score of 1\u0026ndash;2 serving as the outcome of interest. Robust and frail groups were included as controls to help exclude false‑positive proteins potentially arising from non‑specific aging in the older adult population. A total of 26 candidate parameters were considered as independent variables, among which those showing statistically significant differences across the three groups in the validation cohort were selected (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05). These included 11 clinical parameters, namely sedentary behavior, MMSE, MBI, gait speed, grip strength, TUG, SPPB, protein, SLM, CysC, and IGF-1, and 15 differentially expressed proteins, including CP, F8, HPX, VWF, RPLP0, ATP5F1B, LPL, F5, MYH9, HSPG2, SVEP1, BPIFB1, UBE2O, ACTN3, and TREML1.\u003c/p\u003e \u003cp\u003eFirst, the least absolute shrinkage and selection operator (LASSO) regression algorithm was applied for feature standardization by adding an L1 norm to the objective function to penalize the regression coefficients. This resulted in the shrinkage of regression coefficients of redundant variables to zero, thereby achieving feature selection, reducing model complexity, and addressing multicollinearity\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eSubsequently, Kendall's rank correlation coefficient (τ) was used to detect monotonic correlations among features to ensure model parsimony and stability\u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e. Pairwise Kendall correlation coefficients were calculated for all features to generate a symmetric matrix. A threshold of |τ| \u0026ge; 0.8 was set, where feature pairs with absolute correlation values above this threshold were considered redundant. In such cases, the feature with a stronger correlation to the target variable was retained, and the other was removed.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eMachine Learning\u003c/h2\u003e \u003cp\u003eModel development and comparison were performed in the Python 3.7 programming environment using eight machine learning algorithms based on the selected important features. These included ensemble learning algorithms suitable for high-dimensional data: random forest (RF), gradient boosting (GB), and extreme gradient boosting (XGBoost) \u003csup\u003e\u003cspan additionalcitationids=\"CR29\" citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e; linear algorithms with good interpretability: logistic regression (LR) and ridge regression (RR) \u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e,\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e; a probabilistic algorithm: naive Bayes (NB) \u003csup\u003e\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e; an instance‑based learning method: K‑nearest neighbors (KNN) \u003csup\u003e\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e; and a single‑tree algorithm: decision tree (DT) \u003csup\u003e\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e. During the modeling process, five rounds of cross‑validation and grid search for parameter optimization were repeated to prevent overfitting. Hyperparameter tuning for all algorithms was performed on the training set using nested cross‑validation. The final results were summarized across five rounds of validation, and the average performance of each algorithm was calculated. Model performance was evaluated primarily using the receiver operating characteristic (ROC) curve and the area under the curve (AUC) along with its 95% confidence interval (CI), supplemented by accuracy, precision, recall, and F1 score for comprehensive assessment of the classification performance of the eight machine learning algorithms.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eModel Interpretability\u003c/h2\u003e \u003cp\u003eShapley Additive Explanations (SHAP) were applied to interpret the optimal model. Rooted in cooperative game theory, SHAP assigns each feature an importance value (Shapley value) by averaging its marginal contribution across all possible feature subsets. This approach provides a mathematically grounded, transparent explanation of how each feature influences model predictions, thereby enhancing the interpretability and credibility of the model\u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e。\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eClinical Characteristics in Discovery Cohort\u003c/h2\u003e \u003cp\u003eIn the discovery phase, the mean age of the participants was 77.00\u0026thinsp;\u0026plusmn;\u0026thinsp;3.17 years, and 15 participants were male (50%). No statistically significant differences were observed in general demographic characteristics among the three groups (all \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026gt;\u0026thinsp;0.05). In terms of clinical functional performance, Fried score (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001), VAS score (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.010), GDS-15 score (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.013), and TUG time increased with the severity of frailty (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.009). MNA-SF score (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.024), MBI (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.009), gait speed (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.003), grip strength (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.013), and SPPB score (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.010) were all decreased in both the prefrail and frail groups. Circulating levels of CK (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.038) also decreased progressively from the robust to the frail state. No significant inter‑group differences were observed for number of chronic diseases, protein, BFM, SLM, Vit D, CysC, IGF-1, CRP, Cr, or TG (all \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026gt;\u0026thinsp;0.05, Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eClinical characteristics of participants in the discovery cohort\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCharacteristic\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eRobust (n\u0026thinsp;=\u0026thinsp;10)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003ePrefrail (n\u0026thinsp;=\u0026thinsp;10)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eFrail (n\u0026thinsp;=\u0026thinsp;10)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eF/Z/χ\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u003cem\u003eP\u003c/em\u003e Value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"6\" nameend=\"c6\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eDemographic characteristics\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGender (male/female, n)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5/5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e5/5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e5/5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e1.000\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAge(years)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e75.60\u0026thinsp;\u0026plusmn;\u0026thinsp;1.17\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e78.10\u0026thinsp;\u0026plusmn;\u0026thinsp;4.58\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e77.30\u0026thinsp;\u0026plusmn;\u0026thinsp;2.54\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1.697\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.202\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEducation(years)*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e12.00(9.00\u0026ndash;16.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e9.00(8.25-16.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e11.32(9.00\u0026ndash;12.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.476\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.788\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBMI(kg/m2)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e23.86\u0026thinsp;\u0026plusmn;\u0026thinsp;1.01\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e24.07\u0026thinsp;\u0026plusmn;\u0026thinsp;3.46\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e24.16\u0026thinsp;\u0026plusmn;\u0026thinsp;4.80\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.020\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.980\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSmoking history(yes/no)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4/6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2/8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e4/6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1.200\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.549\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAlcohol consumption(yes/no)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e6/4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2/8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e3/7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e3.732\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.155\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eInsomnia(yes/no)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3/7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e7/3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e8/2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e5.833\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.054\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSedentary behavior(yes/no)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2/8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e5/5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e7/3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e5.089\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.079\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMMSE*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e27.50(25.00\u0026ndash;29.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e26.50(25.00\u0026ndash;29.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e27.50(25.00-28.25)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.327\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.849\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"6\" nameend=\"c6\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eClinical characteristics\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFried*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.00(0.00\u0026ndash;0.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.00(1.00-1.25)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e3.00(3.00-3.25)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e27.878\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNumber of chronic diseases*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e2.00(1.00-2.50)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2.22(1.00-4.02)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e3.50(1.75-4.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e2.421\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.298\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVAS*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.00(0.00\u0026ndash;0.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.00(0.00\u0026ndash;5.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e4.00(0.28\u0026ndash;5.25)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e9.280\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.010\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMNA-SF*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e14.00(14.00\u0026ndash;14.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e12.50(10.00\u0026ndash;14.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e13.00(10.75-14.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e7.441\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.024\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGDS-15*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.00(0.00\u0026ndash;2.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.00(1.00\u0026ndash;2.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e3.20(1.75\u0026ndash;6.25)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e8.730\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.013\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMBI*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e100.00(99.50\u0026ndash;100.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e99.00(94.25\u0026ndash;100.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e93.50(86.75\u0026ndash;98.50)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e9.368\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.009\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGait speed\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.45\u0026thinsp;\u0026plusmn;\u0026thinsp;0.56\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.21\u0026thinsp;\u0026plusmn;\u0026thinsp;0.37\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.76\u0026thinsp;\u0026plusmn;\u0026thinsp;0.25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e7.066\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.003\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGrip strength\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e31.93\u0026thinsp;\u0026plusmn;\u0026thinsp;9.46\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e29.75\u0026thinsp;\u0026plusmn;\u0026thinsp;6.42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e22.03\u0026thinsp;\u0026plusmn;\u0026thinsp;5.35\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e5.093\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.013\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTUG*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e8.49(7.07\u0026ndash;10.40)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e8.96(8.10-12.51)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e11.82(10.71\u0026ndash;13.24)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e9.363\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.009\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSPPB*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e12.00(10.75-12.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e11.00(9.75-11.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e9.00(7.75\u0026ndash;10.25)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e9.206\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.010\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eProtein\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e9.16\u0026thinsp;\u0026plusmn;\u0026thinsp;1.58\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e8.75\u0026thinsp;\u0026plusmn;\u0026thinsp;1.64\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e8.21\u0026thinsp;\u0026plusmn;\u0026thinsp;1.58\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.887\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.424\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBFM\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e17.75\u0026thinsp;\u0026plusmn;\u0026thinsp;2.06\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e21.00\u0026thinsp;\u0026plusmn;\u0026thinsp;7.30\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e22.07\u0026thinsp;\u0026plusmn;\u0026thinsp;9.93\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.973\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.391\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSLM\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e44.18\u0026thinsp;\u0026plusmn;\u0026thinsp;7.47\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e42.47\u0026thinsp;\u0026plusmn;\u0026thinsp;7.94\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e40.25\u0026thinsp;\u0026plusmn;\u0026thinsp;7.78\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.648\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.531\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVit D (ng/ml)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e22.55\u0026thinsp;\u0026plusmn;\u0026thinsp;6.52\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e23.19\u0026thinsp;\u0026plusmn;\u0026thinsp;13.19\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e17.68\u0026thinsp;\u0026plusmn;\u0026thinsp;4.21\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1.163\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.328\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCysC\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.09\u0026thinsp;\u0026plusmn;\u0026thinsp;0.01\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.11\u0026thinsp;\u0026plusmn;\u0026thinsp;0.02\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.11\u0026thinsp;\u0026plusmn;\u0026thinsp;0.03\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e2.940\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.070\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIGF-1(ng/mL)*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e13.99(10.90-22.02)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e29.41(18.96\u0026ndash;33.26)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e21.02(15.37\u0026ndash;27.80)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e5.772\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.056\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCRP (mg/L)*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.92(0.13\u0026ndash;1.66)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.49(0.35\u0026ndash;0.63)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.63(0.16\u0026ndash;2.80)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.321\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.852\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCK (U/L)*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e89.91(64.75-110.88)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e87.80(67.10-138.48)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e57.95(53.30-84.37)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e6.521\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.038\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCr (umol/L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e71.99\u0026thinsp;\u0026plusmn;\u0026thinsp;12.11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e70.15\u0026thinsp;\u0026plusmn;\u0026thinsp;15.64\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e70.01\u0026thinsp;\u0026plusmn;\u0026thinsp;13.41\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.064\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.938\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTG (mmol/L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1.10(0.97\u0026ndash;1.86)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.27(0.86\u0026ndash;1.85)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1.18(1.07\u0026ndash;1.71)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.156\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.925\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"6\"\u003e\u003cb\u003eNotes\u003c/b\u003e: *, the Kruskal‑Wallis test was used for non‑normally distributed data.\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e\u003cp\u003e\u003cstrong\u003eAbbreviations:\u0026nbsp;\u003c/strong\u003eBMI, body mass index; MMSE, Mini-Mental State Examination; VAS, Visual AnalogueScale; MNA-SF, Mini-Nutritional Assessment-Short Form; GDS-15, Geriatric Depression Scale (15 items); MBI, Modified Barthel Index; TUG, Timed Up and Go Test; SPPB, Short Physical Performance Battery; BFM, body fat mass; SLM, skeletal muscle mass; Vit D, vitamin D; CysC, cystatin C; IGF 1, insulin like growth factor 1; CRP, C-reactive protein; CK, creatine kinase; Cr, creatinine; TG, triglycerides.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eProteins in the Discovery Cohort\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn the discovery cohort, a total of 1,145 raw proteins were identified at quantifiable levels. Comparative analysis revealed 166 proteins that were differentially expressed among the robust, prefrail, and frail groups (Figure 2A). To elucidate the biological functions associated with frailty progression, enrichment analyses were performed using the GO and KEGG databases. Significantly enriched GO terms included vesicle‑mediated transport, regulation of response to stimulus, regulation of immune system processes, platelet activation, regulation of cell communication, proteolysis, muscle structure development, and muscle cell differentiation (Figure 2B).\u003c/p\u003e\n\u003cp\u003eKEGG pathway analysis revealed that the transition from robust to frail states involved complement and coagulation cascades, ferroptosis, focal adhesion, riboflavin metabolism, platelet activation, cholesterol metabolism, glutathione metabolism, glycerolipid metabolism, and ubiquitin‑mediated proteolysis (Figure 2C).\u003c/p\u003e\n\u003cp\u003ePPI was constructed to visualize functional connectivity among the differentially expressed proteins (Figure 2D). Clustering analysis highlighted 57 key proteins involved in biological processes such as regulation of adaptive immune response, regulation of cytokine production, protein polyubiquitination, regulation of skeletal muscle tissue development, lipid metabolism, and proteolysis.\u003c/p\u003e\n\u003cp\u003eBased on the above bioinformatic analysis of differentially expressed proteins, candidate proteins were prioritized according to their correlation with clinical phenotypes (Figure 2E), expression trends (Figure 2F), and sex specificity (Figure 2G). This led to the selection of 30 candidate proteins as potential frailty biomarkers for subsequent PRM quantitative validation in the verification cohort (all \u003cem\u003eP\u003c/em\u003e \u0026lt; 0.05; Table 2).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 2\u0026nbsp;\u003c/strong\u003eCandidate proteins associated with frailty\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"100%\"\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 15px;\"\u003e\n \u003cp\u003eAccession No.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 11px;\"\u003e\n \u003cp\u003eProtein\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003eName\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 8px;\"\u003e\n \u003cp\u003eF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 9px;\"\u003e\n \u003cp\u003e\u003cem\u003eP\u0026nbsp;\u003c/em\u003eValue\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 15px;\"\u003e\n \u003cp\u003eE9PAV3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 11px;\"\u003e\n \u003cp\u003eNACA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003eNascent-Polypeptide-Associated Complex Alpha Polypeptide\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 8px;\"\u003e\n \u003cp\u003e3.592\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 9px;\"\u003e\n \u003cp\u003e0.043\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 15px;\"\u003e\n \u003cp\u003eP00450\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 11px;\"\u003e\n \u003cp\u003eCP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003eCeruloplasmin\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 8px;\"\u003e\n \u003cp\u003e3.919\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 9px;\"\u003e\n \u003cp\u003e0.030\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 15px;\"\u003e\n \u003cp\u003eP00451\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 11px;\"\u003e\n \u003cp\u003eF8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003eCoagulation Factor VIII\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 8px;\"\u003e\n \u003cp\u003e4.579\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 9px;\"\u003e\n \u003cp\u003e0.031\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 15px;\"\u003e\n \u003cp\u003eP02787\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 11px;\"\u003e\n \u003cp\u003eTF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003eTranscription Factor\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 8px;\"\u003e\n \u003cp\u003e3.923\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 9px;\"\u003e\n \u003cp\u003e0.023\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 15px;\"\u003e\n \u003cp\u003eP02790\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 11px;\"\u003e\n \u003cp\u003eHPX\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003eHemopexin\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 8px;\"\u003e\n \u003cp\u003e3.861\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 9px;\"\u003e\n \u003cp\u003e0.024\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 15px;\"\u003e\n \u003cp\u003eP04275\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 11px;\"\u003e\n \u003cp\u003eVWF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003eVon Willebrand Factor\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 8px;\"\u003e\n \u003cp\u003e4.771\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 9px;\"\u003e\n \u003cp\u003e0.017\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 15px;\"\u003e\n \u003cp\u003eP05156\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 11px;\"\u003e\n \u003cp\u003eCFI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003eComplement Factor I\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 8px;\"\u003e\n \u003cp\u003e3.749\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 9px;\"\u003e\n \u003cp\u003e0.039\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 15px;\"\u003e\n \u003cp\u003eP05388\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 11px;\"\u003e\n \u003cp\u003eRPLP0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003e60S Acidic Ribosomal Protein P0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 8px;\"\u003e\n \u003cp\u003e3.448\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 9px;\"\u003e\n \u003cp\u003e0.048\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 15px;\"\u003e\n \u003cp\u003eP06576\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 11px;\"\u003e\n \u003cp\u003eATP5F1B\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003eATP Synthase Subunit Beta， Mitochondrial\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 8px;\"\u003e\n \u003cp\u003e6.353\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 9px;\"\u003e\n \u003cp\u003e0.003\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 15px;\"\u003e\n \u003cp\u003eP06681\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 11px;\"\u003e\n \u003cp\u003eC2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003eComplement component 2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 8px;\"\u003e\n \u003cp\u003e4.247\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 9px;\"\u003e\n \u003cp\u003e0.043\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 15px;\"\u003e\n \u003cp\u003eP06858\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 11px;\"\u003e\n \u003cp\u003eLPL\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003eLipoprotein Lipase\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 8px;\"\u003e\n \u003cp\u003e4.594\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 9px;\"\u003e\n \u003cp\u003e0.019\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 15px;\"\u003e\n \u003cp\u003eP07585\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 11px;\"\u003e\n \u003cp\u003eDCN\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003eDecorin\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 8px;\"\u003e\n \u003cp\u003e17.270\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 9px;\"\u003e\n \u003cp\u003e0.000\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 15px;\"\u003e\n \u003cp\u003eP08311\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 11px;\"\u003e\n \u003cp\u003eCTSG\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003eCathepsin G\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 8px;\"\u003e\n \u003cp\u003e3.752\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 9px;\"\u003e\n \u003cp\u003e0.027\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 15px;\"\u003e\n \u003cp\u003eP12259\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 11px;\"\u003e\n \u003cp\u003eF5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003eCoagulation Factor V\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 8px;\"\u003e\n \u003cp\u003e4.579\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 9px;\"\u003e\n \u003cp\u003e0.042\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 15px;\"\u003e\n \u003cp\u003eP20742\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 11px;\"\u003e\n \u003cp\u003ePZP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003ePregnancy Zone Protein\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 8px;\"\u003e\n \u003cp\u003e3.927\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 9px;\"\u003e\n \u003cp\u003e0.024\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 15px;\"\u003e\n \u003cp\u003eP28289\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 11px;\"\u003e\n \u003cp\u003eTMOD1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003eTropomodulin-1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 8px;\"\u003e\n \u003cp\u003e3.358\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 9px;\"\u003e\n \u003cp\u003e0.048\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 15px;\"\u003e\n \u003cp\u003eP35443\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 11px;\"\u003e\n \u003cp\u003eTHBS4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003eThrombospondin 4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 8px;\"\u003e\n \u003cp\u003e4.743\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 9px;\"\u003e\n \u003cp\u003e0.025\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 15px;\"\u003e\n \u003cp\u003eP35579\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 11px;\"\u003e\n \u003cp\u003eMYH9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003eMyosin-9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 8px;\"\u003e\n \u003cp\u003e6.566\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 9px;\"\u003e\n \u003cp\u003e0.006\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 15px;\"\u003e\n \u003cp\u003eP51884\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 11px;\"\u003e\n \u003cp\u003eLUM\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003eLumican\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 8px;\"\u003e\n \u003cp\u003e3.758\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 9px;\"\u003e\n \u003cp\u003e0.029\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 15px;\"\u003e\n \u003cp\u003eP55010\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 11px;\"\u003e\n \u003cp\u003eEIF5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003eEukaryotic Translation Initiation Factor 5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 8px;\"\u003e\n \u003cp\u003e6.952\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 9px;\"\u003e\n \u003cp\u003e0.003\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 15px;\"\u003e\n \u003cp\u003eP62256\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 11px;\"\u003e\n \u003cp\u003eUBE2H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003eUbiquitin-conjugating Enzyme E2 H\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 8px;\"\u003e\n \u003cp\u003e5.162\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 9px;\"\u003e\n \u003cp\u003e0.010\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 15px;\"\u003e\n \u003cp\u003eP98160\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 11px;\"\u003e\n \u003cp\u003eHSPG2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003eBasement Membrane-Specific Heparan Sulfate Proteoglycan Core Protein\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 8px;\"\u003e\n \u003cp\u003e4.872\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 9px;\"\u003e\n \u003cp\u003e0.022\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 15px;\"\u003e\n \u003cp\u003eQ08043\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 11px;\"\u003e\n \u003cp\u003eACTN3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003eAlpha-Actinin-3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 8px;\"\u003e\n \u003cp\u003e4.855\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 9px;\"\u003e\n \u003cp\u003e0.013\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 15px;\"\u003e\n \u003cp\u003eQ08495\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 11px;\"\u003e\n \u003cp\u003eDEMA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003eDynamin-Related Protein 1\u0026nbsp;/\u0026nbsp;Dermatological Evaluation of Mucocutaneous Adverse Events\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 8px;\"\u003e\n \u003cp\u003e3.919\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 9px;\"\u003e\n \u003cp\u003e0.027\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 15px;\"\u003e\n \u003cp\u003eQ4LDE5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 11px;\"\u003e\n \u003cp\u003eSVEP1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003eSushi，Von Willebrand Factor Type A，EGF and Pentraxin Domain-containing Protein 1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 8px;\"\u003e\n \u003cp\u003e3.878\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 9px;\"\u003e\n \u003cp\u003e0.023\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 15px;\"\u003e\n \u003cp\u003eQ86YW5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 11px;\"\u003e\n \u003cp\u003eTREML1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003eTrem-like Transcript 1 Protein\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 8px;\"\u003e\n \u003cp\u003e3.542\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 9px;\"\u003e\n \u003cp\u003e0.035\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 15px;\"\u003e\n \u003cp\u003eQ8TDL5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 11px;\"\u003e\n \u003cp\u003eBPIFB1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003eBPI Fold-containing Family B Member 1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 8px;\"\u003e\n \u003cp\u003e6.241\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 9px;\"\u003e\n \u003cp\u003e0.006\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 15px;\"\u003e\n \u003cp\u003eQ8WWA0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 11px;\"\u003e\n \u003cp\u003eITLN1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003eIntelectin 1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 8px;\"\u003e\n \u003cp\u003e3.424\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 9px;\"\u003e\n \u003cp\u003e0.041\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 15px;\"\u003e\n \u003cp\u003eQ9BY43\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 11px;\"\u003e\n \u003cp\u003eCHMP4A\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003eCharged Multivesicular Body Protein 4A\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 8px;\"\u003e\n \u003cp\u003e4.253\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 9px;\"\u003e\n \u003cp\u003e0.015\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" style=\"width: 15px;\"\u003e\n \u003cp\u003eQ9C0C9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 11px;\"\u003e\n \u003cp\u003eUBE2O\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 54px;\"\u003e\n \u003cp\u003e(E3-independent) E2 Ubiquitin-conjugating Enzyme\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 8px;\"\u003e\n \u003cp\u003e4.273\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" style=\"width: 9px;\"\u003e\n \u003cp\u003e0.021\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cstrong\u003eClinical Characteristics in Validation Cohort\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn the validation phase, the mean age of participants was 75.32 \u0026plusmn; 5.09 years, with 27 males (27.3%). The prevalence of sedentary behavior was significantly higher in the pre‑frail and frail groups (47.5% and 95.0%, respectively; \u003cem\u003eP\u003c/em\u003e = 0.001) compared to the robust group (45.0%). MMSE scores decreased progressively with worsening frailty status (\u003cem\u003eP\u003c/em\u003e = 0.004). No significant differences were observed among the three groups in gender, age, years of education, BMI, smoking history, alcohol consumption use, or insomnia (all \u003cem\u003eP\u003c/em\u003e \u0026gt; 0.05).\u003c/p\u003e\n\u003cp\u003eFrailty status differed significantly among groups (\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.001), with Fried scores increasing across the robust, pre‑frail, and frail categories. Compared to the robust group, the frail group showed significantly lower scores in the MBI, gait speed, handgrip strength, and SPPB (all \u003cem\u003eP\u003c/em\u003e \u0026lt; 0.05). TUG time was significantly longer in both the pre‑frail and frail groups than in the robust group (\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.001). Total body protein and SLM were notably lower in the pre‑frail group compared to the robust group (\u003cem\u003eP \u003c/em\u003e= 0.017 and \u003cem\u003eP\u003c/em\u003e = 0.009, respectively). CysC and IGF‑1 levels were significantly altered in the frail group compared with the robust group. (\u003cem\u003eP\u003c/em\u003e = 0.040 and\u003cem\u003e\u0026nbsp;P\u003c/em\u003e = 0.004, respectively). In contrast, no significant inter‑group differences were detected for number of chronic diseases, VAS, MNA‑SF, GDS‑15, BFM, Vit D, CRP, CK, Cr, or TG levels (all \u003cem\u003eP\u003c/em\u003e \u0026gt; 0.05; Table 3).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 3\u003c/strong\u003e Comparison of clinical characteristics in the validation cohort\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"118%\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 168px;\"\u003e\n \u003cp\u003eCharacteristic\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003eRobust(n=20)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003ePrefrail(n=59)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 128px;\"\u003e\n \u003cp\u003eFrail(n=20)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003eF/Z/\u0026chi;\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e\u003cem\u003eP\u0026nbsp;\u003c/em\u003eValue\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" colspan=\"6\" valign=\"top\" style=\"width: 674px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eDemographic characteristics\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 168px;\"\u003e\n \u003cp\u003eGender (male/female, n)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e7/13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e12/47\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 128px;\"\u003e\n \u003cp\u003e8/12\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e3.666\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.160\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 168px;\"\u003e\n \u003cp\u003eAge(years)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e75.55\u0026plusmn;4.20\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e74.71\u0026plusmn;4.97\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 128px;\"\u003e\n \u003cp\u003e76.90\u0026plusmn;6.06\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e1.419\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.247\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 168px;\"\u003e\n \u003cp\u003eEducation(years)*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e11.28(9.00-16.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e12.00(9.00-14.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 128px;\"\u003e\n \u003cp\u003e12.00(9.00-16.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e0.452\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.798\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 168px;\"\u003e\n \u003cp\u003eBMI(kg/m2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e25.70\u0026plusmn;3.65\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e23.91\u0026plusmn;3.02\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 128px;\"\u003e\n \u003cp\u003e23.69\u0026plusmn;3.48\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e2.616\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.078\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 168px;\"\u003e\n \u003cp\u003eSmoking history(yes/no)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e1/19\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e8/51\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 128px;\"\u003e\n \u003cp\u003e2/18\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e1.139\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.566\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 168px;\"\u003e\n \u003cp\u003eAlcohol consumption(yes/no)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e6/14\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e13/46\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 128px;\"\u003e\n \u003cp\u003e2/18\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e2.452\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.293\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 168px;\"\u003e\n \u003cp\u003eInsomnia(yes/no)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e10/10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e32/27\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 128px;\"\u003e\n \u003cp\u003e11/9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e0.129\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.937\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 168px;\"\u003e\n \u003cp\u003eSedentary behavior(yes/no)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e9/11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e28/31\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 128px;\"\u003e\n \u003cp\u003e19/1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e15.106\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 168px;\"\u003e\n \u003cp\u003eMMSE*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e28.00(27.00-29.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e28.00(26.30-29.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 128px;\"\u003e\n \u003cp\u003e26.00(25.00-27.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e10.941\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.004\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" colspan=\"6\" valign=\"top\" style=\"width: 674px;\"\u003e\n \u003cp\u003e\u003cstrong\u003eClinical characteristics\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 168px;\"\u003e\n \u003cp\u003eFried*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e0.00(0.00-0.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e1.00(1.00-2.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 128px;\"\u003e\n \u003cp\u003e3.00(3.00-4.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e61.308\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 168px;\"\u003e\n \u003cp\u003eNumber of chronic diseases*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e2.50(2.00-4.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e2.00(1.00-4.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 128px;\"\u003e\n \u003cp\u003e3.00(2.00-5.50)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e4.454\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.110\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 168px;\"\u003e\n \u003cp\u003eVAS*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e0.00(0.00-1.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e0.00(0.00-5.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 128px;\"\u003e\n \u003cp\u003e3.00(0.00-5.50)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e5.557\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.060\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 168px;\"\u003e\n \u003cp\u003eMNA-SF*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e14.00(13.00-14.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e13.00(12.00-14.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 128px;\"\u003e\n \u003cp\u003e13.00(12.00-14.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e4.895\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.090\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 168px;\"\u003e\n \u003cp\u003eGDS-15*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e1.00(0.00-2.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e2.00(1.00-3.50)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 128px;\"\u003e\n \u003cp\u003e2.00(1.00-4.50)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e5.494\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.060\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 168px;\"\u003e\n \u003cp\u003eMBI*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e100.00(98.00-100.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e100.00(96.00-100.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 128px;\"\u003e\n \u003cp\u003e88.00(83.5-94.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e30.948\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 168px;\"\u003e\n \u003cp\u003eGait speed\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e1.13(0.90-1.40)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e1.09(0.97-1.30)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 128px;\"\u003e\n \u003cp\u003e0.69(0.47-0.81)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e23.608\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 168px;\"\u003e\n \u003cp\u003eGrip strength\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e25.45(23.60-35.35)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e24.10(21.95-29.55)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 128px;\"\u003e\n \u003cp\u003e21.70(18.80-25.10)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e9.785\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.010\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 168px;\"\u003e\n \u003cp\u003eTUG*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e8.74(8.15-9.64)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e9.00(7.80-11.51)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 128px;\"\u003e\n \u003cp\u003e15.42(13.44-25.72)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e24.813\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 168px;\"\u003e\n \u003cp\u003eSPPB*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e11.00(9.00-11.50)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e11.00(10.00-12.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 128px;\"\u003e\n \u003cp\u003e7.00(3.50-8.50)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e28.535\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e\u0026lt;0.001\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 168px;\"\u003e\n \u003cp\u003eProtein\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e8.60(7.80-10.55)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e7.60(7.25-8.50)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 128px;\"\u003e\n \u003cp\u003e8.40(7.30-9.25)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e8.205\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.017\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 168px;\"\u003e\n \u003cp\u003eBFM\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e22.08\u0026plusmn;6.52\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e20.59\u0026plusmn;5.86\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 128px;\"\u003e\n \u003cp\u003e20.23\u0026plusmn;5.67\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e0.589\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.557\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 168px;\"\u003e\n \u003cp\u003eSLM\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e41.70(38.05-51.15)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e37.40(35.00-41.35)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 128px;\"\u003e\n \u003cp\u003e41.15(35.75-45.85)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e9.403\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.009\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 168px;\"\u003e\n \u003cp\u003eVit D (ng/ml)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e17.93(11.26-27.93)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e17.28(13.45-24.64)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 128px;\"\u003e\n \u003cp\u003e16.77(11.04-23.03)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e0.983\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.612\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 168px;\"\u003e\n \u003cp\u003eCysC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e0.10(0.09-0.12)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e0.10(0.09-0.11)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 128px;\"\u003e\n \u003cp\u003e0.12(0.10-0.14)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e6.441\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.040\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 168px;\"\u003e\n \u003cp\u003eIGF-1(ng/mL)*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e20.83(12.92-34.12)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e18.97(13.62-27.64)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 128px;\"\u003e\n \u003cp\u003e38.75(21.71-49.42)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e11.007\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.004\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 168px;\"\u003e\n \u003cp\u003eCRP (mg/L)*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e0.52(0.30-1.33)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e0.54(0.30-1.22)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 128px;\"\u003e\n \u003cp\u003e0.73(0.20-1.22)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e0.072\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.965\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 168px;\"\u003e\n \u003cp\u003eCK (U/L)*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e85.15(72.25-114.30)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e75.80(58.10-100.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 128px;\"\u003e\n \u003cp\u003e62.65(44.05-110.90)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e4.887\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.087\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 168px;\"\u003e\n \u003cp\u003eCr (umol/L)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e65.53(48.20-81.25)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e65.00(53.65-77.55)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 128px;\"\u003e\n \u003cp\u003e70.65(59.85-91.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e4.767\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.092\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 168px;\"\u003e\n \u003cp\u003eTG (mmol/L)*\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e1.55(1.22-2.17)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 132px;\"\u003e\n \u003cp\u003e1.23(0.90-1.63)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 128px;\"\u003e\n \u003cp\u003e1.34(1.12-1.96)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 53px;\"\u003e\n \u003cp\u003e4.859\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd nowrap=\"\" valign=\"top\" style=\"width: 61px;\"\u003e\n \u003cp\u003e0.088\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cstrong\u003eNotes:\u0026nbsp;\u003c/strong\u003e*, the Kruskal‑Wallis test was used for non‑normally distributed data\u003cstrong\u003e.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAbbreviations:\u0026nbsp;\u003c/strong\u003eBMI,body mass index;MMSE,Mini-Mental State Examination;VAS,Visual Analogue Scale;MNA-SF,Mini-Nutritional Assessment-Short Form;GDS-15,Geriatric Depression Scale (15 items);MBI,Modified Barthel Index;TUG,Timed Up and Go Test;SPPB,Short Physical Performance Battery; BFM,body fat mass;SLM,skeletal muscle mass;Vit D,vitamin D;CysC,cystatin C;IGF 1,insulin like growth factor 1;CRP,C-reactive protein;CK,creatine kinase;Cr,creatinine;TG,triglycerides\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTargeted Proteins in the Validation Cohort\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTargeted proteomic analysis via PRM was performed in the validation cohort to quantify the expression levels of 30 candidate proteins previously identified in the discovery phase (Figure 3A). Among these, 15 proteins exhibited statistically significant differential expression across the robust, pre‑frail, and frail groups\u0026nbsp;(\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.05; Figure 3B), including ceruloplasmin (CP), coagulation factor VIII (F8), hemopexin (HPX), von Willebrand factor (VWF), 60S acidic ribosomal protein P0 (RPLP0), ATP synthase subunit beta mitochondrial (ATP5F1B), lipoprotein lipase (LPL), coagulation factor V (F5), myosin‑9 (MYH9), basement membrane‑specific heparan sulfate proteoglycan core protein (HSPG2), SVEP1 (sushi, von Willebrand factor type A, EGF and pentraxin domain‑containing protein 1), BPI fold‑containing family B member 1 (BPIFB1), ubiquitin‑conjugating enzyme E2O (UBE2O), alpha‑actinin‑3 (ACTN3), and trem‑like transcript 1 protein (TREML1).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFeature Selection\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA total of 26 independent variables showing statistical significance in the validation phase were subjected to LASSO regression analysis (Figure 4A). Feature selection was performed using LASSO with cross‑validation, and seven features with non‑zero coefficients corresponding to the optimal lambda on the left side of the curve were selected: MMSE, gait speed, TUG, protein, SLM, UBE2O, and ACTN3 (Figure 4B). Based on the LASSO results, Kendall correlation analysis (Figure 4C) revealed high collinearity between SLM and total protein content (\u0026tau; = 0.95). The redundant feature with relatively lower nonlinear association, \u0026ldquo;protein\u0026rdquo; was removed. The remaining six features, namely MMSE, gait speed, TUG, SLM, UBE2O, and ACTN3, were subsequently used to construct the machine learning model.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePerformance Comparison of Machine Learning\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eBased on the six selected important features, the study systematically evaluated the classification performance of eight widely used machine learning algorithms, including RF, GB, XGBoost, LR, RR, NB, KNN, and DT (Figure 4D). The results showed that ensemble learning methods exhibited the most prominent overall performance. The RF algorithm achieved the highest AUC (0.896, 95% CI: 0.792\u0026ndash;0.971), indicating excellent discriminative ability and stability. GB ranked second (AUC = 0.875, 95% CI: 0.775\u0026ndash;0.983), followed by XGBoost (AUC = 0.792, 95% CI: 0.707\u0026ndash;0.969), which demonstrated relatively narrow CIs and good generalizability. Among traditional models, LR (AUC = 0.844, 95% CI: 0.475\u0026ndash;0.948) and RR (AUC = 0.833, 95% CI: 0.464\u0026ndash;0.948) showed moderate performance, yet their wide CIs suggested sensitivity to data distribution. The probabilistic model NB (AUC = 0.760, 95% CI: 0.448\u0026ndash;0.929) and distance‑sensitive KNN (AUC = 0.745, 95% CI: 0.531\u0026ndash;0.955) performed similarly, with NB displaying greater variability. In contrast, DT yielded the poorest results (AUC = 0.625, 95% CI: 0.599\u0026ndash;0.964), reflecting its instability and limited applicability in this classification task (Table 4).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSHAP Analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe best‑performing RF algorithm was further interpreted using SHAP analysis. Figure 4E illustrates the mean absolute SHAP values for each feature, ranked in descending order of importance: gait speed, SLM, UBE2O, TUG, ACTN3, and MMSE. Figure 4F further reveals the contribution magnitude and direction of each feature to the classification results in the RF model. Gait speed exhibited the strongest predictive influence, with lower speed (blue dots) associated with higher risk of pre‑frailty (positive SHAP values). SLM and the protein markers UBE2O and ACTN3 showed more complex, non‑linear relationships, where both high and low levels could contribute to frailty risk. TUG time, while less influential, was positively associated with pre‑frailty risk, whereas MMSE score only contributed substantially under conditions of pronounced cognitive decline.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTable 4\u003c/strong\u003e Comparison of predictive performance among the eight model\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"609\" class=\"fr-table-selection-hover\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 84px;\"\u003e\n \u003cp\u003eAlgorithm\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 174px;\"\u003e\n \u003cp\u003eAUC(95%CI)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 92px;\"\u003e\n \u003cp\u003eAccuracy\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003ePrecision\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003eRecall\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 93px;\"\u003e\n \u003cp\u003eF1 Score\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 84px;\"\u003e\n \u003cp\u003eRF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 174px;\"\u003e\n \u003cp\u003e0.896(0.792 \u0026ndash; 0.971)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 92px;\"\u003e\n \u003cp\u003e0.750\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.733\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.917\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 93px;\"\u003e\n \u003cp\u003e0.815\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 84px;\"\u003e\n \u003cp\u003eGB\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 174px;\"\u003e\n \u003cp\u003e0.875(0.775 \u0026ndash; 0.983)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 92px;\"\u003e\n \u003cp\u003e0.700\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.714\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.833\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 93px;\"\u003e\n \u003cp\u003e0.769\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 84px;\"\u003e\n \u003cp\u003eXGBoost\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 174px;\"\u003e\n \u003cp\u003e0.792(0.707 \u0026ndash; 0.969)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 92px;\"\u003e\n \u003cp\u003e0.600\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.625\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.833\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 93px;\"\u003e\n \u003cp\u003e0.714\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 84px;\"\u003e\n \u003cp\u003eLR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 174px;\"\u003e\n \u003cp\u003e0.844(0.475 \u0026ndash; 0.948)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 92px;\"\u003e\n \u003cp\u003e0.707\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.694\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.834\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 93px;\"\u003e\n \u003cp\u003e0.732\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 84px;\"\u003e\n \u003cp\u003eRR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 174px;\"\u003e\n \u003cp\u003e0.833(0.464 \u0026ndash; 0.948)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 92px;\"\u003e\n \u003cp\u003e0.658\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.641\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.825\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 93px;\"\u003e\n \u003cp\u003e0.723\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 84px;\"\u003e\n \u003cp\u003eNB\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 174px;\"\u003e\n \u003cp\u003e0.760(0.448 \u0026ndash; 0.929)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 92px;\"\u003e\n \u003cp\u003e0.713\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.696\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.880\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 93px;\"\u003e\n \u003cp\u003e0.778\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 84px;\"\u003e\n \u003cp\u003eKNN\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 174px;\"\u003e\n \u003cp\u003e0.745(0.531 \u0026ndash; 0.955)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 92px;\"\u003e\n \u003cp\u003e0.650\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.667\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.833\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 93px;\"\u003e\n \u003cp\u003e0.741\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 84px;\"\u003e\n \u003cp\u003eDT\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 174px;\"\u003e\n \u003cp\u003e0.625(0.599 \u0026ndash; 0.964)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 92px;\"\u003e\n \u003cp\u003e0.650\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.692\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 83px;\"\u003e\n \u003cp\u003e0.750\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 93px;\"\u003e\n \u003cp\u003e0.720\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e"},{"header":"Discussion","content":"\u003cp\u003eTo our knowledge, this is the first attempt to integrate high‑throughput proteomic data with clinical phenotypes to systematically delineate the molecular trajectory of proteins during the transition from robustness to frailty, with prefrailty examined as a distinct status. The results revealed that among 161 significantly differentially expressed proteins detected in older adults across different health statuses, 15 core proteins associated with frailty were validated by targeted proteomics, including CP, F8, HPX, VWF, RPLP0, ATP5F1B, LPL, F5, MYH9, HSPG2, SVEP1, BPIFB1, UBE2O, ACTN3, and TREML1. Through machine learning combined with SHAP‑based interpretability analysis, UBE2O and ACTN3, together with clinical features including gait speed, SLM, TUG time, and MMSE score, were further identified as the most promising indicators for early frailty detection. Our findings provide new insights into the mechanisms underlying frailty development and offer a reference for precise clinical identification.\u003c/p\u003e\n\u003cp\u003eIntegrating data from two independent cohorts, the results demonstrated that different frailty statuses were accompanied by widespread proteomic alterations. A total of 15 proteins involved in intercellular signaling (SVEP1, HPX, RPLP0), protein metabolism (ATP5F1B, CP, LPL), immune regulation (VWF, F5, F8, BPIFB1, TREML1), maintenance of skeletal muscle structure and function (MYH9, ACTN3, HSPG2), and proteolysis (UBE2O) maintained statistically significant differences. This trend of proteomic changes detected during the transition from robustness to frailty suggests that the onset of frailty typically emerges gradually under the influence of multisystem dysregulation.\u003c/p\u003e\n\u003cp\u003eThe study further combined LASSO regression with Kendall correlation analysis for feature selection. Using eight machine learning algorithms with complementary advantages, namely RF, GB, XGBoost, LR, RR, NB, KNN, and DT, coupled with SHAP‑based interpretability analysis, the classification mechanism of the optimal model was systematically elucidated. Ultimately, an RF‑based feature set comprising gait speed, SLM, UBE2O, TUG, ACTN3, and MMSE score was established. These findings provide a multidimensional perspective on the characteristic patterns of early frailty and offer an interpretable machine learning decision framework to facilitate the clinical translation of protein biomarkers.\u003c/p\u003e\n\u003cp\u003eClinical characteristics have long been strong discriminators of frailty and represent the most intuitive manifestations of frailty in older adults and populations with various chronic diseases \u003csup\u003e37\u003c/sup\u003e。Our findings align with related research on frailty risk factor prediction, where functional indicators (e.g., gait speed, grip strength, balance, mobility), socio-psychological factors (e.g., cognition, loneliness, economic status), and physiological parameters (e.g., BMI, waist circumference, muscle mass, visceral fat area) are the most frequently included risk factors in frailty prediction models and dominate clinical frailty assessment\u0026nbsp;\u003csup\u003e37,38\u003c/sup\u003e. In a large Chinese cohort study with an average follow-up of five years, a prediction model based on nine risk factors was developed and validated. This model incorporated cognitive function, lifestyle behaviors, and body composition as predictive variables, providing significant insights for the early screening and risk stratification of frailty in the elderly\u003csup\u003e39\u003c/sup\u003e. Parameters related to functional status, psychosocial factors, and basic physiological indicators have become core variables in frailty discrimination models due to their operability and well‑defined mechanisms. However, with the advancement of precision medicine, integrating dynamic molecular marker data with clinical observational indicators has emerged as a new direction for personalized early risk assessment of frailty.\u003c/p\u003e\n\u003cp\u003eIn this study, ACTN3 and UBE2O, which were highly expressed in serum, were identified as two key protein molecules for identifying early frailty through machine learning algorithms. ACTN3, which was upregulated in the peripheral circulation of prefrail older adults, is a key structural protein of type II muscle fibers and is predominantly expressed within skeletal muscle cells; the R577X polymorphism in its associated gene is known to influence fast-twitch muscle fiber function, regulating muscle strength, power, and athletic potential\u003csup\u003e40\u003c/sup\u003e. Numerous studies on muscle degenerative diseases report that elevated blood levels of ACTN3 may result from stress-induced activation of the X allele in damaged skeletal muscle cells, leading to its release into the peripheral circulation via extracellular vesicle transport pathways\u003csup\u003e41\u003c/sup\u003e. Furthermore, the compensatory increase in ACTN3 can disrupt the balance between muscle protein synthesis and degradation, negatively regulating muscle mass. Particularly in females, aberrant ACTN3 expression may further impair muscle growth and function by inhibiting downstream effectors of the mTOR signaling pathway, a key regulator of muscle hypertrophy\u003csup\u003e42\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eWhile ACTN3's effect on protein synthesis and degradation is indirect, UBE2O, functioning as a ubiquitin-conjugating enzyme, can directly influence normal skeletal muscle development by ubiquitinating and degrading target proteins (e.g., the AMPKα2 subunit), thereby blunting the mTOR signaling pathway\u0026nbsp;\u003csup\u003e43\u003c/sup\u003e. Although current literature typically discusses the functions of ACTN3 and UBE2O independently, our initial PPI analysis captured an indirect interaction between them. Both proteins may converge on the AMPK/mTOR signaling pathway, where the ubiquitin-proteasome system selectively degrades proteins, disrupting metabolic homeostasis between protein synthesis and breakdown. Studies in animal models of obesity or metabolic syndrome have precisely defined that abnormal increases in UBE2O accelerate protein ubiquitination and hydrolysis, leading to the disruption of skeletal muscle protein metabolic homeostasis \u003csup\u003e44\u003c/sup\u003e. In clinical research, the elevated expression of various ubiquitination-related proteins, including UBE2O and UBE2T, has been implicated in the pathogenesis of aging and cancer and is closely associated with poor prognosis \u003csup\u003e45\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eACTN3 and UBE2O elucidate the complex pathological regulatory network of prefrailty from two perspectives: skeletal muscle structural development and protein ubiquitin-mediated degradation. Their interaction may synergistically exacerbate the onset and progression of frailty. This negative impact could represent a potential pathological basis for early frailty and an important risk factor for the prefrail state. However, it is crucial to acknowledge the potential influence of sample size limitations. First, although FDR correction was applied to mitigate false positives, only 50% (15/30) of the candidate proteins were replicated in the validation cohort, suggesting possible false discoveries. Second, while power analysis indicated 90% statistical power for detecting medium-effect-size differences with the current sample size, proteins with important biological significance might still have been missed.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eLimitations\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study has several limitations. The cross‑sectional design precludes the observation of dynamic progression from robustness to frailty, and the causal relationship between proteomic alterations and early frailty remains unclear. Future longitudinal studies with extended follow‑up or multi‑time‑point sampling may help elucidate the temporal sequence linking protein expression and functional decline. The modest sample size may limit the accuracy and generalizability of the findings; although key results reached statistical significance, subtle changes in secondary indicators may have been underpowered. Furthermore, the single‑center cohort may not fully represent the heterogeneity of older adults across different geographic and cultural contexts. Future investigations could expand sample size, adopt a multicenter design, and conduct stratified analyses accounting for varying levels of frailty, comorbidities, and medication use. Computational constraints also restricted the breadth of cohort coverage, potentially introducing data bias, and limited the depth of algorithmic comparison. Additional external validation is needed to further assess the generalizability of the machine‑learning models.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eThis study revealed widespread proteomic alterations across different frailty statuses and identified 15 proteins potentially associated with prefrailty, which were involved in key pathways including cell signaling, protein metabolism, immune regulation, skeletal muscle function maintenance, and proteolysis. Based on the random forest algorithm, a preliminary combination of clinical indicators and protein markers was established for frailty identification. Among these, ACTN3 and UBE2O demonstrated discriminatory potential as key molecules in early frailty, providing a scientific basis for the clinical translation of omics findings and the development of multidimensional tools for frailty assessment. Future studies could leverage the predictive performance of these markers to expand frailty phenotyping and facilitate their translation into clinical applications.\u0026nbsp;\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and informed consent\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Ethics Committee of the Chinese PLA General Hospital (Approval No. S2019‑140‑03). All participants provided written informed consent for participation and data collection.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSources of funding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was supported by the National Key Research and Development Program of China (Grant No. 2018YFC2002004).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eYY contributed to study design, data collection, data analysis, manuscript drafting, and manuscript editing. JWL and ZZ contributed to study conceptualization, data collection, and analysis. SHC contributed to data visualization. SXX contributed to data curation. MYC, YYZ, FW, and NHZ contributed to study conceptualization. GBW contributed to data analysis, visualization, and conceptualization. NP contributed to study design, conceptualization, supervision, manuscript review, and funding acquisition.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConflicts of interest disclosure\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eOrkaby AR, Schwartz AW, Callahan KE, Frailty. Ann Intern Med. 2026;179(2):ITC17\u0026ndash;32. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.7326/ANNALS-25-04412\u003c/span\u003e\u003cspan address=\"10.7326/ANNALS-25-04412\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCai S, Li J, Fang Y, et al. Frailty and pre-frailty prevalence in community-dwelling elderly with multimorbidity: A systematic review and meta-analysis. Arch Gerontol Geriatr. 2025;132:105782. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.archger.2025.105782\u003c/span\u003e\u003cspan address=\"10.1016/j.archger.2025.105782\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLapi F, Marconi E, Aprile PL, et al. Frailty, comorbidity, and multimorbidity and their relation with medications adherence in primary care older adults. Eur Geriatr Med. 2024. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s41999-024-01098-4\u003c/span\u003e\u003cspan address=\"10.1007/s41999-024-01098-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSato R, Vatic M, Peixoto da Fonseca GW, Anker SD, von Haehling S. Biological basis and treatment of frailty and sarcopenia. Cardiovasc Res. 2024;120(9):982\u0026ndash;98. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/cvr/cvae073\u003c/span\u003e\u003cspan address=\"10.1093/cvr/cvae073\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSerra-Prat M, Fortuny Borsot A, Burdoy E, Lavado Cuevas \u0026Agrave;, Mu\u0026ntilde;oz Ortiz L, Cabr\u0026eacute; M. Frailty reversal and its main determinants: a population-based observational and longitudinal study. Fam Med Community Health. 2025;13(2). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1136/fmch-2024-003250\u003c/span\u003e\u003cspan address=\"10.1136/fmch-2024-003250\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWilkinson DJ, Piasecki M, Atherton PJ. The age-related loss of skeletal muscle mass and function: Measurement and physiology of muscle fibre atrophy and muscle fibre loss in humans. Ageing Res Rev. 2018;47:123\u0026ndash;32. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.arr.2018.07.005\u003c/span\u003e\u003cspan address=\"10.1016/j.arr.2018.07.005\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHodgins ML, Maxwell SP, Howlett SE, Rockwood K. Blood biomarkers of frailty and cognition: A scoping review. Neurobiol Aging. 2026;161:14\u0026ndash;30. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.neurobiolaging.2026.01.003\u003c/span\u003e\u003cspan address=\"10.1016/j.neurobiolaging.2026.01.003\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSciancalepore F, Salemme S, Valletta M, et al. Frailty and the brain: A narrative review of functional and pathological correlates. Ageing Res Rev. 2025;113:102945. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.arr.2025.102945\u003c/span\u003e\u003cspan address=\"10.1016/j.arr.2025.102945\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu G, Jiang S, Xie W, et al. Biomarkers for sarcopenia, muscle mass, muscle strength, and physical performance: an umbrella review. J Transl Med Jun. 2025;12(1):650. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s12967-025-06575-3\u003c/span\u003e\u003cspan address=\"10.1186/s12967-025-06575-3\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMitchell A, Malmgren L, Bartosch P, McGuigan FE, Akesson KE. Pro-Inflammatory Proteins Associated with Frailty and Its Progression-A Longitudinal Study in Community-Dwelling Women. J Bone Min Res. 2023;38(8):1076\u0026ndash;91. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/jbmr.4861\u003c/span\u003e\u003cspan address=\"10.1002/jbmr.4861\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSathyan S, Ayers E, Gao T, Milman S, Barzilai N, Verghese J. Plasma proteomic profile of frailty. Aging Cell Sep. 2020;19(9):e13193. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/acel.13193\u003c/span\u003e\u003cspan address=\"10.1111/acel.13193\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eUbaida-Mohien C, Lyashkov A, Gonzalez-Freire M, et al. Discovery proteomics in aging human skeletal muscle finds change in spliceosome, immunity, proteostasis and mitochondria. Elife. 2019;8. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.7554/eLife.49874\u003c/span\u003e\u003cspan address=\"10.7554/eLife.49874\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePedro KM, Alvi MA, Hejrati N, Quddusi AI, Singh A, Fehlings MG. Machine learning-based cluster analysis identifies four unique phenotypes of patients with degenerative cervical myelopathy with distinct clinical profiles and long-term functional and neurological outcomes. EBioMedicine. 2024;106:105226. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.ebiom.2024.105226\u003c/span\u003e\u003cspan address=\"10.1016/j.ebiom.2024.105226\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNiu L, Thiele M, Geyer PE, et al. Noninvasive proteomic biomarkers for alcohol-related liver disease. Nat Med. 2022;28(6):1277\u0026ndash;87. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41591-022-01850-y\u003c/span\u003e\u003cspan address=\"10.1038/s41591-022-01850-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFried LP, Tangen CM, Walston J, et al. Frailty in older adults: evidence for a phenotype. Journals Gerontol Ser Biol Sci Med Sci. 2001;56(3):M146\u0026ndash;56.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJia X, Wang Z, Huang F, et al. A comparison of the Mini-Mental State Examination (MMSE) with the Montreal Cognitive Assessment (MoCA) for mild cognitive impairment screening in Chinese middle-aged and older population: a cross-sectional study. BMC Psychiatry. 2021;21(1):485. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s12888-021-03495-6\u003c/span\u003e\u003cspan address=\"10.1186/s12888-021-03495-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChe Y, Xin H, Gu Y, Ma X, Xiang Z, He C. Associated factors of frailty among community-dwelling older adults with multimorbidity from a health ecological perspective: a cross-sectional study. BMC Geriatr Mar. 2025;14(1):172. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s12877-025-05777-0\u003c/span\u003e\u003cspan address=\"10.1186/s12877-025-05777-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAnbarasan D, Merchant RA. Association between chronic pain severity, falls, frailty and perceived health in older adults at risk of falls. Eur J Med Res Oct. 2025;31(1):1047. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s40001-025-03304-w\u003c/span\u003e\u003cspan address=\"10.1186/s40001-025-03304-w\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIshiwari M, Kono Y, Togashi Y, et al. Mini nutritional assessment short-form as a risk factor for mortality in patients with respiratory disease undergoing urgent hospitalization. BMC Geriatr Feb. 2025;19(1):113. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s12877-025-05767-2\u003c/span\u003e\u003cspan address=\"10.1186/s12877-025-05767-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePark S-H, Kwak M-J. Performance of the Geriatric Depression Scale-15 with Older Adults Aged over 65 Years: An Updated Review 2000\u0026ndash;2019. Clin Gerontol. 2020;44(2):83\u0026ndash;96. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1080/07317115.2020.1839992\u003c/span\u003e\u003cspan address=\"10.1080/07317115.2020.1839992\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSainsbury A, Seebass G, Bansal A, Young JB. Reliability of the Barthel Index when used with older people. Age Ageing May. 2005;34(3):228\u0026ndash;32. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/ageing/afi063\u003c/span\u003e\u003cspan address=\"10.1093/ageing/afi063\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHouston DK, Marsh AP, Neiberg RH, et al. Vitamin D Supplementation and Muscle Power, Strength and Physical Performance in Older Adults: A Randomized Controlled Trial. Am J Clin Nutr Jun. 2023;117(6):1086\u0026ndash;95. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.ajcnut.2023.04.021\u003c/span\u003e\u003cspan address=\"10.1016/j.ajcnut.2023.04.021\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHorn DB, Linetzky B, Davies MJ, et al. Cardiometabolic Parameter Change by Weight Regain on Tirzepatide Withdrawal in Adults With Obesity: A Post Hoc Analysis of the SURMOUNT-4 Trial. JAMA Intern Med. 2026;186(2):157\u0026ndash;67. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1001/jamainternmed.2025.6112\u003c/span\u003e\u003cspan address=\"10.1001/jamainternmed.2025.6112\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi K, Teo GC, Yang KL, Yu F, Nesvizhskii AI. diaTracer enables spectrum-centric analysis of diaPASEF proteomics data. Nat Commun. 2025;16(1):95. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41467-024-55448-8\u003c/span\u003e\u003cspan address=\"10.1038/s41467-024-55448-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChambers AG, Sweet SMM, Chain D, Kim YJ. Global Extraction from Parallel Reaction Monitoring to Quantify Background Peptides for Improved Normalization and Quality Control in Targeted Proteomics. Anal Chem. 2021;93(40):13434\u0026ndash;40. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1021/acs.analchem.1c01375\u003c/span\u003e\u003cspan address=\"10.1021/acs.analchem.1c01375\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDeng M, Sun S, Zhao R, et al. The pyroptosis-related gene signature predicts prognosis and indicates immune activity in hepatocellular carcinoma. Mol Med. 2022;28(1):16. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s10020-022-00445-0\u003c/span\u003e\u003cspan address=\"10.1186/s10020-022-00445-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTakefuji Y. Reevaluating feature importance in machine learning: concerns regarding SHAP interpretations in the context of the EU artificial intelligence act. Water Res. 2025;280:123514. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.watres.2025.123514\u003c/span\u003e\u003cspan address=\"10.1016/j.watres.2025.123514\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOwens CD, Pinto CB, Mukli P, et al. Neurovascular coupling, functional connectivity, and cerebrovascular endothelial extracellular vesicles as biomarkers of mild cognitive impairment. Alzheimer's Dement J Alzheimer's Assoc. 2024;20(8):5590\u0026ndash;606. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/alz.14072\u003c/span\u003e\u003cspan address=\"10.1002/alz.14072\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYi Y, Tae M, Shin S, Choi S-I. Predicting acute kidney injury in trauma using an extreme gradient boosting model. Clin Kidney J. 2025;18(4):sfaf002. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/ckj/sfaf002\u003c/span\u003e\u003cspan address=\"10.1093/ckj/sfaf002\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiang D, Wang L, Zhong P, et al. Perspective: Global Burden of Iodine Deficiency: Insights and Projections to 2050 Using XGBoost and SHAP. Adv Nutr. 2025;16(3):100384. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.advnut.2025.100384\u003c/span\u003e\u003cspan address=\"10.1016/j.advnut.2025.100384\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDey D, Haque MS, Islam MM, et al. The proper application of logistic regression model in complex survey data: a systematic review. BMC Med Res Methodol. 2025;25(1):15. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s12874-024-02454-5\u003c/span\u003e\u003cspan address=\"10.1186/s12874-024-02454-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAkhtar N, Alharthi MF. Enhancing accuracy in modelling highly multicollinear data using alternative shrinkage parameters for ridge regression methods. Sci Rep. 2025;15(1):10774. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41598-025-94857-7\u003c/span\u003e\u003cspan address=\"10.1038/s41598-025-94857-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSugahara S, Ueno M. Exact Learning Augmented Naive Bayes Classifier. Entropy (Basel). 2021;23(12). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/e23121703\u003c/span\u003e\u003cspan address=\"10.3390/e23121703\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang J, Geng X. Large Margin Weighted k-Nearest Neighbors Label Distribution Learning for Classification. IEEE Trans Neural Netw Learn Syst. 2024;35(11):16720\u0026ndash;32. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1109/TNNLS.2023.3297261\u003c/span\u003e\u003cspan address=\"10.1109/TNNLS.2023.3297261\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHou R, Wang X, Zhang W, et al. Decision Tree Extraction for Clinical Decision Support System With If-Else Pseudocode and PlanSelect Strategy. IEEE J Biomed Health Inf. 2025;PPdoi. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1109/JBHI.2025.3529682\u003c/span\u003e\u003cspan address=\"10.1109/JBHI.2025.3529682\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCao S, Hu Y. Creating machine learning models that interpretably link systemic inflammatory index, sex steroid hormones, and dietary antioxidants to identify gout using the SHAP (SHapley Additive exPlanations) method. Front Immunol. 2024;15:1367340. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fimmu.2024.1367340\u003c/span\u003e\u003cspan address=\"10.3389/fimmu.2024.1367340\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePerdiguero GG, Spina JC, Mart\u0026iacute;nez J, et al. Enhancing ACLF prediction by integrating sarcopenia assessment and frailty in liver transplant candidates on the waiting list. JHEP Rep Mar. 2024;6(3):100985. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.jhepr.2023.100985\u003c/span\u003e\u003cspan address=\"10.1016/j.jhepr.2023.100985\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShah SJ, Oreper S, Jeon SY, Boscardin WJ, Fang MC, Covinsky KE. Social Frailty Index: Development and validation of an index of social attributes predictive of mortality in older adults. Proc Natl Acad Sci U S A Feb. 2023;14(7):e2209414120. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1073/pnas.2209414120\u003c/span\u003e\u003cspan address=\"10.1073/pnas.2209414120\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDong Y, Wang Q, Zhang K, et al. Development and validation of a prediction model of frailty risk in community-dwelling older adults: From a national longitudinal survey. Public Health Jan. 2025;27:240:63\u0026ndash;70. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.puhe.2024.12.055\u003c/span\u003e\u003cspan address=\"10.1016/j.puhe.2024.12.055\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEl Ouali EM, Barthelemy B, Del Coso J, et al. A Systematic Review and Meta-analysis of the Association Between ACTN3 R577X Genotypes and Performance in Endurance Versus Power Athletes and Non-athletes. Sports Med Open Apr. 2024;12(1):37. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s40798-024-00711-x\u003c/span\u003e\u003cspan address=\"10.1186/s40798-024-00711-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGiza S, Mojica-Santiago JA, Parafati M, et al. Microphysiological system for studying contractile differences in young, active, and old, sedentary adult derived skeletal muscle cells. Aging Cell Jul. 2022;21(7):e13650. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/acel.13650\u003c/span\u003e\u003cspan address=\"10.1111/acel.13650\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSeto JT, Roeszler KN, Meehan LR, et al. ACTN3 genotype influences skeletal muscle mass regulation and response to dexamethasone. Sci Adv Jul. 2021;7(27). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1126/sciadv.abg0088\u003c/span\u003e\u003cspan address=\"10.1126/sciadv.abg0088\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuang H, Zhu W, Huang B, et al. Structural insights into the biochemical mechanism of the E2/E3 hybrid enzyme UBE2O. Structure. 2025;33(2). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.str.2024.12.002\u003c/span\u003e\u003cspan address=\"10.1016/j.str.2024.12.002\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVila IK, Park MK, Setijono SR, et al. A muscle-specific UBE2O/AMPKα2 axis promotes insulin resistance and metabolic syndrome in obesity. JCI Insight. 2019;4(13). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1172/jci.insight.128269\u003c/span\u003e\u003cspan address=\"10.1172/jci.insight.128269\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ede Carvalho LGA, Komoto TT, Moreno DA, et al. USP15-USP7 Axis and UBE2T Differential Expression May Predict Pathogenesis and Poor Prognosis in De Novo Myelodysplastic Neoplasm. Int J Mol Sci Jun. 2023;13(12). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/ijms241210058\u003c/span\u003e\u003cspan address=\"10.3390/ijms241210058\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\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":"bmc-geriatrics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"bgtc","sideBox":"Learn more about [BMC Geriatrics](http://bmcgeriatr.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/bgtc/default.aspx","title":"BMC Geriatrics","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Frailty, Pre-frailty, Older Adults, Proteomics, Machine Learning","lastPublishedDoi":"10.21203/rs.3.rs-9388693/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9388693/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cdiv id=\"ASec1\" class=\"AbstractSection\"\u003e \u003cdiv class=\"Heading\"\u003eBackground\u003c/div\u003e \u003cp\u003eFrailty is associated with anomalies in protein metabolism; however, the underlying serum proteomic signatures and their driving mechanisms remain unclear. This study aims to uncover key protein profiles during the transition from robust health to frailty and evaluate their potential for early identification through serum proteomic analysis.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"ASec2\" class=\"AbstractSection\"\u003e \u003cdiv class=\"Heading\"\u003eMethods\u003c/div\u003e \u003cp\u003eWell-matched older adults were enrolled and categorized into robust, prefrail, and frail groups. In a cross-sectional design, untargeted proteomic screening was first performed in a discovery cohort of 30 participants, followed by targeted validation of candidate biomarkers using parallel reaction monitoring (PRM) in an independent validation cohort of 99 individuals. Multidimensional clinical parameters and differentially expressed proteins were integrated within machine learning pipelines to refine the search for characteristic features of prefrailty.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"ASec3\" class=\"AbstractSection\"\u003e \u003cdiv class=\"Heading\"\u003eResults\u003c/div\u003e \u003cp\u003eIn the discovery phase, 166 proteins were found to be differentially expressed across frailty statuses, with 15 significantly frailty-associated proteins (e.g., CP, VWF, UBE2O, and ACTN3) subsequently confirmed by PRM validation. These proteins were functionally enriched in pathways related to inflammation/immunity, coagulation, muscle structure, and protein metabolism. A Random Forest model, further assembled from gait speed, skeletal muscle mass, UBE2O, Timed Up and Go test, ACTN3, and Mini-Mental State Examination score, exhibited the most robust performance for early frailty identification among multiple algorithms compared (AUC: 0.896, 95% CI: 0.792\u0026ndash;0.971).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"ASec4\" class=\"AbstractSection\"\u003e \u003cdiv class=\"Heading\"\u003eConclusions\u003c/div\u003e \u003cp\u003eThis study reveals a panel of serum protein biomarkers closely linked to frailty progression. Through machine learning algorithms integrating clinical indices with circulating UBE2O and ACTN3, we evaluated the discriminatory capacity of these features for frailty, shedding light on the heterogeneity and protein alterations among community-dwelling older adults.\u003c/p\u003e \u003c/div\u003e","manuscriptTitle":"A Preliminary Predictive Panel for Pre-frailty Based on Serum Proteomic Biomarkers: A Two-Phase Cross-Sectional Study","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-05-18 11:22:48","doi":"10.21203/rs.3.rs-9388693/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"reviewerAgreed","content":"215238742572830876738506807816577393713","date":"2026-05-11T13:51:06+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-05-07T09:16:34+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-05-07T09:15:32+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2026-04-27T16:06:28+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-04-24T10:29:24+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Geriatrics","date":"2026-04-24T09:17:32+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-geriatrics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"bgtc","sideBox":"Learn more about [BMC Geriatrics](http://bmcgeriatr.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/bgtc/default.aspx","title":"BMC Geriatrics","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"0d315e23-7378-4e5c-a19f-c7d50207e4a6","owner":[],"postedDate":"May 18th, 2026","published":true,"recentEditorialEvents":[{"type":"reviewerAgreed","content":"215238742572830876738506807816577393713","date":"2026-05-11T13:51:06+00:00","index":34,"fulltext":""},{"type":"reviewersInvited","content":"10","date":"2026-05-07T09:16:34+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-05-07T09:15:32+00:00","index":"","fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-05-18T11:22:48+00:00","versionOfRecord":[],"versionCreatedAt":"2026-05-18 11:22:48","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9388693","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9388693","identity":"rs-9388693","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}