Liquid–liquid phase separation (LLPS) compartmentalizes and concentrates biomacromolecules into liquid-like condensates, which underlies membraneless organelles (MLOs) formation in eukaryotic cells. With increasing evidence of the LLPS concept and methods, this phenomenon as a novel principle accounts for explaining the precise spatial and temporal regulation of cellular functions. Moreover, the phenomenon that LLPS tends to concentrate proteins is often accompanied by several abnormal signals for human diseases. It is reported that multiple metabolic diseases are strongly associated with the deposition of insoluble proteinaceous aggregating termed amyloids. At present, recent studies have observed the roles of LLPS in several metabolic diseases, including type 2 diabetes mellitus (T2DM), Alzheimer’s disease (AD), and metabolic bone diseases (MBDs). This review aims to expound on the current concept and methods of LLPS and summarize its vital roles in T2DM, AD, and MBDs, uncover novel mechanisms of these metabolic diseases, and thus provide powerful potential therapeutic strategies and targets for ameliorating these metabolic diseases.

Background: Objective body-fluid biomarkers may help characterize exercise-induced physical fatigue; however, heterogeneity in exercise protocols, biological matrices, and analytical platforms complicates biomarker interpretation. Purpose: This study aimed to identify candidate acute exercise-responsive metabolic biomarkers associated with physical fatigue assessment by integrating conventional biomarker evidence, exercise-related metabolomics, and exploratory targeted Liquid chromatography–mass spectrometry (LC-MS) analysis. Methods: We conducted three complementary analyses: (1) a systematic review and random-effects meta-analysis of quantitative biomarkers measured in body-fluid matrices, including blood, plasma, serum, saliva, sweat, urine, and interstitial fluid, at pre-exercise and post-exercise time points; (2) a descriptive fold-change-based synthesis of exercise-related metabolomics studies; and (3) an exploratory paired plasma–saliva liquid chromatography–mass spectrometry pilot analysis in seven apparently healthy young male volunteers. Eligible studies included apparently healthy human participants without reported acute or chronic diseases, major injuries, pregnancy, or clinically diagnosed pathological fatigue, undergoing endurance, resistance, high-intensity interval, or mixed exercise protocols. Biomarkers reported in ≥10 independent comparisons were pooled using log-transformed post-exercise/pre-exercise response ratios. Results: The systematic review and meta-analysis included 110 articles, 129 experiments, and 1826 participants. Lactate showed the strongest pooled response after exercise (response ratio = 4.43, 95% CI: 3.74–5.25, p < 0.00001; I² = 98%), followed by IL-6 (2.15, 95% CI: 1.50–3.07, p < 0.0001; I² = 87%), CK (1.70, 95% CI: 1.56–1.86, p < 0.00001; I² = 90%), and LDH (1.23, 95% CI: 1.11–1.35, p < 0.0001; I² = 93%). Metabolomics synthesis identified lactate, pyruvate, hypoxanthine and xanthine as frequently reported exercise-responsive metabolites related to glycolysis and purine degradation. Exploratory LC-MS analysis showed consistent directional changes in these metabolites in paired plasma and saliva samples. Conclusions: Lactate, pyruvate, hypoxanthine, and xanthine represent candidate metabolism-based acute exercise-responsive biomarkers associated with physical fatigue. However, their fatigue specificity remains to be established and requires validation using independent fatigue criteria, non-fatiguing exercise controls, matrix-specific analyses, and adequately powered diverse cohorts.