Metabolomic characteristics and mechanisms of subgingival plaque in MASLD patients with periodontitis.

INTRODUCTION: Metabolic dysfunction-associated steatotic liver disease (MASLD) exhibits a significant comorbidity with periodontitis (PD), yet its molecular mechanisms remain unclear. This study aims to elucidate the metabolic characteristics of MASLD Patients with Periodontitis (MASLD-PD) in a comorbid state through metabolomic analysis, thereby exploring potential biological mechanisms.

OBJECTIVES: To elucidate metabolic characteristics of MASLD-PD patients via metabolomics, explore comorbidity mechanisms, and provide insights for early diagnosis and targeted intervention.

METHODS: Thirty subjects were recruited (15 each in the MASLD-PD group and PD group). Subgingival plaque samples were collected and subjected to non-targeted metabolomic analysis via ultra-high-performance liquid chromatography-tandem mass spectrometry (UHPLC-MS/MS). Differential metabolites were identified through multidimensional statistical analysis (PCA, PLS-DA, OPLS-DA), LASSO regression was further applied to screen core diagnostic features from the top 20 upregulated metabolites, with model performance evaluated using Kappa coefficient, F1 score, Precision, Recall, and Accuracy. followed by ROC curve analysis and KEGG pathway enrichment studies.

RESULTS: A total of 2126 significantly differentially expressed metabolites were identified. Compared with the PD group, 49 metabolites were significantly upregulated, and 2077 metabolites were significantly downregulated in the MASLD-PD group. Multidimensional analysis revealed significant separation of the metabolomic profiles between the two groups. ROC analysis was performed as an exploratory approach to evaluate the discriminatory capacity of the 20 significantly upregulated metabolites between the MASLD-PD and PD groups. The combined model yielded an AUC of 1.0000, providing preliminary evidence that these metabolites may collectively distinguish the two groups in this small cohort and generating hypotheses for further validation. LASSO regression identified 5 core metabolites with non-zero regression coefficients, and the core feature combination achieved perfect discriminatory efficacy (AUC = 1.0000) with high consistency (Kappa = 0.8000), overall accuracy (0.9000), and no missed diagnoses (Recall = 1.0000). Building upon this, we further explored the functions of these top 20 upregulated metabolites through KEGG pathway enrichment analysis. This revealed their specific enrichment in pathways related to lipid metabolism (e.g., Steroid hormone biosynthesis), Signal transduction (e.g., ErbB signaling pathway), and the immune system (e.g., Th17 cell differentiation).

CONCLUSION: This study systematically reveals the unique metabolic phenotype of patients with MASLD-PD, suggesting that immune-metabolic network dysregulation may be involved in the pathophysiological mechanism of this comorbidity. The 5 core metabolites identified via LASSO regression exhibit promising discriminatory potential. The identified differentially expressed metabolites and their enriched pathways generate hypotheses for understanding the comorbidity mechanism, laying a foundation for future studies on early diagnosis and targeted intervention after large-scale validation.