**Background:** Obesity is a chronic disease that can lead to metabolic syndrome (MetS), increasing cardiovascular risk. Some individuals with obesity remain metabolically healthy (OBO), while others develop metabolic complications (OBM). The molecular mechanisms differentiating these phenotypes are not fully understood. This study aimed to identify distinct molecular signatures and metabolic pathways between OBO and OBM using an integrated multi-omic approach.
**Methods:** This cross-sectional case-control study included 39 participants with obesity (BMI ≥ 35 kg/m²) recruited at the Qatar Metabolic Institute. Participants were classified as OBO (n=18) or OBM (n=21) using IDF metabolic syndrome criteria. Groups were age- and BMI-matched. Fasting blood samples were collected. Whole blood was used for miRNA profiling (754 miRNAs via TaqMan OpenArray), transcriptomic profiling (25,682 transcripts including protein-coding genes [PCGs] and non-coding transcripts via Clariom D Assay), and metabolomic profiling (704 metabolites via LC-MS/MS at Metabolon). Differentially expressed (DE) miRNAs, PCGs, and metabolites were identified using LIMMA (p≤0.05, no FDR correction due to small sample size). Metabolite set enrichment analysis (MSEA) was performed using MetaboAnalyst 5.0 with KEGG pathways. miRNA-PCG interactions were mapped using mirDIP (integrated score ≥0.25, ≥9 sources). Single-sample gene set enrichment (GSVA) was used to estimate pathway enrichment per sample, followed by unsupervised hierarchical clustering.
**Key Results:** Clinical characteristics: OBO and OBM groups had similar mean age (38.1 vs. 40.5 years, p=0.283) and BMI (40.0 vs. 39.6 kg/m², p=0.746). Significant differences were observed in HbA1c (5.5% vs. 7.02%, p=0.002), fasting triglycerides (1.39 vs. 2.65 mmol/L, p=0.001), HDL-cholesterol (1.5 vs. 1.0 mmol/L, p=0.008), fasting glucose (5.2 vs. 7.4 mmol/L, p=0.009), and insulin (19.0 vs. 27.6 miU/mL, p=0.053). MSEA identified 12 enriched metabolic pathways (p≤0.1). Of these, 8 pathways contained both DE metabolites and DE-PCGs acting as kinases: amino sugar and nucleotide sugar metabolism, alanine/aspartate/glutamate metabolism, arginine and proline metabolism, D-glutamine and D-glutamate metabolism, galactose metabolism, glyoxylate and dicarboxylate metabolism, lysine degradation, and starch and sucrose metabolism. These pathways involved 8 DE metabolites, 25 DE-PCGs, and 9 DE miRNAs (with 12 interaction edges). Most DE miRNAs were down-regulated in OBM, while their target DE-PCGs were primarily up-regulated, except PFKM (down-regulated). Pathways related to glutamine/glutamate metabolism and associated PCG GOT1L1 and metabolites L-Asparagine and L-Glutamine were down-regulated in OBM. GSVA-based unsupervised hierarchical clustering using the 8 pathways segregated OBO from OBM into two clusters (C1 and C2). Most pathways were positively enriched in the OBM cluster (C2) and negatively enriched in the OBO cluster (C1). Within the OBM cluster, two sub-clusters were distinguished based on blood glucose and insulin levels.
**Clinical Implications:** This integrated multi-omic analysis identified 8 metabolic pathways that can differentiate metabolically healthy from unhealthy obesity. The dysregulated elements include miRNAs (e.g., miR-636, miR-1303, miR-135), PCGs (e.g., GNE, NANS, GALT, GPI, PFKM, ARG1, ALDH2), and metabolites (e.g., L-Glutamine, L-Asparagine, sarcosine, D-glucose). These findings suggest that multi-omic profiling may provide a panel of biomarkers for early detection of metabolic deterioration in obesity and could guide personalized interventions. The study highlights the importance of integrating multiple omics layers to capture the complexity of obesity-related metabolic dysfunction. Limitations include small sample size, lack of FDR correction, and use of whole blood transcriptomics which may not fully reflect tissue-specific changes. Future prospective studies are needed to validate these biomarkers and explore their causal roles.