Pre-operative serum microRNA profiles differ between metabolic phenotypes and predict metabolic response to sleeve gastrectomy in severe obesity

Abstract

Abstract People with severe obesity show marked heterogeneity in cardiometabolic risk, classified as metabolically healthy (MHO) or unhealthy obesity (MUO). In this exploratory pilot study, we performed genome-wide serum microRNA (miRNA) profiling using Agilent SurePrint G3 microarrays (8 × 60 K; 2,549 probes) before and three months after sleeve gastrectomy (SG) in eight people with severe obesity (MHO n = 4; MUO n = 4) classified per Japanese national criteria. Despite comparable body mass index and visceral fat area, people with MUO had significantly higher HOMA-IR, fasting glucose, fasting insulin, HbA1c, and AST at baseline (all p = 0.029). Pre-operative miR-5100 expression correlated with baseline HOMA-IR (ρ = 0.805, p = 0.016) and predicted post-operative HOMA-IR reduction (ρ=−0.878, p = 0.004). Pre-operative miR-5585-3p predicted post-operative CRP reduction (ρ=−0.854, p = 0.007). Set analysis identified ten miRNAs, including miR-22-3p, detected in all MUO but no MHO participants. Welch’s t-test identified 23 miRNAs significantly altered post-operatively in MUO versus 8 in MHO. miR-8069 was consistently down-regulated after SG across all participants (Wilcoxon p = 0.016). Consistency across complementary analytical methods supports reliability despite the small sample. These preliminary findings suggest that pre-operative serum miRNA profiles may differ between MHO and MUO phenotypes and may predict metabolic responsiveness to bariatric surgery, and warrant validation in larger independent cohorts.