Blood-based microRNAs as biomarkers of insulin resistance in childhood obesity: A systematic review and meta-analysis

Childhood obesity is closely linked to insulin resistance, but affected children are metabolically heterogeneous. Blood-based microRNAs (miRNAs) have emerged as potential minimally invasive biomarkers of early metabolic dysfunction. A systematic search of PubMed, Embase, Web of Science, and Scopus was conducted from database inception to April 1, 2026. Eligible studies included children or adolescents with obesity that evaluated blood-based miRNAs, including serum, plasma, or peripheral blood mononuclear cell-derived miRNAs, in relation to IR. Narrative synthesis was the primary approach. Study quality was assessed using the Newcastle-Ottawa Scale, and exploratory quantitative synthesis was performed only for comparable miRNA-outcome pairs with extractable correlation data. Twenty studies met the inclusion criteria; 7 were rated as low risk of bias, 11 as moderate risk, and 2 as high risk. Most studies had cross-sectional or case-control designs, and substantial heterogeneity was present in study populations, biological sample source, insulin resistance definitions, miRNA methods, and outcome reporting. Across the included studies, miR-122 was the most consistent candidate, showing recurrent up-regulation in adverse metabolic phenotypes and positive associations with HOMA-IR (r = 0.53; 95%CI, 0.35-0.67) or related IR markers (r = 0.60; 95%CI, 0.47-0.70). miR-29a also showed relatively consistent positive associations (r = 0.67; 95%CI, 0.59-0.73), particularly in cohorts with more severe metabolic impairment. Other candidates, including miR-146a, miR-34a, miR-191-3p, miR-30d-5p, and miR-221-3p, showed relevant but less consistent findings. Blood-based miRNAs are associated with IR in childhood obesity, with miR-122 emerging as the strongest and most recurrent candidate and miR-29a as another promising marker. However, the evidence remains heterogeneous and insufficient for clinical application. Larger, standardized, and longitudinal studies are needed to validate the most recurrent candidates and define their role in early risk stratification.