Aims/hypothesis: MicroRNAs are being sought as biomarkers for the early identification of type 2 diabetes. This study aimed to synthesise the evidence from microRNA–type 2 diabetes association studies and microRNA-regulated type 2 diabetes pathway delineation studies that met stringent quality criteria to identify and validate microRNAs of both statistical and biological significance as type 2 diabetes biomarkers.

Methods: Eligible controlled studies on microRNA expression profiling of type 2 diabetes were retrieved from PubMed, ScienceDirect and Web of Science. MicroRNA-regulated type 2 diabetes pathway delineation studies were conducted by integrating and cross-verifying the data from miRTarBase, TransmiR, miRecords, TargetScanHuman, the Kyoto Encyclopedia of Genes and Genomes (KEGG) and the Retraction Watch database. Before meta-analysis, quality assessment was performed according to the corresponding reporting guidelines for evidence-based medicine. To select the most statistically significant microRNAs, we conducted extensive meta-analyses according to the latest methodology. Subgroup and sensitivity analyses were carried out to further examine the microRNA candidates for their tissue specificity and blood fraction specificity and the robustness of the evidence. Signalling pathway impact analysis of dysregulated microRNAs identified from meta-analyses was performed to select biologically significant microRNAs that were enriched in our newly built microRNA-regulated pathways.

Results: Of the 404 differentially expressed microRNAs identified in the 156 controlled profiling studies with a combined sample size of >15,000, only 60 were both consistently and significantly dysregulated in human type 2 diabetes. No microRNAs were both consistently and significantly dysregulated in multiple tissues according to subgroup analyses. In total, 58 microRNAs were found to be robust in sensitivity analyses. A total of 1966 pathway delineation studies were identified, including 3290 microRNA–target interactions, which were further combined with KEGG pathways, producing 225 microRNA-regulated pathways. Impact analysis found that 16 dysregulated microRNAs identified from extensive meta-analyses were statistically significantly enriched in the augmented KEGG type 2 diabetes pathway.

Conclusions/interpretation: Sixteen microRNAs met the criteria for biomarker selection. In terms of both significance and relevance, the order of priority for verification of these microRNAs is as follows: miR-29a-3p, miR-221-3p, miR-126-3p, miR-26a-5p, miR-503-5p, miR-100-5p, miR-101-3p, mIR-103a-3p, miR-122-5p, miR-199a-3p, miR-30b-5p, miR-130a-3p, miR-143-3p, miR-145-5p, miR-19a-3p and miR-311-3p.

Registration: PROSPERO registration number CRD42017081659.