Micro RNA as a potential blood-based epigenetic biomarker for Alzheimer's disease.

As the prevalence of Alzheimer's disease (AD) increases, the search for a definitive, easy to access diagnostic biomarker has become increasingly important. Micro RNA (miRNA), involved in the epigenetic regulation of protein synthesis, is a biological mark which varies in association with a number of disease states, possibly including AD. Here we comprehensively review methods and findings from 26 studies comparing the measurement of miRNA in blood between AD cases and controls. Thirteen of these studies used receiver operator characteristic (ROC) analysis to determine the diagnostic accuracy of identified miRNA to predict AD, and three studies did this with a machine learning approach. Of 8098 individually measured miRNAs, 23 that were differentially expressed between AD cases and controls were found to be significant in two or more studies. Only six of these were consistent in their direction of expression between studies (miR-107, miR-125b, miR-146a, miR-181c, miR-29b, and miR-342), and they were all shown to be down regulated in individuals with AD compared to controls. Of these directionally concordant miRNAs, the strongest evidence was for miR-107 which has also been shown in previous studies to be involved in the dysregulation of proteins involved in aspects of AD pathology, as well as being consistently downregulated in studies of AD brains. We conclude that imperative to the discovery of reliable and replicable miRNA biomarkers of AD, standardised methods of measurements, appropriate statistical analysis, utilization of large datasets with machine learning approaches, and comprehensive reporting of findings is urgently needed.