Abstract
AIM: Early-stage diagnosis of diabetes through non-invasive and diagnostic biofluid-like saliva has become a very popular approach to facilitate future preventive interventions and improve patient care. Meanwhile, the alteration of small non-coding RNA in human fluids has been suggested as a probable precedent for the early stages of diabetes. METHODS: In the present study, we checked the expression of miR-320a, 182-5p, 503, and 375 by using quantitative PCR in both stimulated and unstimulated saliva and blood samples of 40 adult patients with type-2 diabetes compared to 40 healthy individuals. In addition, we have sought to understand the possibility that miRNAs could provide new information about the status of type 2 diabetes in salivary samples beyond what can now be identified from blood samples and link their expression to the presence of clinically relevant risk factors. For this purpose, we have used a set of multivariate models. RESULTS: The results showed that three miRNAs were more highly expressed in patients with type 2 diabetes, while miR-320-a was down-regulated in those patients compared to healthy subjects. Furthermore, the data showed that miR-320a was the most reliable predictor for distinguishing diabetic patients from healthy subjects, with AUCs of 0.997, 0.97, and 0.99 (97.4% sensitivity and 100% specificity, p = 0.001) for serum, unstimulated, and stimulated saliva samples, respectively. CONCLUSIONS: Interestingly, the results of this study indicated that the amount of four miRNAs expressed in stimulated saliva was the same as in serum samples, which could conclude that specific miR-320a and 503 in stimulated saliva may introduce credible, non-invasive, and diagnostic biomarkers that can be used to monitor diabetic patients' status, while there is a need to design more research studies to confirm these findings.