Salivary extracellular RNA biomarkers for insulin resistance detection in hispanics
Yong Zhang1, Jie Sun2, Feng Li3
1First Clinical Division, Peking University School and Hospital of Stomatology, Beijing, China; School of Dentistry, University of California Los Angeles, Los Angeles, CA, USA.
Aims:
Insulin resistance (IR) detection is challenging and no test is currently used in clinical practice. We developed salivary biomarkers that could be used for IR detection.
Methods:
We collected saliva from 186 healthy and 276 pre-diabetic participants, divided them into high and low IR groups based on a HOMA cutoff of 2.5. We profiled extracellular transcriptome by microarray in saliva supernatant from 23 high IR and 15 low IR participants, and pre-validated the top ten extracellular mRNA (exRNA) markers in a new cohort of 40 high and 40 low IR participants. A prediction panel was then built and validated in an independent cohort of 149 high and 195 low IR participants.
Results:
Transcriptomic analyses identified 42 exRNA candidates differentially present in saliva of high and low IR participants. From the top ten candidates, six were individually validated (PRKCB, S100A12, IL1R2, CAMP, VPS4B, CAP1) (p<0.01) and yielded AUC values ranging from 0.66 to 0.76. Body mass index (BMI) was significant higher in high compared to low IR group with AUC of 0.66, and showed no correlation with any of candidate biomarkers. The combination of four exRNA markers (IL1R2, VPS4B, CAP1, LUZP6) with BMI achieved excellent results in the prediction panel building dataset (AUC=0.79, sensitivity=79%, specificity=64%). The prediction model was validated in an independent cohort (AUC=0.82, sensitivity=63%, specificity=92%).
Conclusions:
A panel of four salivary exRNA biomarkers (IL1R2, VPS4B, CAP1, LUZP6) and BMI was validated that can distinguish high and low IR participants, overall and in subgroups of healthy and pre-diabetic participants.
