Whole Blood NMR Relaxometry for the Detection of Insulin Resistance




Deodhar, Sneha
Patel, Vipulkumar
Mishra, Ina
Cistola, David


Journal Title

Journal ISSN

Volume Title



Time-domain nuclear magnetic resonance relaxometry (TD-NMR) is a practical method for measuring the physical and dynamical properties of complex, heterogeneous samples. In prior work, we showed that TD-NMR measurements of human plasma or serum report on an individual’s metabolic status, particularly with respect to insulin resistance. The relationship between plasma water transverse relaxation time (T2) and insulin resistance is mediated by subtle subclinical shifts in protein and lipoprotein levels in the circulation. The previous test required separation of blood cells in order to conduct T2 measurements on isolated plasma or serum. We hypothesized that this separation may not be required, as it is conceivable that similar metabolic relationships could be gleaned from T2 measurements on water in whole blood. To test this hypothesis, we conducted an observational, cross-sectional study of over 30 asymptomatic, non-diabetic human subjects who were recruited though an approved IRB protocol. Antecubital venous blood was drawn into lavender-top tubes containing EDTA as the anticoagulant. The blood samples, prior to NMR measurements, were allowed to spontaneously settle in the tube thereby creating two phases: a liquid supernatant (plasma) and blood cell pellet. The NMR relaxation constants T1 and T2 were determined using a Bruker mq20 Minispec NMR instrument operating at 20MHz. The data were collected using inversion recovery and modified Carr-Purcell-Meiboom-Gill pulse sequences, respectively. The NMR time-decay curves were transformed using an inverse-Laplace algorithm in order to extract T2 values. In addition, we measured over 100 diagnostic biomarkers on each subject and correlated the NMR measurements with established markers of metabolic function. The supernatant water T2 values from whole blood were compared to those obtained from fractionated plasma samples and to the other 100-plus biomarkers. The associations were quantified using parametric and non-paramertic correlations and the Student t-test. Statistically-significant bivariate correlations were observed between whole blood water T2 and and lipid biomarkers, which are associated with insulin resistance and metabloic disease.