Use of Frozen vs. Fresh Plasma to Assess Early Insulin Resistance Syndrome

Background: Insulin resistance is defined as the blunted response to insulin by tissues and can progress to prediabetes and type 2 diabetes. About 86 million US adults were identified with pre-diabetes in 2012. By the time prediabetes develops, approximately 70% of β-cell secretory function has been lost irreversibly. Thus, it is imperative to detect insulin resistance at an earlier stage in order to preserve pancreatic function and prevent progression to diabetes. Prior work revealed that water proton transverse relaxation time (T2) measured by NMR relaxometry using fresh human plasma samples provides a sensitive and specific biomarker for early insulin resistance syndrome. Purpose: The objective of this study was to compare T2 values of frozen plasma samples with those from fresh plasma to assess the feasibility of analyzing bio-banked samples from longitudinal population studies. Hypothesis: Freezing at -80oC will have little or no impact on measured T2 values and their correlation with insulin resistance markers. Methods: We recruited 45 asymptomatic, non-diabetic human volunteers though an IRB approved protocol. Blood samples were collected after an overnight fast and were processed and analyzed immediately, with the remaining samples stored at -80oC. In addition, over 200 blood biomarkers were measured on each fresh blood sample – many by outside laboratories including Quest Diagnostics, Inc. and Atherotech, Inc. After several months in the freezer, the NMR measurements were performed on the once-frozen, once-thawed samples. All NMR measurements were performed at 37oC using a Bruker mq20 Minispec instrument and a modified CPMG pulse scheme. The associations between frozen and fresh T2 values and metabolic biomarkers were quantified using the Pearson’s product moment and concordance correlation coefficients. Results: Plasma water T2 from frozen samples showed a strong, statistically significant correlation with fresh plasma water T2 values (Pearson r=0.85, Concordance correlation coefficient=0.74). However, the frozen plasma water T2 were 5% lower, on average, than fresh samples. Nevertheless, this difference did not impact the overall pattern of association between T2 and metabolic biomarkers of early insulin resistance syndrome. Conclusions: These findings establish the feasibility of using frozen bio-banked specimens for the validation of plasma water T2 as a metabolic biomarker and screening tool for diabetes risk assessment.