Introduction: The metabolic abnormalities that precede type 2 diabetes progress slowly and in stages. Current evidence-based diabetes prevention programs target individuals in Stage 2 (impaired glucose tolerance or prediabetes). However, by that stage, 70% of pancreatic beta-cell insulin secretory capacity has been lost irreversibly. Thus, it is imperative to identify individuals in Stage I (early insulin resistance syndrome) in order to preserve pancreatic insulin secretion and prevent both diabetes and prediabetes. Early insulin resistance syndrome is characterized by compensatory hyperinsulinemia, dyslipidemia, sub-clinical inflammation and acid-base abnormalities. The nature of the association between these elements is unclear, and different subtypes of insulin resistance syndrome may exist. Hypothesis: We have developed two novel and unconventional biomarkers for characterizing insulin resistance in non-diabetic subjects. One approach is based on dynamic light scattering (DLS) of human serum. The second approach measures the T 2 relaxation time of water in human plasma using compact time-domain NMR relaxometry (TD-NMR). We hypothesize that these methods can detect subtypes, i.e. , insulin resistance with or without inflammation, hypercholesterolemia, or acid-base abnormalities. Methods: Seventy-two asymptomatic non-diabetic human subjects were recruited through an IRB-approved biomarker discovery protocol. Medical histories, anthropomorphic measurements and fasting blood samples were obtained, and over 1300 blood biomarkers were measured on each subject along with DLS and TD-NMR parameters. Bi-variate correlation analyses and multiple regression models were used to analyze continuous variables, and categorical variables were established for insulin resistance, inflammation, acid-base abnormalities and lipid abnormalities. Multiple means comparisons were performed using one-way ANOVA and Tukey-Kramer testing. Results: Plasma water T 2 from TD-NMR was strongly correlated with markers of early insulin resistance syndrome. Multiple regression analysis showed independent contributions from markers of hyperinsulinemia, hypercholesterolemia and inflammation. Multiple means comparisons showed significant differences for insulin resistance with and without inflammation. By contrast, DLS parameters were strongly correlated with insulin markers, but could not distinguish the inflammatory subtypes. Conclusions: Both TD-NMR and DLS are able to detect early insulin resistance in individuals who do not meet the criteria for prediabetes or metabolic syndrome. However, TD-NMR has the unique ability to distinguish inflammatory vs. non-inflammatory subtypes of insulin resistance. Subtyping and risk stratification are important for the design of personalized interventions to prevent diabetes, prediabetes and cardiovascular disease.
Read full abstract