Multivariate analysis of relationships between insulin sensitivity and blood rheology: is plasma viscosity a marker of insulin resistance?

Abstract

We previously reported in populations exhibiting all the spectrum of insulin sensitivity (SI) values correlations between SI and blood viscosity eta suggesting that high eta is an additional symptom of the insulin resistance syndrome. However, due to the elevation of insulinemia (I) which is usually associated with insulin resistance it remained to determine whether this relationship was explained by SI or I. We analyzed SI with the minimal model procedure in 108 nondiabetic subjects and analyzed correlations of SI with blood rheology (eta, RBC aggregation and rigidity). Across quartiles of SI (defined after log transformation since distribution of SI was not normal), hematocrit and red cell rigidity remained stable, while aggregability and plasma viscosity (etap) increased in the lowest quartile. SI was correlated to only two rheological parameters: etap (r= -0.280, p=0.005) and Myrenne index M1 (r= -0.219, p=0.044). Among SI, I, age and BMI multivariate analysis selected only BMI as a determinant of either whole blood viscosity (etawb: r= -0.301, p=0.004) and RBC disaggregation threshold (gammaD: r= -0.331, p=0.013), only I as determinant of M1 (r=0.254, p=0.03), and a combination of BMI (p=0.009) and SI (p=0.007) for etap. Although age and obesity are factors of hyperviscosity, the hemorheological disturbances found in insulin resistance are not fully statistically "explained" by those two factors. While hyperaggregability (measured with M1) is rather related to hyperinsulinism, etap is influenced by SI and should be further investigated as a simple marker for the follow up of insulin-resistant states.