Insulin Sensitivity Remains Stable in Serial Studies of Healthy Patients.

The standard for measuring whole-body insulin sensitivity is via euglycemic-hyperinsulinemic clamp. In a given individual, it can vary significantly from hour to hour in response to interventions such as exercise, diet, drugs, and stressors. Over longer periods of time, fluctuations in body weight, physical activity, and progressive age, might alter insulin sensitivity. The variability of insulin sensitivity over time in persons at low risk for progression to diabetes is incompletely understood. We measured insulin sensitivities in 4 healthy volunteers with no family history of diabetes, 2M/2F, age 25±6y, BMI 22 ± 2 kg/m², FBS 90 ± 4 mg/dL, and A1c 5.0 ±0.2%. Each person was assessed 2-3 times over the course of 3.5 years via a 4-hour euglycemic-hyperinsulinemic clamp (2.0 mU/kg*min). Insulin levels were at steady-state during the fourth hour of the clamps as measured by RIA (173.6 ± 9.5 µU/mL). At this insulin level, hepatic glucose production is fully suppressed, so glucose utilization (M) was estimated by the glucose infusion rate, which initially was 11.6 ± 0.6 mg/kg*min. M did not change significantly over time, being 11.8 ± 1.2 mg/kg*min at the final evaluation. Over the study period, several metabolic indices varied substantially from the initial measurements in any given subject, with the following mean coefficients of variation (CV): BMI 4.9 ± 1.2%, FBS 4.4 ± 1.4%, fasting plasma insulin 45.0 ± 10.8%, HOMA-IR 46.9 ± 9.8%, total cholesterol 8.2 ± 3.1%, triglycerides 45.5 ± 6.0%, HDL 16.0 ± 5.8%, and LDL 16.4 ± 2.1%. We demonstrated a mean CV for M of 8.1 ± 0.2%. Since steady-state insulin levels achieved during the clamps also varied, we related whole-body insulin sensitivity to exogenous insulin via M/I* 100, whose cohort average at baseline was 14.6 ± 2.2 and at the end of the study 17.5 ± 1.9 (p=ns). The mean CV of the M/I* 100 for each of the subjects, over all of the clamps, was 12.2 ± 6.1%. By comparison, the initial and final HOMA-IR were, respectively, 1.1 ± 0.3 and 1.8 ± 0.4 (p=ns), and the mean CV over all measures was 55.1±13.3%. These results suggest a high degree of reproducibility of clamp-measured insulin sensitivity and that in healthy individuals at low risk for diabetes, insulin sensitivity does not change substantially or disproportionately to variations in other metabolic parameters over the course of several years. The considerably larger CV for HOMA-IR supports that it is not a robust index of insulin sensitivity when assessing individuals or smaller cohorts. [ABSTRACT FROM AUTHOR]