We found a match
Your institution may have rights to this item. Sign in to continue.
- Title
Developement of a Sparse-Sample Method to Estimate Insulin Resistance in Korean Clinical Population.
- Authors
Chul Woo Ahn; Kyung Soo Park; Jong Suk Park; Min Ho Cho; Chul Sik Kim; Hai Jin Kim; Ji Sun Nam; Han Young Jung; Ji Eun Yoon; Young Duk Song; Bong Soo Cha; Eun Jig Lee; Sung Kil Lim; Kyung Rae Kim; Hyun Chul Lee
- Abstract
A measure of insulin resistance is routinely evaluated as the rate constant for plasma glucose disappearance. The purpose of this study was to develop a mixed effect model approach to estimate insulin resistance that can be applied to sparsely-sampled plasma glucose data. Data were taken from 1516 subjects who took an insulin tolerance test. Blood samples were collected at 3, 6, 9, 12, 15 min after intravenous injection of insulin with a dose of 0.1U/kg. Plasma glucose concentrations were analyzed by the glucose oxidase method. A one-compartment pharmacokinetic model consisting of the plasma glucose disappearance rate constant (K) and the baseline plasma glucose concentration (CO) was used incorporating clinical covariates to describe the time course of glucose, within the context of mixed effect model framework. Validity of the model was tested by internal validation. Based on the model so derived, a set of sparse sample times that best describes the original glucose kinetics was determined from the combinations of the 5 sample times. Estimated population means for K and C0 were 0.0362min[sup -1] and 4.8769mmol/L with interindividual variability of 41.71% and 4.83% in CV, respectively. Important covariates found were age, sex and clinical status for K, while C0 dependent only on clinical status. The weighted residuals (WRES) obtained from NONMEM lied approximately within -3 and 3 with a mean of about zero, indicating no bias for the model. For a sparse sample design, a set of samples at 3, 6, and 15 min allowed best prediction of plasma glucose concentration, with no noticeable deviation in the estimate of K. This work suggests the feasibility of applying a sparse-sample design approach to assessing insulin resistance. The classical approach such as the glucose clamp test may be necessary to further validate the accuracy of the model suggested.
- Subjects
SOUTH Korea; INSULIN resistance; MULTILEVEL models; BLOOD sugar; OXIDASES; PHARMACOKINETICS
- Publication
Diabetes, 2007, Vol 56, pA662
- ISSN
0012-1797
- Publication type
Article