减少样本数改良最小模型技术

摘要

Objective Classical minimal model technique (MMT) requires 30 samples and 3 hours.To overcomethe defects that MMT needs so many the numbers of sample and time.MethodsWe chose a sample schedule thatminimized the variance of the parameter estimates and the error in reconstructing the plasma insulin profile, trying toreduce the total number of samples of MMT from 30 to 12 or 10 and the time of MMT from 3 hours to 2 hours.rnRespectively using 30, 12 and 10 samples protocols, insulin sensitivityrn rnrnrespectively) and glucoserneffectivenessrn rnrnrespectively) were estimated in 16 subjects with normal glucose tolerancern(NGT) and 13 patients with type 2 diabetes mellitus.ResultsThere were strong relationships between not onlyrn rnrnin every group ( γ =0.74～0.92, P <0.0001)；rn rnrnhad highly relationships withrn rnrnin each group ( γ = 0.86～ 0.93, P <0.0001).ConclusionThe reformations reducing the total number of samplesrnof MMT from 30 to 12 or 10 and the time of MMT from 3 hours to 2 hours may be feasible.%目的 为克服经典的最小模型技术(MMT)样本数太多(30个)、历时太长(3小时)的缺陷，以减少样本数所获得的信息对估计参数和重建胰岛素曲线的误差降至最小范围。方法试对经典MMT进行样本数减至12个和10个、时间缩至2小时改良，分别用30、12和10个样本三种方案的胰岛素改良MMT(IMMMT)评估16例葡萄糖耐量正常(NGT)者和13例2型糖尿rn病病人的胰岛素敏感性rn rnrn和葡萄糖效应rn rnrn。结果不仅12个样本，而且10个样本的rn rnrn与30个样本的rn rnrn在每个rn组值都相近，相关性高(γ=0.74～0.92，P<0.0001)。结论提示样本数减至10～12个、试验时间缩至2小时改良IMMMT是可行的。