论文部分内容阅读
目的探讨利用口服葡萄糖耐量试验(OGTT)中5个时点的血糖值和胰岛素值,借助线性最小模型(LMM)计算胰岛素敏感指数(ISI)和分泌功能指数(BCI)的可行性。方法对660例天津地区的中国人正常糖耐量组150例、糖调节受损组220例、糖尿病组290例的OGTT资料进行研究,分析比较4种ISI:李氏IAI、DeFronzoISI、HOMAIR、LMMISI,和5种BCI:李氏MBCI、HOMAβ、ΔI30/ΔG30、ΔI60/ΔG60、LMMBCI。结果(1)用HOMAβ调整后,在4种ISI中,LMMISI与G梯形的相关性相对最好(r=-0.6083,P<0.001)。(2)用DeFronzoISI调整后,在5种BCI中,LMMBCI与G梯形的相关性相对最好(r=-0.9092,P<0.001)。(3)LMMISI与LMMBCI相配合可以较好的解释G梯形的变化(决定系数R2=0.836,P<0.001)。结论利用OGTT资料所计算的LMMISI和LMMBCI是一对相对较好的反映胰岛素敏感性和胰岛素分泌功能的指数。
Objective To explore the feasibility of calculating insulin sensitivity index (ISI) and secretory function index (BCI) using linear least squares model (LMM) at five time points in oral glucose tolerance test (OGTT). Methods The data of OGTT in 150 Chinese patients with normal glucose tolerance, 220 patients with impaired glucose regulation and 290 patients with diabetes mellitus were analyzed in 660 patients in Tianjin. The data of four types of ISI were analyzed including ISI, DeFronzoISI, HOMAIR, LMMISI, And 5 BCIs: Lee MBCI, HOMAβ, ΔI30 / ΔG30, ΔI60 / ΔG60, LMMBCI. Results (1) The correlation between LMMISI and G trapezoid was the best among the four ISIs (r = -0.6083, P <0.001) after adjusting with HOMAβ. (2) The correlation between LMMBCI and G trapezoid was the best among the five BCIs (r = -0.9092, P <0.001) after adjusting with DeFronzoISI. (3) The combination of LMMISI and LMMBCI can better explain the change of G trapezoid (the coefficient of determination R2 = 0.836, P <0.001). Conclusions LMMISI and LMMBCI, calculated using OGTT data, are relatively good indicators of insulin sensitivity and insulin secretion.