Leslie Dan Faculty of Pharmacy, University of Toronto, Toronto, Canada.
Cardiovasc Diabetol. 2010 Aug 13;9:37. doi: 10.1186/1475-2840-9-37.
Increased coronary intima media thickness (CIMT) has been associated with adverse cardiovascular outcomes, as have increased glucose levels. The link has not been established between glucose and CIMT; therefore, we sought to assess the relationship between glucose and CIMT.
Medline, EMBASE, Scopus, and Cochrane databases were searched from inception through 2009 for original research reporting both postprandial glucose levels and CIMT measurements. Glucose was classified as normal, impaired, or diabetic. Outputs included inverse variance weighted effect size and also average correlation (using the Wang and Bushman approach). Data were combined using a random effects meta-analytic model. Heterogeneity as assessed using chi(2) and I(2); bias was examined using Egger plots and Begg-Mazumdar tau. Polynomial functions (i.e., linear, quadratic, cubic, quartic) were fit to the data and the Akaike Information Criteria were used to select the optimal model.
We identified 172 papers; 161 were rejected (19 inappropriate design, 8 had selected patients, 101 inappropriate outcomes) leaving 11 accepted. We used data from 15,592 patients (8250 normals, 3013 impaired glucose, 4329 diabetics). There was no evidence of heterogeneity or publication bias. The overall correlation was 0.082 (CI95%:0.066-0.098); the overall effect size was 0.294 (0.245-0.343) between diabetics and normals and 0.137 (0.072-0.202) between normals and those with impaired glucose. The equation of best fit was linear (CIMT = 0.828 + 0.009*glucose).
There is a small but significant relationship between postprandial glucose levels and CIMT, which have both been associated with adverse cardiovascular outcomes.
已发现，增加的冠状动脉内膜中层厚度（CIMT）与不良心血管结局相关，血糖水平升高也是如此。目前尚未确定血糖与 CIMT 之间的联系，因此，我们试图评估血糖与 CIMT 之间的关系。
从建立到 2009 年，我们在 Medline、EMBASE、Scopus 和 Cochrane 数据库中搜索了原始研究，这些研究均报告了餐后血糖水平和 CIMT 测量结果。血糖分为正常、受损或糖尿病。结果包括逆方差加权效应大小，还包括平均相关系数（使用 Wang 和 Bushman 方法）。使用随机效应荟萃分析模型合并数据。使用卡方检验和 I2 评估异质性；使用 Egger 图和 Begg-Mazumdar tau 检验偏倚。对数据拟合多项式函数（即线性、二次、三次、四次），并使用 Akaike 信息准则选择最佳模型。
我们确定了 172 篇论文；161 篇被拒绝（19 篇设计不当，8 篇选择患者，101 篇不合适的结果），剩下 11 篇被接受。我们使用了来自 15592 名患者的数据（8250 名正常，3013 名血糖受损，4329 名糖尿病患者）。没有证据表明存在异质性或发表偏倚。总体相关性为 0.082（95%CI：0.066-0.098）；正常人和糖尿病患者之间的总体效应大小为 0.294（0.245-0.343），正常人和血糖受损患者之间的效应大小为 0.137（0.072-0.202）。最佳拟合方程为线性（CIMT=0.828+0.009*血糖）。
餐后血糖水平与 CIMT 之间存在微小但显著的关系，这两者均与不良心血管结局相关。
