| 摘要: | 研究目的–本研究目的主要有 (1)、以世代研究了解糖尿病的自然史並估計台灣的糖尿病發生率 (2)、探討B、C 型肝炎與糖尿病的相關性 (3)、研究代謝症候群和糖尿病、心血管疾病的關係材料與方法 –本研究團隊已經進行此世代研究兩年(94-95 年度)，研究樣本來源，以自民國93 年1 月1 日起，在雲林縣台大雲林分院接受40 歲以上成人健檢抽血檢查，且空腹血糖低於126 mg/dl 的非糖尿病患者為對象，進一步進行2 小時 OGTT，以連續收樣的方式進行樣本蒐集，兩個年度預計至少可以收到600 個樣本。已經抽血檢查項目主要有OGTT、總膽固醇、高密度膽固醇、低密度膽固醇、三酸甘油脂、胰島素、糖化血色素等。接下來三年計畫的第一年，以強化世代生化基礎資料為主，分析血樣其他重要的生化項目，包含B、C 型肝炎帶原、發炎指標(hsCRP)、肝功能(GOT、GPT、GGT)、脂蛋白-A1、脂蛋白-B。此外，繼續收取100 位OGTT 樣本，使樣本數達到700 人。第二年至第三年度起，開始進行追蹤糖尿病發生與相關因子的評估。統計分析方法-描述性的資料與生理特徵，以平均值與標準差表示， 各組間的差異以變異數分析（ANOVA）、共變數分析(ANCOVA)、t-test 與卡方檢定方法，檢定不同組間的差異。分組的方法有兩種，NFG(Normal Fasting Glucose)、 IFG(Impaired Fasting Glucose)、IGT(Impaired Glucose Tolerance)、IFG+IGT 四組。另一種分組只分為糖尿病與非糖尿病兩組。第一年的資料，以正常組做為參考族群，以單變項與多變項邏輯式迴歸分析（univariate and multivariate logistic regression）方法，估計各因子造成血糖異常的勝算比。以線性迴歸方程式(linear regression)，分析血糖改變量的影響因子。追蹤的第二、三年，進展成糖尿病的相關因子，因屬於世代研究，所以採用Cox’s proportion hazard model 分析導致糖尿病或是胰島素阻抗發生的原因。糖尿病發生率的估計將以世界衛生組之2000 年標準人口為依據，估計本世代的標準化發生率。統計差異顯著水準(p value)為5%。
OBJECTIVE –The purposes of this study are to explore the nature history of diabetes mellitus and to examine the relationship between hepatitis B/C infection and DM by cohort study in Taiwan. We will also study the metabolic syndrome associated with glucose intolerance and cardiovascular disease. RESEARCH DESIGN AND METHODS – About 600 subjects aged ≧ 40 years with fasting glucose under 125 mg/dl were randomly selected from the National Taiwan University Hospital Yun-Lin Branch during January 2004 and December 2005, and they had an oral glucose tolerance test (OGTT) and physical examination from 2005-2006. At baseline (2005-2007), those individuals in the non-diabetic state are considered eligible for further follow up during 2007-2009. Subjects with either FPG ≧ 126 mg/dl or 2hPG ≧ 200 mg/dl are considered to have diabetes. In this ongoing study, we will conduct blood examination for estimation of biochemical, metabolic and anthropometric characteristics, including weight, height, waist circumference, hip circumference, OGTT, TG, high and low density cholesterol, blood pressure, insulin resistance, high sensitivity c-reactive protein (hsCRP), GPT, GPT, GGT, apolipoprotein A1, and apolipoprotein B. Statistical Analysis - Descriptive data will be described as means and SDs for continuous variables, and analysis of variance (ANOVA), ANCOVA and Chi-square tests were used for assessing the significances. Bonferroni method was used for post-hoc comparison in ANOVA. Multiple linear regressions were used for predicting determinants of FPG. We tested the variables scales in this study for multicollinearity by correlation matrix and VIF (variance inflation factor). Odds ratios (ORs) and the 95% confidence interval (CI) were calculated to estimate the relative risk of diabetes mellitus by logistic regression model. Using Cox’s proportion hazard model, the hazard ratios of type 2 diabetes for relative insulin resistance and varying degrees of glucose intolerance to the normal group will analyzed. A p-value below 0.05 was considered significant. The statistical analyses were performed with SPSS statistical Package (SPSS base 10.0, SPSS Inc. Chicago). |
