簡國龍臺灣大學：預防醫學研究所許金旺Hsu, Chin-WangChin-WangHsu2007-11-282018-06-292007-11-282018-06-292005http://ntur.lib.ntu.edu.tw//handle/246246/59221前言與目的：目前已有許多研究有關急性心肌梗塞預後因子，但很少有研究利用急診初始血醣值來評估急性心肌梗塞預後，而急診的初始血醣值可代表早期對急性心肌梗塞壓力性的血醣反應，可能與日後短期及長期預後有相關危險性。我們目前研究目標，評估急診的初始血醣值對第一次急性心肌梗塞病患預後所佔角色。 材料與方法：此研究以醫學中心急診病患族群為主，自民國九十年1月1日起至民國九十二年12月31日止，三年間因第一次急性心肌梗塞至三軍總醫院急診部病患(大於18歲)的臨床資料進行回溯性研究。排除非第一次急性心肌梗塞、急性心肌梗塞超過二十四小時、從其他醫院轉入急性心肌梗塞病患及有接受經皮冠狀動脈介入術或冠狀動脈重建手術。急診初始血醣值分為三種程度，level I急診初始血醣值小於或等於140 mg/dl，level II急診初始血醣值大於140 mg/dl 及小於或等於200 mg/dl，及level III急診初始血醣值大於200 mg/dl。針對初始血醣值三種程度值及其他重要變項導入羅吉斯氏迴歸(Logistic regression)，及使用Cox 迴歸模式分別對短期及長期預後進行多變項分析。 結果：此研究共收集198例，男性有159人，女性有39 人，平均年齡63.1 ± 14.2歲。以急診初始血醣值level I 為基準，在心肌梗塞短期預後結果，在危險比Level II 高出5.1倍(p<0.0001)，Level III 高出6.5倍(p<0.0001)，及長期預後結果，在危害比Level II 高出1.7 倍(p=0.081)，Level III 高出2.9倍(p<0.0001)。在一年長期預後存活分析，急診初始血醣值三種程度有明顯統計差異(logrank test, p=0.0002)。以急診初始血醣值level I 為基準，在調整性別、年齡、有無再灌注治療及心肌梗塞部位後，心肌梗塞短期預後調整危險比，隨著急診初始血醣值高而增加。(Level II: Odds ratio: 3.87; 95% Confidence Interval [CI], 1.71 to 8.78; P=0.001; level III, 5.16; 95% CI, 1.97 to 13.51; p<0.0001)一年長期預後調整危害比，Level III對一年長期預後仍佔重要危險因子(Hazard ratio: 3.08, 95% CI 1.59 to 5.98, p=0.001). 結論：急診初始高血醣值為第一次心肌梗塞後，短期及長期預後的高危險性重要且獨立預測因子。Background & Objectives: Although studies on prognostic factors for acute myocardial infarction (AMI) were available, there were scanty reports on initial glucose levels. Especially initial glucose levels at emergency department (ED) can be presented as early glucose response to stress; it might be associated with short-term and long-term prognosis of first AMI. The aim of the study is to evaluate the prognostic role of initial glucose levels in first AMI patients. Material and Methods: This study design was a hospital-based retrospective cohort study, recruited first AMI patients from one tertiary hospital ED between January 1, 2001, and December 31, 2003. Patients with previous coronary artery disease history, duration AMI more than 24 hours or referral from other hospital were excluded. Initial glucose levels at ED were stratified into three categories (level I: <=140, II: 140~200, III: >=200 mg/dl). Logistic and Cox’s regression models were applied to estimate one-month short-term and one-year long-term outcomes, respectively. Results: Total 198 eligible subjects (men: 159, women: 39, mean age 63.1 ± 14.2 years) were recruited into the study. There was 5.1-fold (p<0.0001) higher risk in level II and 6.5-fold (p<0.0001) in level III for short-term prognosis, and 1.7-fold higher risk in level II and 2.9-fold risk in level III for long-term prognosis in patients with first AMI compared with level I. The estimate of survival function for long-term prognosis among three initial glucose levels shows significantly different (logrank test, P=0.0002). After adjusting for sex, age, reperfusion therapy and infarct subtype, the adjusted odds ratio for short-term outcomes progressively increased with higher levels compared with level I (Level II: Odds ratio: 3.87; 95% Confidence Interval [CI], 1.71 to 8.78; P=0.001; level III, 5.16; 95% CI, 1.97 to 13.51; p<0.0001) and the highest initial glucose level still played an important risk factor for long-term prognosis (Hazard ratio: 3.08, 95% CI 1.59 to 5.98, p=0.001). Conclusions: Higher initial glucose levels in ED were an important and independent predictor of higher risk of short-term and long-term prognosis in patients with first AMI.Contents 中文摘要……………………………………………………………… 5 Abstracts 7 I. Introduction 9 II. Literature Review 10 II-1. Stress Hyperglycemia Following Acute Myocardial Infarction 10 II-2. What is the Stress Hyperglycemia? 11 II-3. The Mechanisms of Stress Hyperglycemia and its Effects on the Adverse Complications of AMI 12 II-3-1 Mechanisms of stress hyperglycemia in acutely ill patients 12 II-3-2 Biologic effect of stress hyperglycemia on the prognosis after acute myocardial infarction 13 II-4. Relationship of Stress Hyperglycemia to the Prognosis of AMI 16 II-4-1 Hyperglycemia is an independent predictor of the prognosis of AMI 17 II-4-2 Hyperglycemia and short-term prognosis of AMI 17 II-4-3 Hyperglycemia and long-term prognosis of AMI 18 II-4-4 Hyperglycemia and the extent of cardiac injury 18 II-4-5 Summary 19 II-5. What is the appropriate level of stress hyperglycemia in AMI? 19 III Limitation of Previous Studies 20 IV. Hypotheses & Aims 22 V. Material and Methods 23 V-1. Study Design and Study Subjects 23 V-1-1. Data collection 23 V-1-2. Exclusion criteria 24 V-2. Exposure and Confounding Factors Assessment 24 V-3-1 Measurements of Glucose level, Total cholesterol, HDL cholesterol, Triglyceride, Uric acid, CK and CKMB 25 V-3-2 Troponin-I measurement 26 V-3-3 AMI-related treatment and infarct site 26 V-4. Outcome Assessment 27 (1) One-month short-term outcome definition 27 (2) Long-term outcome 27 (3) Length of Stay 27 V-5 Statistical Analysis 28 V-5-1 Descriptive statistics 28 V-5-2 Analytic statistics 28 VI. Results 30 VI-1. Demographic Description and Enrollment Flowchart 30 VI-2. Patient characteristics 30 VI-3. Thrombolytic therapy, infarct site and coronary artery stenosis in AMI among three initial glucose level 31 VI-4. Short-term prognosis following AMI and three glucose levels 32 VI-5 Long-term prognosis following AMI and three glucose level 32 VI-6 Univariate analysis of factors associated with the prognosis following AMI 33 VI-7 Multivariate analysis of factors associated with the prognosis following AMI 34 VI-8 Length of hospital and intensive care unit stay 35 VI-9 The short-term and long-term prognosis in diabetic and nondiabetic patients with first AMI 35 VII. Discussion 37 VII-1. Hyperglycemia and short-term prognosis of AMI 37 VII-2. Hyperglycemia and long-term prognosis 38 VII-3 Diabetes status and the prognosis of AMI 40 VII-4 Extent of infarct size and hyperglycemia 41 VII-5 Clinical risk factors and hyperglycemia in predicting the short-term and long-term prognosis in first AMI 41 VIII Limitation 42 IX. Conclusion 44 X. References 45 Tables 49 Table 1. Literature review about the association between hyperglycemia and morbidity/mortality in AMI patients 49 Table 2. Basic characteristics among study population, specified by initial glucose level in emergency department 53 Table 3 Basic characteristics among study population, specified by short-term and long-term outcome 54 Table 4 Acute myocardial infarction (AMI)-related reperfusion therapy, infarct subtype and coronary artery stenosis 55 Table 5 One-month short-term prognosis outcome in the study population, specified by initial glucose levels 56 Table 6 One-year long-term prognosis outcomes in the study population, specified by glucose levels 58 Table 7 Unadjusted odds ratio and respective 95% confidence intervals of various risk factors for short-term and long-term prognosis following first AMI 59 Table 8 Adjusted odds ratio and respective 95% confidence interval of risk factor profiles, under the logistic regression models for the short-term prognosis 61 Table 9 Adjusted hazard ratio and respective 95% confidence interval of risk factor profiles, under the Cox’s regression model for the one-year long-term prognosis 62 Table 10 Length of hospital and intensive care unit stay, specified by initial glucose level 63 Table 11 Short-term and long-term prognosis in diabetic and nondiabetic patients with first AMI 64 Figure 66 Figure 1 The mechanism of hyperglycemia in critical illness and possible effect of hyperglycemia on cardiac outcome following AMI 66 Figure 2 Enrollment flow chart. Patients with an first acute myocardial infarction 67 Figure 3 Scatter plot of initial glucose level at ED and fasting plasma glucose level in AMI patients 68 Figure 4 Kaplan-Meier cumulative survival curves for patients during follow-up period of one year, stratified by three initial glucose levels (the log-rank test P=0.0002). 69 Figure 5 Short-term and long-term prognosis in diabetic and nondiabetic patients with first AMI 70335753 bytesapplication/pdfen-US急性心肌梗塞預後因子初始血醣值prognostic factorsacute myocardial infarction (AMI)initial glucose levelsthesishttp://ntur.lib.ntu.edu.tw/bitstream/246246/59221/1/ntu-94-R91846012-1.pdf