代謝症候群動態流行統計分析

Abstract

背景 代謝症候群是造成心血管疾病的主要原因，並且在臨床科學上得到很好的闡明。對於代謝症侯群在族群中的動態流行病學變化，研究則是相當有限，而對於代謝症候群的動態變化研究，有助於預防心血管疾病。 研究目的 本論文利用社區篩檢資料，藉由盛行率與發生率來預測代謝症候群之自然病史，並且使用四階段隨機過程模式，來估計使用藥物與非藥物介入後，代謝症候群之各階段轉移參數，以評估介入方法之功效。 材料及方法 本研究相關資料是取自1999年至2009年基隆社區整合式篩檢計畫，此計畫從1999年開始，每年於基隆所執行之多種疾病篩檢計畫。總共有118,087名個案納入研究分析，包括47,124名男性與70,963名女性，年齡分佈從20歲至79歲。其中，18,640名男性與35,779名女性追蹤兩次以上，總共追蹤300,329人-年，平均追蹤長度，男性為5.1年，女性為5.7年。 本研究利用貝氏方法估計盛行率與發生率之比例，並且利用四階段隨機過程，來闡明對於代謝症候群之介入方法，在預防心血管疾病上的效果。 結果 代謝症候群之盛行率/發生率之比率，代表疾病存在狀態的時間長短，在男性為3.9年，女性則為3.2年。當以年齡每10歲為分組時，盛行率/發生率之比率在男性從2.9年增加到4.7年，在女性則從2.7年增加至3.8年。兩者盛行率/發生率之最高數值，皆落在50至59歲這一年齡區間。 中等教育程度、肥胖，以及抽煙、喝酒與檳榔等習慣的人會有較高的盛行率/發生率的比值。在多變項分析中，肥胖為影響停留在代謝症候群狀態中，最顯著的因子。代謝症候群的不同階段中，產生心血管疾病的風險，年齡都是重要的因子，在尚未有代謝症候群的階段時，肥胖因子有其影響。 對於代謝症候群之自然病史的轉移參數，從正常轉移到1至2個代謝因子為0.4906，從1至2個代謝因子轉移到正常為0.3348，從1至2個代謝因子轉移到代謝症候群則為0.0478，估計降低35%的心血管疾病風險。 結論 代謝症候群之動態流行病學使用族群基礎之健康篩檢資料，此方法提供了一個有效率的量化方式，來評估介入政策對於預防心血管疾病之效果。

Background: Metabolic syndrome (MetS) is a well-known major cause of cardiovascular disease and has been well elucidated in clinical science. However, information on dynamic epidemiology of MetS at population level is very limited, which precludes one from making good health planning for prevention of cardiovascular disease (CVD). Aims: This thesis aims to make use of prevalence and incidence of MetS from empirical data to elucidate the dynamic epidemiology of MetS and use a four-state Markov process to elucidate the natural course of MetS and to project the risk of CVD before and after introduction of intervention on MetS so as to estimate the expected efficacy of each intervention. Methods: Study subjects were participants during 1999-2009 in the Keelung Community-based Integrated Screening (KCIS) program, a multiple disease-screening program conducted annually with recruitment in the northernmost county of Taiwan since 1999. A total of 118,087 individuals (aged 30-79, extended to 20 years old later) were eligible for the following analysis, including 47,124 men and 70,963 women. Among these, 18,640 men and 35,779 women were recruited more than once, which were made up of total follow-up duration of 300,329 person years. The average follow-up times were 5.1 years in male and 5.7years in female. Bayesian approach to estimate the ratio of prevalence to incidence was applied. This estimate together with incident cohort of MetS forms a four-state stochastic process to elucidate the efficacy of possible intervention programs specified for the prevention of CVD in relation to MetS. Results: The ratio of prevalence to incidence for the overall group yielded the duration of natural course of MetS, approximately 3.9 years in men and 3.2 years in women. The corresponding figures for 10-years age groups ranged from 2.9 years to 4.7 years in men and 2.7 years to 3.8 years in women. The highest duration in both sexes all fell into the 50-59 age group. Medium education level, obesity, and habits including alcohol drinking, cigarette smoking and betel chewing tend to had higher ratio of prevalence to incidence. In multivariate analysis, obesity is the most detrimental factor that influenced the duration of natural course of MetS. Age was an important factor for individuals in each MetS state developing CVD, and obesity had its influence among individuals in normal or 1-2 MetS components state for developing CVD. Transition parameters for the natural course of MetS based on incident cohort were 49.06% from normal to 1-2 individual MetS components, 33.48% from 1-2 individual MetS components to normal, and 4.78% from 1-2 individual MetS components to MetS. If 50% of life-style modification is expected, 35% reduction of CVD-related outcome was expected. Conclusion: Dynamic epidemiology of MetS has been elucidated by using empirical population-based health check-up data. The better use of such information provides an effective quantitative measure for the efficacy of intervention program for the prevention of CVD.