C-反應蛋白、代謝症候群與頸動脈硬化之相關及發生腦中風之危險性: 金山心血管世代研究

Abstract

發炎指標─高敏感度C反應蛋白(high sensitivity C-reactive protein, CRP)為近十年來受研究者關注的新興心血管疾病危險因子，也與代謝症候群有顯著相關。研究者進而建議將CRP作為例行心血管疾病早期預測指標。本研究以參與1994-1995年金山心血管世代追蹤研究的金山地區居民，探討以下三個主要問題。 一、影響CRP濃度的人口學與可修飾的心血管危險因子 (橫斷式研究 1) 背景：CRP濃度為慢性低強度的發炎反應之良好指標。本研究探討人口學因素及心血管疾病的危險因子是否與此程度的發炎反應有關。方法與結果：研究對象為具有CRP與代謝症候群完整資料的916位金山居民。單變項分析顯示，年齡、性別、婚姻狀況、身體質量指數、腰圍、三酸甘油脂、高血壓、糖尿病與代謝症候群與較高的CRP濃度有關。複回歸模式(multiple linear regression model)顯示，年齡、腰圍與代謝症候群與CRP濃度有關。結論：本研究與過去的研究一致，顯示CRP與代謝症候群的相關性。在此族群，中央肥胖與CRP濃度之相關較身體質量指數強。 二、CRP、代謝症候群與頸動脈硬化的相關 (橫斷式研究 2) 背景：過去研究顯示CRP與代謝症候群與頸動脈硬化有關，但其關係在不同硬化指標並不一致且有性別差異。方法與結果：本研究分別探討男性(456人)與女性(354人)居民之血漿(plasma) CRP濃度、代謝症候群與頸動脈硬化之相關。頸動脈硬化的評估指標為頸動脈增厚(thicker intima-media thickness [IMT] on the common carotid artery [CCA], IMT > 1.0 mm)及出現頸動脈粥狀硬化斑塊(presence of plaque on the extracranial carotid artery)。研究對象中，女性的CRP中位數濃度與代謝症候群盛行率高於男性。複回歸分析結果(multiple logistic regression model)顯示，有代謝症候群的女性較傾向頸動脈增厚(odds ratio [OR] = 2.07)，男性則無此現象。然而，相較於CRP濃度小於1 mg/L者，濃度高於3 mg/L的男性出現粥狀硬化斑塊機會較高 (OR=1.99)；若男性同時具有代謝症候群者出現硬化斑塊的機會更高(OR=2.22)。結論：本研究顯示女性的頸動脈增厚與代謝症候群相關；男性之頸動脈粥狀硬化斑塊的出現與CRP濃度有關，且代謝症候群的出現增強此關係。 三、 CRP、代謝症候群與腦中風的相關 (重疊病例對照研究) 背景: CRP對於發生腦中風的影響及其是否改善已知危險因子對心血管疾病的預測能力還未定論。此外，相關研究多數在白人裔族進行，然而華人的CRP中位數濃度顯著低於其他種族。研究亦顯示CRP對於代謝症候群與心血管疾病的關係具有相加效果(additive effect)。方法及結果：本研究以自1994 (或1995)年至2005年為止，65位新發生中風病患為研究組，並隨機選出109位在此期間從未發生中風或冠狀動脈心臟病者為對照組，分析這些研究對象在1994-1995年時的血漿CRP濃度及代謝症候群狀況對其後發生中風的影響。病例組的基線(baseline)血液CRP濃度高於對照組(1.3對1.1 mg/L, p = 0.007)。多變項回歸分析(multiple logistic regression model)結果顯示，相較於對照組CRP濃度分布三分位之最低分位者(<= 0.7 mg/L)，CRP濃度於最高分位者(> 1.3 mg/L)發生中風的危險性為2.55 (95%信賴區間1.05-6.23)，在模式中加入代謝症候群並不影響結果。以Cox回歸模式(Cox proportional hazard model)分析並未改變主要結果。以the receiver operating characteristics curves區域下面積分析結果顯示，相較於已知危險因子(年齡、性別、高血壓、糖尿病、總膽固醇及腰圍)(0.68)，CRP、代謝症候群(皆為0.69)或合併兩者(0.70)皆未顯著增加對於疾病的區別能力。結論：本研究在華人族群印證了較高CRP濃度為發生中風的危險因子，但CRP、代謝症候群及兩者共同對於中風發生的預測能力並未優於已知危險因子。

This study consisted of three sub-projects to investigate the association between C-reactive protein (CRP) levels, the metabolic syndrome (MetS) and cardiovascular risk in middle-aged and older Taiwanese adults. Three major aims were achieved using two types of study designs for the three studies. The nested case-control study evaluated whether CRP levels and MetS predicted stroke risk, in addition to the traditional cardiovascular disease risk factors. The other two parts of the study used a cross-sectional design to examine the association of CRP levels and MetS with carotid atherosclerosis by sex, and investigated the demographic and modifiable correlates for the CRP levels. Study subjects were participants in the 1994-1995 follow-up survey of the Chin-Shan Community Cardiovascular Cohort study. The cross-sectional study included 810 subjects who received measurements for CRP levels, MetS components, and ultrasound examinations for common carotid artery (CCA) intima-media thickness (IMT) and extra-cranial carotid artery plaques. However, for the study aimed to investigate CRP level correlates, participants with complete information on CRP were included regardless of the availability of ultrasound examination data (n=916). The individuals selected for the nested case-control study consisted of 65 first-ever strokes and 109 controls free of stroke during the 10-year follow-up period, from 1994 or 1995 until 2005. CRP levels in the plasma were determined using a high-sensitivity immunoturbidimetric assay. MetS was defined according to criteria from the U.S. National Cholesterol Education Program Adult Treatment PanelⅢ. Results of the cross-sectional study 1: Age, sex, marital status, body mass index, waist circumference, triglycerides, hypertension, diabetes and MetS were associated with CRP levels in univariate analyses. However, age, waist circumference and MetS were remained as the significant predictors of CRP levels in multiple linear regression analyses. Results of the cross-sectional study 2: Women had higher median CRP (1.3 vs. 1.1 mg/L) and higher MetS prevalence than men (58.8 vs. 34.2%). Thicker IMT was associated with MetS in women (adjusted odds ratio [OR] 2.07, 95 % confidence interval [CI] 1.04-4.11) but not in men. Compared to participants with CRP < 1 mg/L, men with CRP > 3 mg/L had an elevated OR with the plaque presence (OR 1.99, 95 % CI 1.10-3.61) but not women. The odds ratio for the artery plague increased to 2.20 (p = 0.046) for individuals with MetS and CRP level > 3 mg/L, compared with men with CRP < 1 mg/L and no MetS. Results of the nested case-control study: The median CRP values in stroke cases and control subjects were 1.5 mg/L and 1.1 mg/L, respectively (p = 0.007). The logistic regression models showed that the multivariate-adjusted stroke OR was 2.55 (95% CI 1.05-6.23) for subjects at the top tertile CRP level compared with the bottom tertile levels in the controls. The risk was not attenuated after further adjustment for MetS or for hypertension. The risk for stroke associated with MetS was eliminated after including hypertension and diabetes in the model. The estimated risks were weaker, but the main results remained, when the Cox proportional hazard model was used for analyses. The area under receiver operating characteristic curves for traditional risk factors (0.68) was improved little by adding CRP or MetS (both 0.69) or the combination of these two variables (0.70). In conclusion, CRP is an independent risk factor associated with stroke. But, both CRP and MetS provide limited improvement for the stroke risk discrimination beyond the traditional risk factors. In the sub-clinical stage of the disease, CRP levels and MetS correlate with distinct phenotypes of carotid atherosclerosis among men and women. MetS is related to thicker IMT in women, while elevated CRP is associated with the presence of plaque in men. This association is enhanced by MetS. Finally, in addition to MetS, central obesity is strongly associated with CRP levels independent of BMI in this population.