中国人群总胆固醇水平与脑卒中及其亚型关系的前瞻性研究暨脂联素与心血管疾病的meta分析

【作者】 张欢；

【导师】 顾东风；

【作者基本信息】 北京协和医学院 ， 流行病与卫生统计学， 2012， 博士

【摘要】 背景与目的：脑卒中是当今世界危害人类健康最重要的疾病之一,具有极高的致残率和较高的致死率。与国外研究相比,中国人群总胆固醇(TC)水平与脑卒中及其亚型,如出血性脑卒中的发病和死亡风险之间关系的大规模队列研究较少且结果不明确。本研究以中国队列研究协作组(China Cohort Studies Collaboration, CCSC)以及十一五代谢综合征队列人群为研究对象,从多个角度探讨基线TC水平与脑卒中及其亚型之间的关系,为更好地预防和控制我国脑卒中的发病和死亡提供一定的科学依据。资料与方法：本研究中CCSC队列纳入分析者62,603人,十一五队列纳入分析者19,796人,共82,399人,其中男性47,875人,女性34,524人。两队列人群根据基线TC水平按照成人血脂异常防治指南(以下简称指南)划分标准分为四组,比较各组间基线特征的差别。采用COX比例风险回归模型,以TC<5.18mmol/L (200mg/dl)为对照组,计算其余各水平组(20mg/dl为步长)脑卒中、缺血性脑卒中、出血性脑卒中发病和死亡的风险比(Hazard Ratio, HR)及相应的95%置信区间(95%CI)。同时计算TC水平每升高1mmol/L相应的结局事件发生的HR(95%CI)。计算人群归因危险度百分比(PARP),分析脑卒中及其亚型归因于高TC血症的比例。结果：CCSC.总随访人年为573,894.92,最终发生脑卒中1,068例(缺血性脑卒中367例、出血性脑卒中361例和未分类脑卒中340例)。十一五队列总随访人年为156,306.26,最终发生脑卒中400例(缺血性脑卒中227例、出血性脑卒中148例和未分类脑卒中25例)。年龄标化后,十一五队列的基线TC平均水平比CCSC增加0.23mmol/L(8.88mg/d1)。脑卒中、缺血性脑卒中以及出血性脑卒中的标化发病率,无论在男性和女性人群中,十一五队列都高于CCSC。在男性人群中,脑卒中、缺血性脑卒中以及出血性脑卒中的标化死亡率均为十一五队列高于CCSC。在女性人群中脑卒中以及出血性脑卒中的标化死亡率为CCSC高于十一五队列,而缺血性脑卒中的标化死亡率在两个队列中基本一致。CCSC中,按照指南分为四组,与对照组相比,多因素调整后,TC在5.18-5.68mmol/L(200-219mg/d1)以及>6.22mmol/L(240mg/d1)组缺血性脑卒中的发病风险分别增加55%和77%,致死性缺血性脑卒中的发生风险分别增加99%和131%(均P<0.05)。TC水平每升高1mmol/L,缺血性脑卒中和致死性缺血性脑卒中的发病风险分别增加21%和25%(均P<0.05)。十一五队列中,按照指南分为四组,与对照组相比,多因素调整后,TC在各水平组发生脑卒中、缺血性脑卒中和出血性脑卒中风险的HR值均没有统计学意义(均P>0.05)。两个队列合并的分析结果显示,按照指南分为四组,与对照组相比,多因素调整后,TC在各水平组缺血性脑卒中的发病风险分别增加31%、45%和61%(均P<0.05)；TC在5.18-5.68mmol/L(200-219mg/d1)和≥6.22mmol/L(240mg/d1)组发生致死性缺血性脑卒中的风险分别增加80%和131%(均P<0.05)。多因素调整后,TC水平每升高1mmol/L,脑卒中的发病风险增加5%,缺血性脑卒中的发病风险增加20%,致死性缺血性脑卒中的发生风险增加24%,均P<0.05。PARP分析结果显示,脑卒中的发病3.6%可归因于TC水平升高,缺血性脑卒中的发病11.1%可归因于TC水平升高。亚组分析结果显示,多因素调整后,TC水平每升高1mmol/L,缺血性脑卒中的发病风险在男性和女性人群中分别增加20%和17%；在年龄<60岁和≥60岁人群中分别增加18%和19%；在SBP<140mmHg和≥140mmHg人群中分别增加14%和25%；在吸烟和不吸烟人群中分别增加17%和21%(均P<0.05)。与对照组相比,多因素调整后,在男性、年龄<60岁、SBP≥140mmHg和吸烟人群中,TC在5.69-6.21(220-239)和≥6.22mmol/L(240mg/d1)组缺血性脑卒中的发病风险均显著增加；在SBP≥140mmHg人群中,TC≥6.22mmol/L(240mg/d1)组脑卒中的发病风险增加32%(P=0.01)；在吸烟人群中,TC≥6.22mmol/L(240mg/d1)组脑卒中和出血性脑卒中的发病风险分别增加33%(P=0.02)和50%(P=0.04)。此外,合并其他危险因素时,随着危险因素数目的增加,脑卒中及其亚型的发病风险均增加。在相同危险因素条件下,与TC<5.18mmol/L(200mg/d1)者相比,TC≥5.18mmol/L(200mg/d1)的个体脑卒中及其亚型的发病风险均增加。结论：近二三十年来,中国人群TC水平呈现升高的趋势。脑卒中、缺血性脑卒中和出血性脑卒中的发病率也不断增加。随着TC水平的增加,脑卒中和缺血性脑卒中的发病风险均不断增加,致死性缺血性脑卒中的风险也显著增加。吸烟人群中,与TC<5.18mmol/L(200mg/d1)组相比,TC>6.22mmol/L(240mg/d1)组出血性脑卒中的发病风险显著增加。合并其他危险因素时,在相同危险因素条件下,高TC水平者脑卒中、缺血性脑卒中和出血性脑卒中的发病风险均增加。鉴于我国人群TC水平和脑卒中发病率不断升高,及TC水平对脑卒中及其亚型发病的危险性,通过平衡膳食,提倡健康生活方式,减少血脂紊乱和降低人群血清TC水平,是减少与控制我国人群脑卒中发病率和危害的重要公共卫生策略和措施之一。背景与目的：既往研究发现脂联素可以调节糖代谢以及脂质代谢,并具有抗炎以及抗动脉粥样硬化的作用。这些结果提示脂联素可能具有降低冠心病发病风险的作用。此外,许多研究探讨了脂联素基因(ADIPOQ)单核苷酸多态性,即+45T>G(rs2241766),+276G>T(rs1501299)以及-11377C>G(rs266729),与心血管疾病风险的关系,然而这些研究并没有得到一致的结果。本研究拟通过meta分析评价ADIPOQ三个多态位点与心血管疾病风险的相关性,并且通过对脂联素水平与冠心病发病关系的前瞻性研究资料进行meta分析,以评估脂联素水平对冠心病发病风险的影响。资料与方法：两名评价者独立系统检索Pubmed、EMbase、CBM、维普、万方等数据库,获取ADIPOQ三个多态位点与心血管疾病风险的关联研究文献,以及脂联素水平与冠心病关系的前瞻性研究文献。在2012年2月之前发表的研究被纳入进行meta分析。本研究中,ADIPOQ与心血管疾病的meta分析主要考虑病例对照组之间等位基因频率比较的比值比(Odds Ratio,OR值)；脂联素与冠心病的1neta分析主要提取各研究的多因素调整的相对风险度(RR)及95%置信区间(95%CI)等数据。采用Q检验及I2统计量评估研究间的异质性,根据异质性选择固定效应或随机效应模型合并RR值或OR值及95%CI；使用Egger线性回归检验估计发表偏倚。本研究使用相关评分标准对所有纳入的研究进行评分。结果：ADIPOQ三个多态位点与心血管疾病风险的meta分析纳入37个病例对照研究,包含6,398个病例和10,829个对照(rs2241766)；8,392个病例和18,730个对照(rs1501299)；以及7,835个病例和14,023个对照(rs266729)。三个多态位点都与心血管疾病存在显著关联。合并OR值分别为：rs2241766,1.22(95%CI：1.07,1.39；P=0.004)；rs1501299,0.90(95%CI：0.83,0.97；P=0.007)；rs266729,1.09(95%CI：1.01,1.17；P=0.032)。rs2241766和rs1501299与冠心病也显著相关,合并OR值分别为1.29(95%CI：1.09,1.52；P=0.004)和0.89(95%CI：0.81,0.99；P=0.025)。rs2241766和rs266729与脑卒中存在显著关联。固定效应模型合并OR值分别为1.28(95%CI：1.12,1.46；P<0.0001)和1.20(95%CI：1.08,1.34；P=0.0001)。三个位点与高血压之间关联都没有统计学显著性。分析发现纳入的研究间存在显著异质性。对rs266729与CVD之间的关系按照种族进行亚组分析可以消除异质性,欧洲和亚洲人群亚组分析的I2分别为26.1%和O。亚组分析发现只有在亚洲人群中的相关性是显著的,合并OR值(固定效应模型)为1.29(95%CI：1.18,1.42；P<0.001)；在欧洲人群中相关性并不显著,合并OR值(固定效应模型)为1.01(95%CI：0.94,1.08；P=0.880)。分析中发现rs2241766与CVD关系的研究存在发表偏倚(Egger’s检验,P=0.007)。脂联素水平与冠心病meta分析共纳入12个前瞻性研究,其中包括8个巢式病例对照研究和4个队列研究。最终共有11,461名研究对象纳入分析,其中4,132名研究对象发生冠心病。结果显示脂联素水平与冠心病的发病风险呈负相关。合并的RR值为0.83(95%CI：0.69,0.98；P=0.031)。亚组分析结果显示在男性、女性以及年龄小于65岁的群体中都存在这种负相关性。本项研究中没有发现发表偏倚(P=0.911)。结论：本项meta分析显示ADIPOQ基因单核苷酸多态性与心血管疾病之间存在显著的相关性。脂联素水平对冠心病的发生具有保护作用。我们仍然需要高质量的大规模研究来证实这些相关性。更多还原

【Abstract】 Background and Objective:Stroke is the most common cause of death and severe disability in the world. The relationships between serum total cholesterol (TC) level and risk of stroke and its subtypes are not well established in Asia. We used individual data from the cohorts of China Cohort Studies Collaboration (CCSC) and the Incidence and Comprehensive Control of Metabolic Syndrome Study (ICMS) to explore the impact of baseline TC level on the risks of total stroke, ischaemic and haemorrhagic stroke and its implication in the population prevention of stroke in China.Materials and Methods:A total of62,603participants from CCSC were included, of which data on cholesterol levels at baseline and the outcome of stroke at follow-up were all available. For ICMS, of27,020participants aged35to74at baseline,21,556participants have completed the follow-up survey and19,796participants with full information were analyzed in this study. Participants in these two cohorts were divided into four groups according to the baseline TC level, that was TC<5.18mmol/L (200mg/dl),5.18-5.67mmol/L (200-219mg/dl),5.68-6.21mmol/L (220-239mg/dl) and>6.22mmol/L (240mg/dl). The differences of mean levels of baseline risk factors were compared. Cox regression models were used to estimate the hazard ratios (HR) and95%confidence intervals (CI) for the incidence and death of stroke and stroke subtypes. Population attributable risk percentage (PARP) was also calculated.Results:Over the course of573,894.92person-years of follow-up in CCSC, a total of1,068incident strokes were collected, including367ischaemic strokes,361haemorrhagic strokes and340unclassified strokes. For ICMS, after156,306.26person-years of follow-up, a total of400incident strokes were collected, including227ischaemic strokes,148haemorrhagic strokes and25unclassified strokes.The age-adjusted mean level of baseline TC in ICMS participants was0.23mmol/L (8.88mg/dl) higher than that of CCSC participants. The age-standardized person-year incidence of stroke, ischaemic and haemorrhagic stroke were all higher for both men and women in ICMS. In male subgroup analysis, the standardized person-year mortality of stroke, ischaemic stroke and haemorrhagic stroke were all higher in men from ICMS compared with that in men from CCSC. However, the age-standardized person-year mortality of stroke and haemorrhagic stroke were higher in CCSC in women. The age-standardized person-year mortality of ischaemic stroke was basically the same in both CCSC and ICMS in women.In CCSC, the multivariate-adjusted HRs (95%CIs) for total (fatal or nonfatal) ischemic stroke were1.55(1.17,2.06) and1.77(1.27,2.45) in TC5.18-5.68mmol/L (200-219mg/dl) and TC>6.22mmol/L (240mg/dl) groups compared with the lowest baseline TC level group, respectively. Accordingly, compared with the lowest baseline TC level group, the multivariate-adjusted HRs (95%CIs) for fatal ischemic stroke were1.99(1.22,3.24) and2.31(1.31,4.08) in TC5.18-5.68mmol/L (200-219mg/dl) and TC>6.22mmol/L (240mg/dl) groups, respectively. Each1-mmol/L increase in TC was associated with a21%greater risk of total ischemic stroke (P<0.0001). In ICMS cohort, the multivariate-adjusted HRs for stroke, ischaemic and haemorrhagic stroke were not statistically significant in TC5.18-5.68mmol/L (200-219mg/dl),5.69-6.21mmol/L (220-239mg/dl) and TC>6.22mmol/L (240mg/dl) groups compared with the lowest TC group.For the combined data of CCSC and ICMS, the multivariate-adjusted HRs (95%CIs) for total ischemic stroke were1.31(1.04,1.65),1.45(1.10,1.91) and1.61(1.24,2.08) in TC5.18-5.68mmol/L (200-219mg/dl),5.69-6.21mmol/L (220-239mg/dl) and TC>6.22mmol/L (240mg/dl) groups compared with the lowest TC group, separately. The multivariate-adjusted HRs (95%CIs) for fatal ischemic stroke were correspondingly1.80(1.16,2.79) and2.31(1.42,3.75) in TC5.18-5.68(200-219) and TC≥6.22mmol/L (240mg/dl) groups. Each1-mmol/L increase in TC was significantly associated with a5%,20%and24%higher risk of total stroke, total ischemic stroke and fatal ischemic stroke, respectively. PARP for total stroke and total ischemic stroke was3.6%and11.1%, respectively.The results of subgroup analyses showed that each1-mmol/L increase in TC was significantly associated with a20%higher risk of ischemic stroke for men and17%for women,18%for age<60and19%for age>60,14%for SBP<140mmHg and25%for SBP≥140mmHg, and17%for smokers and21%for never smokers. The multivariate adjusted risk of ischemic stroke was significantly increased in TC5.69-6.21(220-239) and≥6.22mmol/L (240mg/dl) for men, age<60, SBP≥140mmHg and smoker subgroups compared with the reference group. The multivariate-adjusted HR (95%CI) for stroke was1.32(1.06,1.65) in TC>6.22mmol/L (240mg/dl) group for SBP>140mmHg subgroup accordingly. In smokers, the multivariate-adjusted HRs (95%CIs) for total stroke and total haemorrhagic stroke were1.33(1.05,1.69) and1.50(1.01,2.22), respectively. Besides, an elevated HR of total stroke and its subtypes were found when considering TC with the increased number of other risk factors.Conclusions:In the three decades, the serum TC level in Chinese populations significantly increased, so as the incidence of stroke, ischaemic and haemorrhagic stroke in China. The baseline TC level was strongly associated with the risks of stroke and ischaemic stroke in our study. For smokers, the risk of total stroke and total haemorrhagic stroke significantly increased in TC>6.22mmol/L (240mg/dl) compared with TC<5.18mmol/L (200mg/dl). Higher TC levels were strongly associated with increased risks of stroke, ischaemic and haemorrhagic stroke.Reducing serum TC level in residents through primary prevention and lifestyle modification will have an important impact on population prevention and control of stroke in China. Background and Objective:Adiponectin is the most abundant circulating protein secreted by adipocytes. Previous studies have examined the associations between polymorphisms of adiponectin gene (ADIPOQ) and cardiovascular disease (CVD), but those studies have been inconclusive. There is also uncertainty about the association between concentrations of circulating adiponectin and coronary heart disease (CHD) risk. The aims of this study were to evaluate the relationship between three single nucleotide polymorphisms (SNPs),+45T>G (rs2241766),+276G>T (rs1501299) and-11377C<G (rs266729) in ADIPOQ and CVD, and summarize the effect of adiponectin on the incident CHD.Materials and Methods:A comprehensive search was conducted to identify all studies on the association of ADIPOQ gene polymorphisms with CVD risk and the association of adiponectin levels with incident risk of CHD. The fixed and random effect pooled measures (i.e. odds ratio (OR) and95%confidence interval (CI)) were calculated in the meta-analysis. Heterogeneity among studies was evaluated using Q test and the I2. Publication bias was estimated using modified Egger’s linear regression test.Results:Thirty-seven studies concerning the associations between the three polymorphisms of ADIPOQ gene and CVD risk were enrolled in this meta-analysis, including6,398cases and10,829controls for rs2241766,8,392cases and18,730controls for rs1501299and7,835cases and14,023controls for rs266729. The three SNPs were significantly associated with CVD, yielding pooled ORs of1.22(95%CI:1.07,1.39; P=0.004),0.90(95%CI:0.83,0.97; P=0.007) and1.09(95%CI:1.01,1.17; P=0.032) for rs2241766, rs1501299and rs266729, respectively. Rs2241766and rs1501299were significantly associated with coronary heart disease (CHD), yielding pooled ORs of1.29(95%CI:1.09,1.52; P=0.004) and0.89(95%CI:0.81,0.99; P=0.025), respectively. The pooled OR for rs266729and CHD was1.09(95%CI:0.99,1.19; P=0.090). Rs2241766and rs266729were significantly associated with stroke, yielding pooled ORs (fixed-effects method) of1.28(95%CI:1.12,1.46; P<0.0001) and1.20(95%CI:1.08,1.34; P=0.0001), respectively. The associations between the three SNPs and hypertension were not significant. Significant between-study heterogeneity and evidence of publication bias were observed in the meta-analysis.Twelve prospective studies comprising8nested case-control studies and4cohort studies were included in the meta-analysis. A total of14,960participants were enrolled and4,132incident CHD events were observed. The pooled RR of CHD was0.83(95%CI:0.69,0.98; P=0.031). The inverse association was consistently observed in men and women and in the studies with mean age<65year. No publication bias was found in our study (P=0.911).Conclusions:The present meta-analysis showed that the associations between rs2241766, rs1501299and rs266729in the ADIPOQ and CVD were significant. Higher levels of adiponectin were associated with a low risk of CHD. The protective effect was consistent in men and women and in the middle-aged populations. High quality studies are still needed to confirm the associations.更多还原