Coronary Atherosclerosis and Alcohol Consumption

【摘要】  Objective- Moderate alcohol consumption is associated with reduced cardiovascular disease (CVD) risk. Whether this protection is based on a lesser degree of coronary atherosclerosis has not been established.

Methods and Results- We studied 1676 men and 465 women consecutively undergoing coronary angiography. A score (ATS) was calculated by summing the percent lumen narrowing of all main vessels; alcohol consumption was quantitated by questionnaire. In univariate analysis, ATS was significantly ( P 0.001) associated with male sex, age, familial CVD, smoking, diabetes, hypertension, and serum cholesterol levels; alcohol consumption was associated with less frequent diabetes ( P <0.001) and lower ATS ( P =0.02). By multivariate analysis, alcohol intake was associated with lower ATS ( P <0.01) independently of the other risk factors; the estimated effect size was comparable to that associated with a 1-mmol decrement in serum cholesterol. Over a median follow-up of 93 months, 37 women and 194 men died from a cardiac cause. By Cox analysis, positive predictors for cardiac mortality were male sex (hazard ratio , 1.7; 95% confidence interval , 1.1 to 2.6]), age (HR, 2.1; 95% CI, 1.8 to 2.5 per decade) and diabetes (HR, 1.7; 95% CI, 1.2 to 2.4), whereas alcohol consumption was the only negative predictor (HR, 0.84; 95% CI, 0.71 to 1.00).

Conclusions- In a selected high-risk population, moderate alcohol consumption was independently associated with less coronary atherosclerosis and lower risk for cardiac mortality.

In 2141 consecutive patients undergoing quantitative coronary angiography for the clinical work-up of ischemia, moderate (<231g/wk in men, <154 g/wk in women) alcohol consumption was associated with less coronary atherosclerosis and a lower risk (HR, 0.84; 95% CI, 0.71 to 1.00) of cardiac death at follow-up independently of classical cardiovascular risk factors.

【关键词】  alcohol angiography coronary atherosclerosis

Introduction

Atherosclerotic cardiovascular disease (CVD) still is the leading cause of death and loss of disability-adjusted life years. A large number of epidemiological studies have documented that moderate alcohol use is protective against CVD. 1 In a recent analysis of the modifiable risk factors for acute myocardial infarction in 52 countries, 2 regular alcohol consumption was independently associated with reduced incident myocardial infarction in both sexes, at all ages and in all regions. A number of mechanisms have been postulated to explain this association, including an increase in serum high-density lipoprotein (HDL) and apolipoprotein-A 1 levels, 3 anticoagulant effects-by either direct platelet inhibition or inhibition of the fibrinolytic system 4-6 -a reduction in inflammation, 7 and an enhancement of insulin sensitivity. 8

The physiological basis for the protective effect of alcohol is, however, less certain. The 6 angiographic studies in the English literature 9-14 (Table I, please see http://atvb.ahajournals.org) have included only men (n=4) or only women (n=1), are of limited size, have used variable adjustment for relevant CVD risk factors, and mostly (5 of 6) lack follow-up data. This prompted the present study, in which the association of alcohol consumption with angiographically quantified coronary atherosclerosis was analyzed in a large cohort of men and women, in whom information on classical CVD risk factors and clinical outcome was available.

Materials and Methods

Patients

From 1983 to 1992, 2141 of the 4754 patients admitted to our coronary division underwent coronary angiography as part of the clinical work-up of symptoms or signs of heart disease. The indications for angiography were: acute myocardial infarction (n=43), previous myocardial infarction (n=213), cardiomyopathy (n=118), valvular disease (n=105), arrhythmias (n=30), unstable angina (n=34), acute pulmonary edema (n=6); the remainder of the patients had signs or symptoms of myocardial ischemia. Clinically, 52% of men and 56% of women presented with angina on effort, whereas 75% of men and 26% of women presented with resting angina; 56% of the cases presented with both. Since 1983, the relevant clinical data of each patient have been transferred immediately after discharge to a database created for this purpose. Each patient record consisted of the following blocks of information: (1) demographic data (sex, age, height, weight, years of school education); (2) clinical data (history of resting or effort angina, myocardial infarction, acute cerebrovascular events, coronary bypass surgery, and coronary angioplasty); (3) angiographic data. Any stenosis in a main coronary artery was graded in 1 of 4 levels (50%, 75%, 90%, and 100%); stenoses smaller than 50% were defined as irregularities. Clinically significant obstruction is defined as a narrowing by at least 50% with respect to the prestenotic segment, the grade of the narrowest one was considered. These descriptions constituted the basis for the diagnosis of 1-vessel, 2-vessel, or 3-vessel disease.

Clinical Chart Review

From a detailed review of the clinical charts, the following variables were added. Family history of ischemic heart disease, expressed as present or absent according to whether 1 first-degree relatives were affected. For the first variable, a score was also derived as the sum of all affected relatives (scored as 1 for parent, 1/n for each affected of n siblings, 0.5 for each second-degree relative; values were multiplied by 5 if the subject was affected before age 60 years). Patients were classified as cigarette smokers, past smokers, or nonsmokers. Cigarette consumption (in pack-years) was estimated by multiplying the number of packets per day by the years of smoking. Patients with either a previous diagnosis of hypertension or with blood pressure 140/90 mm Hg on repeated in-hospital measurements were labeled as hypertensive; whenever possible, disease duration was also estimated. Systolic and diastolic blood pressure and heart rate were recorded as the average of five morning values during the first 5 days after admission to hospital. Patients were classified as having known diabetes if they reported a previous diagnosis or were receiving antidiabetic treatment. Patients with type 1 diabetes mellitus (n=12) were excluded from the analysis. Patients with 2 consecutive fasting plasma 7.0 mmol/L and no history of diabetes were classified has having newly diagnosed diabetes. Individual alcoholic beverage consumption was assessed by a standardized questionnaire. Questions about intake of beer, wine, and spirits were included separately. Total alcohol intake was calculated in grams by adding usual intake of alcoholic beverages, assuming the following content: 1 drink of beer, 12.8 g; 1 drink of wine, 11.0 g; and 1 drink of spirits, 14.0 g. From the routine chemistry performed on admission, the following data were taken: plasma glucose, total cholesterol, and triglycerides. In addition to grading coronary disease in terms of clinically significant obstruction, the overall atherosclerotic involvement of the coronary vasculature was quantified by a modification of a method proposed previously. From the original descriptions of the angiograms, an index (ATS score) was calculated by summing the percent narrowing of all stenoses, including multiple stenoses along the same main vessel or those on secondary branches (irregularities were taken as 40% stenosis). This ATS score was calculated for each of the main coronary vessels as well as for the entire visible coronary tree.

Follow-Up

Information concerning death, new acute myocardial infarction, and surgical or transluminal coronary revascularization was collected yearly up to 10 years after discharge from the following sources: outpatient visits, telephone interviews (with the patients, close relatives or the referring physician), questionnaires sent to all patients, and death certificates. Information concerning the events were classified as follows: no event, nonfatal myocardial infarction, cardiac death (which included sudden death, acute left ventricular dysfunction, fatal myocardial infarction and death during or immediately after coronary by-pass surgery), and noncardiac death (which included cancer, accident, stroke, acute pulmonary disease, etc).Of the 2141 patients, 24 subjects were lost to follow-up. Death from all causes was verified in 289 men and 60 women; death from cardiovascular causes was verified in 194 men and 37 women; acute nonfatal myocardial infarction occurred in 144 patients, and an additional 68 patients had a fatal myocardial infarction.

Methods

Cardiac catheterization was performed according to the Judkins technique. 15 Left ventricular end diastolic and aortic pressures were measured routinely during catheterization. Selective left and right coronary angiography and left ventriculography were performed (in that order), with multiple views (2 for the ventriculography and 6 for the coronaries). The images were recorded on 35-mm film and reviewed by the 2 hemodynamists who had performed the invasive study. Echocardiography was performed as described previously.

Statistical Analysis

Values are given as means±SD. Between group mean differences were tested by the unpaired t test (or ANOVA) or the 2 test, for continuous and nominal variables, respectively. Serum triglycerides, cigarette and alcohol consumption, and the ATS score are given as median and and were logarithmically transformed for statistical analysis. Multiple regression and logistic analyses were performed by standard methods; results are given as regression coefficient (mean±SE) and odds ratio (point estimate and 95% confidence interval ), respectively. Multivariate analysis of mortality was performed using Cox proportional hazards analysis; results are expressed as hazard ratio (HR) and 95%CI.

Results

Table 1 shows the clinical characteristics of the study cohort stratified by sex and presence (ATS+) or absence (ATS-) of any angiographically documented coronary narrowing. In comparison with ATS- subjects, ATS+ men and women were older, more often diabetic or hypertensive, or smokers or dyslipidemic (higher serum total cholesterol or triglyceride concentrations). A family history of ischemic heart disease was more prevalent in ATS+ men but not in ATS+ women.

TABLE 1. Clinical Characteristics by Sex and Presence (ATS+) or Absence (ATS-) of Coronary Atherosclerosis

Among men, 24% were nondrinkers as compared with 59% of women ( P <0.0001); however, average alcohol consumption was comparable to that of previous studies ( Table 1 ), and very few subjects 500 g of alcohol per week and men generally consumed more alcohol than women. By using the sex-specific median alcohol consumption (231 g/wk in men and 154 g/wk in women), we stratified subjects into nondrinkers, light drinkers (below the median value) and moderate drinkers (above the median); their clinical characteristics are shown in Table 2. Alcohol consumption was associated with significantly less frequent diabetes in both women and men, and with less frequent hypertension (and lower systolic and diastolic blood pressure values) in women only. Current smoking was more prevalent among moderately drinking men. Alcohol consumption was associated with higher serum total cholesterol concentrations. Serum triglycerides, however, tended to increase in men and decrease in women across drinking status ( P =0.06 for the interaction). The ATS score was smaller in women than men and, in either sex, decreased across drinking category. The percentage of subjects with clinically significant obstruction in a major coronary vessel (single vessel or multivessel disease) generally decreased across drinking status ( 2 =22.3, P =0.0002); this effect fell short of statistical significance when sexes were tested separately.

TABLE 2. Clinical Characteristics by Sex and Drinking Status°

In univariate analysis, the ATS score was significantly ( P 0.001) associated with the classical CVD risk factors (male sex, age, familial ischemic heart disease, smoking, presence of diabetes or hypertension and serum total cholesterol levels); alcohol consumption was associated with a lower ATS ( P =0.02). In multivariate analysis ( Table 3 ), moderate alcohol consumption was associated with a lower ATS score independently of the other CVD risk factors. In this model, the interaction between sex and alcohol consumption was not significant (F=1.62, P =0.2). The predicted effect size was comparable to that associated with a 1-mmol decrement in serum total cholesterol concentrations. When the analysis was performed by using presence or absence of angiographically documented coronary atherosclerosis (ATS+ or ATS-) as the dependent variable, the odds ratio for moderate alcohol consumption was 0.75 (95% CI, 0.56 to 0.99) ( Figure 1 ); adding diabetes treatment had no effect whereas adding antihypertensive treatment yielded an additional independent risk (OR, 1.60; 95% CI, 1.25 to 2.04) without weakening the effect of alcohol use.

TABLE 3. Predictors of Coronary Atherosclerosis *

Figure 1. Logistic regression analysis of factors associated with presence or absence of coronary atherosclerosis (any narrowing on visible vessels by quantitative angiography).

Over a median follow-up period of 93 months, a total of 60 women and 289 men died from all causes; significant predictors of total mortality were male sex (HR,1.5; 95% CI, 1.0 to 2.1), age (HR, 2.0; 95% CI, 1.7 to 2.3 per decade), diabetes (HR,1.6; 95% CI, 1.2 to 2.1), and smoking (HR, 1.4; 95% CI, 1.3 to 1.9). When considering cardiac mortality alone (37 women and 194 men), presence of any degree of coronary atherosclerosis was a strong predictor in both men and women ( P <0.001). When cardiac mortality was related to the number of vessels (0, 1, or 2) with clinically significant stenosis, the protective effect of an alcohol consumption of 200 g per week was only evident in the patients with multivessel disease ( Figure 2 ). By a Cox proportional hazards analysis that included indications for angiography, male sex (HR, 1.7; 95% CI, 1.1 to 2.6), age (HR, 2.1; 95% CI, 1.8 to 2.5 per decade), and diabetes (HR, 1.7; 95% CI, 1.2 to 2.4) were significant positive predictors, whereas alcohol consumption (as a continuous variable) was a significant negative predictor (HR, 0.84; 95% CI, 0.71 to 1.00) ( Figure 3 ). In this model, the sex-alcohol consumption interaction was not significant ( 2 =0,07, P =0,79). In addition, introducing ATS into this model reduced the predictivity of alcohol consumption (HR, 0.86; 95% CI, 0.73 to 1.02).

Figure 2. Cumulative cardiac survival in patients with 0 vessel, 1 vessel, or 2 vessel disease by drinking habits (cutoff of alcohol consumption: 200 g/wk).

Figure 3. Predictors of cardiac mortality by Cox proportional hazards analysis.

Discussion

The main result of the present analysis is that, in a referral population of subjects at high risk for CVD alcohol consumption was associated with less coronary atherosclerosis at angiography and reduced cardiac mortality at follow-up. Specifically, after accounting for all other major CVD risk factors moderate alcohol consumption (in our cohort, <231 g/wk in men, <154 g/wk in women) was associated with a reduction in the ATS score comparable to that predicted by a 1-mmol decrement in serum total cholesterol concentrations. Although the data suggest that lighter alcohol consumption may not carry such benefit, a clear threshold or J-shaped relationship 16 could not be identified because our cohort included too few women (who drank less than men) and very few heavy drinkers. Also, the ATS score decreased across drinking category somewhat, but the change was more evident in women than in men, ie, in the smaller group. Accordingly, when the multivariate analysis of ATS predictors was performed separately in men and women, there was a suggestion that in women the protective effect of alcohol (OR, 0.67) was stronger than in men (OR, 0.87), but this difference was not fully supported statistically. Finally, the effect of alcohol consumption on coronary artery disease was seen both when using ATS, ie, the total extent of lumen narrowing in the visible vasculature and by using the criterion of clinically significant obstruction, ie, the tightest of all visible stenoses along a major branch. Of note is that the effect of alcohol was independent of the confounding influence exerted by much stronger determinants such as age, smoking, and diabetes.

In our cohort, cardiac mortality was predicted by the expected set of risk factors ( Figure 3 ) and was clearly related to the severity of coronary atherosclerosis in men as well as in women. Furthermore, there was a graded increase in cardiac deaths in 1 vessel involvement on the angiogram. This finding implies that, although cardiac mortality can be caused by a variety of causes (myocardial infarction, arrhythmias, and heart failure), the underlying atherosclerotic disease of the coronary vasculature is a common soil for any such cause. By analogy, the observed protection by alcohol consumption against cardiac mortality can be thought to be causally related to the less extensive atherosclerotic disease of the coronary arteries. In fact, when ATS was included among the predictors of cardiac death, the protective effect of alcohol was attenuated. Of note, reduced cardiac mortality with an alcohol consumption of 200 g per week was clearly seen in patients with multivessel disease, whereas this effect was not evident in patients with single vessel disease. Although in general any effect is more readily demonstrable in higher-risk than in low-risk conditions, this finding may also suggest that the mechanism by which alcohol exerts its anti-atherogenic action relates to the progression of the disease rather than its initiation. With regard to this, the only serial angiographic study relevant to this issue, performed in a small cohort of Swedish women hospitalized for an acute coronary syndrome, 14 has documented a slower progression of coronary atherosclerosis in women drinking moderate amounts 35 g/wk) as compared with both lighter drinkers and abstainers.

Collectively, our findings provide strong support for the notion that moderate alcohol consumption is associated with a degree of protection of the coronary vasculature from atherosclerotic involvement, and that this effect translates into a modest reduction in risk of cardiac death over a period of several years. Clearly, this conclusion cannot be readily extrapolated from a Coronary Care Unit referral population to the general population, in which quantitative assessment of coronary atherosclerosis is not feasible (at least by invasive methods). The finding, however, that regular alcohol consumption is independently associated with fewer clinical events (incident myocardial infarction) in both sexes, at all ages and in all regions 2 does imply that reduced coronary atherosclerosis may be the underlying reason also in the general population.

It is of interest that in our cohort there was relatively little difference in the clinical phenotype between drinkers and nondrinkers ( Table 3 ). Thus, age, adiposity, familial ischemic heart disease, and smoking habits were similar across drinking status in both men and women. The only clinically relevant phenotypic differences were the lower prevalence of known diabetes in drinkers versus nondrinkers (in men and women) and the lower prevalence of hypertension (in women only). With regard to this, there is ample evidence that moderate alcohol consumption reduces the risk of both prevalent 17 and incident diabetes. 18-19 Furthermore, moderate alcohol consumption reduces CVD risk in diabetic patients to the same extent than in nondiabetic subjects. 20 Among the mechanisms invoked to explain this protection, enhanced insulin sensitivity and attenuated low-grade inflammation-both predictors of incident diabetes 21 -have been proposed to mediate the effect of alcohol on glucose homeostasis. 22 Notably, the favorable effect of moderate alcohol consumption on insulin action has been particularly well-documented (using a prospective experimental protocol) in nondiabetic postmenopausal women. 23 However, moderate alcohol consumption has been reported to be associated with less hypertension in women. 24,25 This finding has not been confirmed in other studies, 26 and alcohol consumption is associated with obesity, which itself a risk factor for hypertension. 27

With regard to other potential mediators of the effects of moderate drinking, the serum lipid profile has received much attention. 28 Most epidemiological studies concur to show that alcohol consumption is associated with higher serum concentrations of both HDL cholesterol and triglycerides. 3 Likewise, a meta-analysis of short-term alcohol administration protocols 29 shows that alcohol intake raises both HDL cholesterol and triglycerides within a period of 1 to 9 weeks. This dual effect may result from an alcohol-induced increase in postprandial triglyceride-rich lipoprotein flux and removal, which secondarily raises HDL cholesterol and reverse cholesterol transport. 30 In our cohort, serum HDL cholesterol levels were unfortunately missing in many cases, but the increasing pattern of total cholesterol concentrations across drinking status may reflect increasing HDL cholesterol levels. Serum triglycerides, however, increased across category of drinking status in men but not in women, in whom the inverse trend was observed ( Table 3 ). The reasons for the sex difference in the association between alcohol intake and serum triglycerides are unclear and may simply be caused by confounding. However, differences in frequency of alcohol intake, 31 levels of sex hormone-binding globulin, 32 and genetic polymorphism in the apolipoprotein CIII, 33 apolipoprotein E, 34 and other factors controlling lipid metabolism may be involved.

In conclusion, in a selected high-risk population, moderate alcohol consumption was independently associated with less coronary atherosclerosis and reduced cardiac mortality.

Acknowledgments

Disclosures

None.

【参考文献】
  Mukamal KJ, Conigrave KM, Mittleman MA, Camargo C, Rimm EB. Roles of drinking pattern and type of alcohol consumed in coronary artery disease in men. N Engl J Med. 2003; 348: 109-118.

Yusuf S, Hawken S, Ounpuu S, Dans T, Avezum A, Lanas F, McQueen M, Budaj A, Pais P, Varigos J, Lisheng L; INTERHEART Study Investigators. Effect of potentially modifiable risk factors associated with myocardial infarction in 52 countries (the INTERHEART Study): case-control study. Lancet. 2004; 364: 937-952.

De Oliveira E Silva ER, Foster D, McGee Harper M, Seidman CE, Smith JD, Breslow JL, Brinton EA. Alcohol consumption raises HDL cholesterol levels by increasing the transport rate of apolipoproteins A-I and A-II. Circulation. 2000; 102: 2347-2352.

Mennen LI, Balkau B, Vol S, Caces E, Eschwege E. Fibrinogen: a possible link between alcohol consumption and cardiovascular disease? DESIR Study Group. Arterioscler Thromb Vasc Biol. 1999; 19: 887-892.

Zhang QH, Das K, Siddiqui S, Myers AK. Effects of acute, moderate ethanol consumption on human platelet aggregation in platelet-rich plasma and whole blood. Alcohol Clin Exp Res. 2000; 24: 528-534.

Ridker PM, Vaughan DE, Stampfer MJ, Glynn RJ, Hennekens CH. Association of moderate alcohol consumption and plasma concentrations of endogenous tissue-type plasminogen activator. JAMA. 1994; 272: 929-933.

Albert MA, Glynn RJ, Ridker PM. Alcohol consumption and plasma concentration of C-reactive protein. Circulation. 2003; 107: 443-447.

Mukamal KJ, Cushman M, Mittleman MA, Tracy RP, Siscovick DS. Alcohol consumption and inflammatory markers in older adults: the Cardiovascular Health Study. Atherosclerosis. 2004; 173: 79-87.

Barboriak JJ, Anderson AJ, Hoffmann RG. Interrelationship between coronary artery occlusion, high-density lipoprotein cholesterol, and alcohol intake. J Lab Clin Med. 1979; 94: 348-353.

Gruchow HW, Hoffmann RG, Anderson AJ, Barboriak JJ. Effects of drinking patterns on the relationschip between alcohol and coronary occlusion. Atherosclerosis. 1982; 43: 393-404.

Handa K, Sasaki J, Saku K, Kono S, Arakawa K. Alcohol consumption, serum lipids and severity of angiographically determined coronary artery disease. Am J Cardiol. 1990; 65: 287-289.

Ducimetiere P, Guize A, Marciniak A, Milon H, Richard J, Rufat P; for the CORALI Study Group. Arteriographically documented coronary artery disease and alcohol consumption in French men. The CORALI study. Eur Heart J. 1993; 14: 727-733.

Liu Y, Tanaka H, Sasazuki S, Yoshimasu K, Kodama H, Washio M, Tanaka K, Tokunaga S, Kono S, Arai H, Koyanagi S, Hiyamuta K, Doi Y, Kawano T, Nakagaki O, Takada K, Nii T, Shirai K, Ideishi M, Arakawa K, Mohri M, Takeshita A. Alcohol consumption and severity of angiographically determined coronary artery disease in Japanese men and women. Atherosclerosis. 2001; 156: 177-183.

Janszky I, Mukamal KJ, Orth-Gomer K, Romelsjö A, Schenck-Gustaffsson K, Svane B, Kirkeeide RL, Mittleman MA. Alcohol consumption and coronary atherosclerosis progression-the Stockholm female coronary risk angiographic study. Atherosclerosis. 2004; 176: 311-319.

Bush CA, VanFossen DB, Kolibash AJ Jr, Magorien RD, Bacon JP, Ansel GM, Eaton GM, Ramancik MJ, Orsini AR, Palmer SL. Cardiac catheterization and coronary angiography using 5 French preformed (Judkins) catheters from the percutaneous right brachial approach: a comparative analysis with the femoral approach. Cathet Cardiovasc Diagn. 1993; 29: 267-272.

Rouillier P, Boutron-Ruault MC, Bertrais S, Arnault N, Daudin JJ, Bacro JN, Hercberg S. Alcohol and atherosclerotic vascular disease risk factors in French men: relationships are linear, J-shaped, and U-shaped. Alcohol Clin Exp Res. 2005; 29: 84-88.

Solomon CG, Hu F. B, Stampfer MJ, Colditz GA, Speizer FE, Rimm EB, Willet WC, Manson JE. Moderate alcohol consumption and risk of coronary heart disease among women with type 2 diabetes mellitus. Circulation. 2000; 102: 494-499.

Kroenke CH, Naing-Feng C, Rifai N, Spiegelman D, Hankinson SE, Manson JE, Rimm EB. A cross-sectional study of alcohol consumption patterns and biologic markers of glycemic control among 459 women. Diabetes Care. 2003; 26: 1971-1978.

Ajani UA, Gaziano M, Lotufo PA, Liu S, Hennekens CH, Buring JE, Manson JE. Alcohol consumption and risk of coronary heart disease by diabetes status. Circulation. 2000; 102: 500-505.

Weyer C, Bogardus C, Mott DM, Pratley RE. The natural hystory of insulin secretory dysfunction and insulin resistance in the pathogenesis of type diabetes mellitus. J Clin Invest. 1999; 104: 787-794.

Duncan BB, Schimdt MI, Pankow JS, Ballantyne CM, Couper D, Vigo A, Hoogeveen R, Folsom AR, Heiss G. Atherosclerosis Risk in Communities Study. Low-grade systemic inflammation and the development of type 2 diabetes: the atherosclerosis risk in communities study. Diabetes. 2003; 52: 1799-1805.

Howard AA, Arnsten JH, Gourevitch MN. Effect of alcohol consumption on diabetes mellitus: a systematic review. Ann Intern Med. 2004; 140: 211-219.

Davies MJ, Baer DJ, Judd JT, Brown ED, Campbell WS, Taylor PR. Effects of moderate alcohol intake on fasting insulin and glucose concentrations and insulin sensitivity in postmenopausal women: a randomized controlled trial. JAMA. 2002; 287: 2559-2562.

Nanchahal K, Ashton WD, Wood DA. Alcohol consumption, metabolic cardiovascular risk factors and hypertension in women. Int J Epidemiol. 2000; 29: 57-64.

Thadani R, Camargo CA, Stamfer MJ, Curhan GC, Willett WC, Rimm EB. Prospective study of moderate alcohol consumption and risk of hypertension in young women. Arch Intern Med. 2002; 162: 569-574.

Fuchs FD, Chambless LE, Whelton PK, Nieto FJ, Heiss G. Alcohol consumption and the incidence of hypertension. The Atherosclerosis Risk in Communities Study. Hypertension. 2001; 37: 1242-1250.

Breslow RA, Smothers BA. Drinking patterns and body mass index in never smokers: National Health Interview Survey, 1997-2001. Am J Epidemiol. 2005; 161: 368-376.

Wannamethee SG, Lowe GD, Shaper G, Whincup PH, Rumley A, Walker M, Lennon L. The effect of different alcoholic drinks on lipids, insulin and haemostatic and inflammatory markers in older men. Thromb Haemost. 2003; 90: 1080-1087.

Rimm EB, Williams P, Fosher K, Criqui M, Stampfer MJ. Moderate alcohol intake and lower risk of coronary heart disease: meta-analysis of effects on lipids and haemostatic factors. BMJ. 1999; 319: 1523-1528.

Chung BH, Doran S, Liang P, Osterlund L, Cho S, Oster RA, Darnell B, Franklin F. Alcohol-mediated enhancement of postprandial lipemia: a contributing factor to an increase in plasma HDL and a decrease in risk of cardiovascular disease. Am J Clin Nutr. 2003; 78: 391-399.

Hashimoto Y, Futamura A, Nakarai H, Nakahara K. Effects of the frequency of alcohol intake on risk factors for coronary heart disease. Eur J Epidemiol. 2001; 17: 307-312.

Bataille V, Perret B, Evans A, Amouyel P, Arveiler D, Ducimetiere P, Bard JM, Ferrieres J. Sex hormone-binding globulin is a major determinant of the lipid profile: the PRIME Study. Atherosclerosis. 2005; 179: 369-373.

Tilly P, Sass C, Vincent-Viry M, Aguillon D, Siest G, Visvikis S. Biological and genetic determinants of serum apoC-III concentration: reference limits from the Stanislas Cohort. J Lipid Res. 2003; 44: 430-436.

Lussier-Cacan S, Bolduc A, Xhignesse M, Niyosenga T, Sing C. Impact of alcohol intake on measures of lipid metabolism depends on context defined by gender, body mass index, cigarette smoking, and apolipoprotein E genotype. Arterioscler Thromb Vasc Biol. 2002; 22: 824-831.

作者单位：Romana Femia; Andrea Natali; Antonio L?Abbate; Ele FerranniniFrom Department of Internal Medicine and CNR Institute of Clinical Physiology, University of Pisa School of Medicine, Pisa, Italy.