Most large prospective cohort studies show that regular, moderate consumers of alcohol have considerably lower risk for development of coronary heart disease and other types of CVD than do abstainers.
Numerous mechanisms have been demonstrated in basic scientific studies and limited human trials to explain the effects of both alcohol and the polyphenols in some alcoholic beverages on reduction of CVD risk; these include beneficial effects on lipids, coagulation, fibrinolysis, inflammation, glucose metabolism, and endothelial function.
Evidence is accumulating that moderate drinkers are also at lower risk of other chronic diseases (including diabetes, dementia, and osteoporosis) and have lower all-cause mortality rates than those of nondrinkers.
In contrast to potential health benefits from moderate drinking, there are many adverse health and societal effects of the abuse of alcohol. Advice to patients regarding alcohol consumption will vary according to their individual characteristics (age, sex, risk factors) but should always be based on scientifically sound and balanced data.
Any recommendations regarding moderate alcohol consumption for the prevention of CVD for individuals should be based on consultation with the health care provider.
Epidemiologic Evidence Relating Alcohol to Coronary Heart Disease
Starting in the early 1970s, many scientists began to publish data from prospective studies indicating a lower risk for development of coronary heart disease (CHD) for moderate consumers of alcohol. Of hundreds of studies published in the scientific literature during the past few decades, the results have been extremely consistent. Almost uniformly, they have demonstrated that in comparison with nondrinkers, moderate drinkers are less likely to develop CHD.
In a meta-analysis by Corrao and associates based only on high-quality prospective studies, there was a J-shaped curve for the association between alcohol and CHD, with its nadir at about 20 g/day of alcohol, the equivalent of about U.S. drinks (with 12 g of alcohol, the average amount in a “typical drink”). The relative risk crossed 1.0 (the risk for nondrinkers) between 72 and 89 g of alcohol per day ( Fig. 18-1 ).
A lower risk for CHD and other cardiovascular disease (CVD) among moderate drinkers has been shown in populations with a low intake of alcohol, such as the Chinese, and in those with typically high alcohol intake, as in Scandinavia. Data from the latter show that heavy drinking is associated with increases in sudden death. Rimm stated that the evidence to support the hypothesis that the inverse association of alcohol to CHD is causal, and not confounded by healthy lifestyle behaviors, is very strong. This evidence includes the fact that moderate alcohol consumption reduces CHD and mortality in individuals in diverse cultures with varying drinking habits; among subjects with hypertension, diabetes, and existing CHD; and among otherwise healthy individuals. Mukamal and coworkers demonstrated among “very healthy” subjects (who did not smoke, exercised, ate a good diet, and were not obese) that those who drank moderately had a much lower relative risk of CHD (0.38; 95% CI, 0.16-0.89) than that of similar subjects who did not consume any alcohol.
There are notable exceptions to the inverse association between alcohol and coronary artery disease occurrence or mortality in epidemiologic studies. As would be expected, longitudinal studies based primarily on young people tend to have few instances of CVD and may show no protection against CHD. Studies from eastern Europe, where very heavy drinking is the norm, often do not show lower rates of CHD among drinkers. In some studies in which binge drinkers are combined with regular moderate drinkers, lower rates for “moderate” drinking have not been seen. Furthermore, if the range considered to be moderate drinking extends to more than three or four typical drinks per day, an inverse association may not be seen among drinkers. The meta-analysis by Corrao and associates demonstrated that essentially all studies show lower CHD for consumers of up to 12.5 g of alcohol per day, but some fail to show lower rates when consumption of up to 25 or 50 g/day is considered the moderate category.
As described by Wannamethee and Shaper, it is essential that ex-drinkers not be included with lifetime abstainers in the nondrinking category, as the former tend to have higher rates of many types of disease that may falsely increase a putative beneficial effect among moderate drinkers. Poikolainen has pointed out the importance of identifying abusive drinkers in a population before attempting to evaluate the association between moderate drinking and CVD. As will be described later, some studies suggest that African Americans may not show a lower risk of CVD from moderate drinking.
Lack of Randomized Clinical Trials of Alcohol Consumption and Coronary Heart Disease Events
In lieu of randomized clinical trials of alcohol intake and cardiovascular outcomes, physicians are forced to rely primarily on associations shown in observational studies, experiments of the effects of alcohol on risk factors, and animal experiments. Also, results among subjects in prospective studies who change their drinking habits during follow-up may be informative. Several studies have demonstrated that when former abstainers begin to consume alcohol, their subsequent risk of CVD decreases. Gronbaek and colleagues, in a prospective study of more than 14,000 subjects, showed that moderate drinkers had lower risk of total mortality than nondrinkers or heavy drinkers. Furthermore, based on alcohol assessments 5 years apart, they found that in comparison with stable drinkers, subjects who reduced their drinking from light to none increased their mortality risk (RR = 1.40; 95% CI = 1.00-1.95), and those who went from nondrinking to light drinking reduced their risk (RR = 0.71; 95% CI = 0.44-1.14). In the Health Professionals Study, with repeated alcohol assessments, subjects reporting an increase of ≥10 g/day in alcohol had a decrease in risk of subsequent myocardial infarction (RR = 0.55; CI 0.33-0.91), whereas there was a tendency for a slight increase in risk for subjects decreasing their alcohol intake (RR = 1.10, CI: 0.92, 1.31).
Former abstainers who reported moderate intake in follow-up examinations in the Atherosclerosis Risk in Communities (ARIC) study subsequently had a 38% lower chance for development of CVD than did their persistently nondrinking counterparts. In a large population-based cohort of women in Australia, moderate drinkers showed the best self-reported state of overall health, whereas those who stopped drinking during follow-up (even those with no reported comorbidity) showed a decrease in their ratings of overall health ; as in most studies, however, the reasons that the subjects stopped drinking were not known.
Differences in Response to Alcohol by Gender, Age, and Ethnicity
Women are known to show decreased tolerance to alcohol, which is thought to relate to lower levels of alcohol-clearing enzymes (especially alcohol dehydrogenase), their generally smaller body size, and the greater proportion of their body consisting of fat, which metabolizes alcohol less well than muscle does. Both the beneficial and adverse effects of alcohol appear at lower dosages for women. Hence, guidelines generally suggest that moderate drinking for women should be only about half the amount of alcohol for men. From a meta-analysis, Di Castelnuovo and coworkers found that women drinkers showed an increase in mortality over that of nondrinkers above about 2 drinks/day, whereas the increase among men was at about drinks/day. In the Australia Longitudinal Study on Women’s Health, with approximately 12,000 subjects aged 70 years and older, Byles and colleagues found that nondrinkers and women who rarely drink had a significantly higher risk of dying (HR = 1.94; 95% CI, 1.4-2.6) during the survey period than did women who consumed alcoholic beverages at a level of one or two drinks per day; the best general health status and highest levels of physical functioning were among those drinking at this level for 3 to 6 days per week.
Alcohol has been shown to be associated with CHD in the elderly as well as in middle-aged people. In the Cardiovascular Health Study, subjects older than 65 years who consumed 14 or more drinks per week had a risk of incident myocardial infarction or cardiac death about 40% lower than that of abstainers. People tend to drink less as they age, and most current recommendations state that individuals older than 65 years should drink less than what is considered moderate for younger people. On the other hand, two analyses comparing the health effects among elderly subjects consuming one or two drinks per day, in comparison with those consuming only one drink per day or less, have not shown higher morbidity or mortality among those consuming more alcohol.
Tolvanen and colleagues evaluated 10-year mortality according to alcohol intake among subjects initially aged 60 to 99 years, of whom 50% of men and 40% of women died during follow-up. After adjustment for age, sex, educational level, marital status, chronic diseases, functional ability, and smoking, the relative risk for mortality of frequent drinkers (versus nondrinkers) was 0.6 (95% CI, 0.4, 0.8); for occasional drinkers, 0.7 (0.5, 1.0); and for ex-drinkers, 1.1 (0.8, 1.7).
Whereas most of the original studies relating alcohol intake to CHD were in Europeans and European Americans, a large number of studies have found essentially the same effects among Asians. Data on African Americans are limited, and Sempos and colleagues found no inverse association between alcohol and CVD risk among African Americans. Klatsky and associates, however, found similar inverse associations between alcohol and heart failure in the Kaiser Permanente cohort among whites, Asians, and African Americans. Using data from the Women’s Health Initiative, Freiberg and coworkers reported lower total mortality among African American hypertensives who consumed alcohol than among those who did not drink, but a similar relation was not seen among nonhypertensive African Americans.
Importance of Pattern of Alcohol Consumption
It has become very clear that the amount of alcohol consumed (within limits) is not as important in terms of health effects as is the pattern of consumption. The two key aspects of pattern are the frequency of drinking and a large number of drinks on a single occasion (binge drinking). Mukamal and coworkers demonstrated among people who had suffered a myocardial infarction that those who drank moderately without binge drinking had 30% fewer deaths, whereas those who reported binge drinking (defined in this study as consuming three or more drinks within 1 to 2 hours) had death rates even higher than those of teetotalers.
Studies of the effects of alcohol intake on CHD, obesity, cognitive function, and other diseases show that more frequent drinkers have the best outcomes; for subjects consuming the same average amount of alcohol during a week, those consuming alcohol every day may have up to 50% lower risk of disease than those consuming the same amount on only 1 or 2 days. In a meta-analysis based on the only six prospective studies in the literature that permitted such an analysis of pattern of drinking and CHD, Bagnardi and coworkers reported that in comparison with not consuming alcohol, regular drinking (even at fairly high levels of consumption) was associated with lower risk of CHD; irregular heavy consumption, or binge drinking, increased the risk. In a population of light to moderate drinkers, alcohol consumption in general was associated with decreased risk of acute myocardial infarction in women, but episodic intoxication was related to a substantial increase in risk.
Cross-cultural prospective studies have shown that simply correlating the average amount of alcohol consumed with CVD outcomes is inadequate. In some cultures, daily drinking is the norm, but in others, alcohol consumption is generally confined to Friday and Saturday nights; whereas benefits are seen in the former, they are absent in the latter. Data are mixed on whether consumption of alcohol in conjunction with food affects the outcome, but elevated risk for hypertension has been reported in one study only for individuals who drink outside of meals. Furthermore, Gorelik and associates found in an intervention study that red wine served with a high-fat meal inhibited the postprandial increase in serum and urinary levels of cytotoxic lipid peroxidation products.
Is It Alcohol or Associated Lifestyle Factors That Result in Lower Risk for Coronary Heart Disease?
Many lifestyle factors are associated with moderate drinking. In all epidemiologic studies, socioeconomic factors are strongly and inversely related to the risk of CHD; better-educated individuals and those with higher income show lower rates of CHD (and most other diseases), presumably because of their more moderate lifestyles. The extent to which residual confounding by such factors may explain the health-protective effects of moderate drinking shown in observational studies has provoked considerable debate.
As a follow-up to concerns from earlier work by Shaper and colleagues, Fillmore and coworkers have described apparent “errors” in prospective studies that call into question their conclusions about an inverse association between alcohol and CHD. One of their arguments is that exposure measures may not be as accurate in prospective cohort studies as in case-control studies, a point not supported by most epidemiologists. In general, there is greater possibility of misclassification of exposure in case-control studies than in follow-up studies, as the former may “introduce a number of subtleties and avenues for bias that are absent in typical cohort studies.”
The choice of the referent group in an epidemiologic study is important, and the nondrinker category should not mix ex–heavy drinkers (who usually have higher risks for many diseases) with lifetime abstainers. In a population in which abstinence is rare, choosing very light or occasional drinkers as the referent group may be preferable. A key criticism of existing observational studies of alcohol and CHD relates to the inclusion in some of ex-drinkers in the nondrinking category. This is indeed an important problem, but many recent reports have limited the referent, nondrinking group to lifetime abstainers, and the inverse association between moderate drinking and CHD remains, as described by Rimm and Moats and others.
An approach for dealing with potential residual confounding by social class is to limit analyses to subjects who are very similar in socioeconomic terms, as has been done for nurses, health professionals, and business executives. In each of these groups of subjects, moderate drinking was associated with lower CHD risk. Lee and coworkers evaluated a large variety of measures of lifestyle in seeking to determine if the lower total mortality risk in moderate drinkers is due to the alcohol itself or to associated healthy lifestyle factors. Adjusting for a large number of such factors and using sophisticated analytic techniques, the authors found that moderate drinking (versus not drinking) was still associated with a 38% lower mortality risk. Friesema and colleagues reviewed these relations among subjects in a prospective case-cohort, the Lifestyle and Health Study, consisting of 16,210 men and women between 45 and 70 years of age. The authors concluded that the difference in lifestyle between moderate drinkers and both never drinkers and former drinkers was only a partial explanation of the observed inverse relationships between alcohol intake and CVD and all-cause mortality.
Poikolainen and associates evaluated the effects of a large number of lifestyle factors that were considered to increase the risk of CHD; these included hypertension, increased body mass index (BMI), diabetes, depression, sleep disturbances, smoking, physical inactivity, poor life satisfaction, psychological distress, trait anxiety, independent and dependent life events, longer length of working hours, low levels of job control, job strain, and effort-reward imbalance. Most of these conditions were either the same between lifetime abstainers and light to moderate drinkers or more common among drinkers than among abstainers. Thus, these authors concluded that none of the large number of lifestyle factors evaluated is likely to be the reason that abstainers have higher rates of CHD, which supports the theory that it is the alcohol that is associated with less disease.
Rimm summarized the research data on this topic by stating that results from mechanistic studies provide substantial support for the hypothesis that moderate alcohol intake reduces the risk of CHD and that all beverages containing alcohol have shown beneficial effects. He concluded, “The ‘sick-quitter’ hypothesis and the concern that moderate drinkers lead a healthier lifestyle may explain a small proportion of the benefit attributed to alcohol in some studies, but recent studies which have removed sick quitters, updated alcohol and covariate information on diet and lifestyle factors, and separately documented benefits of alcohol among healthy and unhealthy populations further add to the evidence that moderate alcohol consumption is causally related to a lower risk of CHD.” Sorensen and coworkers stated that because alcohol consumption is so common in most Western cultures, it would be exceedingly difficult to carry out a sufficient number of randomized controlled trials to evaluate adequately the effects of alcohol on CVD outcomes. They pointed out that medical decisions must often be made on the basis of observational studies.
Effects of Alcohol on Subjects Already Diagnosed with Coronary Heart Disease
Janszky and coworkers reported results from serial quantitative coronary angiography studies done among middle-aged women at 3 to 6 months and at 3 years after an acute myocardial infarction. They found that moderate alcohol consumption (more than 5 g/day) was protective of coronary atherosclerosis progression. Furthermore, restenosis after stenting has been found to be lower in patients with CHD who continue to consume alcohol. The risk of complications (including death) after an acute myocardial infarction has been shown to be lower for subjects who remain drinkers than for those who abstain. Such a protective effect seems to be lost when the subject consumes alcohol as binge-drinking.
Biologic Mechanisms for Cardiovascular Effects
Box 18-1 lists some of the key mechanisms by which moderate alcohol consumption, as well as the intake of polyphenols in red wine and other beverages, may lead to a reduction in the risk of CHD and other types of CVD.
Improvement of blood lipids (marked increase in HDL, slight decrease in LDL)
Improvement of coagulation, fibrinolysis
Improvement of endothelial function
Activation of genes for fibrinolysis, eNOS
Improvement of ventricular function
Reduction of inflammation
Improvement of glucose metabolism
Adverse effects on blood pressure, especially for heavier drinking
The first recognized mechanism by which alcohol may prevent CVD relates to induced changes in plasma lipid profile, especially increases in high-density lipoprotein cholesterol (HDL-C) and its subtypes (HDL 2 , HDL 3 ). Numerous observational and experimental studies have confirmed a strong, direct association between alcohol intake and HDL. Although it was initially believed that only certain HDL subtypes were increased by alcohol (and that those had little effect on CVD risk), alcohol intake has been shown to increase both HDL 2 and HDL 3 , and both play a role in CVD prevention. In addition, most studies also show slightly lower LDL concentrations, especially small-particle LDL, and non-HDL lipoprotein levels among moderate drinkers. Oxidation of lipids has been shown to be decreased by alcohol or polyphenols in many studies but not in all.
Platelet Aggregation and Thrombosis
As described by Booyse and colleagues, other changes in vascular, myocardial, hemostatic, and endothelial cell functions may be equally important in collectively contributing to the reduction of the risk of CVD. These factors appear to result from alcohol, from polyphenols present in some beverages (especially wine), or from a combination of both, resulting in reduced thrombosis.
The early studies of Serge Renaud and his colleagues pointed out the importance of platelet function in the development of atherosclerosis, thrombosis, and CHD. In an intervention study in humans, Zhang and coworkers showed that alcohol, at physiologically relevant doses, has a dose-dependent inhibitory effect on platelet aggregation. These authors concluded that their findings are consistent with the view that alcohol reduces platelet sensitivity to thrombotic stimuli by inhibition of arachidonic acid release and therefore subsequent thromboxane synthesis. Inhibition of platelet aggregation or function has also been shown by others. Ruf suggested that wine has a greater effect on platelet function than alcohol alone does, stating that the polyphenols in wine provide further reduction in prostanoid synthesis from arachidonate and also decrease platelet activity mediated by nitric oxide. Demrow and colleagues demonstrated that polyphenols from both red wine and grape juice favorably affect platelet aggregation.
Many of the effects of alcohol and polyphenols on coagulation are transient effects, and the beneficial effects may last only 24 to 36 hours after someone has consumed alcohol. Furthermore, after heavy alcohol intake, there may be a rebound with increased platelet aggregation if further alcohol is not consumed, although such rebound may be less after drinking wine than after consuming other beverages. One explanation for the much lower CHD rates in France than in most other countries is that the French have traditionally consumed some alcohol (primarily red wine) with their evening meal every day. Thus, their clotting mechanisms remain in a favorable state all of the time. Unfortunately, many Americans and northern Europeans tend to drink only on the weekend, often consuming a large number of drinks rapidly, a very unhealthy way to drink.
Additional mechanisms for protection of alcohol and polyphenols against CHD and myocardial infarction include decreased myocardial ischemia-reperfusion injury, increased endothelial cell–dependent vasorelaxation, simultaneous activation of endothelial cell antiapoptotic and proapoptotic pathways, decreased plasma levels of factor VII and fibrinogen, increased fibrinolysis and upregulation of fibrinolytic protein gene transcription in cultured human endothelial cells, and increased levels of atrial natriuretic peptide.
There is also strong experimental evidence for additional mechanisms by which alcohol and wine polyphenols affect the initiation or progression of atherosclerotic lesions. These include reduced LDL oxidation and aggregation, foam cell formation, and lesion progression ; inhibition of endothelin 1 synthesis ; downregulation of tissue factor gene transcription in cultured human endothelial cells and monocytes ; and inhibition of smooth muscle cell proliferation. Research studies increasingly show a strong protective effect from alcohol and polyphenols against various indices of inflammation. Wang and colleagues have shown a U-shaped association between alcohol and high-sensitivity C-reactive protein (hsCRP) levels; a proportional odds model analysis showed an odds ratio for increased hsCRP of 0.32 (95% CI, 0.14-0.74) for consumers of 20 to 70 g/day (about to almost 6 drinks) in comparison with nondrinkers. Alpert and coworkers found that CRP was lowest in consumers of 5 to 7 drinks/week.
Booyse and coworkers have demonstrated the sequence of molecular events by which endothelial cell expression of tissue plasminogen activator is increased through common activation of p38 MAPK signaling by alcohol as well as by certain wine polyphenols. Gorelik and associates have shown that red wine polyphenols lower levels of postprandial cytotoxic lipid peroxidation products (malondialdehyde) in humans, another possible mechanism for the benefits of the consumption of wine on risk of disease.
In some epidemiologic studies, the combined beneficial effect of alcohol on HDL-C, fibrinogen, and hemoglobin A1c (the last a marker of glycemic control and insulin resistance) can almost entirely explain the reduced CHD risk among moderate drinkers.
One cardiovascular risk factor that is generally not associated favorably with alcohol consumption is hypertension, and most studies show an elevation of blood pressure with increasing alcohol intake. Some studies show that the effect is mainly from heavier drinking and that light alcohol intake has no effect or even a slight lowering of blood pressure. For example, Klatsky and associates found increases in blood pressure only among subjects reporting three or more drinks per day, a finding similar to the association in all groups except African American men in an ARIC report. In the Luebeck Blood Pressure Study in Germany, Keil and associates reported increases in blood pressure for men consuming more than 40 g of alcohol per day (more than three typical drinks) or women reporting more than 20 g/day. Klatsky and associates have shown that hypertension remains a strong risk factor for CVD regardless of alcohol intake, and hypertensives who are heavy drinkers show a substantial decrease in blood pressure when they decrease their intake. Whereas a direct association between alcohol and blood pressure has been confirmed in most studies, important sequelae of hypertension (notably CHD and ischemic stroke) show an inverse association with moderate drinking.
Genetic Risk Factors
There is no question that individuals differ markedly in their response to alcohol consumption, for its beneficial and adverse health effects. Davey Smith has described how “mendelian randomization—the random assortment of genes from parents to offspring that occurs during gamete formation and conception—provides one method for assessing the causal nature of some environmental exposures.” Such an approach avoids many of the problems of confounding in observational studies. He concluded that “mendelian randomization provides new opportunities to test causality and demonstrates how investment in the human genome project may contribute to understanding and preventing the adverse effects on human health of modifiable exposures.”
At present, however, the association of genes with CHD is poorly understood. The most studied individual genes are those that relate to the metabolism of alcohol, especially those that control alcohol dehydrogenase (ADH) and acetaldehyde dehydrogenase (ALDH). Hines and coworkers showed that moderate drinkers who are homozygous for the slow-oxidizing ADH3 allele had higher HDL levels and a substantially decreased risk of myocardial infarction. However, other studies have failed to show such an association. Jensen and coworkers have found conflicting results across different cohorts in their study relating lipoprotein lipase to CHD.
Current data suggest that there are many genes contributing to the effects of alcohol on the risk of CHD; each tends to play only a small role by itself but a potentially large role in combination with other genes and environmental factors. Furthermore, certain genes seem much more important in certain ethnic groups and cultures than in others. The large number of collaborative genome-wide association studies now being done, with huge numbers of subjects, should help clarify the role that genetic factors play in the potential protection against cardiovascular and other diseases from alcohol use.
Similarly, for cancer, data on genetic factors modifying alcohol’s effects are unclear. Druesne-Pecollo and coworkers published an overview of studies on the combined effects of alcohol drinking and polymorphisms in genes for ADH, ALDH, cytochrome P-450 2E1, and methylenetetrahydrofolate reductase on the risk of alcohol-related cancer. They concluded that current data lend support to a role of polymorphisms ADH1B and ALDH2 combined with alcohol consumption in cancer, but other available data are insufficient or inconclusive.
Summary of Mechanisms
Collins and associates have provided a summary of epidemiologic and mechanistic studies demonstrating how alcohol and polyphenols may be cardioprotective (as well as neuroprotective). In Figure 18-2 , mechanisms of cardioprotection by both alcohol and resveratrol, one of the polyphenols in red wine that has been studied extensively, are illustrated.
Clinical trials in humans are limited, but Leighton and colleagues have carried out a number of intervention studies looking at the effects of wine on CVD risk factors. Their results show that moderate wine consumption may result in an increase in HDL-C, a decrease in the omega-6/omega-3 ratio, and in some cases a slight increase in triglyceride levels. Observed changes in hemostasis include reduced coagulation and increased fibrinolysis; effects on blood pressure have been inconsistent. They have found a reduction in inflammatory markers and an improvement in endothelial function. These investigators concluded that limitation of consumption to below 30 g of alcohol daily (about typical U.S. drinks) for men and half that dose for women is necessary to prevent negative changes in cardiovascular risk factors such as hypertension, hypertriglyceridemia, and hyperhomocysteinemia. Other clinical trials have demonstrated improvements in glucose metabolism from the administration of alcohol. In the future, more such trials as well as genetic studies will help determine the extent to which the beneficial effects of alcohol consumption on CVD risk seen in most observational studies are caused by the alcoholic beverage itself.
Effects of Alcohol on Other Types of Cardiovascular Disease
The third leading cause of death in the United States, and the leading cause of disability, is stroke. It has been shown in many studies that the risk of stroke related to ischemia (atherosclerosis), which is the type of stroke in about 80% of cases in the United States, is reduced by moderate drinking. Moreover, Klatsky and associates found that an increase in risk of hemorrhagic stroke from alcohol is seen only in subjects consuming five or more drinks per day. Others have shown that the risk of peripheral artery disease is reduced among moderate drinkers.
For two other cardiovascular conditions, atrial fibrillation and heart failure, recent scientific evidence suggests a lack of harmful effects from moderate alcohol consumption. Ettinger and colleagues described the “holiday heart syndrome” in 1978. This syndrome often includes atrial fibrillation; the syndrome is usually not associated with long-standing heart disease and tends to resolve when alcohol consumption is ceased. Nissen and Lemberg stated that even moderate drinking can lead to this syndrome, but others have found no effect on the risk of atrial fibrillation for moderate alcohol intake, only for heavy drinking. A recent extensive review of experimental, clinical, and epidemiologic data did not find evidence that alcohol is a factor, certainly not a major factor, in the development of atrial fibrillation.
It is known that excessive alcohol intake can result in alcoholic cardiomyopathy, causing heart failure. It was previously believed that heart failure, in general, would be made worse by any alcohol consumption. As reviewed by Djousse and Gaziano, however, most studies have shown that moderate drinkers are at lower risk for development of heart failure. For example, in a population of more than 100,000 subjects, Klatsky and associates found that alcohol drinking was inversely related to risk of heart failure related to CHD (e.g., at one or two drinks per day; RR, 0.6; 95% CI, 0.5-0.7), with consistency across subgroups of age, gender, ethnicity, education, smoking status, interval to diagnosis, and presence or absence of baseline heart disease or systemic hypertension. For heart failure not associated with coronary disease, moderate drinking was inversely related only in subjects who had diabetes mellitus.
For patients who already have depressed myocardial function, Cooper and associates showed in the Studies of Left Ventricular Dysfunction (SOLVD) that moderate drinkers had lower subsequent all-cause mortality than abstainers did. In contrast, in the Survival And Ventricular Enlargement (SAVE) trial in subjects after myocardial infarction, Aguilar and coworkers did not show a significant effect of moderate drinking on the development of heart failure or on survival.
Effects of Alcohol on Diabetes and Metabolic Syndrome
A major risk factor for the development of CVD not discussed before relates to glucose metabolism, especially the development of diabetes or the metabolic syndrome. Not only do data support lower risk for development of these conditions, but for persons with diabetes or the metabolic syndrome, moderate drinking may lower their subsequent risk of CVD.
Development of Diabetes
Stampfer and colleagues reported in 1988 that moderate drinkers in the Nurses’ Health Study had a lower risk for development of diabetes than did abstainers; compared with nondrinkers, women consuming 5 to 14.9 g of alcohol per day had an age-adjusted relative risk of diabetes of 0.4 (95% CI, 0.3-0.6); for 15 g or more per day, the relative risk was 0.3 (95% CI, 0.2-0.4). The authors reported that a strong inverse association between alcohol drinking and body weight explained much of the apparent protective effect of alcohol. After simultaneous adjustment for Quetelet index (weight [kg]/height [m] 2 ), family history of diabetes, total calorie intake, and age, the relative risk of diabetes for consumers of 5 to 14.9 g/day was 0.8 (95% CI, 0.6-1.2), and for women who drank 15+ g/day, the relative risk was 0.6 (95% CI, 0.3-0.9). Reports in the British Medical Journal in 1995 from two large prospective studies similarly showed a lower risk of diabetes among moderate drinkers. Since then, many epidemiologic studies have presented data supporting such an inverse association. Howard and coworkers have provided a good summary of the research on alcohol and diabetes, giving an estimate of 33% to 56% lower incidence of diabetes for consumers of one to three drinks per day. A meta-analysis by Koppes and associates indicated that for a wide range of alcohol intake (from about to more than 3 drinks/day), the relative risk of diabetes for drinkers is about 30% lower than it is for abstainers.
Mechanisms are unclear for the effects of alcohol intake on diabetes. Some studies have shown lower glucose levels and HbA1c. In an intervention trial, Davies and coworkers found among nondiabetic postmenopausal women that the consumption of 30 g/day of alcohol reduced fasting insulin concentration by 19.2%, reduced triglyceride concentration by 10.3%, and increased insulin sensitivity by 7.2% but did not affect plasma glucose levels. Kroenke and associates found that HbA1c was inversely associated with alcohol intake; among overweight women, there was also an inverse association with insulin. Specifically, these investigators found that insulin levels were lowest for drinkers consuming no more than two drinks per day up to three days per week. In a large heterogeneous group of nondiabetic subjects, insulin sensitivity was lower among abstainers than in all categories of drinkers. Adiponectin serum levels were higher in men consuming alcohol on 2+ days/week than in nondrinkers or occasional drinkers; among women, those consuming all levels of alcohol had higher adiponectin concentrations than nondrinkers did.
There have also been a number of clinical trials relating alcohol intake to diabetic mechanisms. Sierksma and coworkers carried out a randomized crossover study of 23 healthy middle-aged men who were given 40 g of ethanol (four glasses of whiskey, versus no alcohol) on a daily basis for 17 days. They found that alcohol increased plasma adiponectin levels by 11% and increased the insulin sensitivity index in an insulin-resistant subgroup by 21%. A randomized crossover trial by Joosten and coworkers showed that moderate alcohol consumption increased insulin sensitivity and ADIPOQ expression in postmenopausal women.
Shai and colleagues carried out a randomized trial among 91 formerly abstaining diabetics who were randomly assigned to consume 5 ounces of either sauvignon blanc or merlot at dinner every evening or advised to continue to avoid alcohol. Significantly lower levels of fasting blood glucose concentration during a 3-month intervention period were seen among subjects consuming wine than were seen for abstainers. The effect of the wine was variable, with beneficial effects primarily among those with more severe disease.
Effects of Alcohol Intake on Cardiovascular Disease Among Diabetics
Valmadrid and coworkers reported that among persons with older-onset diabetes, the relative risk for death from CHD was 0.44 for subjects consuming 2 to 13 g/day and 0.21 for those consuming 14+ g/day, in comparison with nondrinkers. Solomon and colleagues reported from the Nurses’ Health Study that the risk for development of CHD among type 2 diabetics who were moderate drinkers was much lower than that for nondrinking diabetic subjects: for < drink daily, the relative risk was 0.72, and for ≥ drink daily, the relative risk was 0.45. In the Physicians’ Health Study, increasing levels of alcohol intake were associated with even greater reduction in deaths from CHD among diabetics than among nondiabetic subjects.
A meta-analysis by Koppes and associates indicates that the risk of CHD is 34% to 55% lower among diabetics who are moderate drinkers than in those who consume no alcohol. Among more than 38,000 diabetics observed by the Kaiser Permanente group in California, those who consumed alcohol had evidence of considerably better control of their diabetes than did nondrinkers ; the authors concluded that this finding “supports current clinical guidelines for moderate levels of alcohol consumption among diabetes patients.”
Alcohol has been shown to affect not only the macrovascular sequelae of diabetes but also the microvascular complications of diabetes. In a report from the EURODIAB Prospective Complications Study, involving the follow-up of 3250 type 1 diabetic patients from 16 different European countries, Beulens and colleagues found a lower occurrence of retinopathy, neuropathy, and nephropathy among those individuals who consumed alcohol moderately in comparison with nondrinkers. The association was strongest among wine drinkers, and to some extent among beer drinkers, but was not seen among those consuming spirits.
In summary, evidence from a large number of prospective epidemiologic studies and limited clinical trials strongly suggests that moderate drinkers are less likely to develop diabetes and its sequelae, especially CHD. Although not all of the mechanisms are known, many studies suggest improved insulin resistance or increased adiponectin levels.
Development of Metabolic Syndrome
The findings relating alcohol to metabolic syndrome show associations with alcohol similar to those for diabetes. In the NHLBI Family Heart Study and in NHANES III, we found evidence that moderate drinking is associated with lower risk of most components of the metabolic syndrome (all components except for hypertension). Fan and coworkers evaluated the association between estimates of lifetime alcohol intake and the metabolic syndrome. They found that subjects who averaged more than one drink per day for women or two drinks per day for men had an increase in their prevalence ratio for metabolic syndrome in comparison with drinkers who did not exceed these limits; nondrinkers were not included in these analyses. In a meta-analysis on alcohol intake and metabolic syndrome, based on data from seven previous studies with a total of 22,000 subjects, Alkerwi and coworkers found that the moderate intake of alcohol (defined as ≤40 g of alcohol per day for men and ≤20 g of alcohol per day for women) was associated with 16% lower risk of metabolic syndrome for men and 25% lower risk for women; no significant effects were seen for heavier drinking.
Alcohol as a Component of a Healthy Lifestyle
Alcohol consumption, moderate or otherwise, should not be viewed in isolation but as part of broader social, cultural, and lifestyle issues. We now have good scientific data on which to base our definition of what constitutes a “healthy” diet and lifestyle. The Nurses’ Health Study, the Health Professionals Study, and other research have defined a lifestyle that will lead to >70% fewer myocardial infarctions and >90% fewer cases of diabetes. This healthy lifestyle is shown in Box 18-2 , based on research by Stampfer, Hu, Mukamal, and others.
Avoid obesity (keep the BMI <25).
Consume a healthy diet, high in fiber and unsaturated fat and low in trans -fat and glycemic load (e.g., a Mediterranean-type diet).
Engage in moderate to vigorous physical activity (for at least hour/day).
Unless contraindicated, consume a small amount of an alcoholic beverage regularly.
Akesson and colleagues, in a large prospective study among middle-aged and older women in Sweden, found that those subjects who met all five components of a healthy lifestyle (defined as following what can be described as a Mediterranean-type diet, not smoking, not being obese, getting regular exercise, and consuming at least 5 g/day of alcohol) had a dramatically reduced risk of having a myocardial infarction. The authors suggested that if all women in their population had such a lifestyle, there would have been 77% fewer cases of CHD.
Khaw and associates, in a large prospective observational study from the United Kingdom, assessed the effects on mortality of four “healthy behaviors”: not smoking, being active, having evidence of a high fruit and vegetable intake, and consuming some alcohol but not more than 14 drinks per week (about typical drinks by U.S. standards). Increasing numbers of these behaviors were associated with lower risk of death from all causes as well as from CVD, cancer, and noncardiovascular causes. The authors calculated that the effect of these four healthy behaviors on total mortality, compared with none of them, is equivalent to a 14-year age difference in mortality risk.
Whereas some suggest that we should focus on the first four components of the healthy lifestyle and not be eager to encourage alcohol use, it has been shown that even among very healthy subjects (i.e., who are lean, eat a healthy diet, are active, and do not smoke), those who also consume a little alcohol have much lower risk of heart disease and death. To address the issue of residual confounding by healthy lifestyle among drinkers, Mukamal and coworkers restricted analyses to 8867 healthy men in the Health Professionals Study who adhered to the first four low-risk behaviors (not including alcohol) and examined the association for alcohol consumption with CHD. In this group of “very healthy” men, there were still 106 incident CHD cases; the men who drank moderately (15.0 to 29.9 g/day) had a relative risk of 0.38 (95% CI, 0.16-0.89) compared with abstainers. This strong inverse association between moderate alcohol consumption and CHD in predominantly healthy individuals adds further evidence to support the hypothesis that the inverse association is causal and not confounded by healthy lifestyle behaviors. On the other hand, a paper with a restrictive definition of “healthy” reported no significant effects of alcohol on risk of CHD among the healthiest subjects.
Effects of Alcohol on Noncardiovascular Diseases
Before the encouragement of moderate drinking for the possible reduction in the risk of CVD is even considered, the potential effects on other diseases must be taken into account. In addition to CVD, it has long been known that moderate drinkers are at lower risk of gallstones. There are interesting new data relating alcohol intake to obesity, bone mineral density, and dementia. As will be discussed, an increase in the risk of breast cancer from alcohol is of great concern for women, but recent data indicate a lower risk from moderate alcohol intake for certain other cancers.
An unexpected finding in epidemiologic studies is that obesity is often found to be less common among moderate drinkers than among abstainers. In the prospective Nurses’ Health Study, it was found that light to moderate drinking was not associated with weight gain in women (overall, a 16% lower risk for those reporting 15.0 to 29.9 g of alcohol per day versus abstainers). This potentially beneficial effect on weight gain was not seen in African American women or in heavier drinkers. Arif and Rohrer evaluated alcohol intake and BMI among 8236 nonsmoking respondents who participated in the Third National Health and Nutrition Examination Survey. Current drinkers who reported drinking one drink or two drinks per day had 0.46 (95% CI, 0.34-0.62) and 0.59 (95% CI, 0.41-0.86) the odds of obesity, respectively, than did abstainers.
Tolstrup and colleagues found that among men, the odds ratios for having a high BMI among subjects drinking 1 to 3 days/month, 1 day/week, 2 to 4 days/week, 5 or 6 days/week, and 7 days/week were 1.39 (95% CI, 1.36-1.64), 1.17 (1.02-1.34), 1.00 (reference), 0.87 (0.77-0.98), and 0.73 (0.65-0.82), respectively. Similar associations were found for waist circumference, and corresponding results were found for women. These authors concluded that for a given level of total alcohol intake, obesity was inversely associated with drinking frequency, whereas the amount of alcohol intake was positively associated with obesity. Such an association was supported by Breslow and Smothers among 45,896 never-smoking adults in the 1997-2001 National Health Interview Surveys in the United States. As shown in Figure 18-3 , from that study, more frequent drinking was associated with lower BMI. On the other hand, given drinking of a certain frequency, the number of drinks per drinking day was directly associated with BMI.
These results suggest that the frequent consumption of small amounts of alcohol is the optimal drinking pattern associated with a lower risk of obesity.
Bone Mineral Density and Hip Fracture
A number of epidemiologic studies have shown that moderate drinkers have less osteoporosis and a lower risk of hip fractures than do abstainers. (This goes against “conventional wisdom,” which has always assumed that elderly who drink alcohol may be more unsteady on their feet and at increased risk of falling and sustaining fractures.) Mukamal and colleagues found that among older adults, moderate alcohol consumption had a U-shaped relationship with risk of hip fracture but a graded positive relationship with bone mineral density at the hip. A meta-analysis by Berg and coworkers involving the review of 33 previous studies supports a reduced risk of hip fracture and higher bone density among men and women who consume small to moderate amounts of alcohol (in comparison with nondrinkers). There is probably a J-shaped curve, with increased risk of fractures among heavier drinkers.
Among the most exciting recent scientific findings are that moderate drinkers tend to be less likely to develop Alzheimer disease and other types of dementia. This was reported for wine drinkers in Bordeaux by Orgogozo and colleagues in 1997, and Truelsen and coworkers reported similar protection from wine in Denmark, as did Panza and coworkers in a summary paper.
Britton and coworkers reported from the Whitehall Study in the United Kingdom that there was less cognitive impairment in drinkers for tests of memory (borderline significant), verbal and mathematical reasoning, verbal meaning, and verbal fluency. For men, the risk of dysfunction was lower by about 40% to 50% for drinkers, with the lowest risk at >241 g/week of alcohol, the equivalent of about three drinks per day by U.S. standards. Women drinkers had lower point estimates for most measures, but statistically significant results were seen for verbal and mathematical reasoning (OR = 0.3; CI, 0.2 -0.6) and for verbal fluency (OR = 0.5; CI, 0.3-0.9), both at 49 to 80 g of alcohol per week (the equivalent of about four to seven drinks per week). Lifetime abstaining men and women generally had much greater risk of cognitive dysfunction than did occasional drinkers.
Espeland and colleagues reported from the Women’s Health Initiative Memory Study that compared with no intake, intake of one drink or more per day was associated with higher baseline Modified Mini-Mental State Examination scores ( P < 0.001) and an odds ratio of 0.41 (95% CI, 0.23-0.74) for significant declines in cognitive function. Stampfer and coworkers reported that among women who were moderate drinkers, compared with nondrinkers, the relative risk of impairment was 0.77 for general cognition (95% CI, 0.67-0.88) and 0.81 on the global cognitive score (95% CI, 0.70-0.93). Zhang and coworkers reported that education modifies the putative effects of alcohol on memory, with better results among those with higher education, probably because of more moderate drinking patterns.
Ganguli and colleagues studied men and women aged 74.4 years at baseline who were observed during 7 years with repeated assessments of cognitive functioning; abstainers showed lower baseline scores and a tendency for more rapid deterioration in cognitive functioning than did drinkers. Solfrizzi and colleagues also found less progression from mild cognitive impairment to dementia for light drinkers in comparison with nondrinkers.
Mehlig and coworkers, in a well-done analysis from a long-term prospective study of women in Sweden, reported that the lifetime risk of dementia was associated differently with wine consumption (a decrease in risk of dementia of 70%) and spirits consumption (an increase in dementia risk of about 50%). The authors concluded that the different associations by type of beverage suggest that the nonalcohol components in wine may be an important factor in lowering the risk of dementia. Whereas the investigators adjusted for the usual risk factors for dementia, including education and social class, there is always the possibility that other lifestyle factors that are different between wine drinkers and spirits drinkers may have influenced the results.
Peters and associates carried out a meta-analysis on alcohol and dementia and concluded that moderate drinkers, especially wine drinkers, were at lower risk than abstainers for development of Alzheimer disease or other dementia. The mechanism for such protection is not known but may relate, among other factors, to prevention of cerebral atherosclerosis or decreased inflammation in brain tissue. Collins and associates have reported additional mechanisms for neuroprotection, as described earlier (under mechanisms of alcohol’s effects on CVD).