By Christopher Griffith & Douglas Bogart, MD
Alcohol has been reported to be beneficial in reducing the risk and complications of cardiovascular disease (CVD). This benefit must be balanced with the known risks of alcohol. This review examines the evidence that there is a favorable effect on CVD and attempts to evaluate if the positive effects outweigh the potential pernicious sequellae.
The Centers for Disease Control and Prevention have reported that there were 75,766 deaths attributable to alcohol in 2001 in the United States. Each death results in 30 years of lost life expectancy (2.3 million years).1 In 2005 there were 1.6 million hospitalizations2 and four million emergency department visits3 due to alcohol related problems. The definitions of alcohol use are listed in Table 1.
Most studies relate alcohol consumption and health effects on a J-shaped curve. (See Figure 1.) Abstainers have an increased baseline risk, moderate drinkers have the lowest risk, and heavy drinkers have the highest risk.
The average amount of alcohol in a drink is approximately 14 grams. Table 2 lists the equivalency of the various types of alcohol beverages.
The mechanism through which alcohol acts favorably on the cardiovascular system is not entirely understood. However, there are several proposed theories. Studies have shown that regular moderate alcohol consumption increases HDL, increases fibrinolysis, decreases platelet aggregation, improves endothelial function4 and may decrease insulin resistance.5 A recent Norwegian study has suggested that HDL cholesterol is not a major factor in the observed beneficial effect of alcohol on coronary heart disease.6 Other potential benefits include the reduction in inflammatory markers. Moderate alcohol consumption has been shown to decrease C-reactive protein, tumor necrosis factor-alpha and interleukin-6.7, 8 Two excellent reviews on alcohol use and cardiovascular mortality have been published.9, 10
The beneficial effects of alcohol are not only evident in cardiovascular disease, but also in reducing mortality from all causes. A large meta-analysis by Di Castelnuovo et al. consisting of over one million men and women found a total mortality reduction of 18% in women and 17% in men with moderate consumption.11 They found that up to four drinks/day in men and two drinks/day in women conferred a mortality benefit. The studies included in the meta-analysis had follow-up on the patients from a range of 5.5 to 23 years and most were over 10 years. They did find a J-shaped relationship in that higher alcohol consumption was associated with increased mortality.
The Physicians Health Study observed 89,000 men over a five-year period. They evaluated a wide range of alcohol consumption from < 1 drink per week to ≥ 2 drinks/day. They found a U-shaped relationship between alcohol intake and death. The results showed a risk reduction (RR) for all cause mortality of 0.78 with five to six drinks/week compared with a RR of 0.95 with intake ≥ 2 drinks/day. In this study there was a benefit with reduced CVD mortality even with consumption of ≥ 2 drinks/day.12 The risk of cancer and other mortality increased with one drink or greater per day. In summary they found a U-shaped relationship with the reduction in CVD death at lower levels of consumption and a rise in other causes of death as the consumption increased above one drink per day.
Thus both of these studies found that there was a beneficial effect from alcohol in moderation on both all cause mortality as well as CVD mortality. A J- or U-shaped relationship was present suggesting a zone of benefit from alcohol consumption in moderation but a hazard when taken in excess.
Thun et al. reported on a nine-year study involving 490,000 men and women who self reported their alcohol use at baseline. They found a RR of 0.6 in women who drank at least one drink/day compared to nondrinkers. The RR in men was 0.7. The women in this study who drank more than one drink per day had a 30% higher breast cancer mortality compared to nondrinkers. Mortality from all causes increased with more alcohol consumption and this was particularly true in patients with the lowest cardiovascular risk.13
Mukumal et al. reported on 245,207 participants who completed the National Health Interview Survey on their alcohol intake and found that light drinkers (≤ 3 drinks/week) had a 0.69 RR of cardiovascular mortality. Moderate drinkers (> 3-7 drinks/week in women and > 3-14 drinks/week for men) were found to have a 0.62 RR. The lowered risk existed regardless of gender, age or health status.14 One hundred twenty-eight thousand, nine hundred and thirty-four adults who were in a comprehensive health care program were evaluated for the effect of alcohol consumption and risk of death. Drinking one to two drinks/day resulted in a 0.7 RR of death from coronary artery disease. Greater than six drinks/day resulted in greater risk for non cardiovascular death (cirrhosis, smoking related and unnatural death).15
Costanzo and Castelnuovo performed a meta-analysis of prospective studies in men and women with cardiovascular disease (16,351 patients). The analysis looked at what daily dose of alcohol gave the best protection from cardiovascular mortality and all cause mortality. They also found a J-shaped relationship between events and alcohol consumed. Patients who drank two drinks per day had the maximum benefit at a 22% RR in cardiovascular mortality. The maximum benefit in all cause mortality was an 18% RR in patients who drank one drink per day.16
In summary, four large studies found a reduction in CVD mortality in persons with and without known cardiovascular disease at the onset of the observation period. The magnitude of the benefit was the highest in those at greatest risk for cardiovascular disease.