Date Published: , 2017
Publisher: National Institute on Alcohol Abuse and Alcoholism
Author(s): Mariann R. Piano.
Alcohol use has complex effects on cardiovascular (CV) health. The associations between drinking and CV diseases such as hypertension, coronary heart disease, stroke, peripheral arterial disease, and cardiomyopathy have been studied extensively and are outlined in this review. Although many behavioral, genetic, and biologic variants influence the interconnection between alcohol use and CV disease, dose and pattern of alcohol consumption seem to modulate this most. Low-to-moderate alcohol use may mitigate certain mechanisms such as risk and hemostatic factors affecting atherosclerosis and inflammation, pathophysiologic processes integral to most CV disease. But any positive aspects of drinking must be weighed against serious physiological effects, including mitochondrial dysfunction and changes in circulation, inflammatory response, oxidative stress, and programmed cell death, as well as anatomical damage to the CV system, especially the heart itself. Both the negative and positive effects of alcohol use on particular CV conditions are presented here. The review concludes by suggesting several promising avenues for future research related to alcohol use and CV disease. These include using direct biomarkers of alcohol to confirm self-report of alcohol consumption levels; studying potential mediation of various genetic, socioeconomic, and racial and ethnic factors that may affect alcohol use and CV disease; reviewing alcohol–medication interactions in cardiac patients; and examining CV effects of alcohol use in young adults and in older adults.
There are certain factors that are critically important to understanding and interpreting the data related to the health consequences of alcohol consumption. For example, how was alcohol consumption measured? What were the drink sizes and alcohol concentrations? How often did the subjects drink alcohol? What was the pattern of drinking? And was the study prospective (following subjects over time) or aggregate (pulling together data from several different studies to look for common trends)?
In healthy adults, consuming low-to-moderate amounts of alcohol each day typically has no short-term (i.e., acute) or substantial impact on hemodynamics or blood pressure (BP). However, data suggest that binge drinking (more than 5 standard drinks in a single sitting) is associated with transient increases in BP that range from 4 to 7 mmHg for systolic BP and 4 to 6 mmHg for diastolic BP (Potter et al. 1986; Rosito et al. 1999; Seppä and Sillanaukee 1999).
Several mechanisms may underlie alcohol’s effects on blood pressure. These include impairments in cells that lead to buildup of plaque in arteries (i.e., through alterations in endothelial cell function and nitric oxide availability), and disruptions in arterial-vascular function (i.e., through myogenic mechanisms and changes in baroreceptor function), and hormonal imbalances that control the body’s fluid and BP regulation (through the renin–angiotensin–aldosterone system [RAAS]). Some adverse BP-related mechanisms that may be triggered by alcohol include changes in intracellular calcium levels, baroreflex control, and heart rate and activation of other neurohormonal systems besides the RAAS, such as the sympathetic nervous system (Marchi et al. 2014).
The relationship between and among alcohol consumption, CHD, and stroke has been extensively investigated. Many of these studies have been conducted in middle-aged and older people and across populations. These studies have used ecologic designs (in which at least one risk-modifying health factor is measured at the group level); case–control designs (which compare clinical cases with control subjects to determine if an exposure is associated with an outcome); and longitudinal research designs (which gather data from the same subjects repeatedly over a length of time). Examples include the Health Professionals Follow-up Study (Harvard School of Public Health 2016a), the Nurses’ Health Study (Harvard School of Public Health 2016b), the Framingham Heart Study (National Heart, Lung, and Blood Institute and Boston University 2016), the British Doctors Study, the Physicians’ Health Study,2 the Copenhagen City Heart Study (Schnohr et al. 2002), and INTERHEART (Leong et al. 2014). The availability of these diverse datasets has allowed for completion of several comprehensive systematic reviews and meta-analyses of alcohol, CHD, and stroke relationships. This section summarizes data from meta-analyses, along with data from large international studies such as INTERHEART (Leong et al. 2014) and other recent studies using new methodologies such as Mendelian randomization (reviewed below in “Alcohol Consumption and Total Stroke Incidence and Prevalence”).
In a comprehensive systemic review and meta-analysis, Ronksley and colleagues (2011) analyzed data from several prospective studies (n = 84), of which 40 percent reported on all-male cohorts, 7 percent reported on women only, and 53 percent included men and women. The most data-adjusted analysis, which included both men and women, noted that various alcohol consumption levels (g/day) among active drinkers compared with nondrinkers were associated with a reduced relative risk for CV mortality, incident CHD, and CHD mortality. Alcohol consumption levels between 2.5 g/day and 30 to 60 g/day (<1 standard drink/day to ~5 drinks/day) were cardioprotective for both CV mortality and CHD mortality (figure 2). However, the association between alcohol consumption and CV mortality was insignificant when alcohol consumption was >60 g/day, but remained significantly reduced for CHD mortality.
Many epidemiologic studies also have been conducted to evaluate the association between alcohol consumption and total stroke incidence and prevalence, as well as the separate effects on specific stroke subtypes (e.g., ischemic and hemorrhagic). In the same systematic review and meta-analysis noted above, Ronksley and colleagues (2011) systematically examined the relationships between and among different levels of alcohol consumption and incident stroke and stroke mortality. They found a decrease or no effect on relative risk for incident stroke and stroke mortality, respectively, at <2.5 g and 2.5 to 14.9 g of alcohol/day, and almost no overall associations of alcohol consumption with levels between 15 to 29.9 g and 30 to 60 g of alcohol/day (figure 2). For heavier drinkers (60 g/day) the risk for incident stroke was greater compared with abstainers, and the risk for stroke mortality was about one and a half times greater (figure 2). A subanalysis of stroke subtypes revealed that when pooling the risk among current alcohol drinkers compared with nondrinkers, the risk was actually higher for incident hemorrhagic stroke than for ischemic stroke (n = 12 studies) (Ronksley et al. 2011). Investigators are using new methods to examine the relationship between alcohol consumption and CV outcomes. One such method includes Mendelian randomization, an epidemiologic study design that incorporates genetic information into traditional epidemiologic methods. Mendelian randomization offers an opportunity to test the relationship between a causal factor (e.g., alcohol consumption) and a specific outcome (e.g., CV disease). Holmes and colleagues (2014) used Mendelian randomization to determine if a relationship exists between drinkers with a certain variant in the alcohol dehydrogenase ADH1B gene (rs1229984), which is associated with reduced alcohol consumption, and the likelihood of having fatal or nonfatal CHD and stroke. The investigators found that individuals with the A allele variant ADH1Brs1229984 consumed less alcohol and had a reduced risk of CHD and ischemic stroke compared with noncarriers. This implies that lower alcohol consumption, even in light-to-moderate drinkers, was beneficial for CV health. These results challenge the findings from several of the studies mentioned earlier that support a cardioprotective effect of low-to-moderate alcohol consumption. They also suggest that traditional epidemiologic studies may not capture important nuances related to selection bias or other errors that can affect study results (Chikritzhs et al. 2015; Holmes et al. 2014). Based on these findings in both men and women, alcohol consumption of about 1 to 2 drinks per day is associated with a decrease in CHD. However, alcohol consumption may not have cardioprotective effects in certain racial or ethnic groups, such as in people of South Asian ethnicity living in South Asia (Leong et al. 2014). Several studies and meta-analyses have been conducted to determine the relationship between alcohol consumption and the risk of developing heart failure in healthy subjects, as well as in those with a history of MI or CHD. Heart failure is a syndrome that often results from an MI or CHD. Studies also have examined the “safety” of alcoholic beverage consumption in subjects with heart failure. Compared with CHD and stroke, the relationship between alcohol consumption and PAD has been examined less often, and to date there are no meta-analyses or systematic reviews. PAD is used broadly to refer to pathophysiologic conditions affecting the arterial system. Fifteen years ago, two large prospective studies, one from the United States (the Strong Heart Study) (Fabsitz et al. 1999) and one from Europe (the Rotterdam Study) (Vliegenthart et al. 2002), examined the effects of alcohol consumption on PAD. Both studies used the ankle-to-brachial index (<0.90), which compares BP measured at the ankle with BP measured at the upper arm, as a measure of PAD. Many of the CV conditions outlined above share the pathophysiologic process of atherosclerosis and inflammation. Therefore, alcohol may exert its protective or enhancing effects on these conditions by modifying three broad categories of mechanisms: risk factors (e.g., lipid profiles, carotid intima-medial thickness [cIMT], and insulin sensitivity), hemostatic factors (e.g., fibrinogen levels and platelet reactivity), and inflammation. In addition, and specific to CHD, alcohol consumption may modulate ischemia–reperfusion mechanisms as blood flow is restored to tissues after oxygen deprivation. Several of these potential mechanisms are briefly reviewed below. One common risk factor for CV disease is the composition of the lipids found in the blood, and the effects of alcohol consumption on lipid profiles have been extensively studied. Many researchers have found that alcohol intake increases HDL cholesterol (HDL-c) levels, HDL (“good cholesterol”) particle concentration, apolipoprotein A-I, and HDL-c subfractions (Gardner et al. 2000; Muth et al. 2010; Vu et al. 2016). Findings have been equivocal for other lipids, such as low-density lipoprotein cholesterol (LDL-c) (the estimated amount of cholesterol within LDL particles, or “bad cholesterol”) and triglyceride levels (Rimm et al. 1999; Volcik et al. 2008; Waskiewicz and Sygnowska 2013). High triglyceride levels in the blood stream have been linked to atherosclerosis and, by extension, increased risk of CHD and stroke. However, in a recently conducted Mendelian randomization study, Vu and colleagues (2016) reported that low-to-moderate alcohol consumption reduced triglyceride and LDL-c and increased HDL-c, in particular the HDL2-c subfraction. Interestingly, the researchers found a nonlinear effect of alcohol consumption on HDL2-c levels. This supports the findings from other studies that the alcohol-induced changes in HDL-c do not fully account for the lower risk of CHD in moderate alcohol drinkers (Mukamal 2012). Alcohol consumption can be associated with both a favorable hemostatis/coagulation profile as well as an adverse one (Salem and Laposata 2005). Several epidemiologic and randomized controlled studies have found alcohol consumption decreases coagulation factors such as fibrinogen, which is a CV risk marker at elevated levels (Mori et al. 2015; Rimm et al. 1999). In addition to being essential to the coagulation cascade, fibrinogen also may play a proinflammatory role in the development of certain CV diseases, including vascular wall disease and atherosclerosis (Davalos and Akassoglou 2012). Findings from a meta-analysis of 42 studies by Rimm and colleagues (1999) suggested that 30 g of alcohol/day (2 standard drinks) was associated with a 7.5 mg/dl (−17.7 to 32.7) decrease in fibrinogen concentration. Similarly, the results from the small randomized crossover trial by Mori and colleagues (2015) found that women consuming alcohol (146 to 218 g/week, ~2 to 3 standard drinks/day) for 4 weeks showed a 14 percent reduction in fibrinogen levels. The effects of alcohol consumption on inflammation are twofold. Lower doses are associated with reduced inflammation, as indicated by markers such as C-reactive protein and certain interleukins. Conversely, higher levels induce oxidative stress and a wide variety of inflammatory markers. As reviewed below, oxidative stress in particular is likely a key event in the development of alcoholic cardiomyopathy (discussed in “Acute and Long-term Effects of Alcohol on the Myocardium”). Data from numerous types of research studies show that alcohol may alter levels of antioxidant enzymes and stimulate oxidative damage, and it may therefore be involved in the pathogenesis of many types of alcohol-induced diseases (Ceni et al. 2014; Piano and Phillips 2014). Another mechanism underlying the cardioprotective effects of low-to-moderate alcohol consumption and CHD in particular may be related to a phenomenon known as ischemic preconditioning, which produces resistance to the loss of blood supply (and oxygen) to organs or tissues. If the blood supply is impaired briefly (usually for <5 minutes) and then restored so that blood flow resumes, and the process is repeated two or more times, the cells downstream of the tissue or organ are protected from a final ischemic insult, when the blood supply is cut off. In the heart, this would protect the heart muscle (myocardium) from subsequent, more prolonged episodes of restricted blood flow (ischemia) followed by injury when that blood flow returns to the heart (called reperfusion injury or ischemia–reperfusion injury; Veighey and Macallister 2012). Ischemic preconditioning results in smaller infarct sizes, fewer and less severe arrhythmias, and prevention of endothelial cell dysfunction (Veighey and Macallister 2012). During the ischemic phase, the flow of oxygen and nutrients to the tissues is reduced, most significantly to the heart, brain, and kidneys. In contrast, during the reperfusion phase, despite restoration of blood flow, a series of dysfunctional biochemical and metabolic changes are initiated that lead to extensive accumulation of ROS. ROS induce a number of changes. One is the opening of the mitochondrial permeability transition pore, which is formed in the mitochondria during ischemic incidents, contributing to reperfusion injury and cell death. Others include recruitment of neutrophils (white blood cells that are among the first inflammatory cells to respond during inflammation) and dysfunction of the sarcoplasmic reticulum, which can affect calcium ion storage and release into muscle fibers. Drinking patterns, and in particular a binge pattern of drinking and higher frequency of binge drinking, are associated with a heightened risk of CV conditions such as HTN, stroke, and MI, as well as sudden death or increased mortality after MI (Leong et al. 2014; Marques-Vidal et al. 2001; Mukamal et al. 2005; Sundell et al. 2008; Wannamethee and Shaper 1992). In a systematic review (n = 37 studies), Feigin and colleagues (2005) found that excessive ethanol intake (>150 g ethanol/week) was associated with a doubled risk of subarachnoid hemorrhage. The latter findings may relate to the overall large quantity of alcohol consumed (~12 standard drinks/week) rather than a binge pattern.
Apoptosis also may be an important mechanism in ACM and a consequence of oxidative stress. Studies have found evidence of apoptosis in humans with ACM and in animal ACM models. Fernández-Solà and colleagues (2006) evaluated apoptosis in the hearts of adults with long-term alcoholism (n = 19, drinking for 26 years), adults with long-standing hypertension (n = 20), and those with no known disease (control subjects, n = 7). Apoptosis as evidenced by increased protein expression of two key proteins—one that promotes apoptotic cell death (i.e., BAX) and one that inhibits it (i.e., BCL-2)—was significantly higher in both the alcoholic subjects and in the hypertensive subjects, compared with control subjects (Fernández-Solà et al. 2006). Moreover, apoptosis was of a similar magnitude in the alcoholic and the hypertensive subjects. More recent findings from this research group corroborate that apoptosis occurs in humans who have a long history of heavy alcohol consumption (Fernández-Solà et al. 2011).
Researchers have found evidence of mitochondrial dysfunction or impaired bioenergetics related to alcohol consumption. This is not surprising, because mitochondria are a major target for free-radical injury. Dysfunctional mitochondria are less efficient, can become a source of ROS, and are more likely to initiate apoptosis (Marzetti et al. 2013).
Derangements in fatty acid metabolism and transport and formation of fatty acid ethyl esters (FAEEs) also have been implicated in ethanol-induced cell injury. FAEEs can be formed in the body during ethanol metabolism, when ethanol reacts with fatty acids or triglycerides. FAEEs can attach to mitochondria and disrupt mitochondrial function. Lange and Sobel (1983) were the first to identify an increase in FAEE content in postmortem myocardium samples obtained from those subjects who routinely had used large amounts of alcohol (n = 2) and who had a history of recent alcohol intoxication (n = 4).
Long-term alcohol use decreases myocardial protein expression and synthesis and accelerates protein degradation in the myocardium (Lang et al. 2005). This in turn disrupts myocardium function, including contraction and relaxation of the cardiac walls, impairing the heart’s ability to pump blood. Using mass spectrometric–based proteomic analysis in an animal model, Fogle and colleagues (2010) found that long-term alcohol consumption was associated with decreases of 30 to 54 percent in cell-scaffolding proteins (myofibrillar α-myosin and actin) and mitochondrial proteins (mitochondrial dehydrogenases and electron transport proteins), glycolytic enzymes (glycogen phosphorylase and alpha-enolase), and fatty acid metabolism proteins (fatty acid transport protein and LCFA acyl-CoA ligase). These investigators also found decreases in peroxiredoxin 5, antioxidant protein 2, and glutathione transferase 5—important antioxidant enzymes whose cardiovascular protective functions still are not fully understood. For example, some findings suggest an inverse role between peroxiredoxin 5 and stroke severity. During a severe stroke, peroxiredoxin 5 is consumed and its production impaired (Kunze et al. 2014).
Alcohol consumption remains a major risk factor for global burden of disease (Rehm et al. 2009). Nearly all the data on humans exploring the relationship between alcohol consumption and CV risk—including some indications of potential CV benefits associated with low-to-moderate alcohol consumption— are derived from epidemiologic studies. Therefore, because there are no randomized controlled trials, health care professionals should not recommend alcohol consumption as a primary or secondary lifestyle intervention. Instead, clinicians should continue to recommend strategies such as a healthy diet and exercise. Adults who choose to drink can be encouraged to follow the alcohol consumption recommendations from NIAAA (table 3).