Light-to-moderate alcohol consumption has been associated with a decreased risk of ischemic cardiac events and stroke.¹ The cardioprotective effects of alcohol have been attributed to favorable lipid changes, including lower LDL, increased HDL cholesterol and higher apolipoprotein AI and AII levels, antiplatelet, and anti-inflammatory effects.² Low levels of alcohol consumption have been proven beneficial in providing a protective effect upon the cerebral circulation. However, in heavy alcohol consumption, there is an increased predisposition to hemorrhagic and nonhemorrhagic stroke.³ The coronary protective effects of heavy alcohol consumption have not been well studied. Chronic heavy alcohol consumption is a well-known cause of dilated cardiomyopathy. The deleterious effects of heavy alcohol on left ventricular function and on arrythmogenesis may counterbalance any potential coronary benefits.⁴ Few studies have addressed both CAD and left ventricular function in patients chronically consuming large quantities of alcohol. Accordingly, the objective of this study was to assess the associations between alcoholism, CAD severity and left ventricular systolic dysfunction.

Methods
The local institutional review board of the Department of Veterans Affairs New York Harbor Healthcare System approved the study (VA). The VA internal medicine healthcare database was queried for patients admitted between 1994 and 2002 who underwent cardiac catheterization and had a history of chronic alcoholic pancreatitis or chronic alcoholic liver cirrhosis (diagnosed by ICD-9 codes 303.91, 577.1 and 571.2, respectively). All of these patients had presented to the emergency room because of chest pain and subsequently had a positive stress test. The stress test was considered positive if the results of the test on the chart were documented as “positive”, “equivocal” or “nondiagnostic”. Therefore, the patient had this stress test result followed up with a cardiac angiogram in the VA system. During the index admission, all patients underwent echocardiography and had blood drawn for liver function testing. The patients were then subsequently compared to the first 200 consecutive age-matched controls without alcoholism between 1994 and 2002 from the VA internal medicine healthcare database. These patients were defined as nonalcoholic according to the history and physical exams conducted by their outpatient internal medicine physicians that were documented in the computers of the VA internal medicine healthcare. Similar to the alcoholic patients, these nonalcoholic patients were admitted through the emergency room because of chest pain and had a stress test that was documented as “positive”, “equivocal”, or “nondiagnostic”, thus prompting the hospital to refer the patient for cardiac angiography within the VA system (Figure 3).
The following clinical variables were considered in all patients: family history of CAD, active smoking, hypertension, previous myocardial infarction (MI) and diabetes. In both groups of patients, the serum laboratory values included: total cholesterol, triglycerides and high-density lipoproteins (HDL). Low-density lipoproteins (LDL) were determined by the Frederich calculation in patients whose triglycerides were < 400 mg/dL.
All coronary angiograms were performed at the VA in Brooklyn, New York. CAD was defined visually as an obstructive lesion > 50% stenosis of an epicardial coronary artery. The branches of major coronary arteries were not assessed in this study. The number of stenosed vessels in the major coronary artery distributions — the left main coronary artery, the left anterior descending artery, the left circumflex artery, and the right coronary artery were visualized to determine if they had an obstructive lesion > 50% stenosis to quantify the extent and severity of CAD in the alcoholic and nonalcoholic patients.
LVEF was determined by echocardiography using the Teichholz method. Sensitivities, specificities, positive and negative predictive values of elevated serum SGOT (? 40 IU/L) and SGPT (? 40 IU/L) values were determined. Left ventricular dysfunction was defined as an ejection fraction < 40%.⁵
Statistical analysis was performed using SPSS software (SPSS, Inc., Chicago, Illinois). Continuous variables were expressed as mean ± standard deviation and were compared using the Student’s t-test. Categorical variables were expressed as frequencies and percentages and compared using the Fisher’s exact test. A p-value of 0.05 was considered significant; all tests were two-sided. Clinical, morphological and procedural variables that had demonstrated statistically significant differences among the two groups were included in the stepwise multivariate logistic analysis to determine the effect of alcoholic status as an independent predictor of CAD, left ventricular ejection fraction and various risk factors.

Results
A total of 100 patients were identified as having alcohol-related liver cirrhosis or pancreatitis and had undergone coronary angiography. The mean age was 63 ± 4 years. Ninety-one percent (91%) had chronic liver cirrhosis and 9% had a history of alcoholic pancreatitis. These patients were compared to 200 consecutive nonalcoholic patients who also had undergone cardiac catheterization and had a history of chest pain and a positive stress test. Diabetes was more prevalent in the alcoholic group (59% vs. 31%; p < 0.05). Other baseline demographics were similar between both groups (Table 1). However, the prevalence of CAD was lower in the alcoholic group than in the control group (42% vs. 58%; p = < 0.05). In patients with CAD, the alcoholic group had fewer mean numbers of stenosed vessels (1.6 ± 0.6 vs. 2.3 ± 0.7; p < 0.05), and was more likely to have single-vessel CAD as compared to the control group (64% vs. 8%; p < 0.05).
The alcoholic group had a lower mean LVEF compared to the control group (43 ± 13 % vs. 49 ± 9 %; p < 0.05), and left ventricular systolic dysfunction was more common (37% vs. 13%; p < 0.05) than in the nonalcoholic group. In patients with left ventricular dysfunction, there was a lower prevalence of CAD in alcoholics as compared to nonalcoholics (38% vs. 63%; p < 0.05). In contrast, in patients with preserved systolic function, there was a numerically higher prevalence of CAD in alcoholics vs. nonalcoholics, although this did not reach statistical significance (54% vs. 38%; p = NS).
Sensitivities, specificities, negative, and positive predictive values of elevated SGPT (? 40 IU/L) and SGOT (? 40 IU/L) values for the identification of left ventricular dysfunction are shown in Table 2. In both the alcoholic and nonalcoholic groups, the patients who had normal LVEF were more likely to have normal SGPT measurements (87% vs. 14%; p < 0.05). For the entire group, serum SGPT levels correlated inversely with LVEF (r = -0.21; p < 0.05). The prevalence of SGPT levels and LVEF in the alcoholic and nonalcoholic groups are shown in Figure 2.
In a stepwise multivariate logistic analysis, alcoholic status was also found to be an independent predictor of CAD (r = -0.15; p < 0.05). Univariate predictors of CAD and left ventricular dysfunction were entered into the multivariate model. In a stepwise multivariate logistic analysis using, alcohol status, diabetes, SGPT, and CAD, the independent predictors of left ventricular dysfunction were alcohol status and SGPT (r = 0.24; p < 0.05).
Discussion
This study showed less prevalent and less extensive CAD in alcoholic male patients as compared to the age-matched controls with a history of chest pain and a positive stress test. In addition, left ventricular dysfunction was more prevalent and more severe in the alcoholic group (Figure 2). The study extends the observations of epidemiological studies which have shown light to moderate alcohol ingestion associated with fewer cardiovascular events, including MI, cardiovascular mortality, and ischemic stroke.^6,7 These studies evaluated the effect and dose response of alcohol on a number of clinical endpoints. Proposed mechanisms for decreased cardiovascular events include a favorable impact in lipids including: increase of HDL, and lowering of LDL by alcohol.¹ There is also an inhibition of platelet adhesion, increased prostacyclin: thromboxane ratio, increased endogenous tissue type plasminogen enhanced anti-oxidant effect, and anti-inflammatory effects with the exception of a higher prevalence of diabetes in the alcoholic group.⁸ These factors may act collectively to inhibit plaque progression and promote stabilization. While cardiovascular outcomes were not evaluated, the present study suggests that the beneficial effects of alcohol may be related to a reduced atherosclerotic plaque burden.
Prior studies of alcoholics or heavy drinkers are lacking. In an autopsy study, Thompson et al. found less CAD in alcoholics as compared to nonalcoholic men but similar degrees of aortic atherosclerosis.⁹ Therefore, the coronary protective effect of alcohol does not seem to diminish even with heavy alcohol consumption.
The prevalence of traditional cardiovascular risk factors for CAD was similar in the two groups with the exception of diabetes. There were also no significant differences in the mean concentrations of total cholesterol, HDL, LDL, triglycerides between the two groups. This would suggest that the protective effects of alcoholism in the prevalence and severity of CAD may be unrelated to lipids.
The alcoholic patients who have developed cirrhosis of the liver may have developed decreased testosterone levels through the destruction of their hepatocytes.¹⁰ The decrease in testosterone levels in the alcoholic group who developed cirrhosis may directly contribute to the increase in CAD in the alcoholic patients with preserved left ventricular function. Testosterone replacement has been associated with a delay time to ischemia, lowering cholesterol, and a lowering serum tumor necrosis factor-alpha, a proinflammatory cytokine.¹¹ Serum tumor necrosis factor alpha has been found to be elevated in patients with liver cirrhosis.¹² Oxidative stress and inflammation such as elevated levels of tumor necrosis factor-alpha are associated with the propagation of coronary artery disease and congestive heart failure.^13–15 As compared to nonalcoholic patients, alcoholic patients showed a higher prevalence of left ventricular dysfunction and a lower mean LVEF. The findings are consistent with other studies showing an association between heavy alcohol consumption and dilated cardiomyopathy.⁴ Although diabetes was more common among alcoholics and has been associated with dilated cardiomyopathy, diabetes was not an independent predictor of left ventricular dysfunction on multivariate analysis in the present study.
A novel finding is the high negative predictive value of a normal SGPT in excluding left ventricular dysfunction. It has been suggested that patients develop either alcoholic related cardiomyopathy or cirrhosis in a mutually exclusive manner.¹⁶ To date, no study has evaluated specific serum liver enzymes as a marker of left ventricular dysfunction. Although elevated serum SGPT and SGOT levels do not appear to be useful in differentiating patients with or without left ventricular dysfunction, normal SGPT values have a high negative predictive value and therefore are useful in excluding left ventricular dysfunction.
Study limitations. This study is subject to the inherent limitations of a retrospective analysis. All patients were male and underwent cardiac catheterization after abnormal stress testing. ICD-9 codes of alcoholic pancreatitis and alcoholic liver cirrhosis were used as surrogates for heavy alcohol ingestion rather than an actual measure of alcohol consumption. The authors felt that that the presence of end-organ damage from alcohol abuse was a surrogate measure of the degree of alcoholism since the actual alcohol consumed was unavailable. Since not all alcoholics experience end-organ damage this may represent the worst case scenario and a skewed patient population. These results are not necessarily applicable to women because all of the study subjects were male. These groups of patients constitute a small percentage of patients who consume alcohol and are subject to the limitations of small groups. Despite the limitations, it is concluded that there appears to be an inverse relationship between CAD and left ventricular dysfunction in alcoholic patients.