The November 1991 segment of CBS's 60 Minutes introduced millions of Americans to the cardio-protective effects of moderate alcohol consumption. But the phenomenon has been known since 1926 when a study of tuberculosis patients in a sanatorium found that patients who drank alcohol moderately had approximately half the death rate as those who abstained.
The U.S. government and the American medical establishment didn't
rush to embrace this finding, conducted as it was, during the midst of Prohibition.
Medical science has progressed enormously since then with scores of university and hospital studies in the United States and abroad offering not just theories and hypotheses, but hard,
reliable facts. Abstainers and heavy drinkers die sooner of all causes and are hit with crippling or lethal heart attacks
at almost twice the rate as their moderately sipping neighbors.
Again, it's important to remember that alcohol abuse -- heavy, addictive consumption --is a serious problem for society. An estimated 7 to 10 percent of Americans cannot consume alcohol moderately and should drink none. It's also important to remember that while moderate drinkers have a lower death rate from all causes, heavy drinkers are at increased risk for cancer, liver diseases, and accidents. Chronic heavy drinking also causes high blood pressure with a resulting increase in risk for heart disease.
"The epidemiological community has known for many years that moderate alcohol consumption is associated with lower cardiovascular mortality rates but has not publicized this fact, perhaps because of the fear that making a positive statement about drinking would lead to greater abuse of alcohol," said Dr. Curtis Ellison, Chief of the Section of Preventive Medicine and Epidemiology at the Boston University School of Medicine in an editorial in the September 1990 issue of Epidemiology.
This fear, Dr. Ellison said, is not supported by scientific or medical fact. "People who abuse alcohol are addictive personalities and they will abuse regardless of the medical advice given," he said.
Researchers point to alcoholics who will drink perfume, mouthwash and other sources of alcohol when beverage alcohol is unavailable.
But for the other 90 to 93 percent of Americans, proper knowledge and application of the scientific facts might save more than 200,000 people every year from dying of heart attacks and save another 500,000 people from even having a heart attack.
We now know as much or more about the cardio-protective effects of moderate alcohol consumption as we do about aspirin's role in reducing heart attacks or fiber's role in decreasing colon cancer.
Numerous studies, including several conducted in the early 1990s, reveal the existence of a J-shaped curve (sometimes called a U-shaped curve) which graphs death rates against alcohol consumption.
The J-shaped curve shows that overall death rates begin to drop in people who drink moderately. The curve reaches bottom at some point and then begins to rise again as alcohol consumption increases. With increasing consumption, the death rate eventually draws even with abstainers and then continues upward, with heavy drinkers having the highest death rates of all.
This J-shaped curve is a fairly common one in medicine and is associated with a wide variety of substances which become harmful if too little or too much of it is consumed. Vitamin A, for instance, causes disease when it becomes scarce; yet it is toxic when too much is consumed. Most life-saving pharmacueticals exhibit this same J-shaped behavior: too little is unhealthy, too much is dangerous. This is true whether the pharmacueticals are antibiotics, cholesterol-lowering drugs, hypertension medicine, ulcer medication and so on.
The J-shaped curve means that the effects -- good or bad -- are dose-related: the right amount is good for you while too little or too much can be unhealthy.
While alcohol is not a pharmacuetical to be prescribed as medicine, you need to recognize that its effects are dose related. To obtain the healthiest results, your consumption needs to be moderate. As you will learn, the definition of "moderate" drinking varies from study to study, ranging from one to five "drinks" per day (a drink being equivalent to 10 to 12 grams of pure alcohol -- roughly a four to five-ounce glass of table wine, one beer or one mixed drink). Most medical experts say that one to three drinks per day is the most responsible and acceptable definition of "moderate."
More than two dozen hospital and university studies confirm the cardio-protective effects of moderate alcohol consumption in both men and women. One of the most recent was conducted by a research team headed by Eric Rimm at the Harvard University School of Public Health. The study, published in August 1991 in the respected British medical journal, Lancet, found that male physicians who drank, on average, one-half to one drink per day had 21 percent less coronary artery disease (CAD) than abstainers. Put another way, their relative risk (RR) was 0.79 compared to abstainers at 1.00.
The relative risk continued to drop with increased consumption. Men who consumed one to one and a half drinks per day reduced their CAD risk by 32 percent (RR=0.68; one and a half to three drinks per day reduced the risk by 27 percent (RR=0.73); three to four and a half per day reduced it by 43 percent (RR=0.57) and those drinking more than four and a half drinks per day reduced CAD risk by 59 percent (RR=0.41).
While the heart attack risk at the higher levels may seem attractive to some, this level is the beginning of heavy drinking (and perhaps abuse) and carries with it an increasing risk of other diseases.
Another landmark study of the health benefits of moderate alcohol consumption was published in the September 1990 issue of Epidemiology by Paolo Boffetta of the American Cancer Society and Lawrence Garfinkel of the International Agency for Research on Cancer, Lyon, France. Their study confirmed the J-shaped curve, but found somewhat lower protective effects from moderate consumption (perhaps due to under-reporting of consumption by participants or because they used a different way of categorizing consumption).
Using abstainers as a reference (RR=1.00) they found that the RR for heart disease was 0.86 for occasional drinkers, 0.79 for those drinking one drink per day, 0.80 for those drinking two drinks per day and 0.83 for those drinking three drinks per day. The cardio-protective effects continued with increased drinking but began to decrease again for people drinking five or more drinks per day.
The relative risk (RR) of all causes of death was 0.88 for occasional drinkers, 0.84 for those drinking one drink per day, 0.93 for those drinking two drinks per day and 1.02 for those drilling three drinks per day.
One reason why the protective effects were smaller in their study than in other research can be found in Boffetta and Garfinkel's discussion of their research subjects. They emphasized that their study was of middle-aged men (40 to 59 years old) aging over time. "Therefore it can be expected that alcohol intake was also reduced during the follow-up period; if this is true.. the present estimates will represent an underestimation of the true effects of alcohol drinking on coronary heart disease mortality,"
The Boffetta and Garfinkel study also found slight decreases in total cancers and accidents as well as the expected increases in suicide and liver cirrhosis. It's significant that some studies do not find an increase in suicides while others do.
Unfortunately, it's the nature of scientific research to find conflicting data. This is because small (or unaccounted for) variations in the methods of doing the research or analyzing the data can bias the outcome. Because of this, an intelligent decision on whether something is fact or hypothesis can only be made by considering the overall body of evidence - which is just another way of saying that a single study does not make for scientific fact.
This book will present as fact only those conclusions based on multiple confirming studies conducted by respected organizations and a variety of different researchers. To do otherwise -- to try and base a conclusion on a single study -- is naive, scientifically unsound and intellectually dishonest. The studies presented in this book to illustrate fact (such as the ones you've read so far in this chapter) will be those representative of the overall body of evidence.
Some of the older studies have been attacked by critics who charged that the results were biased because the abstainers studied included many people who were not drinking because they were already in bad health, or were possibly abstaining alcoholics.
The Rimm study at Harvard was one of the first (but not the only one) to study this possibility and to correct for it in the analysis. The study concluded that excluding abstainers with pre-existing conditions did not alter the J-shaped curve or the conclusion that "the inverse relation between alcohol consumption and risk of coronary artery disease is causal."
Other critics of the J-shaped curve have suggested that people who drink alcohol are all somehow different or all engage in some other activity that accounts for the cardio-protective effect.
Numerous studies have eliminated education, race, cigarette smoking habits, coffee drinking and others as possible factors. Several of the important studies in this regard have been done by Dr. Arthur L. Klatsky, M.D., at the Kaiser Permanente Medical Center in Oakland, Calif. One of his major articles, published in the Ame rican Journal of Cardiology in 1986, corrects for these factors and "supports the view that alcohol protects against CAD."
Additional proof regarding other socio-economic factors, such as access to medical care, diet, income and personal habits, comes from laboratory experiments conducted with rabbits at The Wistar Institute in Philadelphia. They found that in a genetically matched group of animals, those fed alcohol had up to 60 percent less coronary artery disease than rabbits who had no alcohol.
As you read in Chapter One, the French (and most people around the Mediterranean) flout the laws of accepted medical knowledge with heavy cigarette smoking, high fat consumption and a relative lack of exercise.
Yet they live longer and suffer only one half as many heart attacks as Americans, despite the fact that their cholesterol levels are approximately equal to ours. That's the paradox that flies in the face of conventional medical wisdom.
Conventional medical wisdom's reluctance to cope with the beneficial effects of moderate alcohol consumption is also the reason that it took so long for scientific and medical research institutions to accept alcohol's role in keeping people heart healthy despite other bad habits.
Another key study linking moderate alcohol consumption, particularly wine, with cardio-protective effects, was published in 1979 in the Lancet and has been subsequently corroborated by data gathered over the past 12 years. That study, led by A.S. St. Leger of the British Medical Research Council's Epidemiology Unit in Cardiff, studied ischemic heart disease (lAD) deaths in 18 developed countries to determine what factors were associated with the deaths.
As expected, they found significantly strong positive associations between smoking and deaths from all causes. (A positive association means that as the behavior or other measured factor increases, the death rate rises; a negative association means that the death rate decreases as the measured behavior increases.)
They also found strong positive associations between IHD deaths and total fat intake and also total calorie intake.
Surprisingly, they found no strong association between IHD deaths and such health-service factors as the number of doctors and nurses per 100,000 population.
The St. Leger team also found a number of strongly negative associations with IHD deaths. The IHD deal rates were lower as the per capita gross national product and population density increased. But the most surprising finding was that the strongest negative association in the entire study was the link with alcohol consumption. In other words, of all the factors considered, increased alcohol consumption had a stronger link to decreasing heart attack deaths than did decreased fat consumption or cigarette smoking.
This is not to say that smoking or high fat consumption is safe if you drink alcohol because people who engage in those risky behaviors have a significantly higher risk of dying sooner and having more heart attacks than moderate drinkers who eat smart and don't smoke.
The St. Leger study is one of the few to distinguish among consumption of wine, beer and distilled spirits. While their study strongly linked total alcohol consumption with decreased IHD) deaths, according to the published study, "This is shown to be wholly attributable to wine consumption."
This observation has been somewhat inconsistent in the total body of research where some studies have given the edge to beer or spirits and others to wine. If wine is found to have an edge in this regard, scientists say it may not be due to an intrinsic difference among the beverage types, but be due more to the fact that wine is more often consumed with meals than the other two types of beverage.