Sunday, February 27, 2011

Would you prescribe alcohol to your patients?

Identified in 2005 as the most highly cited researcher in clinical medicine, Meir Stampfer, MD D.Ph., Professor of Nutrition and Epidemiology at Harvard School of Public Health is a leading expert in the etiology of chronic disease - particularly Cancer and Cardiovascular disease. He's also an expert in one other thing: Alcohol's protective effects on the heart.

I invited Meir Stampfer to speak at UMass about The Role of Alcohol in a Healthy Diet after an interesting question was proposed in one of our classes at UMass - would you give the OK to a middle aged man with significant cardiac risk and no history of alcohol abuse the green light to knock back a beer or two? (The patient was asking because he heard it was good for his heart).

To my surprise only 2 people in the group (out of 16 or so) said they would give the green light.

No one is arguing the harms of excessive drinking. Granted, the stakes can be high if a patient is instructed by their physician to drink alcohol and develops dependence - but balancing the risks and benefits, to me it seems like more times than not, the risk is actually higher to not recommend moderate alcohol for the patients in our cardiovascular disease prone society. This point was driven home when Dr. Stampfer showed a slide framing not drinking versus being a cardiovascular risk factor (You have a relative risk of about 1.6 if you are an abstainer of developing a cardiac endpoint versus moderate drinkers)

  • Although there is always a risk of developing dependence, alcoholism usually presents at a relatively young age, so the likelihood of precipitating abuse in a middle aged person is not very likely.
  • Over 100 prospective studies show an inverse association between moderate drinking and CVD (including MI, Ischemic stroke, PVD, Sudden Cardiac Death, and death from all cardiovascular causes)
  • This reduction has been found in both men and women, in those who do not have clinically evident heart disease, in those at high risk for dying of heart disease including people with Type II diabetes, high blood pressure, and existing cardiovascular disease
  • It doesn't matter what type of alcohol you drink (Red Wine, White Wine, Sake, Martinis) - all had equal benefit. (Resveratrol found in Red Wine has been found to be beneficial in mice studies, but you would have to drink 1.5 gallons of wine in order to get this experimental amount.
  • Cause and effect, while not proven in a double blind study - is essentially proven. Many studies have adjusted for confounding variables, and the measurable effects of moderate alcohol consumption (increased HDL levels, decreased hemoglobin A1c, effects on plasminogen activator, fibringen, Factor VII, and von Willebrand) can essentially account for the cardiovascular benefits alone
  • The best correlation with benefits is actually in the number of days one drinks per week, not number one drinks per day. (eg. 1, 2 or even 3 drinks/day on > or = 5 days/week has better outcomes than drinking this many per day for less than 5 days/week
  • The negative effects of alcohol at moderate doses include a modestly increased risk for breast CA, a modestly increased risk in blood pressure. This is a consideration, but studies have shown that all cause mortality is still decreased.
  • A cohort from Harvard School of Public Health was stratified in to a "healthy group" who were nonsmokers, have a healthy diet, are physically active, and have a low BMI. The benefits of Alcohol were still found even among this group.
Despite this, over 65% of men, and an even greater percentage of women report never drinking at all. (I am working on specifics of this statistic- I believe it is from one of the HSPH cohorts)

Harvard's Health Eating Pyramid actually includes alcohol (red wine) in the pyramid. (1 or less/day for women, 2 or less/day for men) The USDA's pyramid is next to it for comparison.

So.... would you ever consider giving the green light to a patient to drink alcohol? How about recommending or prescribing it? What if Pfizer marketed it? Leave your comments below!

U.S.D.A's "Food" Pyramid vs HSPH's "Healthy Eating Pyramid"


  1. hey mitch- great article. Alcohol is one of the fewest ways to raise HDL, along with exercise and also a high fat diet, but of course that also raises LDL.

  2. Thanks for posting Mitch! IIRC it was dark beers and red wine that show this effect... Is that correct? What about my favorite? Wheat Beer!!!!

  3. Or tequila or laboratory ethanol! No association was found with the TYPE of alcohol, as long as it was the "beverage of moderation" of whatever culture was studied. Red wine has resveratrol, but you would have to drink several gallons of wine/day to get the beneficial effect. I've heard IPA's may have some similar properties. If anything darker beers (stouts, porters) may have carcinogens similar to those found in bbq.

    So enjoy your wheat beer!