If two drinks are good, why not three?

It seems like there is always another media grabbing study relating alcohol consumption and our health; but most are left more confused about just how much and what is good for them. Why is reaching a consensus on alcohol consumption and our health such a difficult task? The USDA’s current Dietary Guidelines for Americans recommend that if you choose to drink, you should not exceed 1 drink per day for women and 2 drinks for men. Drinking alcohol in moderation has been linked to a lower risk of heart disease, increased HDLs (“good cholesterol”), lower all cause mortality, lower risk of diabetes, and may reduce risk of dementia. The keys to the beneficial aspects of drinking are regularity and moderation. Overindulgence and binge drinking (consuming more than 5 drinks per day) is hazardous and associated with increased risk of liver cirrhosis, hypertension, cancers of the upper gastrointestinal tract, injury, violence, and death. A big part of the confusion stems from the manner in which these conclusion have been drawn; so lets define the different types of epidemiological studies, or scientific studies, that look at disease and illness and how they relate to the alcohol issue. Randomized controlled trial (RCT) (intervention), the gold standard in clinical research. RCTs are studies in which people are assigned at random to receive one of several clinical interventions- including the standard of comparison (usually placebo or current standard). RCTs look to measure and compare outcomes after participants receive the interventions. Because the outcomes are measured, RCTs are quantitative studies resulting in finite results. Governmental regulating organizations, such as the Food and Drug Administration, require all pharmaceuticals to undergo RCTs to prove effective in treating a specific disease. In terms of studying alcohol intake in the general population, RCTs are virtually useless as precisely controlling units of alcohol consumed is not representative of daily life thus limiting application.

It seems like there is always another media grabbing study relating alcohol consumption and our health; but most are left more confused about just how much and what is good for them.  Why is reaching a consensus on alcohol consumption and our health such a difficult task?

The USDA’s current Dietary Guidelines for Americans recommend that if you choose to drink, you should not exceed 1 drink per day for women and 2 drinks for men.  Drinking alcohol in moderation has been linked to a lower risk of heart disease, increased HDLs (“good cholesterol”), lower all cause mortality, lower risk of diabetes, and may reduce risk of dementia.  The keys to the beneficial aspects of drinking are regularity and moderation.  Overindulgence and binge drinking (consuming more than 5 drinks per day) is hazardous and associated with increased risk of liver cirrhosis, hypertension, cancers of the upper gastrointestinal tract, injury, violence, and death.

A big part of the confusion stems from the manner in which these conclusion have been drawn; so lets define the different types of epidemiological studies, or scientific studies, that look at disease and illness and how they relate to the alcohol issue.

Randomized controlled trial (RCT) (intervention), the gold standard in clinical research.  RCTs are studies in which people are assigned at random to receive one of several clinical interventions- including the standard of comparison (usually placebo or current standard).  RCTs look to measure and compare outcomes after participants receive the interventions. Because the outcomes are measured, RCTs are quantitative studies resulting in finite results.  Governmental regulating organizations, such as the Food and Drug Administration, require all pharmaceuticals to undergo RCTs to prove effective in treating a specific disease.  In terms of studying alcohol intake in the general population, RCTs are virtually useless as precisely controlling units of alcohol consumed is not representative of daily life thus limiting application.

Case-Control or Retrospective studies (observational) compare two groups of people: those with the disease/condition of interest (cases) and a virtually identical group of people who are disease free (controls).  Medical and lifestyle histories of each group are studied to learn what factors may be associated with the disease.  In terms of studying the effects of, for example alcohol and blood pressure, one can choose a group of people with high blood pressure (cases), and a similar disease free group (controls) and compare alcohol consumption; cases and controls might be middle classed men, who live in cities.  Results may find that those with high blood pressure are not regular exercisers.  There are several limitations of case-control studies with respect to studying alcohol intake.  The major limitation in looking retrospectively at lifestyle histories is, recall bias; most people will only be able to give a general quantitative estimate- this is therefore not helpful in forming finite guidelines for alcohol intake.

Cohort or Prospective studies (observational) follow a group of healthy people who share a common experience with a certain exposure and assess their health over time.  The exposure in this case would be alcohol consumption.  A cohort study is not a suitable study to look at the health effects of alcohol, as monitoring and controlling for frequency and amount drunk are nearly impossible and if possible, would not be characteristic of real life.  The studied population could keep a diary of alcohol consumption, but as you can imagine intake diaries are not always accurate, therefore contributing a large amount of bias.  A cohort study is usually conducted on more measurable terms such as exposure to radiation due to geographical location during pregnancy (exposure) and resulting birth weight.  Cohort studies tend to be expensive, time-consuming and logistically difficult (follow up time could be years!)

Cross-Sectional or prevalence studies (observational) look at a groups of people, their attitudes, practices, disease states, etc., at a specific point in time.  Surveys are a type of cross sectional study.  Bias, response and non-response, are major limitations of cross sectional studies, which inevitably slant the results of the study in a certain direction- limiting application to precisely studying and forming finite recommendations regarding alcohol intake and health outcomes.  Benefits of cross sectional studies include: a large amount of information, from a large group can be obtained relatively quickly and inexpensively.

A major limitation of observational studies (not including RCTs) is the ecological fallacy.  This results from interpreting general data too specifically. An example would be taking a generalization that may apply to a population and using this to make individual recommendations.  The most common example of this fallacy is, the French Paradox- the typical diet of those in Southern France includes a high proportion of cheese, butter, eggs, organ meats, and other foods high in fat and cholesterol, yet this population, compared to the U.S., has a much lower level of heart disease- the problematic generalization is that this population also consumes a greater amount of wine, and thus wine was pinned as the reason for lower rates of heart disease.  No other lifestyle factors were considered.

So, since it seems that reaching a universally agreed, evidence based conclusion regarding amount, type, and frequency of healthful (or not) alcohol intake is far off (or nearly impossible…).

There are individuals who should not drink according to the USDA, including: those who are not able to limit the amount they drink or who plan to drive, pregnant women, individuals taking certain medications or with certain medical conditions and those with alcoholism.  And of course, those who do not currently drink are not recommended to take up drinking for health benefits.