I have heard that moderate alcohol consumption may prevent cardiovascular disease, but I have also heard that even moderate alcohol consumption can increase the risk of cancer (particularly breast cancer in women). Does current evidence suggest that the benefits of alcohol outweigh the risks in terms of mortality, and what is the optimal level of alcohol consumption for maximizing life expectancy?


2 Answers 2


Another answer did a great job laying out some of the difficulties plaguing research on this topic. Observational data in the area are (hopelessly?) confounded. I would like to point out two alternative lines of research that may leave the reader with some hope that we’ll eventually have a good answer to this question.

  1. Surrogate markers: A major difficulty with research in this area is that it’s not practical to randomize an adequate number of people to alcohol vs no alcohol1 for a long enough period of time to see statistically significant differences in outcomes like heart attacks, strokes, and death. However, in many areas — most prominently cardiovascular disease — researchers have used surrogate markers that tend to evolve more quickly and have been shown to correlate with the real endpoint that we are interested in. This approach allows us to:

    • Acquire randomized data due to the shorter duration of study required, avoiding all of the mess of observational confounds or;
    • (Less optimal but still something): use observational methods but analyze endpoints that are less likely to be confounded.

    I include two papers that review interventional data suggesting a benefit of alcohol on surrogate endpoints:

    Effect of alcohol consumption on biological markers associated with risk of coronary heart disease: systematic review and meta-analysis of interventional studies.2

    The Effect of Alcohol Consumption on Insulin Sensitivity and Glycemic Status: A Systematic Review and Meta-analysis of Intervention Studies.3

    Despite the feasibility benefits of surrogate endpoints, these “markers of malign” carry certain caveats and can not be regarded as providing a definitive answer about whether alcohol should be recommended to reduce the risk of chronic diseases.

  2. Randomized, double-blinded, placebo-controlled trials (with real endpoints) in the making: This is what we want. There is an NIH U34 active for planning a multi-center long-term randomized controlled trial to investigate the effect of alcohol on the development of chronic diseases. You may be wondering how you randomize and blind people to alcohol consumption. Although the methodology of that large trial is not yet published (to my knowledge), the Mukamal lab has a history of putting tasteless grain alcohol in Crystal Light (see also, WallStreet Journal blog piece). That feasibility study was completed; I look forward to updating this answer in a few years when the data from the large RCT are published.

1. Note that although resveratrol, a component of red wine, has been publicized as the possibly beneficial agent (partly because it can be put in supplements and sold...), many researchers in the field consider ethanol itself to be of more interest. The data in this answer all investigate ethanol rather than other components of wine or beer, several of which may carry additional benefits.

2. Brien, SE et al. BMJ.2011;342:d636

3. Schrieks IC et al. 2015 Apr;38(4):723-732.

  • I wouldn't say the confounds are hopeless, especially when we can study different cultures with different drinking customs. In terms of social relationships, I'm sure there are places where gatherings involve heavy drinking, others where they involve light drinking or no drinking, and others where they're more likely to involve other drugs (tobacco, coffee, khat, marijuana, etc.).
    – octern
    Commented Apr 21, 2015 at 15:30
  • @octern Thanks, that's an interesting approach. I tend to be fairly negativistic about observational data in general for anything beyond exploratory analyses. We don't and can't have randomized controlled trials about seatbelts, but we can and will have randomized controlled trials about alcohol. :-)
    – Susan
    Commented Apr 21, 2015 at 15:35
  • What is the short answer? "Maybe"? Commented Mar 18, 2016 at 9:49
  • Nice answer sweetie... +1 see this in chat here on Health chat.stackexchange.com/rooms/65011/ask-a-health-moderator when you get a chance.
    – Facebook
    Commented Sep 3, 2017 at 0:41

With many of the studies on alcohol consumption and heart health, they look at people who drink moderately, generally a glass of red wine (resveratrol). Some researchers hypothesis that it is the resveratrol that leads to reduce atherosclerosis. This thought comes from the French paradox or the resveratrol paradox [1]. The paradox is why do the French live so long with a diet high in saturated fats while they also consume a lot of red wine (since alcohol is a poison).

However, correlation doesn't equal causation, i.e, there can be third a factor. Researchers from Harvard found that being stressed and unhappy can have lasting negative effects on the body [2]:

A vast scientific literature has detailed how negative emotions harm the body. Serious, sustained stress or fear can alter biological systems in a way that, over time, adds up to “wear and tear” and, eventually, illnesses such as heart disease, stroke, and diabetes. Chronic anger and anxiety can disrupt cardiac function by changing the heart’s electrical stability, hastening atherosclerosis, and increasing systemic inflammation.

How does this relate to moderating drinking? It relates to moderate drinking through research on social circles and prolonged life spans. A team of Australian researchers found that individuals with stronger and bigger social circles have a longer life span [3].

Survival time may be enhanced by strong social networks. Among older Australians, these may be important in lengthening survival.

How do we parse this information? Many who drink moderately tend to go out for that glass of wine and when they go out there is a social aspect to the setting. Husband/wife, children, dinner with other couples, etc. That is, there is a continual social bound occurring. So is it alcohol that is benefit or social circle? Does a positive social circle have an effect on stress and longevity? Luckily Dr. Martin Seligman from the University of Pennsylvania has already done the research. His work as well as many other researchers on the subject are cited, quoted, and expanding on in Perceptions of Longevity and Successful Aging in Very Old Adults by Katie E. Cherry, Loren D. Marks, Tim Benedetto, Marisa C. Sullivan, and Alyse Barker:

Ample experimental evidence documents the association between religious participation and mortality, where religious individuals live longer than do their non-religious counterparts (e.g., McCullough, Friedman, Enders, & Martin, 2009). For example, Hummer, Rogers, Nam, and Ellison (1999) have shown that Americans who attend worship services twice a week live an average of 7.6 years longer than their non-attending counterparts (a figure that nearly doubles to 13.7 years among African Americans; see Marks et al., 2005). Hummer et al.’s study was based on a national sample of 20,000+ that included a variety of faiths (Christian and non-Christian) and received the highest possible rating (‘10’) for methodological rigor (Koenig et. al., 2001, p. 562). The “key” to this longevity difference in Hummer et al. appeared to be frequency of worship service attendance, not world faith or denomination (see also Gillum, King, Obisesan, & Koenig, 2008; Hill, Angel, Ellison, & Angel, 2005; Musick, House, & Williams, 2004) [4].

Hummer et al. noted that the key to longevity was the frequency of attending a social event in this case church, but if you read the entire pub med paper in [4], you will see this extends to non-church social event frequency as well.

So researchers have shown that social circles lower stress, keep people happy, and promote longevity. Many of the researchers who look at moderator drinkers don't look at the social aspect of the group just the fact that they (A) consume moderate amount of alcohol and (B) haven't died of a heart attack. Therefore, they conclude A causes B but A could be causing more social get togethers where the moderate consumption of alcohol occurs and these social settings will help reduce stress, promote community, and belonging which promotes longevity which leads to lower rates of atherosclerosis.

From the American Red Cross circulation Red Wine and your Heart [5]:

Despite considerable data from epidemiological studies and strong suggestions from experimental research, the evidence is still insufficient to encourage patients who do not drink to start consuming red wine as part of a strategy to protect against atherosclerosis. Too much alcohol consumption has been shown repeatedly to contribute to cardiovascular disorders such as alcoholic cardiomyopathy (which develops when the heart muscle becomes too weak to pump blood effectively), high blood pressure, and certain electrical disturbances of the heartbeat. Excessive alcohol use can lead to liver cirrhosis, cancers, pancreatitis, neurological disorders, motor vehicle accidents, and addiction. Individuals with a personal or family history of alcohol abuse or liver disease should avoid drinking alcohol. Nevertheless, many medical societies view light use of ethanol as potentially beneficial to the cardiovascular system, although no formal recommendations for light alcohol consumption have been made.

That is, alcohol has no formal recommendation and light consumption may be beneficial, but again there is no mention of environment. Moreover, the Mayo Clinic makes the same statement but never analyzed the persons environment as well [6].

Until more studies start to look at social environmental factors and not just the consumption point of view, I don't think ingesting a poison to maximize life expectancy is the way to go since there is no medical recommendation on doing so; hence, it is still called the Resveratrol paradox and not a theory or fact.

  • Can you clarify whether there is any published research on the hypothesis that moderate drinking is a proxy for social activity? Or is this your own theory? It's an interesting idea, but I'd like to see it tested. The fact that social relationships predict health and longevity has been accepted in epidemiology for at least 40 years, so I'd be surprised if none of the studies of alcohol and health had ever thought to adjust for it.
    – octern
    Commented Apr 21, 2015 at 15:24
  • @octerm yes a studies do exist on it which I have read some time ago but couldn't find. If you read the studies on alcohol, they don't factor in social factors. They use surveys to ask about alcohol consumption and nothing else.
    – user139
    Commented Apr 21, 2015 at 16:05
  • What is the short answer? "Maybe"? Commented Mar 18, 2016 at 9:50

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.