Over the last five years, the health benefits of moderate drinking have been widely celebrated in the headlines. To those who think everything enjoyable must be bad for you, this news might seem like a dream come true.

Of course, there are many caveats – and these studies don’t indicate that teetotalers should take up drinking or that infrequent drinkers should start drinking more. The operative word here is drinking in moderation.

Studies show, for example, that health benefits only come with moderate drinking and are greatest for older men. And even moderate drinking is not recommended for women who are pregnant or thinking of becoming pregnant, or for people who are under 21.

The strongest medical evidence exists for the link between moderate drinking and a reduced risk of heart disease.

Dr. Kenneth Mukamal, an internist at Beth Israel Deaconess Medical Center and assistant professor at Harvard Medical School in Boston, was the lead author of a New England Journal of Medicine study examining the roles of drinking patterns and heart disease that found, after 12 year of follow-up, that men who consumed alcohol between three and seven days a week had fewer heart attacks than men who drank once a week.

Below, Mukamal discusses the risk and benefits of moderate drinking.

We think that a lot of the benefits of alcohol are on the blood vessels and on blockages in the arteries to the heart and to the brain. This might be related to alcohol’s effect on the good cholesterol, the HDL cholesterol.

In fact, alcohol affects HDL levels just about as strongly as any other lifestyle factor. People also think that alcohol may lower heart attack risk by acting as a blood thinner.

A wide variety of health effects have been attributed to moderate drinking. A lower risk of diabetes has been seen in women and men.

There actually have been experiments done in which alcohol was administered over a couple of months to people without diabetes. In those studies, most of which have been conducted in women interestingly, it looks like moderate drinking improves the body’s sensitivity to insulin.

It may actually lower insulin levels altogether and may prevent diabetes through that mechanism.

More recently we’ve done some work on moderate drinking and dementia. We looked at a group of older adults in the United States – average age was in the mid-70s – and found a reduced risk.

There has been some more work in slightly younger populations from Europe, and those studies have fairly consistently suggested that older adults who were drinking moderately may have a lower risk of dementia. We’re not exactly sure what the mechanisms may be behind that.

Some of it may very well be because drinking tends to occur in social settings and just the process of getting out and socializing may be an important way to prevent dementia.

There is also evidence that moderate drinking may prevent silent strokes or other subtle types of brain injury that we know over time can predispose to dementia. I think it’s still an area where we need some more investigation.

In most of the studies that look at this issue, people have been asked ‘How much alcohol do you usually drink?’ When that question is asked, people take an average. For example, I drink 10 drinks a month. But 10 drinks a month is very different for someone who has them all on one night vs. someone who has them on 10 different nights of the month.

That kind of detail surprisingly hasn’t been available in most of the studies that have been devoted to this topic. In our study we tried to figure out the drinking pattern that’s most closely tied to lower heart attack risk.

What we found in a study of about 38,000 men was that the key factor wasn’t what men were drinking, or frankly even so much how much they were drinking at a time, but how frequently they were drinking alcohol.

We found that men who were drinking at least three to four days a week or more had lower heart attack risks than people who had one drink a week.

We also have some very strong studies showing that heart disease risk, while lower amongst moderate drinkers, can be substantially higher among people who drink to excess even occasionally.

They don’t have to be drinking excessively every single night to potentially have a greater heart attack risk.

Many of the effects of moderate drinking, such as acting as a blood thinner, are only true at moderate levels of drinking. Those effects actually go away and reverse if people drink too much.

What doctors usually consider a drink is basically a medium glass of wine, a 1.5 oz shot of spirits, or a can or bottle of beer. All of those have roughly similar amounts of pure alcohol in them.

We usually define moderate drinking as up to one drink per day for adult women who aren’t pregnant and up to two drinks per day for adult men. Some guidelines recommend that moderate drinking among adults over 65 be limited to one drink per day.

In general, when we’re thinking about the putative health benefits of moderate drinking, they mostly apply to older people and to men. Issues for women and for younger individuals are much more difficult to sort out.

The role of alcohol consumption in heart disease varies strongly by gender. The reason for that is twofold. On the one hand, women at any given age tend to have lower risks of heart disease than men do.

As a result, the benefits of moderate drinking accrue disproportionately to men. At the same time, there are some particular risks of drinking for women that don’t exist for men.

There is some evidence that women may be particularly prone, for example, to liver disease related to drinking. Even moderate drinking may increase breast cancer risk.

And, while the effects on heart attack risk are roughly similar in men and women, I think it’s even more difficult to determine what the ideal level of drinking ought to be for women than it is for men.

I think it is fair to say that if young women in general are drinking with the expectation that there is some health benefit to it for them, they’re probably mistaken. Young women are a group of people for which, as of now, we basically have no clear proof that the overall balance of alcohol’s risks and benefits is going to work in their favor.

There is fairly consistent evidence that breast cancer rates are higher among women who drink moderately. I think that’s important because obviously breast cancer is very common disease. I certainly think women at high risk for breast cancer should talk with their doctors about whether they should be drinking any alcohol.

Another important risk, which is unrecognized for many people in this country, is that even moderate drinking among people with hepatitis C may increase their risk of permanent liver damage. Anybody who is known to have hepatitis C shouldn’t be drinking any alcohol at all.

People who have risk factors for hepatitis C ought to be tested because it will very substantially impact what the potential risks are related to moderate drinking.

In addition, although we don’t think moderate drinking necessarily clouds our judgment, it turns out that it probably does. In simulated driving tests that were done as far back as the 1950s, people have realized that at very low blood alcohol levels, simulated driving performance is impaired. When I say low blood alcohol, what I’m talking about is as low as .02 percent.

Some studies, for example, the analysis of the National Alcohol Survey, showed something similar. You begin to see higher risks of injury even when people are reporting one drink a day.

That’s why we still recommend that even moderate drinking occur in the home, preferably tied to meals.

That is not so much because we find that that drinking with a meal is more likely to lower heart disease risk, for example, but because it’s the safest way to prevent high blood alcohol levels that can get people into accidents.

Although it has been bantered back and forth, most people think that people who have a personal history of alcoholism very rarely can return to social drinking. People who, for personal or family reasons have never had alcohol before, at least as of now, probably shouldn’t start drinking for any health reason.

It’s hard to give any single piece of advice because of all the things we’ve learned about moderate drinking. The potential risks and benefits are going to vary by a person’s health history, their age, sex and family history.

The number of factors that would have to go into the decision is really very substantial. As a primary care doctor myself, these are long discussions that people should have with their doctor. I would not recommend that anybody go out tomorrow and start drinking alcohol simply on the basis of results that we and others have presented.

I would say that for people who are drinking moderately and are able to control it and don’t have any of the absolute reasons why they shouldn’t be drinking alcohol, that there is no evidence now that that’s a bad thing to do.

Beyond that, I don’t think right now we have enough evidence to say that anybody should take up drinking just for any particular benefit unless their doctors recommend that they do so.