Even moderate alcohol consumption may put older adults with heart disease or diabetes at risk of developing a common form of arrhythmia, a new study says.

The work suggests that people who are in their 60s or beyond and who have had a heart attack, stroke or have hardening of the arteries or Type 2 diabetes should be careful about how much they drink.

One of the authors admitted Monday that this message might be perplexing for people who have been told for years that a daily drink or two may be good for heart health.

"It is in a way confusing because if I was 65 or 70, I'm used to having two drinks a day because my doctor says it's good, now this new study says 'You know, it maybe it's not so good,"' said Dr. Koon Teo, a professor of medicine at McMaster University in Hamilton.

But based on these findings, Koon, who is a cardiologist, said he would tell patients "two drinks or less may be better than two drinks or more."

The study was published Monday in the Canadian Medical Association Journal.

The findings are drawn from an analysis of data from two large trials designed to study treatment regimens for controlling high blood pressure and congestive heart failure. The studies enrolled a total of over 30,000 adults in 40 countries. The median age of subjects was 66 and they were followed for four and a half years on average.

Built into the design of the studies were questions aimed at trying to tease out risk factors for atrial fibrillation, the most common form of arrhythmia. Arrhythmia is a condition in which a sufferer's heart beat is irregular.

People who suffer from atrial fibrillation often complain of the sensation that their heart is racing, or that they are dizzy or breathless, even when they aren't exerting themselves. The Canadian Heart and Stroke Foundation estimates that about 350,000 Canadians have atrial fibrillation.

The condition puts sufferers at risk of experiencing a stroke. In fact, people with atrial fibrillation are three to five times more likely to have a stroke than people who don't have the condition.

In this study, when moderate and heavy drinkers were compared to people who drank lightly, higher rates of atrial fibrillation were seen.

The researchers used standard measures to classify subjects as low, moderate or high drinkers. Low-level drinkers consume less than a drink a week. Moderate drinkers imbibe between one and 14 drinks a week for women and one to 21 for men. Heavy drinkers consume more than 14 and 21 drinks a week for women and men respectively.

Atrial fibrillation was seen at a rate of about 14.5 cases per 1,000 people per year in low-level drinkers. Among moderate drinkers, that rate rose to 17.3 cases and among heavy drinkers it was 20.5. If the researchers are correct in their conclusions, alcohol consumption may account for the differences in the rates.

The study also suggests binge drinking -- imbibing more than five drinks a day -- was associated with a higher risk of atrial fibrillation.

But Dr. David Juurlink, an internal medicine specialist in Toronto's Sunnybrook Health Sciences Centre, warned there is an important caveat people need to keep in mind in assessing the findings of this study.

This type of study can only identify associations. It cannot prove that the one thing (moderate or heavy drinking) caused another (higher rates of atrial fibrillation).

Only randomized controlled trials can prove causality. In those studies, people who are similar to each other in age, health status and other factors are divided into two groups. One gets an intervention, the other gets a placebo, and the results are compared.

No ethics committee will ever greenlight a study to see what happens to heart health when people are give two drinks (or more) a day. So looking at data like the McMaster one will have to do, Juurlink said.

Still, he cautioned that sometimes association studies get it wrong. At one point, studies seemed to point to an increased rate of pancreatic cancer among coffee drinkers. It turned out that smoking -- which often goes hand-in-hand with coffee drinking -- was the behaviour that increased the pancreatic cancer risk.

Likewise people who drink moderately or heavily may be different in other important ways than people who don't, and that other factor or factors may be responsible for the increased rates of atrial fibrillation in these people, Juurlink said.

"I think it's fair to say increased alcohol intake does appear to be associated with an increased risk of atril fibrillation. It may or may not be the cause and perhaps additional research in the future will help sort that out," he said.

In the meantime, less is probably more when it comes to alcohol consumption by older adults with heart problems.

"It's hard to make sweeping pronouncements from a single study, but there is a compelling commonsense argument for moderation, and this study supports that," Juurlink said.

"If someone who drinks heavily needs one more reason to cut back, this is it. But as we all know there are plenty of other reasons to moderate one's alcohol intake."