Rod Jackson writes how a change of heart led him to modify his drinking for the better.

Five years ago I inadvertently removed myself from an international network of academic researchers studying alcohol and coronary heart disease.

As a result I missed out on the associated regular invitations to speak at conferences and invitations to join writing groups for international publications on alcohol and heart disease. What had I done wrong?

For over 10 years before this, I had been convinced by the evidence suggesting that light to moderate drinking saved lives due to protective effects of alcohol on the arteries of the heart (the coronary arteries).

My PhD research showed a strong protective association between light to moderate drinking and the risk of heart attacks in Aucklanders.

The protective effect of alcohol on coronary arteries appeared to start at just a few standard drinks per week and the well-known harms of drinking (for example, on risk of injury, cancer, stroke, liver disease and heart muscle) didn't appear to outweigh the coronary artery benefits until people drank more than about three standard drinks per day.

However, about five years ago my colleagues and I decided to reassess this evidence, in the light of new studies suggesting that we are not very good at separating the heart benefits of one factor, like light to moderate alcohol consumption, from associated healthy lifestyle factors.

Our re-analysis suggested that the apparent benefits of light to moderate drinking on coronary arteries were more likely due to the many other healthy behaviours that light to moderate drinkers engage in.

To confuse the picture a little, the evidence still supports a protective effect of alcohol on coronary arteries if you drink a lot of alcohol - more than two to three drinks per day.

Unfortunately it appears that by the time you drink enough to get the coronary artery benefits, you are drinking enough to cause more overall harm than benefit. Many alcoholics, for example have surprisingly clean coronary arteries.

So how did I lose out on my regular invitations to international meetings?

Well, my colleagues and I made the mistake of publishing our reanalysis in the Lancet, an international medical journal, in a paper entitled "Alcohol and ischaemic heart disease: probably no free lunch".

Although I thought the academic meetings I attended and the academic groups I joined were independent of the alcohol industry, in retrospect I think the industry must have been indirectly supporting many of these meetings about the benefits of light-moderate drinking.

The academics involved have not been "bought" by the industry, they truly believe in the cardio-protective effects of light to moderate drinking - I know I did!

However, it is perhaps not surprising that the alcohol industry would do everything they could to support researchers, conferences and publications presenting information on the benefits of drinking.

But when did you last read a headline about the harms of light to moderate drinking? A recent large French study supported our reanalysis but perhaps because there is no large, well-funded industry that would benefit from a study suggesting we should drink less, it didn't make many headlines.

My change of heart has led me to modify my previously moderate alcohol consumption. I used to drink two to three glasses of wine with dinner most nights, mainly because I liked it, but also because I liked the idea of partaking in something that tasted good and was also good for my health.

Now I try to drink about one to two glasses every other night and just for the pleasure rather than my heart health.

So, how do I explain the French paradox? The French have a high fat diet and are heavy smokers but have relatively low heart attack rates.

French men do have fewer heart attack deaths than New Zealand men probably because they drink more alcohol, but New Zealand men live longer probably because they drink less alcohol.

* Professor Rod Jackson is at the School of Population Health, University of Auckland.