In 1970, the seven countries study1 showed there was a relatively lower rate of coronary heart disease in Mediterranean countries and that this difference was attributed to diet. In 1980–95 many studies demonstrated a favourable effect of moderate alcohol consumption on the incidence and secondary prevention of coronary heart disease. If one considers that wine consumption may have some protective effects against coronary heart disease, a number of questions arise:

• is there any specific property of wine compared with other alcoholic beverages?

• does wine have an advantage over fruit and vegetables that are also rich in flavonoids?

• if so, should we recommend red or white wine or, perhaps a wine from a precise appellation, and what might be the daily “dose”?

 To answer the first question, only transverse or retrospective studies are possible; a prospective study seems impractical—imagine a cohort of many thousands randomised for many years to water or to a daily amount of some alcoholic beverage without any flavour, or to the “wrong” wine associated with the same food in all groups! What about dropout, cost of drinking, addictions, overdose, cross over, and side effects? Furthermore, it would be difficult to select wine drinkers and abstainers on a random basis—for example, in the case of our hospital, volunteer wine drinkers might belong to the red wine group because the hospital is located along the vineyards of chateau Haut Brion and many of the (voluntary) abstainers might be slim vegetarians or addicted to rigorous physical training and by no means representative of a control group.