In France deaths from heart attacks are about a quarter of the UK's. Epidemiologists have struggled to account for this, particularly as the usual culprits (high levels of animal fat consumption and high cholesterol) cannot explain the difference. Wine and alcohol seemed to offer a plausible solution to the paradox.
This fits with the observation that across Europe there is lower heart disease, the more wine that is drunk. But, as a recent paper in the British Medical Journal shows, this may be an unreliable guide.
The key feature of the new analysis is time. Consumption of animal fats and cholesterol concentrations is now similar in the two countries. But this has been true for only a relatively short time - 15 years or so. Until about 1970, the French had a lower consumption of animal fats and lower levels of cholesterol. Only in the decade from 1970 to 1980 did their eating habits and cholesterol levels become similar to those in Britain.
So the new idea is that there is a long time-lag between eating a particular diet and its effect on heart disease. It seems likely that a high-fat diet that starts in childhood shows itself mainly in the over-fifties. As with smoking, it can take up to 50 years to cause serious illness. Thus, the prediction goes, the French will eventually acquire death rates from heart disease that are similar to ours.
But does alcohol play no role at all? Anecdote is insufficient; large- scale studies are essential. In the US, when 18,000 deaths from heart disease were investigated it was found that those who drank one unit of alcohol a day had a 20 per cent reduction in risk from a heart attack. Drinking more than one unit did not reduce the risk further.
There is no evidence that red wine has any specific protective components. But there is the other side to drinking alcohol; in excess, it kills. Mortality of French males from all causes is similar to that of British males. They die less frequently of heart disease, but three times more often from illnesses such as cirrhosis of the liver and alcohol-related cancers.
Can we be sure that this new analysis is finally the right answer? Heart disease has a complex set of causes whose interactions are still not fully understood. For example, one research group has suggested that the French have a lower rate of heart disease because of their higher intake of nuts and fibre. In one study, survivors of a heart attack were randomly assigned either a normal or a "Mediterranean-style" diet - rich in fruit, vegetables and monosaturated fats. Over two years the levels of cholesterol and blood pressure in the two groups were similar. The risk of a further heart attack should have been similar, according to the standard view. But it was found that the Mediterranean diet group had a much reduced incidence of heart attacks and mortality. Overall, their death rate was reduced by more than 70 per cent.
Yet another complication is the possible influence of the mother while the baby is still in the womb. There is a strong correlation between under- nutrition of a foetus and heart disease in adult life, though the reasons are unclear. But there is no good evidence that French mothers are better fed than their British counterparts.
So I shall continue to drink my malt whisky, and take miniatures to dinners where only wine is served. (It seems to be regarded as acceptable to drink from these, whereas a hip flask would not be.) Maybe, to be on the safe side, I should also start carrying nuts?