Then came a study suggesting that alcohol may reduce susceptibility to the common cold. What could be nicer than a wee dram to make winter more bearable?

The research was done at the Medical Research Councils Common Cold Unit in Salisbury, England, by members of that unit and of the MRCs Addiction Research Unit, and by investigators from Carnegie Mellon University in Pittsburgh, Pennsylvania, and the University of Wales College of Cardiff. It sought, under controlled conditions, to test the view that both smoking and drinking increase susceptibility to viral infections. Instead, while smoking was indeed incriminated, consumption of alcohol was found to reduce the risk of infection, particularly among nonsmokers.

The experimental procedure is reminiscent of work done by Louis Pasteur over a century ago. Pasteur, tired of hearing his mother admonish him not to get his feet wet lest he catch le rhume or, worse, la grippe, kept one-half of a flock of chickens warm and dry, but forced the rest to stand around in their bare feet in cold water. He sprayed both groups with cold viruses. Each group contracted the same number of colds. So much for Pasteurs mothers advice! (It is not recorded whether any of the chickens smoked. It is certain that none were in contact with wine until in the pot.)

The Medical Research Council studied 154 men and 263 women, all healthy volunteers, who were between 18 and 54 years of age. Pregnant women were excluded.

The subjects were given nasal drops containing one of several cold viruses (rhinovirus types 2, 9 or 14, respiratory syncytial virus, or coronavirus 229E). Some were given saline as controls. They were quarantined from two days before until seven days after the viral exposure. The study was conducted in the usual "double-blind" fashion to ensure objectivitythat is, neither the subjects nor their examiners knew who got what. Measured results included recovery of viruses from subjects (evidence of active infection), measurements of virus-specific antibody levels (indicating an immune response to a recent infection), and examination for evidence of clinical infections (colds). These meticulous investigators went so far as to weigh daily mucus production.

As expected, cigarette smokers had increased risk of infection and increased symptoms if infected.

The authors were surprised to discover alcohol consumption to be associated with reduced risk of infection. Moreover, this observed benefit was dose related: each increase in drinking, up to three or four per day, was associated with a decreased risk of illness.

The benefit of drinking was modified by smoking. Smokers derived little if any benefit from drinking. Nonsmokers who did not drink experienced nearly the same frequency of infections as did smokers. As nonsmokers consumed more alcohol, their risks progressively diminished. There was suggestive evidence that drinkers actually became infected, but that they resisted the viral attack, and so remained well.

No one has any sound theory explaining alcohols protective effects (which appear not to discriminate among various beverages). It is established that chronic heavy drinking may adversely affect a number of bodily functions related to immunity. Moderate drinking has no such effect. Conjectures of the MRC researchers include possible limitation of viral replication or of the inflammatory process by alcohol.

Others speculate that antioxidant properties of the polyphenolic flavonoids suspected of protection against heart disease and cancer might also be operative against infections. This could not apply to all beverages, for beer and wine, red wine in particular, is relatively rich in these compounds, but spirits ( with the exception of cask aged spirits) contain little, if any.

To my knowledge, information of interest on interrelationships of alcohol and viral infections have been limited to two reports, both thus far unsubstantiated and of uncertain import. Wines and spirits, but not beer, consumed with raw oysters contaminated with hepatitis A were associated with a reduced risk of infection. Alcohol appeared to speed the replication of the human immunodeficiency virus of AIDS in vitro (in laboratory test tubes). This has no relevance to catching the disease.

The take-home message about colds is clear: drink moderately; dont smoke cigarettes. And chicken soup wouldnt hurt. But I must caution against relying upon alcohol as prophylactic or curethe data remain ambiguous and the prospective consequences possibly mixed.

Harvey Finkel is Clinical Professor of Medicine at The Boston University Medical Centre, and Chairman of the Committee on Health for The Society of Wine Educators.