Cancer researchers at the University of California at Berkeley report what could be the first laboratory evidence showing how moderate levels of alcohol may trigger cancer-causing changes in human cells.

The study, reported in the most recent Proceedings of the National Academy of Sciences, emphasizes that drinking has not been found to directly cause cancer, but the research may provide a plausible explanation for apparent links between drinking and some cancer."Alcohol has been implicated in a very indirect way for a very, very long time with cancer, but nobody could show exactly how it could be responsible," Bea Singer, a molecular biologist, said.

Singer and her husband, molecular biologist Heinz Fraenkel-Conrat, were trying to find logical explanations for the connection between alcohol and cancer.

The rate of breast cancer, for instance, was found to be two times higher among women who drink compared with those who do not imbibe. Higher-than-average rates of liver, rectal and oral cancers also have been found among people who drink moderately.

Most carcinogens alter nucleic acids, the building blocks of genes, leading to cancer-promoting mutations in the DNA instructions that control gene growth. Alcohol alone, however, has no effect.

But the new studies show that a breakdown product of alcohol when it is metabolized, called acetaldehyde, may work in concert with alcohol to attack genes.