[The cancers for which increased risk was alleged were cancers of breast, liver, rectum, and "upper aero-digestive tract." Not one of the cable journalists mentioned that the study had also shown a decreased risk of thyroid cancer, non-Hodgkin lymphoma, and renal cell carcinoma.]
See for example, "Moderate Alcohol Intake and Cancer Incidence in Women," by Naomi E. Allen et al., at http://jnci.oxfordjournals.org/cgi/content/abstract/djn514, and "Million women study shows even moderate alcohol consumption associated with increased cancer risk," at http://www.physorg.com/news154715618.html .
Basically, the claim is that the expected rate of cancers is 118 per 1000 middle-aged women, but with one drink a day you get 15 "extra" cancers of the varieties listed.
I haven't been able to get access to the original article, which was in the February 24, 2009 online issue of the Journal of the National Cancer Institute; it may be that it takes a more conservative position. Nevertheless, I would very much like to have your input on this, especially with regard to the statistics. The Allen et al. article tells us that "Cox regression models were used to estimate cancer risks associated with alcohol consumption after adjustment for other risk factors," but doesn't provide any information about what those risk factors were. And I don't pretend to understand Cox regression models.
What this most reminds me of is the Nurses Cohort Study that set off a similar endless series of media reports warning women that even a few pounds of weight gain put them in serious danger of a heart attack, without so much as mentioning, even once, the environment of incessant stress and hostility that nurses typically have to work in.