Fellas, if you ever want to have a relaxed drink with your wife or sweetheart again, you’ll need to put a stake through the heart of this junk science.
A new study by Harvard researchers in the Journal of the American Medical Association reports that moderate alcohol consumption (3-6 drinks per week) increases the risk of breast cancer by 20 percent — i.e., from 2.8% to 3.5% over a 10-year period.
This study is entirely unreliable for the simple fact that the researchers have no idea how much alcohol any of the women actually consumed.
It is another from that perpetual junk science machine known as the Nurses Health Study. Ever since the mid-1970s, Harvard researchers have surveyed 121,700 registered nurses on their dietary and health habits every two years.
The data is self-reported, not validated and, hence unreliable. This is especially in evidence with its alcohol consumption data.
The survey questionnaire asks for daily, weekly and monthly alcohol consumption in terms of number of cans/bottles of beer, glasses of wine and drinks of liquor.
Even if someone could remember precisely how many cans, bottles, glasses and drinks they consumed, the outstanding question would still be the amount of alcohol in all those beverages.
Beer, wine and liquor all have varying percentages of alcohol and people consider different volumes as one drink. And the more one drinks the less reliable a recordkeeper one is. Survey respondents may also tend to underreport drinking for fear of social stigma.
Considering that the study result is a weak association — i.e., in the statistical noise range — the unreliability of the alcohol consumption data makes this study another for the circular file.
The researchers also acknowledge that they have no idea what the biological explanation for alcohol-cauing breast cancer might be. They hypothesize hormone changes, but they really don’t know anything with anything approaching certainty.
Even the accompanying JAMA commentary merely concludes that more research in a broader context of all-cause mortality is needed. Moderate alcohol consumption, after all, may also have beneficial health effects that might be lost by cutting back.
But none of this will stop junk scientists Walter Willet, Graham Colditz et al. from using this dog to scare women.