Saying something is ok in moderation always defines moderation relative to the harm. This study puts moderation very low ~1-2 glasses of wine a month. Studies focused on different forms of harm have different thresholds, but you get all forms of harm not just what a specific study looks at.
Saying I accept the harm for this benefit is fine, but pretending it does not exist is irrational.
Do you have the study? Not some newspaper articles. It feels very unlikely to me that they were able to prove harm from 3 glasses a month, I'd like to read that.
[edit] In the article cited in NPR, https://www.thelancet.com/action/showPdf?pii=S0140-6736%2818... , which makes the claim "zero consumption minimises risk", you can see that ischemic heart disease & diabetes risks actually go down for 1 glass a day, while cancer & tuberculosis go slightly up. Their own chart of "relative weighted risk from all attributable causes" is flat-line for the first 1 glass/day. And that's from the paper that's cited to say "no level of alcohol consumption is safe!"
Look at the "light drinking"[] result in figure 2 - most whiskers cross the 1.0 line when it's not downright in the left-hand side! (e.g. lung cancer relative risk - if you trust that result, it means light drinking _improves_ your lung cancer prospects). They flat out state that lymphoma risk is inversely correlated with drinking. AND it's only about cancer - no word about other diseases (e.g. ischemic heart disease, where there are studies showing light wine drinking improve your odds).
Overall, your meta-study hardly proves that "any quantity of alcohol is bad for your health"; quite the contrary in fact, I'd say.
[] 1 glass of wine a day according to their definition.
Each cancer type is independently analyzed independently. 1 wine a day measurably increases overal cancer risks as stated by the authors. Rather than assuming you have a better understanding after a quick analysis you may want to rethink your analysis.
Hint the base rate for each cancer is not equivalent.
Beyond the original statement. “Every category of drinking was associated with an increased risk of cancer.”
“In the present study, alcohol drinking was associated withcancer of the oral cavity and pharynx, oesophagus (SCC) andfemale breast even at low doses. These results represent an updateand corroboration of previously published findings on the linkbetween light alcohol drinking and cancer (Bagnardi et al, 2013).Given the high proportion of light drinkers in the population andthe high incidence of these tumours, especially breast cancer(Ferlay et al, 2010), even small increases in cancer risk might be of great public health relevance. Approximately 5000 deaths from oraland pharyngeal cancer, 24 000 from oesophageal SCC and 5000 from breast cancer were attributable to light drinking in 2004worldwide (Bagnardi et al, 2013).”
Yes, some cancers where negatively associated with drinking, but the net effect was negative over the range of cancers studied.
As to light drinking and heart disease, that link is under some question as to the relative gain from drinking vs associated lifestyle factors. Short term controlled studies suggest some benifit, but it’s less clear for younger drinkers and people with good cardiovascular heath.
At a guess, the benifit of moderate drinking might win somewhere between 40 and 70 depending on specific risk factors.
Ignoring the harm of the alternatives is also irrational. If you remove the wine you replace the calories with something else, and that will have its own risk numbers. Some studies sort of average that in but an average isn't nearly as good as numbers for specific items. You can't reasonably extrapolate total risk from one or two diseases either.
Dealing with tiny risk factors is very difficult, and it's easy to make the wrong conclusions. See also the vast majority of reporting on whether x food is good or bad for you.
Saying I accept the harm for this benefit is fine, but pretending it does not exist is irrational.