The cited study count two factors:
>Alcohol attributable deaths(AAD) 28.5 per 100,000 population
>Years of potential life lost (YPLL) 823 per 100,000 population
Thats like 3% of the 10% that maybe actually died directly from an alcohol related cause.
My first question is what constitutes AAD? I could find no definition but given that this appears to be a study mainly directed at demonizing alcohol, use I'll make my own assumption that it is alcohol poisoning, drunk driving accidents, accidental deaths which alcohol is mentioned etc.
YPLL? Heck this is even a more vague category. Most likely overall longevity among drinkers compared to non drinkers. Not related directly to drinking, but hey you died so it counts. Got hit by a bus stone sober crossing the street, but you were a drinker? You still add to the YPLL total just like the guy with liver failure.
Then to take it one more step removed. The Alzheimer's research folks are now claiming that about 1/3 of deaths in senior citizens are Alzheimer related. John Doe died of heart failure, but they also want to count it as Alzheimer's. Why? Because John had Alzheimer's. Not that the condition had any direct causal relationship, he had it, so we can count it in our total. Cancer folks do the same, along with a whole host of other groups pushing for awareness and funding. If you add them up I would not be surprised if they add up to 300% of deaths caused by one thing or another.
If you have cancer, drink alcohol and have Alzheimer's, does that count as a triple word score?