Not to mention it doesn't say anything about when. Is that an alcohol related death in their 30s, 50s, or 80s? What is the impact of removing DUI-related deaths for drinkers who don't drive? Hell, does it count people who drink more than a drink per day but get killed by someone else driving drunk?
I gotta say that this falls below the risk threshold that I care about. Actually, I wonder if that is the intent, like a bit of reverse psychology by saying "don't do this dangeous thing, or you risk increasing split ends in your hair by 12%!" when it is actually trying to discredit people talking about the legitimate risks by treating trivial ones as serious. Though probably not because most people would probably just take the main message rather than look closely enough to see the misdirection.