Here's the latest on the relationship between all-source mortality and alcohol consumption: a U-curve rather than a J. That is to say, heavy drinkers and abstainers have roughly comparable mortality rates, with moderate drinkers doing best. Compared to moderate drinkers, and correcting for a host of confounding variables including whether current abstainers are former drinkers, they find abstainers have a 49% increased mortality risk while heavy drinkers have a 42% increased mortality risk. So it's worse to be an abstainer than a heavy drinker, in their sample.
It's just one study though; I'll stick with the meta-study consensus of a J-curve rather than a U-curve, with a best estimate of about a 16% reduction in all-source mortality with moderate drinking but elevated risk for folks having more than about 4 standard drinks per day. But I'll look forward to estimates from future meta-studies that include this new work.
Will the Ministry of Health ever acknowledge the existence of the J-Curve? Unlikely.
HT: Time, via @mitchellhall