A new review of 87 studies linking moderate drinking to health benefits has concluded that many were flawed, with designs suggesting benefits where there were likely none.
A key problem was how "abstainers", the typical control group, were defined. As “current abstainers”, they can include people in poor health who've cut out alcohol. Unfortunately, only 13 of the 87 studies avoided biasing the abstainer comparison group (65 studies included former drinkers in the abstainer group, while 50 studies included occasional drinkers). Those 13 showed no health benefits for moderate drinkers.
Even without correcting for this, it was actually "occasional" drinkers — people who had less than one drink per week; a biologically insignificant amount of alcohol — who lived the longest.
http://www.eurekalert.org/pub_releases/2016-03/joso-imd031616.php
Stockwell, T., Zhao, J., Panwar, S., Roemer, A., Naimi, T., & Chikritzhs, T. (March 2016). Do "moderate" drinkers have reduced mortality risk? A systematic review and meta-analysis of alcohol consumption and all-cause mortality. Journal of Studies on Alcohol and Drugs, 77(2), 185-198. http://dx.doi.org/10.15288/jsad.2016.77.185