My opinion: I think there are some hidden reasons that may more easily explain this phenomena. Drinking probably doesn't help longevity directly, but it can in other manners. I assume that this study has documented the lives of people who are middle-aged or older. These people have had a long time when they were legally allowed to drink, as opposed to younger adults. I also would assume that drinking may become addictive early on in life, rather than later. Thus, the people who were moderate drinkers have probably always been moderate drinkers - they have the ability to control the amount that they drink. Hence, because of this control, they may also be able to control other lifestyle choices that prevent them from dying early. Drinking may also give them a sense or relieve, reducing stress, which in turn may lead to sicknesses. Non-drinkers, on the other hand, may not have such experience with control and are unable to find other ways to relax. Does this explanation sound reasonable, or does alcohol directly benefit body chemistry? Feel free to comment.
Alcohol and Life Expectancy: Unraveling the Mystery of Why Nondrinkers Have Higher Risk of Premature DeathPatrick M. Krueger
Prior research has documented that both heavy drinkers and nondrinkers have higher risks of premature death than their peers who drink in moderation. Heavy drinkers have elevated rates of death from accidents, suicides, homicides, liver disease, and some cancers. But the reason for the elevated rates of death among nondrinkers is less well understood. Some researchers* have advocated for national guidelines that discourage nondrinking and encourage moderate alcohol consumption. But physicians are reticent to suggest that their nondrinking patients drink more, because alcohol is a nonessential part of a person’s diet, is disallowed by many religions, and can have adverse consequences for health if consumed to excess.
Our findings show that diverse groups of nondrinkers also have diverse rates of death. For example, adults who quit drinking because of their histories with problem drinking or for health reasons have among the highest rates of death among the nondrinkers. As such, the subgroups of nondrinkers who have the worse mortality outcomes would likely have even higher rates of death if they were to begin drinking. In contrast, adults who have consumed very little alcohol throughout their lives, due to interests in being responsible family members or for moral reasons, have mortality risks that are as low as those who drink in moderation. Thus, nondrinkers who avoid drinking for the most positive reasons may have little to gain, in terms of further reducing their mortality risk, if they were to begin drinking in moderation.