For years we've known that people who drink no alcohol at all have worse outcomes than people who drink some alcohol. We also thought that drinking a very small amount of alcohol - a glass or two of wine per week - was protective but we didn't think more alcohol than that was protective.
Recently a study was released and heavily reported.
One reason we thought that tee-totalers had worse health outcomes was the number of people with alcoholism in that group or people with other severe health problems. This new study claimed to have corrected for that and only included healthy people in the teetotal group. The study then said that drinking no alcohol was associated with dying sooner than drinkin alcohol. The curve they released showed most benefit to people drinking one drink a day, but showed benefits over not drinking at upto about 5 drinks a day.
That was different to what we thought before.
It turns out this new study has a bunch of flaws. There are some interesting effects. Older women who drink do see some protective benefits. But the health benefits of moderate drinking for most people are not at all clear.
It's pretty important to publicise the corrections because of the heavy reporting of that flawed research, and because of the misunderstanding (that you repeat) that moderate drinking provides health benefits.