[vanusa]

Interesting graphs.

What's interesting in particular is that for several categories (such as ischemic heart disease) the relative risk does seem to go down rather decisively for up to 6(!) standard drinks per day (which they note is "offset" by the increased risk in other categories -- including cancer).

So going by their own combined/weighted risk in Figure 5 - we see a curve that is essentially flat in the region between 0 and 1 - that is, showing negligible risk (asymptotically zero) at 1 drink per day, rising to more concerning levels starting at 2 or 3 per day.

And that's the decisive issue here: "Is having an average of 1 drink per day really any more risky than 0 per day?" The only take-way I can get from the article is that there's essentially no difference in risk level in this interval.