Not sure how much science there is, but I suspect that a lot of drinking turns out to be self-medication for depression (and/or anxiety). If so, it's obviously not ideal, but it may be that in some cases there is no clear superior alternative.