My understanding is that the decent longitudinal studies fail to take account of the fact that a fair proportion of people who don't drink AT ALL are ex-alcoholics. But I haven't seen these studies in particular.

To be honest, in the UK, not drinking at all is a very strong signal for either being a Muslim or having some reasonably serious personal whackiness and non-conformism. Both of those seem likely to me to be at least as likely to explain health effects as the booze itself.