My takeaway is that whatever effect there is must be so small as to be meaningless. The issue has been so thoroughly studied for so many decades. Even the cancer risks associated with moderate alcohol barely exceed the confidence interval - we're just collecting noise along the way.