This study has a 95% uncertainty interval, and their conclusion is that alcohol consumption increases your risk from: 918 per 100,000 people from 914 per 100,000.

..I'd argue their conclusion is meaningless given the uncertainty interval.