If person died from diabetes, but drank, this goes into the "alcohol deaths" count, even if the person would have suffered from it without the alcohol.

If person died from diabetes, but drank, this goes into the "alcohol deaths" count, even if the person would have suffered from it without the alcohol. That would be a monumentally stupid way to conduct a research. Until you can show me evidence of this level of incompetence I'll assume you are mistaken.