Review on 30146330

This is review on https://pubmed.ncbi.nlm.nih.gov/30146330/

Sometimes the harm of moderate alcohol consumption is justified with article https://pubmed.ncbi.nlm.nih.gov/30146330/
However it's possibly wrong for the following reasons:

I didn't find any information in this article about all-cause mortality.

They don't study it. They only study the effects of alcohol on 4 things:
- coronary heart disease risk (10 to 30g of ethanol a day is optimal... and even 60g ethanol a day is as harmful as 0g ethanol a day (p.10).
- Risk of diabetes (10 to 20g ethanol per day is optimal... at 35g the risk is about the same as at 0g)
- risk of oral cancer - yes, it goes up with increased alcohol consumption
- Risk of tuberculosis - yes, also goes up.

And then they take those 4 risks and add them up. Most likely, they simply add up mechanically, without taking into account the fact that the risks of CHD and diabetes are much more important and contribute more to mortality than the risks of tuberculosis or oral cancer (well, at least in developed countries). And they get that their "overall" risk increases monotonically with increasing alcohol consumption.

And they themselves write in the Discussion section that earlier studies showed a positive role of alcohol, but more modern studies increasingly show either a statistically insignificant positive role of alcohol, or no effect of alcohol on all-cause mortality ("More recent research, which has used methodologies such as mendelian randomization, pooling cohort studies, and multivariable adjusted meta-analyses, increasingly shows either a non-significant or no protective effect of drinking on all-cause mortality"). So no claims on negative role of moderate alcohol consumption.

Another drawback is that they give their general main graph (Fig.5) without separating males and females. So perhaps the harm of 20g of ethanol a day could be because of its harm to women, who can be really harmed by that much amount (omitting the question of what's drawn in Figure 5, it's not all-cause mortality).