This kind of thing is not science, but digging through data to look for “associations”. There are associations in any big dataset. There is no reason to infer from this that there is any cause-and-effect relationship whatsoever. But it can be an amusing exercise to think up confounders. Are patients with incipient dementia advised by their doctors to abstain from alcohol? Probably.

The limitations of prospective observational studies will generally be the same regardless of the topic, there are few controls, a lack of randomization, and mechanisms of action are only proposed and not directly tested for (in this case the paper doesn't suggest any mechanism of action at all). This paper simply states that there is a correlation with alcohol abstinence and all-cause dementia. That claim alone is probably true, but without randomization or controls dementia can not be directly attributable to alcohol consumption because there could be all types of confounding factors. As an extreme hypothetical imagine if in a fictional society only the wealthiest .01% could afford alcohol, nearly any observational study will show better health outcomes for alcohol drinkers if you don't control for wealth. In reality observational studies have to control for many, many more factors to be considered strong evidence of causation.

From the paper: "It is difficult to provide evidence of causality in observational research, but triangulation of evidence from alternative study designs and across different research settings has been identified as a promising avenue ... Among current older drinkers, the current study failed to find a consistent relationship between alcohol use and dementia across different contexts, some of which are likely to be subject to quite different confounding structures Meanwhile, Mendelian randomisation studies to date have generally failed to demonstrate a causal relationship between alcohol use and dementia, except perhaps with respect to age of onset for dementia. While there are limitations associated with the current study, as well as the studies based on Mendelian randomisation, strong evidence for the causal link between alcohol use and dementia remains elusive."

Alcohol itself lowers blood pressure. Alcohol in various types of drinks comes along with other compounds that are variously healthy and unhealthy. There's been a fair amount of research published showing moderate alcohol intake is healthy, and unhealthy. It's likely some kind of local optimum and there are better choices to be made regarding health, health span, and longevity.

There's fairly solid evidence that mild to moderate drinkers live longer. This is usually attributed to better socialization, which is a major factor in longevity, i.i. people with more robust social networks live longer and people who are lonely tend to die.

https://www.alcoholproblemsandsolutions.org/drinkers-live-lo...

Another possibility is that healthier people tend to drink, though overall health is usually controlled for in studies like these.

The data is observational, which generally means you should ignore it. It is too noisy.

There were those studies that showed moderate alcohol use improved health and only heavy drinkers saw detrimental health effects. The problem was that "no drinking" group included people that weren't drinking because of poor health. Later studies compared drinking vs. a "no drinking" sample of people that drank around two glasses of wine per year. The improved health effects completely disappeared. The more you drink, the worse it is for your health.

So this study is like that in using a potentially unhealthy comparison group. They try to offset that a little by also throwing in people that quit drinking. But it is likely that some people quit drinking because of health problems. So I'd guess that this study has the same problem with an unhealthy comparison group. The study probably can't tell you what the actual relationship between alcohol use and dementia with any authority.

As someone who has never even tasted alcohol, I am here to say I hope you are correct.

At face value, it seems pretty likely that people with dementia are actively discouraged from drinking by those around them, so an association seems very likely to me, though causation would be in the opposite direction. Didn't read enough of the study to see if they attempted to control for this in any way.