Hacker News new | ask | show | jobs
by Conlectus 1426 days ago
This is irresponsible reporting. The authors themselves have clarified that this is an _association_ study, whereas the above article claims that isolation is _causal_ to increased dementia risk.

From the authors:

> We agree with Dr. Kawada that no casual conclusion should be made in the present study, as it is an association study. We avoided using terms that imply causal inferences in the paper.

5 comments

Indeed. The notion that decreased cognitive function is the cause of social isolation, not the result, seems equally plausible.
Or entirely correlated like ice cream sales and drowning.
Or quite well correlated like flying kite with metal strings during a storm and getting hit by lighting.

After all we know that people are social animals, that we acquire language and reasoning abilities by interacting with others in childhood (and children that are prevented from doing that have developmental problems), and that forced isolation in adults (in prison, lost in some forrest, or whatever) also brings several psychological and mental problems, and can even be used as punishment and torture.

If "decreased cognitive function" was the cause, then it should have been there before (or at the very beginning) of the social isolation (to cause it).
Direct link to the actual study:

https://pubmed.ncbi.nlm.nih.gov/35676089/

Aside from these issues in the reporting, I think we would need to control for active reading, and for active remote interactions with other people, whether via pen, telephone, or video call. Also whether the social isolation is being caused by the dementia.

Not that this is an easy topic to study, but although the scientists are experts, at least the article seems to show some large holes in the reasoning.

Right, association just means you find two things together, not that one drives the other.
While I agree, this caveat is true for a huge number of studies posted on HN, if not most of them.
And it should therefore be pointed out every time.
Or taken for granted and spare us the triviality
That's why (good) science is difficult and why good study designs matter.