The interesting thing about dementia is that it is not normally distributed whatsoever. Pick basically any characteristic imaginable - environmental, behavioral, or genetic, and you're going to find a difference, often very significant, between groups.

Everything from your occupation, to your diet, to martial status, to hobbies, and much more have been shown to have significant relationships with dementia rates. The problem you obviously run into here is that a person's approach towards healthcare is itself a major behavioral characteristic and so seeing varying rates of dementia based on this characteristic alone would be very unsurprising.

> seeing varying rates of dementia based on this characteristic alone would be very unsurprising

Sure. That's why the antibody cross-reactivity is intriguing.

You're begging the question. While endless things are associated with dementia (or its absence) nobody knows what causes it, and so looking for causal reasons with behavioral characteristics is going to mislead without carefully controlled experimental (and not observational) studies.

The shingles vaccine is another promising one[1]

> Researchers found that compared to those who didn't get the shingles vaccination, those who received it...

> were 3.5% less likely to develop dementia over seven years (a 20% reduction)

These types of findings are problematic for anti-vaxxers, however they seem likely to overcome this through their wholesale butchery of US research.

[1] https://www.health.harvard.edu/blog/can-a-routine-vaccine-pr...

I think it's going to be very interesting to see what kind of effects long term antiretrovirals for HIV treatment and prevention have. Surely they're not so tightly targeted that they only affect HIV, so it'll be interesting to see which conditions they prevent and which they cause, since we've got tons of people on them for a quarter century now or so.

I also wonder this kind of thing about other long term treatments - perhaps Prozac prevents dementia, or causes it.