The arrow of time really does help an inference of causation. However with Alzheimer's patients we're not generally taking brain biopsies of old people to surveil for Alzheimer's plaques, so we don't have the arrow of time in either direction.

A valid conclusion can come from studies for which, in this case, there was no funding because of poor rhetoric like this. Correllations which merely suggest causation are indeed sufficient as the focus of hypotheses. Scientists too often forget that a hypothesis is a valid part of the scientific process, and the aversion to adoption of one without an attempt to gather further data about it is shoddy, unprofessional, and more reminiscent of dogmatic orthodoxy than of science. The next time you think to say "correllation does not imply causation", maybe think to ask if anything more can or should be learned about either the correllation or the cause being investigated, and consider that such obstruction of efforts to collect more evidence is precisely why the present data is inconclusive. The fact here, from TFA, is that there was no such problem with the studies' design, but rather with the myopic cognition of those who obstructed them.

>The next time you think to say "correllation does not imply causation"

I didn't write that phrase precisely because people turn off their brains and learn nothing from it! It's become worthless.

The alternative to A -> B is B -> A, or C -> A and B.

You seriously think that running through those possibilities in your head is "dogmatism"?

Who then is the authority on which one of those is privileged and which require further research to validate, when they are all consistent with the evidence so far?