[Alex3917]

> The point is that there's no clear causation in either direction

C.f. my other comment. I think there is good evidence for causation in both directions.

> In fact, the DSM definition of major depressive disorder/episode

I wouldn't use the DSM definitions of depression due to, among other reasons, the fact that there is very little iter-rater reliability:

http://behaviorismandmentalhealth.com/2013/03/30/dsm-5-inter...

I think the article I linked to in my other comment has it correct, where depression should be considered as something that exists within all people on a spectrum rather than as a mental illness. Now obviously if depression gets to the point where it's causing health problems and interfering with your ability to work and enjoy life then that may be a diagnosable issue, but it's a mistake to think of depression as being something that you either have or don't have in the more general case.