Nope, I wasn't trying to assert that they are. My point was just that you can 'predict' depression to a limited extent by knowing where people live, and that t

Nope, I wasn't trying to assert that they are. My point was just that you can 'predict' depression to a limited extent by knowing where people live, and that the features of (outdoor, unfiltered) photos are a proxy for where people live. That's a pattern that applies at the country level, without controlling for anything. S.A.D. is a thing, and there's an argument that depression risk is higher in cold, sunless places, but I'll leave more-knowledgeable people than me to have that discussion.