Activities like mass observation (uk diary program run out of Sussex university, dates back to 1937 with interruptions) can help with trend analysis.

I used to complete a national happiness survey for Australian social science research, from memory it was annual for a while.

The main problem I feel is the confounding behaviours and like problems. Even being asked "are you lonely" can incur burdens. FOMO can feed it, just as much as being online all the time can feed it.

That. And the immensely subjective nature of how people speak, report, ideate how they feel.