One thing you learn from reading depression papers is that there are a lot of things that supposedly have significant effects on depression according to some papers, but then fail to work when people try them or they’re tested in trials.

The linked paper is pretty obvious AI paper mill content, so I wouldn’t take anything it says as directly true without checking citations. The citation is a meta-analysis so you’d have to check the criteria and the studies.

The most common explanation for the “everything treats depression” result is inclusion of studies that have no control group. The placebo response rate in depression studies is very high as many patients revert to the mean of being not-super-depressed after they are so depressed that they enroll in a study for depression treatments. Paper writers seeking a positive result will abuse this to their benefit by omitting the control group. They’ll collect depressed patients, give them some treatment, and when the average improves they’ll publish a paper saying that the treatment has positive effects in depression. So after reading a lot of papers about depression, you start dismissing claims by default unless you can confirm they came from a properly powered, placebo-controlled, double-blind study.

But hey, if you’re already drinking coffee and enjoying it, continue to enjoy it!