This seems like a very easy one to point out that they may be getting the cause and effect backward.
They've done the analysis: meta-analysis of random controlled trials , see https://www.bmj.com/content/384/bmj-2023-075847 "Exercise is an effective treatment for depression, with walking or jogging, yoga, and strength training more effective than other exercises, particularly when intense. Yoga and strength training were well tolerated compared with other treatments. Exercise appeared equally effective for people wi…
A.) you have to actually do the exercise, which means the people who don’t may simply not do so because their mental illness is worse or at least different than the ones who do.
2. If you’ve ever read the attempts to quantify various mental illnesses, they’re extremely unscientific and borderline laughable if you’ve ever known a few people who suffer from them. What we call “depression” affects so many people in so many ways that in the future, it’s likely it will become a category of diseases with many different individual names. One of my exes medicated her depression by working, one by calling off work, for example. One would have (and did) do lots of exercise, one did not, and there’d be no way to quantify either on any simple scale.
3.) It’s obviously neither double nor single blind. You can’t not know if you’re exercising, and the researcher can likely tell which are which to a decent degree just by looking at them.
All a meta analysis of a bunch of non-controlled studies really tells you is the prevailing pre-sentiment among the researchers of a topic. Which, really, isn’t nothing (they’re more likely to guess correctly than a random person) but it’s far from an actual controlled study.
And I still stand by that there’s no way you could realistically say this study indicates causation with any certainty. Though that never stops people in the social sciences I guess.