Unironically, I'd love to see a sourced meta study from ideally the last 3 years that concludes that, which is based on long-term studies, which were properly peer reviewed, and not funded by the drug manufacturers themselves.

[–] zhengman777 0 points 2 days ago (1 children)

Sure, this one probably comes closest to what you’re looking for. Looks like modest benefit. Nothing super effective but enough to make a positive difference generally. I definitely recognize that they can have side effects, so decisions should be personalized.

https://onlinelibrary.wiley.com/doi/10.1111/acps.13541

There are some other older ones done before 2022 as well like this one.

https://psychiatryonline.org/doi/10.1176/appi.focus.16407

[–] [email protected] 1 points 2 days ago* (last edited 1 day ago) (1 children)

For this one:

https://onlinelibrary.wiley.com/doi/10.1111/acps.13541

We selected 46 RCTs out of 1807 titles and abstracts screened...

There was no indication of subtstantial small study effects, but 36 RCTs had a high or uncertain risk of bias, particularly maintenance trials.

Going into the actual paper...

10 low bias studies, 2935 subjects.

12 high bias studies, 3547 subjects.

24 unclear bias studies, 9689 subjects.

Cool, so the vast majority of analyzed subjects were not from studies that could be established as having a low bias.

Oh hey, remember when I asked for a meta study that didn't include studies done by or funded by drug manufacturers?

All but four studies were funded, partly or wholly, by drug manufacturers.

Ok, so you obviously either did not read what I asked for, or you didn't read the paper.

Further, I said long-term, only 7 of the 42 studies are about maintenence stage, you know, long term.

And...

We did not carry out a RoB analysis in maintenance RCTs because all carried a high risk of bias.

Wow! Amazing!

This study does not even kind of come close to the conditions I specifically laid out.

... Onto study number 2.

https://psychiatryonline.org/doi/10.1176/appi.focus.16407

46 (9%) of 522 trials were rated as high risk of bias, 380 (73%) trials as moderate, and 96 (18%) as low;

Cool, 18% of the studied trials were low bias this time, roughly in line with the other meta study.

409 (78%) of 522 studies were funded by pharmaceutical companies.

Awesome.

[–] zhengman777 1 points 1 day ago

At least in the first study, they did an analysis of the non-pharm funded studies and saw some good results. Unfortunately that’s pretty much the best we can go off of.