I doubt this study very much. It’s a meta-analysis that found that the therapeutic effect of antidepressants is much stronger than the effect demonstrated by individual studies that showed a very weak effect at best. Meta-analyses can confirm the results of a series of individual studies. For example, if individual studies show a weak correlation between some input and some output, meta-analysis should show the same thing and not a strong correlation.

I write this as someone who tried antidepressants and found that their effect was very weak and the side-effects very strong. Interestingly, patients who dropped out of some of these double-blinded studies due to side-effects were excluded from the studies which enabled drug companies to sell a stronger therapeutic effect than is deserved:
https://feelinggood.com/tag/placebo-effect/