FWIW findings about adverse events vary considerably:

> The proportion of patients reporting adverse events ranged from 16% to 98% (median 80%) under SSRI treatment and from 17% to 94% (median 70%) under placebo [1]

So the answer IMO depends on whether we can attribute adverse effects to treatment, which is difficult to disentangle from adverse effects of disease

I’m open to being shown more certainty if available from the literature but meta-analyses are high-strength evidence

[1] https://www.cambridge.org/core/journals/the-british-journal-...