Across 26 trials in 2,160 people taking antidepressants too, omega-3 modestly beat placebo, and nearly all of it came from capsules built on EPA rather than DHA.
Every trial gave the capsules alongside antidepressants rather than instead of them, so what was measured is an add-on effect. The EPA-heavy figure rests on a smaller set of trials than the overall one, which makes it the top of a range rather than a settled number. A correction to this analysis was published two years afterwards, so its exact figures are worth re-checking. Depression trials also carry a strong placebo response, which pushes small differences around.
