A comparison of 20 antidepressants across 115 trials found three did best at keeping depression from coming back, and the most famous one wasn’t among them

Most comparisons of antidepressants look at the first few weeks of treatment: who gets better, and how fast. But many people stay on them for years to keep depression from returning, and little was known about which do that best. An international team led from Radboud University Medical Center in the Netherlands pooled 115 randomized trials with more than 30,000 participants. They compared 20 antidepressants with each other and with fluoxetine, better known as Prozac, over both short- and long-term use.

Three antidepressants did better than fluoxetine at getting people well and keeping them well over the long term: escitalopram, agomelatine and duloxetine.

The study also confirmed that antidepressants in general help sustain improvement during long-term use, not only in the first weeks.

“Escitalopram is already widely used,” said first author Paul van Haaren. “Agomelatine works well, but one of its side effects is the risk of liver failure. Duloxetine is less well known and is prescribed less frequently as a first-line treatment.” Agomelatine is also not approved in the United States.

Escitalopram, sold as Lexapro and Cipralex, belongs to the SSRI class that most guidelines recommend as a first choice. Duloxetine, sold as Cymbalta, is an SNRI, a class that also acts on noradrenaline and is used for some kinds of pain as well as depression.

“Of these SSRIs, escitalopram now appears to be the most effective, also for long-term use. This study may have implications for future treatment guidelines,” said Eric Ruhé, professor of difficult-to-treat depression at Radboud university medical center.

The best-known earlier comparison looked only at the short term. In 2018, a network meta-analysis in The Lancet led by Andrea Cipriani compared 21 antidepressants across 522 trials and more than 116,000 people, and found that all of them worked better than placebo over roughly eight weeks, with escitalopram among the more effective and better-tolerated options. The new study asks the question that matters for anyone who stays on medication: what happens after that.

The method was a network meta-analysis, which combines direct comparisons from trials that tested two drugs head to head with indirect comparisons through shared comparison groups, so that drugs never tested against each other can still be ranked.

The second finding is about how to use the results. Ruhé stresses that this does not mean everyone with depression should switch to escitalopram. “As a prescriber, you consider the balance between a medication’s benefits and side effects. That balance differs from person to person,” he said. “Nevertheless, our study can help inform the decisions that prescribers and patients make together: if you start an antidepressant, choose one that not only helps you to get well, but also keeps you well.”

For anyone taking an antidepressant that works for them, the findings are not a reason to switch, and stopping or changing medication should only happen with a doctor because of the risk of relapse and withdrawal effects. For people starting treatment, long-term effectiveness is now a reasonable question to raise. New tools are also changing how mental health risks are spotted, from AI that predicts PTSD after injury to targeted treatments for binge drinking.

There are limits to what this shows. Network meta-analyses depend on the quality of the trials they combine, many of which were funded by drug makers and lasted for limited periods. Differences between drugs were average effects and may not apply to any one person. The press materials do not report the size of the differences, so how much better the top three performed is not clear from them.

Source

van Haaren P, et al. “Comparative efficacy and acceptability of antidepressants in the long-term treatment of major depressive disorder: a systematic review and network meta-analysis.” The Lancet Psychiatry, 2026.
DOI: 10.1016/S2215-0366(26)00326-3

Also cited

Cipriani A, Furukawa TA, Salanti G, et al. “Comparative efficacy and acceptability of 21 antidepressant drugs for the acute treatment of adults with major depressive disorder: a systematic review and network meta-analysis.” The Lancet, 2018.
DOI: 10.1016/S0140-6736(17)32802-7