last updated 4/26/2024
You’ve read about antidepressants’ potential risks, and you’ve taken the MoodCheck questionnaire, so you know what kind of depression you have. Now you’re ready to learn more about each antidepressant: how to choose?
Here is a quick comparison of advantages and disadvantages to get you started. I’ve mentioned features that generally agreed upon. Details for each medication then follow below.
generic name/ U.S. trade name | Advantage | Disadvantage |
bupropion/Wellbutrin | no sexual problems, no weight gain, least withdrawal | only use extended release version |
fluoxetine/Prozac | less withdrawal | sexual effects |
sertraline/Zoloft | few interactions with other meds | sexual effects |
citalopram/Celexa | fewer side effects, maybe? | sexual effects |
escitalopram/Lexapro | even fewer side effects, maybe? | sexual effects |
venlafaxine/Effexor | worst withdrawal; should almost never use this one | |
duloxetine/Cymbalta | also severe withdrawal risk | |
mirtazapine/Remeron | usually helps sleep, a lot | much more weight gain |
new stuff | advertising makes them sound great | less experience so long term risks unknown |
bupropion/Wellbutrin
No sexual side effects like most of the others, no weight gain, much less withdrawal: why isn’t this one used most often? That’s a great question, an important one. some experts think bupropion should be the most frequently prescribed antidepressant.
Why not? Short answer: bad marketing (really), and a sense that maybe bupropion does not help anxiety and sleep as much as fluoxetine and its cousins. That’s not clear.
Although I took fluoxetine years ago when needed, if I took an antidepressant now, I think it would likely be bupropion. For more, see the bupropion page.
fluoxetine/Prozac
I took this one, during my residency. It helped a lot. I stopped it after about 3-4 months. No withdrawal (but withdrawal risk goes up the longer one stays on a medication). No weight gain. Didn’t notice any problems with sexual function. Pretty much a perfect experience.
But this doesn’t happen for everyone, that’s sure. First off, as you’ve read elsewhere on this site, you have to figure out what kind of depression you have. Mine was a single-episode type.
Fluoxetine’s biggest advantage is a lower risk of severe withdrawal symptoms when you taper it off, compared to almost all the other antidepressants (except probably bupropion).
That’s because fluoxetine has a very long lifetime in the bloodstream. So it sort of tapers itself when you stop, far more than any of the other antidepressants. People can still have withdrawal though. Each time the dose goes down, there’s a delay while the blood level goes down. So if withdrawal symptoms are going to happen, they tend to come on about a week after a dose reduction, sometimes even longer.
sertraline/Zoloft
This seems to be one of the most commonly used antidepressants. Frankly, I’m not sure why. Maybe a little less potential for side effects than fluoxetine. It has a reputation for having “fewer drug interactions”, which is technically true but the interactions that other antidepressants have are usually pretty minor.
citalopram/Celexa
For a while citalopram was definitely the “fewer side effects” choice — until its close cousin escitalopram went generic and thus became similarly inexpensive. Citalopram can cause a heart rhythm change when its dose is over 20 mg daily, so nowadays people often use escitalopram instead.
escitalopram/Lexapro
Escitalopram is less likely to affect heart rhythm even at its higher doses. What else to say about it? Only to remind that escitalopram, like all the above except bupropion, works via serotonin, so has the typical potential serotonergic side effects and risks: withdrawal, sexual dysfunction, emotional blunting, and sexual dysfunction.
venlafaxine/Effexor
Venlafaxine causes more severe withdrawal symptoms, more often, than any other antidepressant — so often that there’s basically no disagreement about this problem. (“No disagreement” doesn’t happen often in psychiatry).
A few studies once seemed to show that venlafaxine worked better than other antidepressants. But most were sponsored, and hyped, by the manufacturer. Few mood specialists now believe venlafaxine is significantly better than any other antidepressant.
So why use it? At one time it had a unique “mechanism of action”, affecting two neurotransmitters (norepinephrine too, not just serotonin). Because of this difference, people hoped that venlafaxine could work when other antidepressants had not. But now there are several antidepressants that affect multiple neurotransmitters, and none of them clearly works better than any other.
duloxetine/Cymbalta
Duloxetine came along after venlafaxine. It too affects both norepinephrine and serotonin. Its manufacturers made a big deal about its ability to treat pain as well as depression. But other antidepressants also treat pain (at least the serotonin specific ones, SRIs like escitalopram or sertraline or fluoxetine). In “head-to-head” comparisons of duloxetine with an SRI, duloxetine comes out slightly ahead on some measures, in some studies. Is this enough to justify duloxetine’s greater risk of severe withdrawal (not as bad as venlafaxine but worse than SRIs generally and definitely worse than fluoxetine)? I don’t think that’s clear.
mirtazapine/Remeron
Mirtazepine has the advantage of helping people sleep, but the disadvantage of causing more weight gain than other commonly used antidepressants. Since it is not clearly more effective against depression than the rest; and since any antidepressant that actually improves depression often improves sleep; and since weight gain can worsen depression in the short and long term; mirtazapine deserves to stay in the background. Of course like any antidepressant, some patients respond extremely well to it even after not benefiting from other antidepressants, so it’s good to have mirtazapine as an option.
new stuff
As of this writing, I am no longer treating patients myself. So I don’t have direct experience with the newer antidepressants. For a while, my semi-retirement didn’t matter much, because the new ones were expensive and no one had shown they were clearly better than the old ones.
But now the “new ones” are no longer so new! As they become generic, they may be worth considering. But two things should be kept in mind.
- New medications haven’t been around as long as old ones. Sometimes it takes years before the risks of a medication really become clear (e.g. slow long-term weight gain on SRIs). So the newer the medication, the more unknown are its risks.
Even when they cost about the same, then, the newer ones should have evidence that they work better than the old ones. - A company that makes a new antidepressant has to compete against all the old ones. So they have a very strong incentive to find something special about their drug. Then they will make a lot of noise about that special thing.
For example, there are new medications for women who have become depressed just after having a baby (“post-partum depression”). One of them required a series of injections, but now there’s a pill version. They’re extremely expensive. But sure enough, already there’s a company-sponsored study of zuranolone in depression (not just post-partum depression).
Just wait, the company will soon be declaring a “brand-new mechanism of action” in expensive advertising. Never mind that their drug has not been compared to existing antidepressants. And it was only barely better than placebo.