Which antidepressants are least likely to cause weight gain?

After almost 2 years of discussing pharmacotherapy, a 48 year old woman with severe generalized anxiety disorder (GAD) agrees to take medication. She starts taking escitalopram and increases the dosage. After 8 weeks on a daily dose of escitalopram 20mg, her anxiety symptoms are in remission. She has no side effects. However, when you see her 6 months later, she has gained 10lbs. She’s been careful with her diet for the last 6 weeks and has been more active, but it has not made a difference. She gained 3lbs in the last month. You discuss how weight gain is a common side effect of long-term treatment with SSRIs like escitalopram, and that it now makes sense to change the medication. She is open to trying something else and asks, “Are there any SSRIs that are less likely to cause weight gain?”

The figures below are from a 2010 meta-analysis by Seretti et al published in The Journal of Clinical Psychiatry. They illustrate the average weight change associated with antidepressant use during acute treatment (4-12 weeks, Figure 2) and maintenance treatment (>4 months, Figure 3).

In the short-term, SSRIs and SNRIs are associated with mild weight LOSS (Figure 2). In the long-term, SSRIs and SNRIs are associated with mild weight GAIN (Figure 3). One hypothesis is that maintenance antidepressant treatment leads to the downregulation of serotonin receptors over time, causing cravings for carbohydrate-rich foods, increased food intake, and ultimately weight gain. At a population level, fluoxetine is the least likely SSRI to cause weight gain, while paroxetine is the most likely.  Limited evidence suggests that citalopram and escitalopram are more likely to cause significant weight gain than sertraline or fluoxetine (Maina, 2004). During maintenance treatment, most individuals gain a little weight on SSRIs and SNRIs. Some individuals don’t gain any, and some experience substantial ongoing weight gain, like the patient in our case.

There are two other clinically important things to notice in these figures.

1) Bupropion is the only antidepressant associated with long-term weight neutrality or weight loss. A pharmaceutical company harnessed this side effect when creating Contrave, a weight-loss combo pill that contains bupropion and naltrexone. If a patient takes bupropion for depression or smoking cessation, don’t add in Contrave for obesity. The extra bupropion puts the patient at risk for seizures from (accidental) bupropion overdose (See our previous PsychSnap for a review of bupropion dosing). If you want to treat someone for obesity with Contrave and maintain them on bupropion for depression, you could add naltrexone 25mg BID to bupropion SR 150-200mg BID to approximate the standard Contrave dose (naltrexone 16mg BID + bupropion SR 180 mg BID). 

2) Mirtazapine caused substantially more weight gain than the SSRIs or SNRIs during acute treatment (Figure 2), with an average weight gain of 1.7kg (3.8 lbs). Over time, however, the difference in average weight gain between the SSRIs and mirtazapine narrows.  During maintenance treatment, the average weight gain on mirtazapine (2.6 kg or 5.8 lbs) was less than the average weight gain on paroxetine, though the difference was not statistically significant. Some data suggest that higher doses of mirtazapine (30mg+) are less likely to cause increased appetite and weight gain than lower doses (Stahl, 1997).  

We can use average population data like this to make thoughtful choices about medications. However, our best data on the efficacy and side effects of a medication for a specific patient comes from that patient—the expert on their own experience.

Let’s return to our patient. She had an adequate trial of escitalopram which helped her GAD at 20mg but caused ongoing weight gain, necessitating discontinuation. Are there any SSRIs that are less likely to cause weight gain?

Fluoxetine is the SSRI that is least likely to cause weight gain, and would be a reasonable next choice for this patient. When switching antidepressants due to side effects when the patient has benefited from treatment, I aim to get the patient to a roughly equivalent dose and complete the switch within 2 weeks. 

Cross-tapers are confusing, so I always write them out in addition to reviewing them with patients.

To switch from escitalopram 20mg daily to fluoxetine 40mg daily
You need:
– The new prescription for fluoxetine 20 mg capsules.
– The escitalopram 20mg pills that you have been taking. You will cut them.
Week 1
– escitalopram 10mg (take ½ of your 20mg pill) + fluoxetine 20mg daily (take one 20mg capsule)
Week 2
– STOP escitalopram. Increase to fluoxetine 40mg daily (two 20mg capsules) 
Continue fluoxetine 40mg daily until our next appointment.

If the patient wants to avoid the risk of weight gain altogether, buspirone could also be used as monotherapy for the treatment of GAD (albeit with a smaller average effect size than SSRIs).  See our previous PsychSnap for a review of buspirone and other non-SSRI/SNRI treatments for anxiety.

Key Points

  1. On average, people gain weight on SSRIs and SNRIs in the long term, but there is wide variation among individuals. 
  2. Among the SSRIs, paroxetine is the most likely to cause weight gain, and fluoxetine is the least likely to cause weight gain.  
  3. Bupropion is the only antidepressant that is associated with weight loss. 
  4. Mirtazapine’s reputation for causing weight gain is well-deserved, particularly during acute treatment. Higher doses of mirtazapine may cause less weight gain than lower doses.


Maina, Giuseppe, Umberto Albert, Virginio Salvi, and Filippo Bogetto. “Weight gain during long-term treatment of obsessive-compulsive disorder: a prospective comparison between serotonin reuptake inhibitors.” Journal of Clinical Psychiatry 65, no. 10 (2004): 1365-1371.

Serretti, Alessandro, Laura Mandelli, and M. Laura. “Antidepressants and body weight: a comprehensive review and meta-analysis.” The Journal of clinical psychiatry 71.10 (2010): 979.

Stahl, S. M., C. M. E. Kremer, and R. M. Pinder. “Tolerability of mirtazapine used in high or low initial dose.” European Neuropsychopharmacology 1002.7 (1997): S157-S158.

Send a Comment