Serious But Rare Antidepressant Side EffectsWe typically warn our patients about common side effects, such as headache, GI distress and sexual dysfunction when we start SSRIs. Here are some of the side effects that we don’t necessarily talk to patients about, but should be able to recognize and discuss in particular contexts.


SSRIs famously inhibit the reuptake of serotonin in the brain, but they also inhibit serotonin reuptake in other parts of the body, notably in the platelets. When platelets have less serotonin they don’t function as well as they should, potentially leading to bleeding problems.

How clear is the evidence that SSRIs actually cause bleeding? A meta-analysis found that SSRIs roughly double the risk of upper GI bleeding, while combining SSRIs with NSAIDs increases the risk 5-fold (Loke YK et al., Aliment Pharmacol Ther. 2007;27(1):31-40). This may sound scary, but according to the article, the baseline risk of upper GI bleeding is only about 1/1000, so a doubling of this risk still only brings you to a 1/500 chance of having an event. On the other hand, adding NSAIDs to the mix brings the risk up to 1/200.

Another study found that the bleeding risk of antidepressants is roughly proportional to the degree of inhibition of serotonin reuptake. These researchers identified 196 patients in the Netherlands who had received new antidepressants and had then been hospitalized for bleeding problems (Meijer WE, et al., Arch Intern Med. 2004;164(21): 2367-2370). The single most common type of bleeding was uterine bleeding, followed by upper GI bleeding. The antidepressants most associated with bleeding were also those that were the most serotonergic: Prozac, Paxil, Zoloft, and clomipramine. (The authors did not report the absolute risk of bleeding for each medication, only the relative risks). Luvox, Celexa, Effexor, amitryptiline, and imipramine were intermediate in risk, while Wellbutrin, nefazodone, nortriptyline, and desipramine incurred the lowest risk. Lexapro was not examined in this study, but it has a high affinity for serotonin transporters and so would presumably be in the higher bleeding risk group.

The bottom line is that bleeding is probably too rare a side effect to require informing younger and otherwise healthy patients. However, for patients who have a history of GI bleeding or who are taking NSAIDs, you might consider choosing a non-SSRI antidepressant.


An excellent and comprehensive review of the literature on the association between SSRIs and hyponatremia (defined as a serum sodium concentration below 130 mEq/L) was published in 2006 (Jacob S and Spinler SA, Ann Pharmacother 2006;40:1618-1622). These researchers reviewed all the existing literature on the topic, including about 50 case reports and a dozen or so chart reviews, case-control studies, and observational trials.

One collection of 706 cases of spontaneous reports of hyponatremia compiled by the World Health Organization and Canadian authorities found that most victims were 65 years or older and that the median time of onset was 13 days after starting an SSRI. Of course, these spontaneous reports cannot tell us anything about the degree of risk. Case series have generally been of elderly patients in psychiatric hospitals, and the incidence of hyponatremia of patients on SSRIs has varied from 17.5% to 25.7% of patients. Elderly patients on SSRIs are about three times more likely to develop hyponatremia than patients on other antidepressants.

Thus, the incidence is greater than most people realize. While the authors do not recommend routine monitoring for sodium in all patients on SSRIs, they do recommend to be watchful of common hyponatremia symptoms, such as lethargy, delirium or nausea, during the first 2 weeks of treatment. The definitive treatment is to stop the medication and to restrict fluids to about 1 or 1 1/2 liters per day. Most cases resolve within 2 weeks.

Suicidal Ideation

In October of 2004, the FDA required antidepressant drugmakers to carry blackbox warnings about the risk of suicidal ideation in children and adolescents started on antidepressants. This was based on an FDA analysis of placebo-controlled, randomized trials showing that patients assigned to antidepressants had roughly double the risk of suicidal ideation – from 2.5% for the placebo group to 5% for the drug group – based on trials lasting two to three months (Newman TB, NEJM 2004;351:1595-1598). While there were no completed suicides among any of these subjects, the FDA was concerned about the longer-term implications of these numbers. In the words of the FDA panel member who wrote the article cited above, “Does a five percent risk over the course of three months become a 20% risk over the course of a year, or does the benefit-risk balance improve over time?”

Then the FDA got busy analyzing similar data for adults, and in May of 2007, they determined that there was a similar risk for young adults (18-24 years old), and required black box warnings for this population as well. The increased risk was smaller than for children – roughly 1.5 times the risk vs. double the risk – and, unlike the case for children, it was not statistically significant. So why the black box warning?

In a recent commentary in the American Journal of Psychiatry, Dr. Andrew Leon, the biostatistician appointed to the FDA’s Psychopharmacologic Drugs Advisory Committee, explained the decision-making process (Leon A, Am J Psychiatry 2007;164:1786-1789). While acknowledging that the meeting which led to the black box warning was “emotionally charged” (including heart-wrenching testimony from survivors of committee’s final decision was based on solid evidence. Two independent meta-analyses reviewed data from 77,382 participants in 295 randomized controlled trials. In a fascinating graph, in which the risk of suicidality is broken down by age group, you see that the highest risk is in the youngest group (<18), with the risk progressively decreasing with increasing age. For patients 65 and older, antidepressants were actually protective, lowering the risk of suicidality by 61%. For ages 25-64, antidepressants neither worsen nor protect against suicidality.

The bottom line is that any suicidality risk of antidepressants seems confined to people 24 years old and younger, and these are the patients we should be most closely monitoring.


This article originally appeared in:

The Carlat Psychiatry Report
Click on the image to learn more or subscribe today!

This article was published in print 4/2008 in Volume:Issue 6:4.