Do SSRIs and SNRIs cause bleeding? Several review articles have been published about it, and patients are beginning to ask us about it. What’s the scoop? First, let’s talk mechanisms. Only a minority of serotonin receptors live in the brain, and in fact platelets contain more than 90% of circulating serotonin. Serotonin promotes platelet aggregation and therefore blood clotting. SSRIs and SNRIs inhibit serotonin reuptake and therefore deplete platelets of serotonin, which is the leading theory for how these antidepressants cause bleeding. There is a second possible mechanism, which is that SSRIs increase gastric acidity, potentially causing ulcers and GI bleeding (Andrade C et al, J Clin Psychiatry 2010;71(12):1565–1575).
Obviously, SSRI-induced bleeding is not common, or most of our patients would come into the office with bruises and bloody noses. While the initial clinical trials of SSRIs did not report any increased incidence of bleeding events compared to placebo, such rare side effects usually do not show up in the initial trials. The best evidence would be a randomized double blind controlled trial specifically designed to detect SSRI-induced bleeding, but in the absence of such gold standard studies, researchers have had to resort to less robust research designs.
The most common one is the “case control” design. You identify a bunch of patients on SSRIs who had, for example, a GI bleed (these are the “cases”), and you compare them with a control group of similar patients on SSRIs who did not have bleeding (the “controls”). A recent review identified 14 case control and other retrospective studies published since 1999, involving hundreds of thousands of patients (Andrade ibid). These data suggest that serotonergic ADs are, indeed, associated with an increased risk of bleeding, especially from the upper GI tract (usually involving either stomach or esophageal ulcers). The overall risk is low, with one study suggesting that roughly one upper GI bleed will develop for every 8000 SSRI prescriptions (deAbajo FJ et al, BMJ 1999;319(7217):1106–1109).
Another review suggests that 411 patients would need to take an SSRI for a year for one extra patient to develop a GI bleed (Loke YK et al, Aliment Pharmacol Ther 2008;27(1):31–40). The severity of GI bleeds varies—sometimes it presents as a medical emergency, but often it presents more chronically, with symptoms such as dizziness or shortness of breath due to anemia and black tarry stools.
In addition to GI bleeding, SSRIs are associated with increased blood loss during surgical procedures. In one study of 66 patients undergoing total hip replacement while taking SSRI medication, the mean blood loss was 95 ml which was a 17% increased amount compared to controls (vanHaelst LMM et al, Anesthesiology 2010;112(3):631–636). A smaller study of 26 patients undergoing various orthopedic procedures reported a 75% increase of blood loss (average of just over a liter) and a four-fold increased frequency of transfusion compared to antidepressant non-users (Movig KLL et al, Arch Intern Med 2003;163:2354–2358).
In contrast, two studies that looked at SSRI-associated bleeding and transfusions in patients undergoing coronary artery bypass grafting (CABG) did not find an increased bleeding risk (Andrade op.cit). Given these minimal and conflicting data, it’s unclear what we should tell our patients who are about to go under the knife. Given that most patients can tolerate stopping an SSRI for a few days before surgery, this is probably the safest course—unless your patient has a history of rapid decompensation off meds, or is on high dose venlafaxine Effexor) or paroxetine (Paxil), both notorious for causing severe discontinuation symptoms.
Aside from upper GI bleeds and perioperative bleeding, there have been reports of other kinds of bleeding. These include bruising, nosebleeds, internal hemorrhoids, and menorrhagia (abnormally heavy or prolonged menstrual periods). It’s not clear how commonly these occur, but you should be aware of them in case a patient reports one of these symptoms to you.
While there are not enough data to know for sure, it appears that some antidepressants may be more likely to cause bleeding than others, with SSRIs posing a greater risk than SNRIs. Furthermore, the higher the dose, the higher the bleeding risk. Reassuringly, antidepressants with little or no serotonin receptor effect, such as nortriptyline (Pamelor), desipramine (Norpramin), mirtazapine (Remeron) and bupropion (Wellbutrin) have not been associated with bleeding episodes. Combining NSAIDs such as ibuprofen with SSRIs increase the bleeding risk as much as seven to 15-fold, depending on the study (Andrade ibid). The risk of abnormal bleeding is also increased when SSRIs are used in combination with the antiplatelet treatment clopidogrel (Plavix) and the anticoagulant warfarin (Coumadin). On the other hand, adding a proton pump inhibitor (such as omeprazole) to an SSRI lowers the bleeding risk to an insignificant level (Andrade ibid).
What’s the bottom line? For the typical healthy non-elderly patient, SSRI-induced bleeding is likely to be a non-issue, and may not even require that you mention it in your discussion of side effects as it occurs relatively infrequently. However, you should mention this risk in the following situations:
Patients with a history of stomach ulcers or bleeding disorders.
Patients about to have surgery.
Patients taking NSAIDs, aspirin, warfarin, or antiplatelet drugs.
In these patients, we recommend you say something like, “Although it seems to be a rare effect, your antidepressant may affect the way your blood naturally clots. If you notice any increased bruising, bleeding, or burning stomach pain or if you plan to have surgery or major dental work, you will need to contact me or your primary care physician. Also, if you begin taking any new medications, particularly pain medications (even over the counter ones) you will need to let me know.”