An important adverse effect of SRIs: their association with increased bleeding.
“Mr Hardy,” a 67-year-old man, sought psychiatric evaluation for increased forgetfulness and word-finding trouble. Prior to this evaluation, he had been prescribed escitalopram 10 mg/d for mild anxiety and low mood. Although his MMSE score at initial assessment was in the normal range at 28 out of 30 points, his growing need for reminders and other compensatory supports suggested the presence of mild neurocognitive disorder in this highly educated individual. His medical health was considered good except for mild, untreated obstructive sleep apnea. His laboratory evaluation included normal range results on CBC, CMP, thyroid functions, B12, folate, ESR, and RPR. His MRI showed a chronic right frontal lacunar infarction and white matter changes consistent with small vessel disease.
Although his functional status suggested only mild cognitive compromise, more extensive neuropsychological testing identified significant impairment of contextual and noncontextual memory. Mr Hardy was started on donepezil 5 mg/d with a provisional diagnosis of early Alzheimer disease. He experienced this medication as helpful, prompting a dose increase to 10 mg/d after several months. He felt that his medications reduced his anxiety and improved his cognition, though his occasional bouts of diarrhea may have been an adverse medication effect.
Several months after beginning donepezil, his cardiologist diagnosed atrial fibrillation and prescribed apixaban 5 mg/d. Mr Hardy did not discontinue his previously prescribed ASA 81 mg/d. Five months after adding apixaban to his regimen, Mr Hardy complained of new dizziness, headache, and worsening confusion. He reported no head trauma, although his wife recalled that he had fallen from his bicycle the previous week and could have hit his head. On interview, he seemed to have trouble with verbal communication. CT scan showed a large acute right subdural hematoma with mass effect and midline shift. Apixaban was held and vitamin K–dependent coagulation factors plus antithrombotic proteins C and S were administered in the emergency room. After consultation from neurosurgery, craniotomy was performed with evacuation of the subdural hematoma. Postoperatively, apixaban was held. Donepezil and escitalopram were continued.
Four months postoperatively, Mr Hardy demonstrated more significant cognitive impairment. His MoCA-BLIND score was 15 out of 22 points. He was attending speech therapy for help with focus and communication. His cognition improved gradually, and after 4 additional months his MoCA-BLIND score had increased to 18 out of 22. One year after neurosurgery, his MoCA-BLIND score had risen to 20 out of 22, although he still reported impaired memory, complex attention, and executive function. He was at that time taking aspirin 81 mg/d, donepezil 10 mg/d, and escitalopram 10 mg/d. In retrospect, his clinicians appreciated how confluent multiple factors contributed to his hemorrhagic event and wondered which medications or factors were most relevant. They wondered, too, how bleeding risk might be minimized when serotonergic antidepressants and blood thinning agents are coprescribed.
Serotonergic Antidepressants and Bleeding Risk
Serotonergic antidepressants (SRIs) are among the most commonly prescribed medications and play important roles in treating depression, anxiety, agitation, and other behavioral syndromes. Data from 2015 to 2018 published by the National Center for Health Statistics show that during those years, 13.2% of adults aged 18 years and over reported using an antidepressant medication within the preceding 30 days. Women aged 60 years and older reported a rate of 24.3%.1
The SRIs include selective serotonin reuptake inhibitors (SSRIs) and serotonin norepinephrine reuptake inhibitors (SNRIs). Although the SRIs’ adverse effect profiles are milder than those of the earlier tricyclic antidepressants and monoamine oxidase inhibitors, SRI treatment is nonetheless associated with several potentially problematic adverse reactions including effects on sexual function, sleep, appetite, and sodium regulation. Another important adverse effect of SRIs is their association with increased bleeding. Their serotonin reuptake blocking effect not only improves depression through its activity in the brain, but also interferes with normal hemostasis by diminishing platelets’ accumulation of serotonin, which cannot be synthesized by a platelet. Serotonin is released by platelets to promote vasoconstriction when a bleed is detected, hence decreased serotonin uptake decreases platelets’ capacity to decrease bleeding. The resulting anticoagulant effect is warned against in SRIs’ prescribing information (see, for example, the prescribing information for sertraline).2
Prescribed alone, SRIs have been reported to increase a variety of hemorrhagic events including gastrointestinal and intracranial bleeding. In anested case-control study looking at the risk of abnormal bleeding in a cohort of more than 64,000 new antidepressant users, SRI use was associated with an increased risk of hospitalization due to abnormal bleeding. Higher risk was linked with use of the SRIs with intermediate and high degrees of serotonin reuptake inhibition.3 In another large study, spontaneous intracranial hemorrhage (ICH) was linked with SSRIs possessing greater affinity for the serotonin transporter.4
Many reports address bleeding associated with the SSRIs. Bleeding associated with SNRIs has been less fully investigated, although their prescribing information contains a similar warning regarding their coprescription with anticoagulants. An important consideration with the SNRIs may be the serotonin-to-norepinephrine (NE) reuptake inhibition ratio. The medications with greater relative norepinephrine reuptake blocking activity produce less marked serotoninergic effects.5 Further research may shed light on whether SNRIs with greater NE reuptake activity and less dominant effect on serotonin reuptake might have less impact on bleeding risk than the selective SSRIs or the SNRIs with lower NE to 5HT reuptake blocking ratios.
Anticoagulants and Adverse Effects Related to Bleeding
Anticoagulant medications, prescribed for prevention of pathological thrombosis in a variety of medical conditions, are another category of frequently prescribed medications associated with potentially harmful bleeding. About 4.2 million Americans aged 18 years or older take anticoagulants, and the rate is higher among older adults.6 Recent advances in anticoagulant pharmacology have provided clinicians with new treatment options. Many prescribers, however, have limited knowledge of the mechanisms of action, risks, and benefits of this varied class of drugs.
Warfarin and other coumarins work by inhibiting vitamin K’s role in the biosynthesis of prothrombin and are collectively known as vitamin K antagonists (VKAs).7 Treatment with most VKAs is typically safe and effective as long as the international normalized ratio (INR), monitored through periodic blood testing, is maintained within a safe therapeutic range. Achieving a safe and stable range, however, can be challenging, especially in the elderly where polypharmacy is common and can result in drug-drug interactions. In addition, the effect of VKAs can be affected by certain foods.8
In contrast to VKAs, low-molecular-weight heparins (LMWHs) such as dalteparin and enoxaparin act by inhibiting the final common pathway of the coagulation cascade, thereby impeding the conversion of fibrinogen into fibrin. Patients self-administer these medications by subcutaneous injection. Dosage is adjusted in individuals with chronic kidney disease, which is associated with a risk of accumulation and increased risk for problematic bleeding.9
The third and most recently introduced group of anticoagulants, collectively referred to as direct oral anticoagulants (DOACs), works by yet another mechanism: the direct inhibition of specific proteins within the coagulation cascade. Dabigatran works by reversibly binding to the active site on the thrombin molecule, thereby preventing thrombin-mediated activation of coagulation factors, while several others (rivaroxaban, apixaban, edoxaban, and betrixaban) inhibit factor Xa. DOACs have been shown to be as effective as VKAs with a safer adverse effect profile.10 Advantages of DOACs include their more immediate drug onset and offset effects, fewer drug and food interactions, and fewer monitoring requirements.11
Antiplatelet medications such as aspirin and clopidogrel, although not formally classified as anticoagulants, are widely used for the prevention of vascular events and are associated with an increased bleeding risk. Antiplatelet medications inhibit platelet cyclooxygenase, thereby blocking the formation of thromboxane A2. This impairs thromboxane-dependent platelet aggregation and consequently prolongs bleeding time. Many guidelines now caution against their use among those who are at increased risk of bleeding because the risks associated with use of these medications can overshadow their benefits.12
Enhanced Bleeding Risk With Coprescribed Anticoagulant and SRI
For a variety of reasons, patients prescribed an SRI may receive concurrent medically appropriate treatment with an anticoagulant. In many cases, these medications are prescribed independently by clinicians who are not in communication with each other. Each of these medication classes increases bleeding risk, and their concurrent use can amplify this risk. SRI package inserts warn clinicians of this enhanced risk when SRIs are coprescribed with aspirin, nonsteroidal anti-inflammatory drugs, warfarin, or other anticoagulants.13-15 Prescribing clinicians may be unaware of the potential risk associated with this medication combination or even unaware that the combination has been coprescribed. Upper gastrointestinal tract bleeding and intracranial hemorrhage can have serious consequences.16,17 We will selectively summarize evidence for these adverse drug interactions and suggest an approach to optimizing safety when both an SRI and an anticoagulant are medically necessary. When an SRI and anticoagulant are coprescribed, bleeding risk most likely is affected both by the anticoagulant’s mechanism of action and the pharmacokinetic/pharmacodynamic profile of the SRI.
For VKAs, data regarding the risk of coprescription with SRIs are not all consistent. In 1 large cohort of patients treated with warfarin for atrial fibrillation and followed for a median duration of 6 years, for example, the increase in GI and IC bleeding risk conferred by concurrent use of SRIs was explored.18 The SRIs prescribed in this study were paroxetine (51.6% of SRI users), fluoxetine (49.4%), sertraline (13.9%), venlafaxine (4.6%), citalopram (3.8%), and fluvoxamine (0.5%). After adjusting for underlying bleeding risk and time at international normalized ratio range (INR) >3, SRI exposure was associated with an increased rate of hemorrhage compared with warfarin exposure alone (adjusted relative risk 1.41, 95% confidence interval 1.04 to 1.92, p = 0.03). Although not highlighted in this study, the strong CYP450 1A2 inhibiting effect of fluvoxamine, which could impair warfarin’s deactivating metabolism, suggests that the fluvoxamine/warfarin combination should be regarded as one to avoid when possible. Another study, interestingly, followed a cohort of patients on coumarins for the first time and reported that the risk for non-gastrointestinal tract bleeding, including ICH, was increased among concurrent SSRI users (adjusted OR, 1.7; 95% CI, 1.1-2.5). In this study, however, concurrent use of coumarins with SSRIs did not increase the risk for gastrointestinal tract bleeding (adjusted OR, 0.8; 95% CI, 0.4-1.5).19
With regard to DOACs, 1 review of studies published between 2014 and 2016 concluded that DOACs were safer in combination with SRIs than warfarin as a result of the DOACs’ more limited pharmacodynamic interactions.20 This result is consistent with outcome data from the ROCKET AF study, which compared the safety and efficacy of combining SSRI treatment with either the factor Xa inhibitor rivaroxaban or warfarin in 737 patients followed for a mean period of 1.6 years. The authors concluded that the apparently small increase in bleeding risk added by an SSRI to rivaroxaban or warfarin treatment was not statistically significant. There was,however, a nonsignificant trend suggesting slightly greater risk for severe bleeding (intracranial and fatal bleeding) with warfarin than with rivaroxaban.21
With regard to low molecular weight heparins (LMWH), findings have not supported a synergistic increased risk of bleeding. The authors of a retrospective cohort study of 575 patients with a primary or secondary diagnosis of an acute venous thromboembolism started on full-dose enoxaparin with or without an SSRI examined 5 years of follow-up data. Concurrent use of an SSRI was not associated with an increased bleeding risk in patients on this LMWH. The concurrent use of alcohol, however, did significantly increase the rate of major bleeding events.22 Alcohol use by itself, incidentally, increases bleeding risk directly by suppressing the bone marrow and therefore hematopoietic stem cells, and indirectly by promoting nutritional deficiencies that impair the production and function of various hematopoietic cells, most importantly platelets, which are essential in the formation of clots.
Nonsteroidal anti-inflammatory drugs (NSAIDs), or aspirin—which produce reversible inhibition of platelet cyclooxygenase with effects on bleeding that vary depending on the NSAID’s dose, serum level, and half-life—are often used by individuals taking an SRI. The clinical risk of bleeding with NSAID use is enhanced by concomitant use of an SRI, or by concurrent use of alcohol or other anticoagulants, as well as by associated medical factors such as advanced age, liver disease, and other coexisting coagulopathies.23 Clopidogrel may be coprescribed with an SRI, but there is limited information about the effect of this combination on bleeding risk. One retrospective cohort study reported an increased risk of bleeding when this combination of drugs was prescribed to patients at high risk for stroke (HR 2.35, 95% CI 1.61–3.42).24
Although the SRIs share overlapping pharmacodynamic effects, they differ in their pharmacokinetic interactions with the cytochrome P450 system. The differences between specific SRIs can affect drug-drug interactions with anticoagulants. Warfarin is a racemic mixture of S-warfarin and R- warfarin, for example, and although both isomers are clinically active, the S-isomer is more potent than the R-isomer.25 The S-isomer is metabolized by the 2C9 isoenzyme, whereas the R-isomer is metabolized by the 1A2 isoenzyme as well as the 2C19 and 3A4 isoenzymes. Among SSRIs, fluvoxamine (which also inhibits 1A2) and fluoxetine appear to have the highest potential for inhibiting warfarin metabolism through the 2C9 isoenzyme, thereby decreasing its clearance and enhancing anticoagulation. Paroxetine coadministered with warfarin has been shown to mildly increase bleeding risk. The DOACs are predominantly metabolized in the liver, mainly by cytochrome P450 enzymes 3A4 and 2J2, and other CYP-independent pathways.26 A retrospective study looking at drug-drug-interactions between DOAC and SRIs showed that 80% of individuals in the study who had suspected interactions were taking SRIs that are known CYP3A4 substrates along with the DOAC.27 The SRIs with strongest CYP3A4 inhibition are fluvoxamine and, perhaps to a lesser extent, fluoxetine. Awareness of the possibly greater bleeding risk when these antidepressants are combined with a DOAC should be factored into the clinician’s prescribing decisions.
Discussion and Recommendations
Many older adults stand to benefit from treatment with an anticoagulant or antiplatelet medication, yet some of these individuals are already taking an SRI or might come to need one. Similarly, many individuals who are receiving acute or long-term SRI treatment might develop the need for short-term or chronic treatment with an anticoagulant or antiplatelet medication. As it currently stands, the warnings about coprescribing members of these medication classes might deter a clinician from treating 1 of 2 comorbid serious conditions. Rather than avoiding use of any SRI in all patients on anticoagulants or antiplatelet medications, a more nuanced and patient-centered approach can be adopted. We propose the following recommendations for addressing this clinical dilemma.
Although several studies have found an association between SRI use and increased bleeding risk, from our review, we show that not all SRI/anticoagulant combinations substantially increase clinical risk; furthermore, risk appears to differ on the basis both of anticoagulant mechanism of action and pharmacokinetic/pharmacodynamic SRI profiles. We recommend that physicians take a complete inventory of each patient’s medications, including nonprescription drugs such as NSAIDs. Meticulous medication reconciliation and communication among prescribing treaters is encouraged so that all prescribers are aware of all medications coadministered. Ongoing assessment of need for coprescribing and elimination of unnecessary medications, especially considering the risk of polypharmacy in older adults, is paramount. A clinician may opt for the safer combinations (Figure) or in some cases choose a non-SRI antidepressant. The risk must be evaluated with special care for patients with a history of bleeding, cerebral amyloid angiopathy, or greater alcohol consumption. Other factors to be considered are comorbid clotting disorders, such as von Willebrand disease, or genetic polymorphism in the serotonin transporter gene, which is known to be associated with platelet dysfunction.28 With regard to GI bleeding, if an SRI is required in an individual who is already at high risk of an upper GI bleed, the use of a gastro-protective agents should be considered.
Additionally, physicians should do a thorough risk-benefit analysis and involve the patient in shared decision-making. Patients who take multiple drugs should be educated against the regular use of nonprescription drugs which increase bleeding risk such as aspirin or ibuprofen, as well as on how to recognize symptoms of GI bleeding.
With attention to the considerations and choices outlined here, the risk of bleeding can be diminished for patients who need to use an SRI and an anticoagulant concurrently. This combination can be safely managed in order to diminish the “hemorrhagic hazard” of combined treatment while allowing management of comorbid medical needs. There is a need for further studies to supplement and refine current knowledge, as well as to provide more comprehensive and specific clinical guidance.
Dr Husain-Krautter is an attending psychiatrist at Zucker Hillside Hospital. Dr Ellison is a geriatric psychiatrist and director of the Swank Center for Memory Care and Geriatric Consultation, ChristianaCare.
1. Pratt LA, Brody DJ, Gu Q. Antidepressant use in persons aged 12 and over: United States, 2005-2008. NCHS Data Brief. 2011;(76):1-8.
2. Highlights of prescribing information: Zoloft. FDA. Accessed November 8, 2022. https://www.accessdata.fda.gov/drugsatfda_docs/label/2016/019839S74S86S87_20990S35S44S45lbl.pdf
3. Meijer WEE, Heerdink ER, Nolen WA, et al. Association of risk of abnormal bleeding with degree of serotonin reuptake inhibition by antidepressants. Arch Intern Med. 2004;164(21):2367-2370.
4. Renoux C, Vahey S, Dell'Aniello S, Boivin JF. Association of selective serotonin reuptake inhibitors with the risk for spontaneous intracranial hemorrhage. JAMA Neurol. 2017;74(2):173-180.
5. Raouf M, Glogowski AJ, Bettinger JJ, Fudin J. Serotonin-norepinephrine reuptake inhibitors and the influence of binding affinity (Ki) on analgesia. J Clin Pharm Ther. 2017;42(4):513-517.
6. Hylek EM, Evans-Molina C, Shea C, et al. Major hemorrhage and tolerability of warfarin in the first year of therapy among elderly patients with atrial fibrillation. Circulation. 2007;115(21):2689-2696.
7. Kasperkiewicz K, Ponczek MB, Owczarek J, et al. Antagonists of vitamin K-popular coumarin drugs and new synthetic and natural coumarin derivatives. Molecules. 2020;25(6):1465.
8. Goldman M. The innovative medicines initiative: a European response to the innovation challenge. Clin Pharmacol Ther. 2012;91(3):418-425.
9. Solari F, Varacallo M. Low molecular weight heparin (LMWH). StatPearls. September 4, 2022. Accessed November 8, 2022. https://www.ncbi.nlm.nih.gov/books/NBK525957/
10. Hirschl M, Kundi M. Safety and efficacy of direct acting oral anticoagulants and vitamin K antagonists in nonvalvular atrial fibrillation - a network meta-analysis of real-world data. Vasa. 2019;48(2):134-147.
11. Chen A, Stecker E, Warden BA. Direct oral anticoagulant use: a practical guide to common clinical challenges. J Am Heart Assoc. 2020;9(13):e017559.
12. Aspirin and gastrointestinal bleeding risk in older people. Drug Ther Bull. 2021;59(8):117.
13. Skop BP, Brown TM. Potential vascular and bleeding complications of treatment with selective serotonin reuptake inhibitors. Psychosomatics. 1996;37(1):12-16.
14. Li N, Wallen NH, Ladjevardi M, Hjemdahl P. Effects of serotonin on platelet activation in whole blood. Blood Coagul Fibrinolysis. 1997;8(8):517-523.
15. Paton C, Ferrier IN. SSRIs and gastrointestinal bleeding. BMJ. 2005;331(7516):529-530.
16. Anglin R, Yuan Y, Moayyedi P, et al. Risk of upper gastrointestinal bleeding with selective serotonin reuptake inhibitors with or without concurrent nonsteroidal anti-inflammatory use: a systematic review and meta-analysis. Am J Gastroenterol. 2014;109(6):811-819.
17. Hackam DG, Mrkobrada M. Selective serotonin reuptake inhibitors and brain hemorrhage: a meta-analysis. Neurology. 2012;79(18):1862-1865.
18. Quinn GR, Singer DE, Chang Y, et al. Effect of selective serotonin reuptake inhibitors on bleeding risk in patients with atrial fibrillation taking warfarin. Am J Cardiol. 2014;114(4):583-586.
19. Schalekamp T, Klungel OH, Souverein PC, de Boer A. Increased bleeding risk with concurrent use of selective serotonin reuptake inhibitors and coumarins. Arch Intern Med. 2008;168(2):180-185.
20. Boguta P, Juchnowicz D, Wróbel-Knybel P, et al. Safety of concomitant treatment with non-vitamin K oral anticoagulants and SSRI/SNRI antidepressants. Curr Probl Psychiatry. 2018;19(4):267-278.
21. Quinn GR, Hellkamp AS, Hankey GJ, et al. Selective serotonin reuptake inhibitors and bleeding risk in anticoagulated patients with atrial fibrillation: an analysis from the ROCKET AF Trial. J Am Heart Assoc. 2018;7(15):e008755.
22. Samuel NG, Seifert CF. Risk of bleeding in patients on full-dose enoxaparin with venous thromboembolism and selective serotonin reuptake inhibitors. Ann Pharmacother. 2017;51(3):226-231.
23. Gunasekaran K, Rajasurya V, Devasahayam J, et al. A review of the incidence diagnosis and treatment of spontaneous hemorrhage in patients treated with direct oral anticoagulants. J Clin Med. 2020;9(9):2984.
24. Labos C, Dasgupta K, Nedjar H, et al. Risk of bleeding associated with combined use of selective serotonin reuptake inhibitors and antiplatelet therapy following acute myocardial infarction. CMAJ. 2011;183(16):1835-1843.
25. Sansone RA, Sansone LA. Warfarin and antidepressants: happiness without hemorrhaging. Psychiatry (Edgmont). 2009;6(7):24-29.
26. Weinz C, Schwarz T, Kubitza D, et al. Metabolism and excretion of rivaroxaban, an oral, direct factor Xa inhibitor, in rats, dogs, and humans. Drug Metab Dispos. 2009;37(5):1056-1064.
27. Forbes HL, Polasek TM. Potential drug-drug interactions with direct oral anticoagulants in elderly hospitalized patients. Ther Adv Drug Saf. 2017;8(10):319-328.
28. Abdelmalik N, Ruhe HG, Barwari K, et al. Effect of the selective serotonin reuptake inhibitor paroxetine on platelet function is modified by a SLC6A4 serotonin transporter polymorphism. J Thromb Haemost. 2008;6(12):2168-2174.