CPCP: SSRIs and Risk of Stroke

We had a complicated Clinical Problems in Clinical Psychiatry (CPCP) recently presented by two talented psychiatry residents about selective serotonin reuptake inhibitors (SSRIs) and the risk for bleeding. Dr. Michele Morais leads off with the presentation of literature which tends to show there’s very low to no risk for bleeding with use of SSRIs. Dr. Adam Nardini follows with the opposing position from other studies.
It’s almost like dueling banjos.
I think this calls for remembering the importance of having full informed consent discussions with patients pointing out that studies about this issue use differently methodologies which leave room for uncertainty and that risks and benefits of treatment in psychiatry, just as in general medicine, need to be carefully weighed.

Dr. Michele Morais

1. The use of antidepressants and the risk of hemorrhagic stroke: a nested case-control study

Douglas, I., L. Smeeth, and D. Irvine, The use of antidepressants and the risk of haemorrhagic stroke: a nested case control study. British Journal of Clinical Pharmacology, 2011. 71(1): p. 116-120.

  • Selective serotonin reuptake inhibitors are associated with an increased risk of bleeding disorders at a number of sites.
  • It is currently unclear whether there is increased risk of hemorrhagic stroke with conflicting results reported.
  • This study found no association between SSRI use and hemorrhagic stroke.
  • There were a large number of patients involved which allowed them to rule out any substantial effect.
  • The results were similar in people with and without a previous history of a cerebrovascular accident.
  • Aim of this study is to investigate whether there is an association between the use of SSRI an increased risk of hemorrhagic stroke.
  • It is a case-control nested study which was done within a cohort of antidepressant users in the United Kingdom general practice research database. A group of 365,195 patients prescribed either an SSRI or TCA antidepressant between 1992 and 2006 were identified, 357 cases of hemorrhagic stroke were observed, and 1631 control patients without hemorrhagic stroke were selected.
Study Design:
  • Patients registered at the GPRD who had been prescribed an antidepressant at some time were selected, inclusion criteria: cohort registered between January 1, 1998 to December 31, 2006, patients who were given two or more prescriptions for either SSRI or TCA, people had to be registered at least 12 months before the first SSRI/TCA prescription was given.
  • People receiving both SSRI and a TCA, receiving anticoagulant at any point and patients with evidence of trauma within 30 days before a hemorrhagic stroke were excluded.
  • The start date was the earliest recorded SSRI or TCA prescription and the end date was the earliest of leaving the practice, and hemorrhagic stroke or 31st of December 2006.
Cases and controls:
  • Cases had a diagnosis of hemorrhagic stroke between January 1, 1998 and December 31, 2006. Controls were matched on age, gender, and practice and had to be registered on the date of diagnosis of their matched cases.
  • Cases and controls were excluded if they had a record of ischemic stroke, unspecified stroke, transient ischemic stroke or other cerebrovascular events within 30 days before the index date but patients with a history of cerebrovascular events earlier than 30 days before the index date were retained in the study which allowed them to investigate cerebrovascular events as possible effect modifiers.
  • SSRI and TCA exposure was classified as “current” if the prescription was dated within 30 days before the index date (which was the day the cases were diagnosed as having a stroke and were matched with a control), “recent” was defined if the prescription was dated 31-60 days before the index date. Prescriptions that were dated more than 60 days of the index date were considered unexposed to the antidepressant.
  • Conditional logistic regression was used to adjust for confounders like smoking, alcohol, BMI, prior TIA or other stroke, hypertension, diabetes, NSAID, aspirin, Clopidogrel, and dipyridamole.
  • The cohort of 396,195 patients was identified. 124,373 received two or more TCA prescriptions and 271,822 received two or more SSRI prescriptions. Within the cohort, 365 cases of hemorrhagic stroke were identified, for 8 cases no controls could be matched leaving 357 cases for whom 1631 controls were included.
  • The median age was 75 years for cases and 73 years for controls.
  • 65% of the cases were women compared with 66% of controls.
  • Among cases there was higher prevalence of hypertension, history of TIA/ischemic/unspecified stroke and a higher proportion of cases received aspirin in the past.
  • Among SSRIs fluoxetine was most commonly prescribed, among TCAs it was amitriptyline.
  • 102 cases were current users of SSRI at the index date compared with 386 controls.
  • 67 cases were current TCA users at the index date compared with 379 controls.
  • After adjustment for smoking, alcohol, observation time in the database, prior CVA, BMI, hypertension, diabetes, aspirin, Clopidogrel, dipyridamole, NSAID use and calendar year of first SSRI or TCA prescription the odds ratio was 1.11 (95% confidence interval).
  • For both users of TCA and SSRIs there was no evidence of effect modification by prior cerebrovascular events.
Analysis showed no evidence of an association between current SSRI or TCA use and hemorrhagic stroke.
  • Current use of SSRI compared with no use at the time of hemorrhagic stroke was associated with an adjusted odds ratio of 1.11 (95% confidence interval), for current TCA users the equivalent odds ratio was 0.73.
  • The study found no evidence of increased risk of hemorrhagic stroke associated with current or recent users of either SSRI or TCA.
  • Exposure to drugs was determined using prescription data; there was no guarantee that the drug was ever collected from the pharmacy or actually taken. This could be a potential source of misclassification. To minimize this they exclude patients with only a single TCA or SSRI prescription assuming that repeat prescriptions were more likely to indicate that the person was actually using the drug.
  • They believe that in this study they were better able to control key confounding factors compared with previous studies.
  • In contrast with the results seen in this study, The Women’s Health Initiative study reported a hazard ratio of 2.12 (95% confidence interval) for hemorrhagic stroke comparing prior SSRI users with nonusers in postmenopausal women; however, they considered exposure to the antidepressant at a single time point, and it was unknown whether there was association with current use.
  • This study has limitations in terms of generalizability because patients with recent trauma and prior anticoagulant treatments were excluded so these results may not apply for these patient groups.


  • No evidence was found that SSRIs are associated with an increased risk of hemorrhagic stroke regardless of prior history of cerebrovascular events.
2. Selective serotonin reuptake inhibitors and the risk of stroke: The population based case control study

Bak, S., et al., Selective serotonin reuptake inhibitors and the risk of stroke: a population-based case-control study. Stroke, 2002. 33(6): p. 1465-73.

  • This population-based data was used to estimate the risk of hemorrhagic and ischemic stroke in users of SSRIs.
  • Selective serotonin reuptake inhibitors are widely used in the treatment depression because of their tolerability and safety. However treatment with SSRIs is being associated with bleeding complications and it was recently shown that treatment with SSRIs increases the risk of upper GI bleeds. It is a major concern whether SSRIs also increase the risk of hemorrhagic stroke. This study found no association between the use of SSRIs and the risk of intracranial hemorrhage. Bleeding complications during SSRI treatment could be due to an impaired platelet response and conversely it would be conceivable that SSRIs might reduce the risk of thrombus formation and the risk of ischemic stroke, but this has not been previously studied.
  • Potential confounders: age, sex, hypertension, diabetes, smoking and other stroke risk factors are all possible confounders of the association between antidepressant use and stroke. Information on diseases influencing the risk of stroke was not available from the patient registries where data was collected for the study. Instead they retrieved information on the use of drugs which were regarded as proxy measures for certain risk factors. For example the use of prescriptions for diuretics, beta blockers, calcium channel blockers, ACE inhibitors, antiarrhythmics, and anti-anginal agents were proxy measures for hypertension, cardiac arrhythmias, and heart disease. Antidiabetics and lipid lowering agents were proxies for diabetes and hyperlipidemia.
  • Current use of anticoagulants: low-dose aspirin, and other NSAIDs were defined as for antidepressants.
  • This was a nested case-control study in Funen County Denmark. All patients with a discharge diagnosis of first-ever stroke in the period of 1994-1999 were identified and a validated diagnosis of stroke was reached; 4765 cases and 40,000 controls were randomly selected from the background population.
  • Information on antidepressant drug use for cases and controls was obtained from a prescription registry.
  • Odds ratios were adjusted for age, sex, calendar year and use of other medications.
  • Separate analysis on data from two large population-based surveys (the Middle-aged Twins survey and a longitudinal Study of Aging Danish Twins) which allowed them to study the association of use of SSRIs with various potential confounders not included in the nested case-control study.
  • People >20 years of age were identified with a discharge diagnosis of hemorrhagic stroke, ischemic stroke, subarachnoid hemorrhage, or unspecified stroke from January 1, 1994 to December 31, 1999 and was free of such diagnoses during the period of January 1, 1973 December 31, 1993. The data first admission for stroke was defined as the index date.
  • In some cases because of insufficient diagnostic workup patients could not be classified according to one of the above diagnoses. Patients with clinical signs of stroke for whom neuroradiological examinations or autopsy had not been performed were labeled as clinical stroke cases. Patients with some clinical signs of stroke but inadequate information for fulfillment on WHO definition relabeled as possible cases of stroke. Patients with register diagnoses of stroke that did not fulfill any of the criteria for a diagnosis were excluded. Cases of stroke occurring during hospitalization, subdural or epidural hematomas and head injuries were also excluded. So a total of 4765 cases were included in the case-control study.
  • All residents of Funen County during the period of 1994 to 1999 who were about 20 years of age with no hospitalizations for stroke between 1973 and 1993 were identified. They were given a random index date for the period of January 1, 1994 to December 31, 1999 and 40,000 controls with no stroke discharge diagnoses before the index date were chosen at random.
Exposure definition:
  • All available information was retrieved from the prescription registry on the use of antidepressants and other drugs and cases and controls before the index date. Antidepressants were classified according to their mechanism of action in three groups: the first group was composed of SSRIs including citalopram, fluoxetine, sertraline, paroxetine, and fluvoxamine; the second group was composed of amitriptyline, imipramine, and venlafaxine. The third group consisted of nortriptyline, mirtazapine, doxepin, desipramine, and others (for example mianserin, dosulepin, opipramol), MAOIs were not included.
  • A person was defined as a current user of antidepressants if the supply of the prescription ended after 30 days before the index date, recent users if the supply of prescription ended between 31 and 60 days before the index date, past users if the supply ended before 61 days before the index date, and never users if no prescriptions of antidepressants before the index date were recorded.
Data from Twin Surveys:
  • The data from this nested case-control study did not allow for sufficient control of the number of potential confounders. To estimate the magnitude and direction of association between the patient confounders and use of antidepressants was used from these two large population-based surveys.
  • Participants completed a comprehensive structured interview.
  • A total of 4556 participants were identified. 145 were excluded because of the diagnoses of stroke in a national patient register recorded before the interview date.
  • The same diagnostic codes were used as previously described to identify incidence of stroke in a national patient register and the exclusion procedure for this study was the same. Self-reported medication use was coded by a pharmacist according to the anatomical therapeutic chemical system (ATC). The participants were asked if providers had ever told them that they suffered from a number of diseases including hypertension, diabetes, myocardial infarct. Lifestyle questions included information about smoking, alcohol intake per week (no regular intake, less than one drink, moderate intake for men, 1-21 drinks, for women 1-14 drinks, high intake, men more than 21 drinks and women with 14), self-reported height and weight were used to calculate BMI.
  • To study the effect of treatment duration, never users of SSRI were compared with users of SSRI treated for less than 91 days and users treated for more than 91 days.
  • The effect of dose was estimated by comparing risk of stroke in never users of SSRIs, current users of low daily doses of SSRIs and current users of high daily doses.
  • To study the interaction between SSRIs and NSAIDS all persons who were never users of either drug were the reference group. The risk of stroke was independently assessed on SSRI and NSAID users and compared with the reference group. Finally the risk of stroke for joint use of both drugs was assessed and compared with current users of both drugs. Calculations were performed separately for hemorrhagic and ischemic stroke.
  • Unconditional logistic regression was used to study the association between patients’ stroke risk factors and current use of SSRIs. The adjusted for age, sex, hypertension, diabetes, myocardial infarction, smoking, alcohol intake, and BMI.
  • 4765 patients with a diagnosis of first-ever stroke during the period of January 1, 1994 to December 31, 1999 with identified.
  • 659 had intracerebral hemorrhage, 2717 at ischemic stroke.
  • Ever use of antidepressants was registered in 610 cases and 2910 controls.
  • The most commonly prescribed antidepressant was citalopram.
  • Among patients with intracerebral hemorrhage 21 were current users of SSRIs compared with 742 controls.
  • The relative risks of intracerebral hemorrhage in current users of SSRIs was 1.0 (95% confidence interval) compared with never users after adjustment for confounders.
  • Current use of anticoagulants and low-dose aspirin was independently associated with increased relative risk for intracerebral hemorrhage and drugs that acted as proxy measures of risk factors also increased relative risk of intracerebral hemorrhage.
  • Treatment duration and daily dose of SSRIs have no influence on the risk of ischemic or hemorrhagic stroke.
  • A trend towards increased risk of hemorrhagic stroke and increased risk of ischemic stroke was observed for persons with current exposure to both SSRIs and NSAID use; however, because of the small number of cases with concomitant exposure to both drugs, the results should be interpreted with caution.
  • The results suggest that exposure to SSRIs is not a risk factor for intracerebral hemorrhage and there was no evidence that current exposure is associated with a decreased risk of ischemic stroke.
  • Current anticoagulation therapy and medications acting as proxy measures for hypertension are associated with an increased risk of hemorrhagic stroke.
  • In both ischemic and hemorrhagic stroke they found that advancing age and male sex increase the risk of both hemorrhagic and ischemic stroke.
  • Past use of SSRIs was associated with a small increased risk of ischemic stroke and it is speculated that this is an indication that prior depression might be a risk factor for stroke, which is a finding reported in a number of studies. However, it is still unknown if depression is an independent risk factor for stroke or a consequence of silent strokes.
  • The antiplatelet effect of SSRIs may occur only during treatment with high doses and after long-term exposure. It was found that the average daily dose and treatment duration has no effect on the risk of hemorrhagic or ischemic stroke.
  • The small additive effect that they found in intracerebral hemorrhage in concomitant users of SSRIs and NSAIDS was not statistically significant and may well be a chance finding.

Potential limitations:

  • Misclassification of prevalent cases of stroke with incident cases because post stroke depression occurs frequently and is primarily treated with SSRIs. This was addressed in this study by exclusion of all potential cases and controls with stroke discharge diagnoses from 1973-1993.
  • Information on potential confounders was not directly available to this case-control study.

Dr. Adam Nardini

SSRIs and Bleeding

Halperin, D. and G. Reber, Influence of antidepressants on hemostasis. Dialogues Clin Neurosci, 2007. 9(1): p. 47-59.

  • SSRIs interfere with the platelet binding cascade in two proven ways.
    • Overall decrease in intra-platelet serotonin, which directly affects the ability of the platelets to bind with each other and withstand shear stress placed on formed clots.
    • Lowers platelet activation in response to thrombin receptor peptide, via decrease in the expression of the platelet activation marker CD63.

Of note, the study was done to determine the best treatment for post-MI depression and to minimize the risk of bleeding. It found mirtazapine 30-45 mg a day was superior to placebo in decreasing depression symptoms, without increasing the risk of bleeding.

Two Studies Arguing Against SSRI Usage:

1. Cochran, K.A., et al., Bleeding incidence with concomitant use of antidepressants and warfarin. Ther Drug Monit, 2011. 33(4): p. 433-8.

Materials and Methods: Retrospective, single-Center study of warfarin treated patients both prescribed (n=46) and not prescribed (n=54) any antidepressant over a six-month. Tracking was done with INRs, both outpatient and inpatient, if the patient happened to be admitted.
Results: Non-SSRI antidepressant with warfarin was not associated with significant shifts in the INR or adverse bleeding events. Use of an SSRI with warfarin, however, was associated with both these findings (p=0.04).
Discussion: Authors state the results should be considered a hypothesis and not a hard and fast result, given the retrospective nature of their study.
Of note, they quote a few studies with confirming and contradicting results.
Wallerstedt et al (2009)–Increased hospitalization in patients with atrial fibrillation on both SSRI (Zoloft or Celexa) and warfarin.
Hauta-Aho et al (2009)–Patients already hospitalized and taking SSRI and warfarin have a higher risk for upper GI bleeding.
Kharofa et al (2007)–No increased risk of new intracranial or subarachnoid hemorrhage with concomitant SSRI and warfarin.

2. Hackam, D.G. and M. Mrkobrada, Selective serotonin reuptake inhibitors and brain hemorrhage: a meta-analysis. Neurology, 2012. 79(18): p. 1862-5.

Materials and Methods: Meta-analysis of controlled observational studies (total 2493 citations reviewed, 81 selected, 506,411 patients total) comparing SSRI therapies with a control group not receiving SSRIs.
Results: New intracranial and intracerebral events were related to SSRI exposure in both adjusted and unadjusted analyses. SSRI exposure in combination with or oral anticoagulants (across-the-board) were associated with increased risk of bleeding compared to oral anticoagulants alone.
Discussion: SSRI exposure largely increases the risk of brain hemorrhage due to intracerebral hemorrhage, At an incidence of approximately 24.6 per 100,000 person-years. The authors say that an extra one patient in 10,000 per year would be expected if these medications were used in combination. The authors admit they did not control for many confounders such as hypertension hyperlipidemia, and renal function, but they feel strongly enough to recommend avoiding SSRIs and choosing other antidepressants in patients on oral anticoagulants, patients with cerebral amyloid angiopathy, or severe alcohol abuse.
Of note, with regard to the comment about SSRIs and severe alcohol abuse, the authors quote Ariesen et al, from Stroke (2003) who point out in their paper, Ariesen, M.J., et al., Risk factors for intracerebral hemorrhage in the general population: a systematic review. Stroke, 2003. 34(8): p. 2060-5, that severe (or chronic) alcohol abuse is a significant risk factor for brain hemorrhage, due to prolonged bleeding times and reduced platelet adhesion.dueling_banjos


  1. parentsfriend says:

    Pinned this on my Health and mental health board. The dueling banjos made a nice touch http://www.pinterest.com/pin/147141112799791525/


%d bloggers like this: