Updated: Nov 7
Abstract and Introduction
Purpose of the Review: To examine recently published results of randomized placebo-controlled trials investigating the clinical effects of selective serotonin reuptake inhibitors on the prevalence of clinically significant symptoms of depression and suicidal ideation after an acute stroke.
Recent Findings: The prevalence of poststroke depression varies markedly according to the approach used to define depression, with recently published data suggesting that about one in every three stroke survivors will experience clinically significant symptoms of depression over a period of 12 months. The proportion of stroke survivors with clinically significant symptoms of depression decreases progressively with time, but in 30% of them symptoms persist or recur over 12 months. Routine daily treatment with 20 mg of fluoxetine for 6 months does not affect the prevalence of depression in this population, nor is it effective at treating or preventing poststroke depressive symptoms. Treatment discontinuation, gastrointestinal adverse effects, seizures and bone fractures are more frequent among stroke survivors treated with antidepressants than placebo. Moreover, current data show that thoughts about death or suicide are more frequent among adults who had a stroke than the general population, although recurring suicidal thoughts are uncommon. Routine daily treatment with 20 mg of fluoxetine for 6 months does not change the proportion of people who disclose suicidal thoughts over a period of 12 months after an acute stroke.
Summary: Current evidence raises concerns about the efficacy and safety of antidepressants for the management and prevention of poststroke clinically significant symptoms of depression. It is unclear if these findings can be generalized to people with severe strokes or to stroke survivors with moderate to severe major depressive episodes.
Stroke is a leading cause of disability and health burden worldwide. Poststroke rehabilitation and quality of life are often hindered by the presence of depressive symptoms, so that the effective management and prevention of depression has become an integral part of strategies designed to improve the clinical outcomes of stroke survivors. Antidepressants have been used to prevent and treat clinically significant symptoms of depression after a stroke, and more recently to promote neuro-regeneration and stroke recovery. The results of several randomized controlled trials have been published over the past couple of years, and many of them have examined the effects of antidepressants on the mental and physical outcomes of people who survive a stroke. This paper will summarize the key-findings of these studies.
The literature search was restricted to the years 2021 to January 2023 and made use of the following search terms in PubMed: stroke AND "clinical trial" AND depression AND antidepressant. A total of 39 papers were retrieved, of which 7 reported original information and were relevant. Cited bibliography and forward citations were then reviewed, and another 33 papers were identified. Of the 40 papers available, 23 papers were deemed relevant and were included in this review.
Stroke and Depression
Epidemiology of Poststroke Depression
Depression is common after stroke, although prevalence estimates vary widely according to the setting of the study and the approach used to define depression. Existing systematic reviews suggest that one in three stroke survivors will experience clinically significant symptoms of depression, and that the proportion of those affected does not vary much with time. In contrast, the results of the Assessment of Fluoxetine in Stroke Recovery (AFFINITY) trial found that about 19% of 1221 participants showed evidence of clinically significant symptoms of depression at the time of recruitment 2–15 days after a mild to moderate acute stroke (either ischaemic or haemorrhagic). The study defined clinically significant symptoms of depression as a score of 9 or greater on the Patient Health Questionnaire (PHQ-9). In addition, the proportion of those affected by clinically significant symptoms of depression decreased progressively with time to about 10% after 3 months and 8% after 6 months. The results of AFFINITY also showed that although the proportion of people affected by depression decreases with time, one in three survivors experienced clinically significant symptoms at some point over a period of 12 months, and in 30% of these symptoms were present both during the sub-acute (≤ 3 months) and recovery phases (3–12 months). In other words, depression was persistent or recurring in 30% of the stroke survivors who had experienced clinically significant symptoms of depression.
The Efficacy of Fluoxetine on Outcomes at 12 Months After Acute Stroke (EFFECTS) recruited 1500 adults after an acute stroke. The investigators defined clinically significant depression as a physician-based diagnosis that led to the prescription of a nontrial antidepressant. At 6 months, 6% of those recruited had been offered treatment with an antidepressant, and at 12 months 8.8%. These results were remarkably like those of AFFINITY regarding the prescription of novel antidepressants over a period of 12 months.
The discrepancy between the reported proportion of stroke survivors affected by clinically significant symptoms of depression and those who receive treatment is striking and raises questions about the validity of the approach most studies use to define depression (i.e., specific cut-points of various depression scales) or, alternatively, about the poor sensitivity of health professionals to identify stroke survivors with depression. A recent study of 343 stroke survivors living in Korea showed that less than half of those with depression were aware of their clinical status, and only 20% of these were receiving antidepressant treatment.
History of depression before the stroke, severity of stroke-related functional impairment, and older age seem to be the most robust measures associated with the presence of clinically significant symptoms of depression, whereas sex, living arrangements, and treatment with antidepressants do not seem to mediate or modulate the emergence of depressive symptoms after a stroke.
Antidepressants for the Treatment and Prevention of Poststroke Depression
Two Cochrane systematic reviews and meta-analyses published in 2020 produced equivocal results about the potential benefits of antidepressants for the treatment and prevention of depression after a stroke. Novel data from the AFFINITY and EFFECTS trials have raised additional questions about the use of antidepressants for the management of stroke survivors.
EFFECTS randomly assigned 1500 stroke survivors to 6-months daily treatment with 20 mg of fluoxetine or placebo. Both participants and investigators were blind to treatment assignment. After 6 months, 36 of 750 participants treated with fluoxetine and 51 of 750 treated with placebo had been prescribed treatment with an open-label antidepressant – differences between the groups were not statistically significant. By 12 months, the numbers had increase to 65 of 725 of those treated with fluoxetine and 64 of 728 of participants treated with placebo (P = 0.907).
The investigators of the AFFINITY trial reported data for 607 and 614 stroke survivors randomly assigned to daily treatment with placebo or 20 mg of fluoxetine in a double-blind fashion for 6 months. By the end of this period, 8% and 7% of the participants treated with placebo and fluoxetine showed evidence of clinically significant symptoms of depression – the difference between the groups was not statistically significant (P = 0.467). In addition, of the 228 participants with clinically significant symptoms of depression at the time of enrolment, 20% and 19% of those treated with placebo and fluoxetine were still experiencing clinically significant symptoms of depression after 6 months of treatment (P = 0.840). Conversely, of the 993 participants who were free of clinically significant symptoms of depression when they joined the study (498 assigned to treatment with fluoxetine), 15% and 13% showed evidence of clinically significant symptoms of depression over the 6 months of treatment with placebo and fluoxetine (P = 0.426). Taken together, the results of the AFFINITY trial showed that routine daily treatment with 20 mg of fluoxetine is no more effective than placebo at reducing the 6-month prevalence of poststroke depression, or at treating clinically significant symptoms of depression, or at preventing the emergence of clinically significant symptoms of depression. A similar lack of clinical benefit associated with fluoxetine treatment of stroke survivors was observed in Tanzania according to the results of a small open label trial (n = 59). Of note, a systematic review of randomized controlled trials produced results suggesting that treatment with fluoxetine is associated with a small reduction in the proportion of stroke survivors who were told by a doctor that they had depression during the respective studies.
A secondary analysis of a randomized, double-blind. placebo-controlled trial of treatment of stroke survivors treated with 10 mg of escitalopram daily for 3 months yielded results suggesting that scores on the Montgomery-Åsberg Depression Rating Scale (MADRS) declined more markedly among those using the antidepressant, although MADRS scores were well below a level of severity that would be considered clinically significant. The clinical relevance of this study was further compromised by the loss of over 25% of participants during the study, and the prior publication of trial data showing that treatment with escitalopram had no obvious effect on the proportion of participants affected by clinically significant symptoms of depression (MADRS ≥ 16).
Other Clinical Effects of Antidepressant Treatment After a Stroke One of the driving forces supporting the running of recent trials of antidepressants was the expectation that medications such as fluoxetine could promote neuronal regeneration and recovery after a stroke. Both the EFFECTS and the AFFINITY trials reported data showing that fluoxetine did not improve functional outcomes at 6 and 12 months after 6 months of treatment (function was measured using the modified Rankin Scale). A Cochrane systematic review and meta-analysis of 6 trials that used fluoxetine for the management of stroke survivors confirmed that treatment had no effect on motor function compared with placebo. The results of the review also showed that treatment with an antidepressant increased by 57% the proportion of participants leaving the study early and was associated with a 40% increase in the risk of seizures and 71% increase in the risk of gastrointestinal complaints. Death was no more frequent among people treated with antidepressants than placebo. The trials included in the review were rated as having low risk of bias. Another systematic review and meta-analysis designed to investigate the risk of bone fractures among stroke patients treated with a selective serotonin reuptake inhibitor (SSRI) identified four randomized placebo-controlled trials that had used either fluoxetine (n = 3) or citalopram (n = 1). The review found that treatment with an SSRI for 6 months more than doubled the risk of fractures among adults recovering from a stroke (risk ratio = 2.36). In this instance, the number needed to treat to cause one additional fracture was 58. The potential mechanisms supporting this increase in risk of fractures are not yet clear. There is also some concern that the introduction of antidepressants for the management of adults who had a stroke may increase the risk of stroke recurrence, although the AFFINITY trial found no evidence that treatment with fluoxetine for 6 months changed the risk of recurrence over 12 months. In fact, a slightly lower proportion of participants treated with fluoxetine than placebo had a new ischaemic stroke during follow up.
A posthoc analysis of the EFFECTS trial examined the association between treatment with fluoxetine and the evolution of individual items of the MADRS over 6 months. Antidepressant treatment had no obvious effect on total MADRS scores compared with placebo, but was associated with increased lassitude, which was considered indicative of greater apathy.
Stroke and Suicidal Behaviour
Thoughts about death or self-harm seem to be relatively common among stroke survivors compared with the general population, with a recent systematic review of 21 studies that included 17 189 participants reporting a pooled prevalence of 12%. However, there was marked heterogeneity between the studies included in the review, in addition to differences in the time between the stroke and the assessment (from 4 days to 12 months), inconsistency in the definition of suicidal ideation, and varying study designs. Some of these shortcomings were highlighted by a secondary analyses of the AFFINITY trial, which showed that the prevalence of suicidal ideation varied according to the approach used to define its presence: 3% if these thoughts had to be recurrent and present for most of the time compared with 14% if these thoughts had to be present at least once over a period of 52 weeks. The presence of thoughts about death or self-harm was not affected by the sex of participants, but clinically significant symptoms of depression were present in nearly all of those with recurring thoughts over a 2-week period (95%). Of note, only one the participants in the AFFINITY trial died by suicide, and this person had denied suicidal thoughts at study entry and again 4–6 weeks after the stroke. These results suggest that suicidal thoughts are not infrequent, but they are not robust predictors of suicide attempt or completion when taken in isolation. In addition, if poststroke suicidal thoughts are recurrent in nature, clinicians should complete a detailed mental state examination to clarify if clinically significant symptoms of depression are also present.
However, thinking about death or self-harm is not the same as attempting to kill oneself or completing suicide. A systematic review and meta-analysis of 23 observational studies found that stroke increased the risk of suicide completion by 61% and more than doubled the risk of attempted suicide. However, there was marked heterogeneity between the studies included in the analyses, as well as some evidence of bias favouring the publication of studies reporting a positive association between stroke and suicide (reported in the supplemental material of the relevant paper).
Trial data to guide the potential use of antidepressants to decrease suicidal behaviour after a stroke are limited. The investigators of the AFFINITY trial used the last item of the PHQ-9 ('better off dead or wanting to self-harm') to assess suicidal ideation and found that over the 12 months of the study the cumulative proportion of participants who had reported recurring thoughts of death or self-harm was 15 of 607 and 23 of 614 of those treated with placebo and 20 mg of fluoxetine for 6 months (P = 0.200). Among participants treated with placebo, 85 (14%) reported thinking about death or self-harm at least once during the 12 months of the study compared with 90 (15%) of those treated with fluoxetine (P = 0.744). No other trial data are available to guide the management of poststroke suicidal thoughts and behaviours, but a review of all available data on this topic suggested that treating clinicians should consider the effective management of clinically significant symptoms of depression, the use of interventions that promote rehabilitation and reduce disability, and restriction of access to means that may be used in suicide attempts. Extra-vigilance regarding the use of antidepressants in this population may be required because observational data suggest that exposure to antidepressant medications may increase the risk of suicide attempts in later life.
Stroke is associated with high morbidity and mortality, including mental health morbidity. Clinically significant symptoms of depression are common among stroke survivors, but the assessment of their presence has relied almost exclusively on the use of rating scales, rather than structured interviews leading to a clinical diagnosis according to accepted classifications systems, such as the International Classification of Diseases 11th revision (ICD-11) or the Diagnostic and Statistical Manual of Mental Disorders 5th edition (DSM-5). Consequently, the quality of available data examining the association between stroke and depression are, in this respect, sub-optimal.
The results of recent large randomized controlled trials that used SSRIs for the management of stroke survivors showed that daily treatment with 20 mg of fluoxetine (which was used in most of the trials reviewed) does not change the prevalence of clinically significant symptoms of depression over 6 or 12 months. Treatment with fluoxetine is no more effective than placebo at treating clinically significant symptoms of depression in this population, or in preventing the emergence of depressive symptoms among those who were not depressed at study entry. However, the trials reviewed in this paper were not designed specifically to treat people with poststroke major depressive episodes, so that it may be premature to dismiss the potential benefits of fluoxetine, or of other antidepressants, for the management of poststroke depression. What is clear is that fluoxetine, and possibly other antidepressants, increases the risk of adverse events among stroke survivors, so that its use should be limited to those likely to benefit the most: i.e., people with moderate to severe depressive episodes.
Future research should seek to determine the prevalence of depressive disorders in large community-representative samples of stroke survivors. It should also seek to establish the safest and most effective approach to managing and preventing poststroke depressive episodes and self-harm.
Related Topic: One Life Psychiatry