Next Section 2. Vascular Cognitive Impairment
Table of contents / Mood, Cognition and Fatigue following Stroke
Table of contents / Mood, Cognition and Fatigue following Stroke

1. Post-Stroke Depression

Fifth Edition, 2015 UPDATE


The Canadian Stroke Best Practice Recommendations for the Mood, Cognition and Fatigue following Stroke, 5th Edition (2015) is published in the International Journal of Stroke (IJS) and available freely online. To access the specific recommendations for Post-Stroke Depression and all other sections of the Mood, Cognition and Fatigue module, please click on this URL which will take you to the recommendations online in the IJS.

For the French version of these recommendations, open the appendix at this link.

All other supporting information, including performance measures, implementation resources, evidence summaries and references, remain available through this website, and not through the IJS.  Please click on the appropriate sections on our website below for this additional content.

Rationale +-

Approximately one-third of all individuals who experience stroke will exhibit symptoms of depression at some time following the stroke event (acute, sub-acute and at long-term follow-up). The majority of cases of post-stroke depression (PSD) may develop in the first 3 months of stroke, and incidence rates of depression tend to decline over time, although research reports have indicated symptoms have emerged up to two years after index stroke. In one study, post-stroke depression (PSD) also was reported in 48% of 71 young stroke patients after at least 1 year of follow-up. PSD may prove to be persistent for as many as one-half of the individuals identified as depressed soon after stroke. Severity of functional limitations, stroke severity, cognitive impairment and a previous history of depression have all been identified as important risk factors for the development of PSD.

PSD is associated with poorer functional recovery, increased risk for dependence, poorer cognitive function and reduction in social participation. In addition, the presence of PSD has been associated with increased risk for mortality. Appropriate identification, diagnosis and treatment of PSD have been associated with improved outcomes.

Families and caregivers of patients who experience a stroke are also at risk for depression, with the reported incidence as high as 30% to 60% of caregivers experiencing depressive symptoms.

System Implications +-

The findings of this review lead to several implications for the healthcare system as follows:

  1. Screening tools should be available that are sensitive to unique patient circumstances such as patients with communication deficits and tools that may be culturally appropriate.
  2. Education for primary care practitioners and healthcare providers across the continuum of stroke care on recognition, assessment, and management of post-stroke depression.
  3. Timely access to appropriate mental health specialists as needed who are able to diagnose and evaluate severity of depression, and provide guidance for ongoing management.
  4. Timely access to and availability of specialized therapies to manage post-stroke depression, including medication, counseling and psychotherapy as required.
  5. Mechanisms to ensure good communication and information flow between the range of specialists and programs beyond the core stroke care providers to meet the varied needs of individuals post stroke (e.g., mental health specialists, cognitive specialists, geriatric programs).
  6. Process for ongoing monitoring of any patient with positive screening for depression during screening and assessment process.
  7. Education and support for caregivers of stroke survivors
  8. Processes should be in place to provide education and ensure that the caregivers’ emotional needs are monitored and addressed, ideally through involvement of the primary health care team.
  9. Optimization of strategies to prevent the recurrence of stroke.
Performance Measures +-
  1. Proportion of acute stroke patients with documentation indicating initial screening for post-stroke depression was performed (either informally or using a formal screening tool) in the acute care, rehabilitation, long-term care and community settings (e.g., homecare) setting. (Core Indicator)
  2. Proportion of acute stroke patients referred for additional assessment or intervention for a suspected diagnosis of depression.
  3. Proportion of stroke patients diagnosed with PSD and treated with antidepressants and/or psychotherapy at one month, three months, six months, and one year following the initial stroke event.

Measurement Notes

  • Recommendations for screening and assessment of PSD and corresponding performance measures apply across the continuum of stroke care and should be considered in the acute, early rehabilitation and longer-term recovery phases and in all healthcare settings.
  • When monitoring these performance measures it is important to record when and in what context (continuum of care) the measurements were conducted.
  • Data for measurement may be found through primary chart audit. Data quality will be dependent on the quality of documentation by healthcare professionals.
  • For patients referred to psychiatry, information may be available through provincial physician billing databases.
  • For persons over 65 years old, information on medication prescriptions may be available through provincial and territorial senior drug benefit plan databases.
  • For performance measure 3, the issue addressed is to increase the number of patients with PSD who are adequately treated, and reduce the number of patients with depression who are untreated (depressive disorder + no antidepressant medication) and undertreated (depressive symptoms + antidepressant medication + ongoing symptoms). This should be considered in the measurement and analysis plan.
Summary of the Evidence +-

Post-Stroke Depression and Mood Evidence Tables and Reference List

Post-stroke depression is a common consequence of stroke. In a systematic review of 51 prospective, observational studies of post-stroke depression conducted in hospital-, rehabilitation-, and population-based settings, Hackett et al (2005) estimated that approximately one-third of all individuals who experience stroke exhibit depressive symptoms at some point following the event (i.e., at acute, sub-acute or long-term follow-up). The authors suggest that this proportion is likely an underestimation of prevalence, given possible under-reporting of unusual mood, difficulty in the assessment of depression within the stroke population, and the variability in the methods used to assess and define cases of depression within the literature (Hackett et al. 2005).

In a systematic review intended to establish risk factors for the development of PSD, Hackett and Anderson (2005) included data from a total of 21 studies. Of the many different variables assessed, physical disability, stroke severity and cognitive impairment were most consistently associated with the development of depression following a stroke event. The authors noted that major methodological limitations within the available literature made it difficult to form a definitive conclusion. As part a multi-centred, observational study of depression in stroke, Paolucci and colleagues (2005) identified female sex (OR=1.49), previous stroke (OR = 1.55), previous depression (OR=3.97) and severe disability (Modified Rankin Scale score >3, OR = 2.70) as variables associated with increased odds for the development of PSD. In addition, the risk for post-stroke depression was found to increase exponentially in individuals with more than one risk factor (Paolucci et al. 2005).
Depression and anxiety may also be increased in parents of children with stroke (Goodman 2000; Gordon, 2002). PSD in survivors of childhood stroke is not well studied but appears to be common based on preliminary evidence (Elbers, 2013). In young adults with stroke, PSD risk factors are localization of the infarct in a carotid territory, severe disability, bad general outcome and absence of return to work (Neau et al., 1992).

In a report from the Auckland Regional Community Stroke Study (Hackett and Anderson 2006), the authors described an attempt to create a simple, predictive tool for the identification of individuals most at risk for abnormal mood. Of the factors included in the model, only two were significant predictors of mood; prior treatment for depression (OR = 2.4, 95% CI 1.34 – 3.43) and requiring “much help” with activities of daily living (OR = 2.35, 95% CI 1.33 – 4.14).The ability of the model to predict risk for depression might be increased by the inclusion of other factors such as fatigue and performance of instrumental activities of daily living. However, Van de Port et al. (2007) demonstrated that use of these two predictors (prior treatment for depression and requiring much help with ADLs) in a multivariate model could correctly classify depression in 76% of patients 3 years post stroke.

Examination of the results of multiple meta-analyses (Mitchell and Kakkadasam 2011; Mitchell et al. 2011; Mitchell et al. 2010; Mitchell et al. 2009; Cepoiu et al. 2007) revealed that non-psychiatric physicians, nurses, and therapists demonstrate poor sensitivity relative to gold standard psychiatric interviews and standardized formal rating scales when using clinical observation to identify depression in individuals who have experienced stroke in both inpatient and community-based settings. By comparison, use of standardized screening tools is associated with significantly greater sensitivity (Lowe et al. 2004). As in the recent CANMAT task force recommendations, it is noted that the use of formal instruments is a key component in the diagnostic process required to promote early detection of depression.

The best time to screen formally for the possible presence of PSD is not certain and incident rates of post-stroke depression are not stable over time. Bour and colleagues (2011) reported a decrease in incident cases of depression over the course of the first year following the stroke event. Cumulative incidence of PSD was 18.8% at 1month and 23.1%, 26.7%, 31%, and 36.2% at 3, 6, 9 and 12 months, respectively. Although incident rates decline over time and a general trend toward improvement in depressive symptomatology is evident in the first year post stroke, PSD may prove to be persistent for a longer duration for a significant proportion of individuals identified as depressed (Ostir et al. 2011, Ayerbe et al. 2011, Farner et al. 2010, Berg et al. 2003).
In 2006, the Canadian Stroke Strategy and Heart and Stroke Foundation of Ontario supported a consensus process to identify a standardized basket of outcome assessment tools that could be used to across the stroke continuum of care. Included in the resulting collection of measures were the following screening tools for the identification of possible depression in individuals with stroke: The Hospital Anxiety and Depression Scale (HADS), the Geriatric Depression Scale (GDS) and the Stroke Aphasic Depression Questionnaire (SADQ-10). Additional tools for consideration include the 9-item Patient Health Questionnaire (PHQ-9), and the Children’s Depression Inventory (CDI).   A brief description of these and several additional tools may be found in the accompanying summary table.

Once possible depression has been detected via formal screening and a diagnosis made by an experienced healthcare professional, treatment via pharmacotherapy has been associated with reduction of depressive symptomatology (Chen et al. 2006, Hackett et al. 2008). However, these findings should be considered in light of reports of adverse effects associated with the use of antidepressant medications. Whereas selective serotonin reuptake inhibitors are among the most commonly prescribed pharmacological agents for the treatment of PSD, they may also be associated with increased risk for mortality and stroke (Coupland et al. 2011, Wu et al. 2011). In addition, pharmacodynamic interactions between antidepressants and cardiovascular agents have been noted for several antidepressant classes including SSRIs, which may result in significant adverse events (Tuunainen et al. 2009).

Whereas psychotherapeutic approaches such as cognitive behavioural therapy or interpersonal therapy may be accepted treatment approaches for acute major depressive disorders in general, based on the results of available meta-analyses (Wilson et al. 2009, Hackett et al. 2008), there is insufficient evidence to support their application as monotherapy for the treatment of PSD. However, talk-based therapy, such as problem-solving therapy, when used in combination with pharmacotherapy may be an effective means to promote the reduction of symptoms of depression (Mitchell et al. 2009, Alexopoulos et al. 2012, Lincoln and Flanaghan 2003). The use of other, non-pharmacological strategies, such as light therapy (Tuuainen et al. 2009, Sondergaard et al. 2006), physical exercise (Graven et al. 2011), music therapy (Jun et al. 2012, Sarkamo et al. 2008), and acupuncture (Zhang et al. 2010) have all been demonstrated to have a positive effect on mood within the population of individuals with stroke.

For individuals who, following screening and appropriate follow-up assessment, experience mild symptoms of depression, watchful waiting may be the most appropriate strategy. Although there is no direct evidence of the effectiveness of this approach as an intervention within the stroke population, it has been used with good results as part of interventions, such as stepped care management (van’t Veer-Tazelaar et al. 2009, Dozeman et al. 2012), undertaken in populations of older individuals. A single randomized controlled trial examined the use of pharmacotherapy for individuals with mild depression and found a significant reduction in depressive symptomatology associated with the use of antidepressant therapy; however, study participants were not drawn from the stroke population (Williams et al. 2000).

Prevention of PSD 

Given the high prevalence of PSD and the negative consequences associated with it, there has been increasing attention paid to strategies for its prevention. The most commonly studied strategy has been universal pharmacologic prophylaxis, although discussions have expanded to include selected and/or indicated prophylaxis (Salter et al. 2013), problem-solving therapy (Robinson et al. 2008) and motivational interviewing (Watkins et al. 2007, 2011). Results of recent meta-analyses have provided encouraging results in favour of pharmacotherapy for the prevention of post-stroke depression (Salter et al. 2013, Yi et al. 2010, Chen et al. 2007); however, the individual studies within these analyses were small and provided conflicting results. In addition, little is known about how many individuals in these studies might have been experiencing subsyndromal depression, or had the greatest risk for PSD over time, and what the optimal timing and duration of preventative interventions might be whether using a universal, selected or indicated approach. In each case, the potential benefits associated with the use of prophylactic prevention of PSD must be weighed against the risks associated with the use of antidepressive agents, particularly among older individuals (Coupland et al. 2011).

Pseudobulbar Affect (Post Stroke Emotional Incontinence-PSEI)

Reported frequency of pseudobulbar affect (PBA) post-stroke ranges from 11%-35% depending upon the criteria used to define the condition and time elapsed since stroke onset (House et al. 1989, Kim and Choi-Kwon 2000). Although an association has been reported between PBA and PSD, many individuals with emotionalism may not end up with significant or diagnosable depression (Kim and Choi-Kwon 2000, Tang et al. 2004). In a recent Cochrane review (Hackett et al. 2010), pooled analyses were reported for data gathered from 5 randomized controlled trials examining antidepressant therapy for post-stroke PBA. Large treatment effects in terms of emotionalism, reduced tearfulness, clinical impressions of change and Lability Scale scores were reported in favour of antidepressant treatment when compared to control conditions. Given methodological limitations, the authors conclude that the existing literature does not provide definitive evidence that pharmacotherapy is effective in the treatment of PBA post-stroke. The combination medication dextromethorphan/quinidine (DM/Q) is available and approved in the US for the treatment of PBA.  The combination is not approved in Canada, but both components are available as they are approved for other indications.  As yet, DM/Q has not been evaluated in post-stroke PBA, and tolerability, particularly cardiac toxicity must be considered (Patatanian & Cassellman 2014 PMID:24704895; Schoedel et al 2014 PMID:25061302).

Anxiety Following Stroke

Anxiety following stroke occurs more often in women than in men. In the Perth Community Stroke Study, it was reported that 20% of women who experienced stroke developed symptoms of anxiety following the event while, in the same sample, only 9% of men experienced post-stroke anxiety (Burvill et al. 1995). However, individuals who experienced post-stroke anxiety often reported having anxiety or depression at the time of the stroke event (Burvill et al. 1995). Individuals with generalized anxiety disorder (GAD) after stroke may often experience co-morbid depression. Castillo et al. (1995) reported that, in a sample of individuals with post-stroke GAD approximately 75% were also depressed. Despite the prevalence of post-stroke anxiety, very few studies have included evaluation of the effectiveness of potential treatments. A recent Cochrane review (Campbell Burton et al. 2011) identified only 2 RCTs suggesting a positive effect associated with the provision of pharmacotherapy with or without the addition of psychotherapy. Further research in this area is certainly indicated in this post-stroke population.

Stroke Resources