Post-Stroke Depression and Mood Evidence Tables and Reference List
Post-stroke depression (PSD) is a common consequence of stroke, although reported estimates may be unreliable given possible under-reporting of unusual mood, and the variability in the methods used to assess and define cases of depression within the literature. In a systematic review of 61 prospective, observational studies of post-stroke depression conducted in hospital-, rehabilitation-, and population-based settings, Hackett & Pickles (2014) estimated that approximately one-third of all individuals who experience stroke exhibited depressive symptoms at some point following the event (i.e., at acute, sub-acute or long-term follow-up). The overall pooled frequency estimate of PSD was 31% (95% CI 28% to 35%). Salinas et al. (2017) reported that of 1,424 postmenopausal women included in the Women’s Health Initiative who experienced a first-ever stroke, new-onset PSD occurred in 21.4% of participants, at an average of 16 months post stroke. Jorgensen et al. (2016) reported the incidence of persons developing depression was significantly higher compared with those of the general population matched for age and sex. During a two-year observation period, the incidence of depression was 25.4% vs. 7.8% (adj HR=4.09, 95% CI 4.00-4.18). In the prospective Depression Predictors after Ischemic Stroke study (DEPRESS), Guiraud et al. (2016) reported that among 251 patients with new onset stroke, the incidence of depression was 19% at two months and 24.3% at six months. Risk factors for the development of PSD include increasing age, living alone, high levels of comorbidity, a history of depression, female gender, physical disability (mRS score >2 at discharge), increased initial stroke severity, cognitive impairment and prior history of stroke. (Guiraud et al. 2016, Jorgensen et al. 2016, Kutlubaev & Hackett 2014, Ayerbe et al. 2013b).
The best time to screen formally for the possible presence of PSD is not certain. Although incident rates decline over time and there is a general trend toward improvement in depressive symptomatology during the first-year post stroke, PSD may prove to be persistent for a longer duration for a significant proportion of individuals. Screening for depression should be considered during the acute inpatient stay, at the point of transition to, or during inpatient rehabilitation, upon discharge to the community and during periodic health assessments. Swartz et al. (2017) describes the feasibility of using the 2-item version of the Patient Health Questionnaire during routine clinical practice using 1,500 outpatients attending a stroke prevention clinic. All patients were able to complete the screen, 89% of whom did so in less than 5 minutes. Karamchandani et al. (2015) reported that 70% of patients were eligible for depression screening prior to hospital discharge or transfer to another service. The remaining patients were not eligible due to aphasia, other medical condition, hospice/comfort measures, or prolonged intubation.
The diagnostic accuracies of several post-stroke depression screening and assessment tools have been examined. Meader et al. (2014) included the results of 24 studies and evaluated the performance of 18 previously-validated scales. The three best performing scales for the identification of any depression included Center of Epidemiological Studies-Depression Scale (CES-D) with a sensitivity and specificity of 75% and 85%, the Hamilton Depression Rating Scale (HDRS, sensitivity 84%, specificity 83%) and the 9-item version of the Patient Health Questionnaire (PHQ-9, sensitivity 86%, specificity 79%). The best two performing scales for the identification of major depression were HDRS and the PHQ-9. In a Canadian study (Prisnie et al. 2016) including 122 outpatients attending a stroke prevention clinic, the diagnostic accuracies of the PHQ-9 and PHQ-2 were evaluated. Using a cut-point of 13, the sensitivity and specificity of the PHQ-9 was 81.8% and 97.1%, and 75.0% and 96.3%, for PHQ-2, using a cut point of three.
Once possible depression has been detected via formal screening using a validated screening tool and the diagnosis confirmed by an experienced healthcare professional, treatments may be initiated. Pharmacotherapy with antidepressants has been associated with a reduction of depressive symptomatology. Xu et al. (2016) included the results from 11 RCTs of patients with a clinical diagnosis of post-stroke depression. Treatment with an antidepressant was associated with a significant reduction in depression scores (SMD=-0.96, 95% CI -1.41 to -0.51, p<0.0001), and better response to treatment (RR=1.36, 95% CI 1.01-1.83, p=0.04). A Cochrane review authored by Hackett et al. (2008), also reported the odds of remission of depression (i.e. a reduction of ≥50% in depression scale scores) were significantly higher with pharmacotherapy. Most of the agents evaluated in these reviews were selective serotonin reuptake inhibitors and tricyclic antidepressants. A systematic review by Chen et al. (2006) identified a relationship between duration and benefit of treatment. Analysis of studies with treatment durations of one and two weeks revealed no significant treatment effects; however, when treatment lasted for three weeks or more, the effects were greater. Many adverse events were associated with the use of pharmacotherapy in these studies.
Antidepressants have also been shown to improve functional recovery and reduce dependency in a person post stroke, both with, and without post-stroke depression (Mead et al. 2012, Chollet et al. 2011). The use of antidepressants has also been associated with reductions in emotional lability (Hackett et al. 2010), a common consequence of stroke. Pooling the results from 3 trials, the odds of improvement (i.e., reduction) in tearfulness were significantly increased in the treatment group (OR=9.35, 95% CI 4.26 – 20.54).
Non-pharmacological approaches to the treatment of post-stroke depression include different forms of psychotherapy, physical activity, non-invasive brain stimulation, and acupuncture. Psychotherapy (including problem solving therapy, cognitive behavioural therapy and motivational interviewing), has not been shown to be an effective treatment for depression in person recovering from stroke when used in isolation (Hackett et al. 2008), however, these same techniques may be effective when used in combination with pharmacotherapy (Mitchell et al. 2009). Behavioral therapy was shown to be effective for the treatment of post-stroke depression in persons with aphasia (Thomas et al. 2012). Acupuncture was shown to be superior to pharmacotherapy in the treatment of post-stroke depression. In a meta-analysis including the results of 15 RCTs of persons with post-stroke depression (Zhang et al. 2012), treatment with acupuncture was associated with improved odds of recovery/remission compared with pharmacotherapy (OR=1.48, 95% CI 1.10-1.97). Non-invasive brain stimulation using either repetitive transcranial magnetic stimulation (rTMS) or transcranial direct current stimulation (tDCS) has been shown to improve symptoms of depression. Significant reductions in Hamilton Rating Scale for Depression (HAD-D) scores were reported following two to eight weeks of rTMS therapy, compared with sham treatment in a meta-analysis including the results from 22 RCTs (Shen et al. 2017). At the end of treatment, the mean reduction in HAM-D scores was significantly greater for the rTMS group (MD=-6.09, 95% CI -7.74 to -4.45, p<0.0001). The ability of persons to perform ADLs was also significantly greater in the rTMS group (SMD=1.20; 95% CI 0.68-1.72, p<0.001). Treatment with 12, 30-minute sessions of tDCS (2 mA) in persons with post-stroke depression, treated an average of 15 months post stroke has also been associated with significant reductions in HAD-D scores (Valiengo et al. 2017). Physical activity has been associated with a small, but significant reduction in depression scores in a meta-analysis authored by Eng & Reime (2014) including the results from 13 RCTs (SMD=−0.13, 95% CI −0.26 to −0.01, p=0.03).
Prevention of Post-Stroke Depression
Given the high prevalence of post-stroke depression and the negative consequences associated with it, there has been increasing attention paid to strategies for its prevention. Pharmacologic prophylaxis, using many of the same agents, used for treatment, has been most commonly evaluated. In a pooled analysis based on 776 observations from 12 RCTs, Salter et al. (2012) reported the odds of developing post-stroke depression were reduced significantly with the use of prophylactic pharmacotherapy (OR=0.34, 95% 0.22-0.53, p<0.001). Similar effects have been reported in other systematic reviews (Yi et al. 2010, Chen et al. 2007). Non-pharmacological approaches have also been evaluated for the prevention of post-stroke depression. A Cochrane review (Hackett et al. 2008) included four trials that evaluated psychotherapeutic interventions, including problem-solving therapy (PST), home-based therapy and motivational interviewing. The odds of developing depression were significantly lower for participants in the active intervention groups (OR= 0.64, 95% CI 0.42 to 0.98, p=0.04), while psychological interventions were associated with a significant improvement in General Health Questionnaire (GHQ)-28 scores from baseline to end of treatment (MD= -1.37, 95% CI -2.33, -0.40, p=0.006). In a trial that included pharmacological and non-pharmacological study arms with long-term follow-up, Robinson et al. (2008) randomized 176 patients without depression to receive escitalopram, problem-solving or placebo, which was provided for 12 months. At one year, in the per-protocol analysis, adjusted for previous history of mood disorders, patients assigned to the placebo condition were significantly more likely to develop depression compared with those receiving either therapy with escitalopram (adj. HR= 4.5, 95% CI 2.4-8.2, p<0.001) or problem-solving therapy (adj. HR=2.2, 95% CI 1.4-3.5, p<0.001). In a follow-up study, Mikami et al. (2011) reported that when escitalopram was discontinued at the end of the study period, persons were more likely to develop major depression and had increased Hamilton Depression Rating Scale HDRS scores during the next six months, compared with those given placebo or PST. Finally, after a mean duration of eight years of follow-up, Robinson et al. (2017) reported that participants who received PST were significantly less likely to have died, compared with the combined group of escitalopram + placebo. Increasing age and the development of depression were found to be significant predictors or mortality.
Treatment of Anxiety Following Stroke
People with depression may also have a comorbid generalized anxiety disorder (GAD). Anxiety following stroke occurs in 20-25% of patients and is more common in women (Campbell et al. 2013). Despite the prevalence of post-stroke anxiety, very few studies have included evaluation of the effectiveness of potential treatments. A Cochrane review (Knapp et al. 2017) identified only three RCTs examining pharmacotherapy (paroxetine, buspirone) and a self-help autogenic relaxation CD). While the results from individual trials were positive, the results could not be pooled. The authors concluded there was insufficient evidence to guide treatment. Non-pharmacological approaches to the treatment of anxiety that have been reported to reduce anxiety symptoms include a self-help program (Golding et al. 2016 a,b), multidisciplinary in-home visits from rehabilitation therapists (Ryan et al. 2006) and acupuncture (Ping & Songhai 2008).