Cerebrovascular disease is a leading cause of death and disability in developed countries. During the last decade, the increasing number of stroke survivors with residual impairments and disabilities1 has driven interest in interventions that might optimize rehabilitation. In this context, treatment of poststroke depression (PSD), which is considered among the most frequent yet treatable neuropsychiatric consequences of stroke,2 plays a crucial role in the rehabilitation process.3 Findings suggest that PSD is associated with adverse cognitive effects, impaired recovery in terms of activities of daily living,4 and increased mortality.5 Recent studies highlight the co-occurrence of anxiety disorders and depression in poststroke patients.6 Although the prevalence of anxiety disorders is as high as that of depression, and approximately half of the patients with poststroke anxiety experience major depression, these are often overlooked and consequently not addressed during the rehabilitation process, worsening the functional outcome of stroke sufferers.7
Nearly half of stroke survivors experience poststroke fatigue, which may occur independently of PSD,8 it manifests as both physical and mental lack of energy. Many patients describe fatigue as one of the most debilitating sequelae they have to cope with after stroke because it interferes with rehabilitation and impairs their ability to return to their previous activities;9 unfortunately, treatment with antidepressants is not promising.
Several studies have shown that, in most cases, PSD can and should be rigorously treated to alleviate depressive symptoms and improve neurological deficits.10,11 Various biological treatments, such as antidepressant medication, electroconvulsive therapy (ECT), and repetitive transcranial magnetic stimulation (TMS), have been shown to be of value.12 The use of selective serotonin reuptake inhibitors (SSRIs) is the first-choice treatment for patients with PSD; however, recent studies have shown that serotonin-norepinephrine reuptake inhibitors (SNRIs) can also be used.13
The aim of the present naturalistic study was to determine the relative efficacy and tolerability of duloxetine versus two frequently used SSRIs in the treatment of poststroke depression and anxiety. A second objective of the study was to investigate duloxetine’s effect on fatigue-reduction.
Study Design and Participants
The present study was an open-label, controlled clinical trial of 3-months’ duration. Participants were patients who suffered an acute ischemic stroke or intracerebral hemorrhage followed by PSD. They were recruited from a pool of follow-up patients at the outpatient stroke unit of the Athens University Medical School, Department of Neurology, at the Eginition Hospital, Athens, Greece.
Inclusion criteria were the following:
Diagnosis of the first-ever stroke within the last 12 months, based on clinical history, physical examination, and findings of brain MRI.
Diagnosis of PSD was based on the fulfillment of the criteria for a mood disorder, depressive episode, according to the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV).
Exclusion criteria were as follows:
History of a major psychiatric disorder during the last 5 years before the stroke.
Atherosclerotic disease (i.e., myocardial infarction) or a history of angioplasty or bypass surgery.
Another major medical illness (i.e., severe diseases of the heart, lung, liver, kidney, endocrine system, or metabolic illness).
Degenerative, progressive neurological disease (e.g., dementia, multiple sclerosis, hydrocephalus).
Severe cognitive impairment, as assessed by the Mini-Mental State Exam (MMSE <24).
Also, patients enrolled in the study were required not to have severe comprehension deficits, as indicated by their diminished ability to comprehend and respond to the demands of the psychometric assessments.
Participants were randomly assigned to three age-, sex-matched study groups: the duloxetine group (N=20), which included patients treated with duloxetine (60–120 mg/day) in two divided doses, mean dose (SD): 99.0 (27.7) mg/day; the citalopram control group (N=20), which included patients treated with citalopram (20–40 mg/day; mean dose 24.0 (6.8) mg/day); the sertraline control group (N=20), which included patients treated with sertraline (50–200 mg/day; mean dose 126.0 mg/day (41.6) mg/day). Duloxetine was initiated with 30 mg/day and was gradually up-titrated to 60–120 mg/day over the 3-month study period.
The study was approved by the local ethics committee of the Eginition Hospital and observed the Helsinki Declaration for clinical research. Written informed consent was obtained from all participants.
Sociodemographic data (age, sex, socioeconomic status, marital status, level of education) and a detailed medical history of all participants were recorded.
The Mini-Mental State Exam (MMSE) was used for assessment of cognitive functioning. The degree of disability or dependence in daily activities was measured by the modified Rankin Scale.14 Symptoms of depression and anxiety were assessed through the Hamilton Rating Scale for Depression (Ham-D) and the Hamilton Anxiety Rating Scale (Ham-A), and poststroke fatigue through the Fatigue Severity Score (FSS),15 which is an instrument that differentiates between fatigue and clinical depression. The Systematic Assessment for Treatment-Emergent Events (SAFTEE) scale,16 a structured instrument for recording adverse events, was used for monitoring treatment-related adverse events.
Brain imaging, as well as hematologic, thyroid, and biochemical laboratory tests were available for all participants.
Assessments were done by two experienced psychiatrists at four time-points; baseline assessment was performed at the time of enrollment (time-point: 0), and, thereafter, three more assessments were obtained at monthly intervals over the 3-month study period.
ANOVA post-hoc tests (Bonferroni test for multiple comparisons) were used to compare the mean values of numeric variables within and between groups at the different time-points. Chi-square tests were used for comparisons of categorical data. All statistical inferences were based on two-tailed probabilities. Data analysis was performed using the SPSS Statistical Package.
No significant differences were observed between the three study groups in terms of their sociodemographic characteristics, their mental status (MMSE score), and their disability in daily activities (Rankin Scale score; Table 1).
TABLE 1.Sociodemographic Characteristics, Mental State, and Degree of Disability in Daily Activities of the Sample
| Add to My POL
|Duloxetine Group (N=20)||Citalopram Group (N=20)||Sertraline Group (N=20)|
|Mean age, years||51.1 (13.4)||54.3 (12.5)||52.4 (11.4)|
|Family status, NS/M/D)||S: 6, M: 12, D: 4||S: 4, M: 14, D: 2||S: 5, M: 14, D: 1|
|Education, years||11.1 (4.1)||9.1 (4.1)||9.6 (3.6)|
|Mean Mini-Mental State Exam||28.1 (1.4)||27.9 (1.3)||28.2 (1.3)|
|Mean Rankin score||1.4 (0.5)||1.3 (0.47)||1.20 (0.67)|
The three groups did not differ in terms of Ham-D, Ham-A, and FSS scores at baseline (time-point: 0). A marked improvement in symptoms of depression (p<0.01) and anxiety (p<0.01), but not fatigue, was observed throughout the three consecutive assessments in all study groups (Table 2). At the 2nd and 3rd assessment, however, patients on duloxetine improved significantly more than the control subjects on all measures of psychopathology. At the end of the study after 3 months of treatment, duloxetine-treated patients achieved a higher remission in anxiety than those on SSRIs (p<0.01; Table 2). Side effects of treatment are shown in Table 3. No significant differences were recorded between the three treatment groups during the 3-month study.
TABLE 2.Mean (SD) Scores on Ham-D, Ham-A, and FSS at the Different Time-Points of Assessment in the Three Groups
| Add to My POL
|Treatment Group||1st Assessment (Time 0)||2nd Assessment (Time 1)||3rd Assessment (Time 2)||4th Assessment (Time 3)|
|Ham-D||Duloxetine||24.5 (7.5)||8.7 (5.1)+||3.1 (1.6)†||3.4 (1.5)|
|Citalopram||23.7 (6.7)||16.0 (6.9)*+||8.1 (2.8)*†||3.9 (2.0)×|
|Sertraline||23.8 (7.3)||16.9 (7.0)**+||8.1 (3.4)**†||4.2 (1.7)×|
|Ham-A||Duloxetine||11.9 (2.1)||8.0 (1.2)+||6.3 (2.1)†||4.3 (1.7)×|
|Citalopram||11.6 (2.5)||9.5 (1.5)*+||9.2 (1.6)*†||8.4 (0.6)*×|
|Sertraline||12.1 (2.9)||9.4 (1.5)**+||9.5 (1.3)**†||8.7 (1.5)**×|
|Fatigue Severity Scale||Duloxetine||4.5 (1.4)||4.2 (1.3)||4.0 (1.3)||3.7 (1.1)|
|Citalopram||4.5 (1.5)||4.2 (1.4)||4.0 (1.4)||3.9 (1.3)|
|Sertraline||4.6 (1.6)||4.3 (1.5)||4.1 (1.4)||4.0 (1.4)|
TABLE 3.Adverse Events by Treatment Group
| Add to My POL
|Duloxetine Group||Citalopram Group||Sertraline Group|
|N (%)||N (%)||N (%)|
|Nausea||3 (15)||4 (20)||5 (25)|
|Somnolence||3 (15)||4 (20)||3 (15)|
|Insomnia||1 (5)||0||4 (20)|
|Dry mouth||2 (10)||2 (10)||3 (15)|
|Diarrhea||0||2 (10)||4 (20)|
In the present naturalistic study, we compared the efficacy of duloxetine, sertraline, and citalopram in terms of their antidepressant and anxiolytic effects, as well as their impact on fatigue, in patients with PSD. The main finding was that duloxetine was well tolerated and significantly more effective than citalopram and sertraline for the treatment of anxiety symptoms. On the other hand, symptoms of depression significantly improved regardless of the type of drug treatment received; however, patients on duloxetine improved more and faster than those treated with SSRIs; this improvement was evident by the first month of treatment.
Controlled studies on the effectiveness of antidepressants in PSD are relatively few and are limited to the use of heterocyclic antidepressants or SSRIs.13 The benefits of antidepressant therapy in PSD are well-acknowledged,10 although there is insufficient evidence to support either their routine use for the prevention of depression or their positive effect on recovery from neurological impairments.17 The present study corroborates previous studies showing duloxetine's clinical efficacy in depressed patients with medical comorbidity.18,19 Furthermore, it provides evidence for an early response in PSD, which is of particular importance because early improvement of mood in depressed patients may lead to better treatment compliance and predict a better outcome.20 However, our findings are not in full agreement with those of a recent metaanalysis, where it was reported that duloxetine is less effective, albeit nonsignificantly, than the SSRIs in the treatment of major depression;21 also, significantly less well-tolerated than the SSRIs.22 This discrepancy may be caused by several reasons besides the biases inherent to the naturalistic design of the study, such as the particular clinical manifestations and pathophysiological characteristics of PSD. Thus, our sample comprised PSD patients who fulfilled the criteria for a “mood disorder due to a general medical condition, depressed type.” In poststroke depression and anxiety, somatic symptoms, such as reduced appetite, psychomotor retardation, pain, physical impairment, and medical comorbidity are highly prevalent, which considerably differentiates this population from people with primary depression. Duloxetine has been shown to be effective in alleviating such symptoms23,24 and consequently may improve overall symptoms/ratings of depression and anxiety. Furthermore, the underlying pathophysiological mechanisms of poststroke mood disturbances have been postulated to be considerably determined by biological factors, reflecting a neurophysiological response to brain injury, and psychosocial factors probably contribute less to their onset,25,26 which may be another reason for the differential response to various types of antidepressants.
In spite of the high prevalence (around 25% of poststroke patients) and persistence of anxiety symptoms/disorders after stroke,27 the efficacy of anxiolytic treatment has not been sufficiently investigated. Treatment is essentially based on the use of SSRIs for their better side-effect profile; the use of benzodiazepines is generally discouraged because of their abuse potential and their negative effects on cognitive functioning.28 To the best of our knowledge, duloxetine's efficacy has never been previously investigated in the treatment of poststroke anxiety. In the present study, duloxetine was shown to be clearly superior to SSRIs in terms of an early and significant reduction of anxiety symptoms, which can be attributed to its dual selective reuptake-inhibiting action on serotonergic and noradrenergic neurotransmission.29,30
Regarding poststroke fatigue, which is a major concern after stroke, a dually-acting antidepressant, such as duloxetine, might be a promising agent. However, despite a slight improvement in fatigue, no significant changes were observed after antidepressant treatment. Most probably, since poststroke fatigue is reportedly a multidimensional constellation of symptoms, its pharmacological treatment may be problematic.31,32 Finally, the safety and tolerability of duloxetine, a prerequisite for the treatment of poststroke patients, was comparable to the SSRIs. Nausea, somnolence, insomnia, dizziness, dry mouth, and headache were the most common reported adverse events,33 but none of the patients dropped out of the study.
The main limitations of the present study were its small sample size and its naturalistic design. However, the inclusion of two treatment control groups validates our findings. Further placebo-controlled studies are warranted to provide clear evidence supporting duloxetine's efficacy in PSD. In spite of these limitations, our results reinforce the idea that duloxetine administration after a stroke is well tolerated and effective in reducing poststroke anxiety and depression; however, it does not reduce poststroke fatigue. In conclusion, duloxetine can serve as an alternative option to the already-approved agents for the treatment of poststroke depression and anxiety.