Post-stroke depression: Rapid action helps restore lost function
More than 60% of depressed stroke patients respond to antidepressants and tolerate them well
Patients with post-stroke depression (PSD) pose many clinical dilemmas: Is their depression a psychological reaction or a biological event? Are antidepressants effective for either type? Should antidepressants be given prophylactically after a stroke, even if the patient is not depressed?
Although the answers are not clear, this article describes a practical approach to stroke patients referred for psychiatric evaluation, including:
- strategies to distinguish reactive from endogenous depression
- issues that guide antidepressant selection
- benefits and risks of using medication to prevent depression after an acute stroke.
Reactive or endogenous depression?
Each year 500,000 to 600,000 Americans suffer strokes.1 Depression is the most common emotional sequela, reported in up to 40% of survivors within several months of an acute stroke.
Figure Possible mechanism for endogenous post-stroke depression
Anterior cerebrovascular lesions may block serotonergic and noradrenergic projections into the superficial cortex. The closer the lesion to the nuceli, the greater the pathway interruption, and the more severe the depression may be. Drawing represents nuclei in the brainstem, slightly enlarged, with their projections greatly simplified.
Source: Reference 8
Illustration for Current Psychiatry by Marcia Hartsock, CMIPSD is characterized as reactive (related to physical and psychosocial losses of stroke) or endogenous (a biologic consequence of stroke).
Reactive depression. Patients exhibit a constellation of emotional symptoms while attempting to cope with a new physical or cognitive deficit. This “catastrophic reaction”2 includes anxiety, crying, aggressive behavior, cursing, refusal, displacement, renouncement, and sometimes compensatory boasting.
In 62 stroke patients evaluated with the Catastrophic Reaction Scale:
- approximately 20% had a catastrophic reaction
- the reaction was significantly associated with major depression.
Anterior subcortical damage may be the common mechanism underlying both catastrophic reaction and major depression in stroke patients.3
Post-stroke emotional lability resembles PSD. This “pathologic emotion” or “emotional incontinence” can manifest as sudden, frequent, easily-provoked episodes of crying that are generally mood-congruent. Affected patients may also respond to nonemotional events with outbursts of pathologic crying or laughing.
The pathogenesis of post-stroke emotional lability is unclear, although biogenic amines may play a role. In a 6-week, double-blind trial, 28 patients with post-stroke pathologic laughter and crying were treated with nortriptyline or placebo. Symptoms improved significantly more with nortriptyline in both depressed and nondepressed patients, indicating that the response was not related simply to improved depressive symptoms.4
Endogenous depression. Robinson et al5 propose a neuroanatomic PSD model. They contend that major—but not minor—PSD correlates with the stroke lesion’s proximity to the left anterior frontal pole or underlying basal ganglia. Other investigators, however, question this anatomic distinction between major and minor PSD.
For example, Gainotti et al6 used their own Post-Stroke Depression Rating Scale to compare stroke patients without depression, with minor depression, or with major depressive disorder (MDD) and a group of nonstroke patients with functional MDD. They found that:
- the phenomenology of patients with major PSD was more similar to that of patients with minor PSD than to that of patients with functional major depression
- major and minor PSD were much more likely to be associated with reactive depression than with the endogenous form.
Other researchers disputed the neuroanatomic model after failing to confirm a correlation between PSD and the location of lesions in the left hemisphere.7
Most recently, a meta-analysis by Narushima et al8 suggested a moderately strong correlation between depressive symptom severity and the distance between the anterior border of a left-hemispheric lesion and the frontal pole during the first 6 months following a stroke. This group hypothesizes that more-anterior lesions interrupt the brain’s serotonergic and noradrenergic pathways (Figure) at a site closer to their origin—before they branch posteriorly into the superficial cortex. This interruption presumably increases serotonin and noradrenergic depletion, which is manifest as depression.
A common denominator? Two other recent studies suggest that depression may be a significant independent risk factor for stroke:
- A prospective study has assessed stroke risk factors in 2,800 Australians since 1988. Depression has been a significant stroke risk factor in men and women ages 70 and older.9
- A population-based study showed that depression predicted stroke across all strata in a cohort of 6,000 men and women ages 25 to 74. Subjects were stroke-free at enrollment and followed for up to 22 years.10
How to evaluate a patient for post-stroke depression (PSD)
These observations—plus the fact that depression is a common sequela of stroke—suggest that stroke and depression may result from a common pathophysiology. Some evidence suggests that depression is a cerebrovascular disease. For example, stroke and depression are both associated with increased platelet reactivity (stickiness). However, platelet reactivity does not appear greater in patients who develop depression after a stroke than in those who do not.11
Diagnostic criteria for mood disorder due to stroke*
With depressive features: if the predominant mood is depressed but the full criteria are not met for a major depressive episode.
With major-depressive-like episode: if the full criteria are met (except criterion D) for a major depressive episode.
*DSM-IV-TR diagnostic criteria for mood disorder due to a general medical condition
Source: Adapted and reprinted with permission from the Diagnostic and Statistical Manual of Mental Disorders, 4th edition, text revision. Copyright 2000. American Psychiatric Association.
A study of antidepressants’ effects on platelet dysfunction found that the selective serotonin reuptake inhibitors (SSRIs) paroxetine and sertraline may reverse platelet abnormalities, whereas the tricyclic nortriptyline does not.12
Summary. A biopsychosocial model helps explain the pathogenesis PSD, which may present as:
- a syndrome similar to MDD
- a relatively minor depression similar to dysthymic disorder
- a discrete phenomenon such as catastrophic reaction and emotional lability.
Evaluating patients for PSD
In the absence of PSD diagnostic guidelines, Table 1 provides a clinically useful approach. A comprehensive workup may suggest whether an individual’s depressive symptoms are endogenous or reactive, although this distinction may be subtle. Gathering information from other physicians, the patient, family, and caregivers may require more than one session.
Diagnostic criteria. Consult DSM-IV criteria for “mood disorder due to general medical condition” as they pertain to patients “with depressive features” (minor depression) or “with major depression-like episodes” (Table 2). In the initial evaluation, do not rely exclusively on instruments such as the Hamilton Depression Rating Scale for measuring depression, as patients with emotional lability may not meet the cut-off scores for depression. These scales may help monitor progress later during therapy.
Risk factors. Interview the patient, family, and caregivers to determine if the patient has possible risk factors for PSD:
- history of stroke
- personal or family history of depression
- loss of social activities
- major life event within 6 months of stroke
- cognitive impairment 1 month post-stroke.
Determine whether the patient’s stroke was precipitated by cocaine or other drug abuse. If so, address the cause.
Medical comorbidities. Consider the effect of medical comorbidities such as Parkinson’s disease, heart disease, or diabetes on the patient’s mood. Also rule out other metabolic causes of depression such as thyroid abnormalities, medication side effects, and vitamin B12 deficiency.
Treating post-stroke depression with antidepressants
Proven in double-blind, placebo-controlled trials
Greater anxiolytic effect than TCAs (?)
Alpha-adrenergic blockade, anticholinergic, antihistaminic, and cardiac effects
Drug interactions related to cytochrome P-450 isoenzyme inhibition
Safe in overdose
Onset of action
Slower than SSRIs (?)
More rapid than TCAs (?)
Less expensive than SSRIs
More expensive than TCAs
(?) Not supported by controlled clinical trials
TCAs: Tricyclic antidepressants
SSRIs: Selective serotonin reuptake inhibitors
Cognitive changes. Quantify any cognitive deficits with neuropsychological testing, such as the Mini-Mental State Examination. Also learn all you can about the patient’s premorbid personality for comparison with post-stroke behavior. Ask the nursing or rehabilitation staff about inpatients’ motivation and participation in care.
Social support. Determine if caregivers can provide transportation, medication monitoring, and other social support.
PSD calls for rapid, comprehensive treatment with antidepressants, psychotherapy, and help reintegrating into the community. Untreated PSD is associated with increased morbidity and mortality, whereas effective treatment improves functional outcomes. 13
Antidepressants. Tricyclic antidepressants (desipramine, imipramine, and nortriptyline), SSRIs (citalopram and fluoxetine), and trazodone have been used to treat depression in stroke patients (Table 3) 14 Controlled studies suggest that:
- >60% of patients with PSD respond to medication
- they tolerate antidepressants well
- no antidepressant class has a distinct therapeutic advantage over others.
Antidepressants in other classes—such as venlafaxine, bupropion, and mirtazapine—have not been studied in patients with PSD.
Without strong evidence to guide the initial antidepressant choice, a pragmatic approach is to start with one or two agents with which you are most familiar. Consider side effects, potential drug-drug interactions, cost, and available formulations. A patient with post-stroke swallowing difficulties, for example, may benefit from a liquid form.
We do not have good data regarding the optimum starting dose and duration of therapy for any antidepressants in PSD. To minimize side effects, I recommend starting with low dosages, such as:
- fluoxetine, 10 mg/d
- sertraline, 25 to 50 mg/d
- nortriptyline, 10 to 25 mg/d.
Increase dosages gradually, watching for side effects and symptom improvement.
If treatment is effective, continue the antidepressant for at least 1 year. In patients with a history of depression, continue treatment longer to prevent depressive relapse.
Small, double-blind, controlled trials have used citalopram, fluoxetine, nortriptyline, or sertraline to treat post-stroke emotional lability.14 Compared with placebo, these agents all significantly reduced post-stroke emotionalism.
Psychostimulants such as methylphenidate and dextroamphetamine might be an effective alternative to antidepressants.15 They have a morerapid onset of action, better tolerability, and may be more effective in alleviating post-stroke apathy. Disadvantages include risks for tolerance, dependence, and psychiatric side effects.
As with antidepressants, start low and go slow to minimize side effects. You could start methylphenidate at 5 mg in the morning and increase to 20 to 30 mg/d before you decide—based on response—to continue or discontinue. After the dosage is stabilized, you could switch to a controlled-release formulation.
ECT. Although no controlled trials of electroconvulsive therapy for PSD have been reported, ECT can be an effective option for patients with treatment-resistant depression. Some retrospective studies16 have shown a good response among patients with PSD, although ECT may worsen stroke-related cognitive deficits.
rTMS. University of Iowa researchers recently completed a double-blind controlled trial evaluating the efficacy of repetitive transcranial magnetic stimulation (rTMS) in PSD. R.G. Robinson, MD (personal communication, 2003), reported that preliminary results are encouraging.