Clinical update • CLINICAL PRACTICE
Fary Khan, MBBS, FAFRM (RACP), is Lecturer, Rehabilitation Studies, Department of Medicine,
University of Melbourne, neuro-rehabilitation physician, the Melbourne Extended Care and
Rehabilitation Centre, the Royal Melbourne Hospital, and Head, Orthopaedic and Musculoskeletal
Unit, Caufield General Medical Centre, Victoria.
BACKGROUND S troke is the leading cause of disability in PSD and functional recovery
Poststroke depression (PSD) is common adults and is frequently associated with neu-
and often unrecognised. The diagnosis ropsychiatric symptoms such as depressed Poststroke depression is associated with
can be difficult due to deficits of stroke mood, generalised anxiety and apathy. The poor functional and psychosocial outcome.8
such as impaired self reporting and prevalence of poststroke depression (PSD) Although there is no long term data on the
cognition, poor insight and dysphasia. varies from 25–79% due to the differences in direct effect of PSD on cost of care there are
Untreated PSD can interfere with recovery various study diagnostic criteria, selection of reports of:
and adversely affect functional and social patients, and the time elapsed since the • prolonged inpatient hospital length of stay4
outcomes. stroke. • greater disability with activities of daily
Clinical depression is a common complica- living9
OBJECTIVE tion, and in some long term studies was • severe physical impairment10
This article outlines the diagnosis, shown to persist up to 3 years following • poor cognitive function11
pathophysiology and treatment for PSD. stroke.1,2 Depression decreases patients’ par- • poor participation in rehabilitation11
ticipation in rehabilitation and impairs • reduced social activity1
DISCUSSION functional recovery, community reintegration • poor language function1,2
The natural history of PSD suggests that and long term outcomes.3 It increases the cost • failure to return to work8, and
most PSD is not immediate but develops of treatment and burden of care to families. • a higher mortality at 10 years.12
over months with peak prevalence The diagnosis of PSD can be difficult,
between 6 and 24 months, and in some and is reported to have been missed in
What causes PSD?
cases persists up to 3 years following 50–80% of cases by nonpsychiatric physi- The aetiology of PSD is not well understood.
stroke. General practitioners and treating cians.4 Longitudinal studies on the natural Different mechanisms for PSD may be
specialists need to actively monitor history of PSD progression show that most involved in the aetiology of stroke over time
patients for PSD. While antidepressant depression is not immediate but develops and this has implications for treatment.
medication is the mainstay of treatment over months with peak prevalence between Depression may be a result of:
for PSD, psychotherapeutic interventions 6–24 months poststroke. 5 Most major • a biologic effect of brain damage
are important. Treatment should include depression remits in the 1–2 years following • a reaction to the losses caused by stroke,
patient and family education, re- stroke, either spontaneously having run its or
establishment of sleep pattern, addressing course or through treatment – it is rarely • a combination of these factors.13
functional difficulties, increasing chronic.6 Depression also occurs in 40% of Initially PSD may be caused by a neurophysio-
community participation, improving diet primary care givers of patients with stroke logic imbalance and depression that develops
and regular exercise. and is easily missed.7 later may be caused by psychological factors.1
Reprinted from Australian Family Physician Vol. 33, No. 10, October 2004 831
Clinical practice: Poststroke depression
Biological effects of stroke found to be more depressed overall,2 younger be independent of the presence of depres-
During the acute brain infarction there is patients tended to be depressed in the acute sion. However, depressed patients perform
decreased monoamine synthesis (enzyme early stages following stroke.28 poorly in areas of cognition2 and have poorer
inhibition during ischaemia) resulting in verbal abilities.29 Apathy is also seen in stroke
decreased 5-HT levels. These have been
Diagnosis patients and may coexist with emotional and
implicated in altered mood, 14 sleep, and The diagnosis of PSD may be difficult due to cognitive poststroke disturbances.
appetite.15 The use of serotonergic agents deficits in limited patient self report, impaired Symptoms of depression may be due to
has therefore been suggested to augment cognition, poor insight and aphasia. Features an underlying medical condition or cognitive
stroke recovery. 16 Stroke patients with such as anosognosia, fatigue, emotionalism, deficits rather than an underlying mood disor-
depression may also have: apathy and intellectual decline can also limit a der. Differential diagnoses include organic
• altered cortical receptor activity17 patient’s ability to express themself. This brain syndrome, side effects of medications,
• altered concentration of cerebrospinal may result in discrepancy between self rated sepsis and hypothyroidism.
fluid neurotransmitter metabolites18 depression and observer rated depression. A number of standardised tests are used
• electrophysiological abnormalities19, and Distinguishing between cognitive decline to screen patients for depression or to
• decreased cerebral blood flow.20 due to stroke and poor cognitive function monitor their response to treatment, but
Lesions in the left frontal lobe or basal secondary to depression can be difficult in these should not be used for diagnosis in iso-
ganglia are reported to cause more PSD than this group of patients. Stroke is followed by lation (Table 1). The Hospital Anxiety and
other brain areas.21 There is no clear associa- decline in cognitive function that appears to Depression Scale (HADS) and the General
tion between the volume of the lesion22 and
cortical/subcortical atrophy23 with the type or Table 1. Screening tools for depression and recommendations33
severity of depressive symptoms.
Hospital Anxiety and Depression Scale
Psychosocial effects of stroke
Seven items each measuring depression and anxiety. Originally used in 100 patients
Despite successful rehabilitation for mobility 16–65 years of age in outpatient settings. It has high reliability and validity, and only
and self care skills, social reintegration and 1% false positives and negatives. Correlations with psychiatric ratings were 0.79 for
depression and 0.54 for anxiety. Recommended by the British Stroke Research group
life satisfaction remain an issue for many
stroke patients. Stroke is associated with sig- Beck Depression Inventory
nificant psychosocial difficulties that can Recommended scale in integrated pathway of PSD. Twenty-one item self report
impact on the development of depression instrument with low reliance on somatic items. Respondents are asked how they
have been feeling over the past 2 weeks. Each item is rated on a 4 point scale,
ranging from 0–3. A cut off of 14/15 is indicative of depression. It has good reliability
• grief at loss of function, loss of indepen- and validity and has a positive predictive value of 0.54 and negative predictive value
dence or loss of employment of 0.99, and few false negatives. It has high internal consistency. Factor analysis can
• financial difficulties discriminate between those with cognitive and noncognitive dimensions. The
• social isolation advantage in stroke patients is the low demand on memory
• poor self esteem, and General Health Questionnaire-28 (GHQ-28)
• relationship or sexual difficulties. Uses 28 items for somatic symptoms, anxiety, insomnia, social dysfunction and
Early factors predictive of PSD include: severe depression. It is widely used, acceptable for the elderly patients, sensitive to
• aphasia at 3–12 months poststroke the effects of intervention and recommended by the British Stroke Research group. It
• older age has good reliability and validity. Coefficient correlations between GHQ-28 and
interview measures are 0.67 and 0.83 with a median of 0.76. The sensitivity ranged
• limited social supports
from 0.44–1.0 and specificity ranged from 0.7–0.93
• living alone, and
Depression, Anxiety, Stress Scale (DASS)
• a previous history of psychiatric
This dimensional scale contains 42 questions, with 14 items subdivided into 2–5
items. It assesses self reports on depression, hopelessness, anxiety and stress levels.
There is conflicting evidence regarding DASS has high internal consistency and yields meaningful discriminations in a
gender and PSD. In one study, women were variety of settings
found to be more depressed after stroke,25
but another found otherwise. 26 When fol- Other depression screening tools include: Hospital Stroke Aphasic Depression
lowed up, the male PSD patients had a Questionnaire, Brief Assessment Schedule for Depression Cards, Geriatric Depression
poorer prognosis compared with women Scale, Signs of Depression Scale, and Visual Analogue Mood Scale
patients. 27 Although older patients were
832 Reprinted from Australian Family Physician Vol. 33, No. 10, October 2004
Clinical practice: Poststroke depression
Health Questionnaire (GHQ) are the best work support; they have not demonstrated coping skills as detailed below.
validated scales in patients without commu- any clear benefit on mood state.31 Cognitive
nication problems and are probably the most behaviour therapy (CBT) results are encour-
useful for screening stroke patients without aging but these treatments are not always Selective serotonin reuptake inhibitors
communication difficulties in the rehabilita- accessible. Language limitations and cogni- (SSRIs) are the most commonly used drugs
tion or community setting. Those who score tion issues in stroke patients can interfere for PSD. Tricyclic antidepressants (TCAs) are
more than eight for depression on the HADS with psychotherapy interventions including less commonly used and inhibit the uptake
should be further assessed. CBT. In the weeks and months after the of both serotonin and noradrenaline. There
Full history, examination, and patient and stroke, the residual cognitive and behav- is no high level evidence to suggest SSRIs
family reports are important. The DSM IV ioural deficits are common causes for failure are superior to TCAs. In a double blind,
criteria for major depression are listed in to return to premorbid levels of interper- placebo controlled study, fluoxetine was
Table 2. The normal exclusion of symptoms sonal, vocational and recreational compared with nortriptyline – the response
due to an underlying medical condition is functioning. Patients with behavioural prob- was adequate with both drugs, but only the
waived in PSD.30 Older patients may experi- lems that are difficult to manage may be nortriptyline group improved functional inde-
ence depression without sadness, but they treated with success in specialised care pendence. 32 However, SSRIs have a safer
may present with anxiety and a general where behaviour is monitored, assessed side effect profile, a relatively quick onset of
sense of hopelessness. and subjected to positive or negative rein- action of 7–10 days, and good anxiolytic
forcers. In such patients, giving feedback effects making them first line antidepres-
Treatment and support assists in engaging the patient sant, especially in the elderly.
Treatment for PSD involves a combination in rehabilitation treatment and improves Within a class, there is no hard data to
of patient education, antidepressant medica- compliance. recommend one drug over another for PSD,
tion, behavioural strategies and In practice, antidepressants are often the however, a recent review 32 suggested an
psychotherapy. most pragmatic solution, with more involved optimum regimen for sertraline with a start-
A number of psychosocial interventions psychotherapy reserved for antidepressant ing dose of 50 mg per day, which can be
have been studied. These interventions resistant cases or those who are unable to increased to 100 mg per day in 2 weeks.
included controlled trials of counselling, tolerate medication. Patients still benefit There is little benefit from a higher dosage,
occupational therapy for leisure, and social from pyschoeducation and learning new and if there is no benefit after 4–6 weeks on
100 mg, then it should be discontinued. It
Table 2. DSM IV criteria for major depression should be continued for at least 6 months
and slowly weaned but can continue if
The American Psychiatric Association Diagnostic and Statistical Manual of Mental symptoms recur. Sertraline is also effective
Disorders (4th edn) (DSM-IV) list the following criteria for diagnosis of depression: in the treatment of emotionalism following
Five or more of the following symptoms persisting over a 2 week period causing stroke. Electroconvulsive therapy has been
clinically important distress or impairing work, social or personal functioning (with used to treat refractory severe PSD but is
depressed mood or decreased interest or pleasure as one of the five): limited due to side effects.
• depressed mood most of the day, occurring most days (subjective or observed)
• markedly diminished interest or pleasure most of the day, nearly every day
Psychosocial management strategies
• significant weight or appetite change
• insomnia or hypersomnia • education about depression to the
• psychomotor agitation or retardation (observable by others) patient, carer and family
• managing sleep-wake cycles
• fatigue or loss of energy
• improving nutrition
• feelings of worthlessness or inappropriate guilt • anxiety management techniques
• diminished ability to concentrate or make decisions • problem solving techniques
• recurring thoughts of death or suicide plans • behavioural programs to target specific
issues (eg. apathy, aggression)
Note: the normal exclusion of symptoms due to an underlying medical condition is waived in PSD30 • cognitive remediation (Table 1)
• compensatory rehabilitation strategies
for specific physical problems (eg. spas-
Reprinted from Australian Family Physician Vol. 33, No. 10, October 2004 833
Clinical practice: Poststroke depression
ticity, pain) patient. Gen Hosp Psychiatry 1992;14:69–76. Sunnybrook Stroke study: a prospective study
10. Parikh RM, Robinson RG, Lipsey JR, et al. The of depressive symptoms and functional out-
• addressing relationship difficulties (eg. impact of poststroke depression on recovery in comes. Stroke 1998;29:618–624.
altered family role, intimacy issues, self activities of daily living over a 2 year follow 26. Berg AL, Palomaki H, Lehtihalmes MLP, et al.
image) up. Arch Neurol 1990;47:785–789. Poststroke depression: 18 month follow up.
11. Sinyor D, Amato P, Kaloupek DG, et al. Stroke 2003;34:138–143.
• provision of aids and adaptive equipment Poststroke depression: relationships to func- 27. Burvill PW, Johnson GA, Jamrozik KD, et al.
to optimise physical fitness and maximise tional impairments, coping strategies and Prevalence of depression after stroke: the
function and community participation rehabilitation outcomes. Stroke Perth Community Stroke Study. Br J Psychiatry
• daily activity planning and reduction of
12. Morris P, Robinson R, Andrzejewski P, et al. 28. Robinson RG, Kubos KL, Rao K, Price TR.
social isolation Association of depression with a 10 year post- Mood changes in stroke patients: relationship
• improving patient and carer support, and stroke mortality. Am J Psychiatry to lesion location. Compr Psychiatry
• fatigue management and energy conser- 1993;150:124–129. 1983;24:555–566.
13. Roth EJ. Stroke. In: PM & R Secrets. O’Young 29. Berg A, Palomaki H, Lehtihalmes M, et al.
vation strategies. B, Young M, Steins S, eds. Philadelphia: Hanley Poststroke depression in acute phases after
& Belfus, 1997;253–262. stroke. Cerebrovasc Dis 2001;12:14–20.
Conclusion 14. Meltzer HY, Lowy MT. Serotonin hypothesis of 30. Rigler SK. Management of poststroke depres-
depression. In: Meltzer HY, ed. sion in older people. Clin Geriatr Med
General practitioners play a pivotal role in the Psychopharmacology: the third generation of 1999;15:765–783.
ongoing care and management of patients progress. New York, NY: Raven Press, 31. Turner-Stokes L, Hassan N. Depression after
recovering from stroke – and their families. 1987;513–526. stroke: a review of the evidence base to inform
15. Sonbrie P. Reconciling the role of central sero- the development of an integrated care
Early screening, monitoring and treatment for
tonin neurons in human and animal behaviour. pathway. Part 1: diagnosis, frequency and
depression after stroke can facilitate a Behav Brain Res 1986;9:319–364. impact. Clin Rehab 2002;16:231–247.
patient’s functional recovery, social reintegra- 16. Ramasubbu R, Flint A, Brown G, et al. 32. Robinson JR, Schultz SK, Castillo C, et al.
tion and improve their quality of life. Diminished serotonin mediated prolactin Nortriptyline versus fluoxetine in the treat-
responses in nondepressed stroke patients ment of depression and in the short term
compared with healthy normal subjects. Stroke recovery after stroke: a placebo controlled
Conflict of interest: none declared. 1998;29:1293–1298. double blind study. Am J Psychiatry
17. Barry S, Dinan TG. Alpha 2 adrenergic receptor 2000;157:351–359.
References function in poststroke depression. Psychol 33. Bennett HE, Lincoln NB. Screening for mood
1. Astrom M, Adolfson R, Asplund K. Major Med 1990;20:305–309. disorders after stroke. Stroke Association and
depression in stroke patients: a 3 year longitu- 18. Bryer JB, Starkstein SE, Votypka V, et al. the NHS Changing Workforce Program. School
dinal study. Stroke 1993;24:976–982. Reduction of CSF monoamine metabolites in of Psychology, University of Nottingham UK.
2. Kauhanen ML, Korpelainen JT, Hiltunen P, et poststroke depression: a preliminary report. J March 2004; in press.
al. Poststroke depression correlates with cogni- Neuropsychiatry 1992;4:440–442.
tive impairment and neurological deficits. 19. Kapen S, Greiffenstein M. Stroke and depres- AFP
Stroke 1999;30:1875–1880. sion: the cholinergic REM sleep induction test.
3. McGuire JR, Harvey RL. The prevention and In: Horne J, ed. Sleep 1988. New York, NY:
management of complications after stroke. Gustav Fischer, 1989;238–240.
Phys Med Rehab Clin North Am 20. Grasso MG, Pantano P, Ricci M, et al. Mesial
1999;4:857–874. temporal cortex hypoperfusion is associated
4. Schubert D, Burns R, Para W, et al. Increase of with depression in subcortical stroke. Stroke
medical hospital length of stay by depression 1994;25:980–985.
in stroke and amputation patients: a pilot 21. Starkstein SE, Cohen BS, Federoff P, et al.
study. Psychother Psychosom 1992;57:61–66. Relationship between anxiety disorders and
5. Robinson RG, Bolduc PL, Price TR. Two year depressive disorders in patients with cere-
longitudinal study of poststroke mood disor- brovascular injury. Arch Gen Psychiatry
ders: diagnosis and outcome at 1 and 2 years. 1990;47:246–251.
Stroke 1987;18:837–843. 22. Herrmann M, Bartels C, Wallesch CW.
6. Parikh RM, Lipsey JR, Robinson RG, Price TR. Depression in acute and chronic aphasia:
A two year longitudinal study of poststroke symptoms, pathoanatomo clinical correlations,
mood disorders: prognostic factors related to 1 and functional implications. J Neurol
and 2 year outcome. Int J Psychiatry Med Neurosurg Psychiatry 1993;56:672–678.
1988;18:45–56. 23. Starkstein SE, Robinson RG, Price TR.
7. Kotila M, Numminen H, Waltimo O, Kaste M. Comparison of patients with and without post-
Depression after stroke: results of the stroke major depression matched for size and
FINNSTROKE study. Stroke 1998;29:386–372. location of lesion. Arch Gen Psychiatry
8. Neau JP, Ingrand P, Mouille-Brachet C, et al. 1988;45:247–252.
Functional recovery and social outcome after 24. Anderson G, Vestergaard K, Ingemann-Nielson
cerebral infarction in young adults. M, Lauritzen L. Risk factors for poststroke
Cerebrovasc Dis 1998;8:296–302. depression. Acta Psychiatr Scand
9. Schubert D, Taylor C, Lees S, et al. Physical 1995;92:193–198. Correspondence
consequences of depression in the stroke 25. Herrmann N, Black SE, Lawerence J, et al. The Email: email@example.com
834 Reprinted from Australian Family Physician Vol. 33, No. 10, October 2004