Reviews of the literature on poststroke depression (PSD) report that it is a common emotional outcome of stroke and that its prevalence ranges from 25% to 79%. This variability is attributed to differences in the populations studied (for example, in- or outpatient, hospital, rehabilitation unit, or community-based); different criteria for stroke and depression (for example, ischemic, hemorrhagic, cerebral, noncerebral, unilateral, bilateral, minor, or major); time since stroke onset (acute or chronic); instruments used to assess depression; and inclusion or exclusion of aphasia, dementia, past psychiatric history, or history of substance abuse.
The literature reports morbidity attributable to PSD: studies have found significant correlations with activities of daily living (ADLs), social handicap, functional impairment, and cognitive impairment. Associations with longer hospital stay, intellectual impairment, and significant reduction in speed and success of rehabilitation were also studied, with correlations reported to vary over the course of months and years poststroke. More strikingly, authors specify how difficult it is to determine whether these impairments result from depression or cause depression.
At least 2 studies report that depression or depressive symptoms are associated with increased risk of poststroke mortality Of 91 patients followed up after 10 years, 53% had died poststroke. Major and minor depression were associated with a 3.4 times higher risk of dying, compared with patients without depression. The relation was independent of other measured risk factors such as age, sex, social class, type of stroke, lesion location, and level of social functioning. Of the patients with few social contacts, 90% had died. Over the longer period of the second follow-up study, it was found that 5 or more depressive symptoms at baseline were associated with a significantly increased risk of stroke mortality (P < 0.006) after adjusting for age, sex, and ethnicity, and after additional adjustments for education, alcohol consumption, smoking, body mass index (BMI), hypertension, and diabetes (P < 0.02).
Considering the prevalence, morbidity, and mortality associated with PSD, it is important to be aware of its risk factors so that it is not overlooked—especially because it can often be treated effectively with antidepressants, stimulants, or electroconvulsive therapy (ECT), either combined with psychotherapy or alone.
During the late 1970s and early 1980s, researchers described a clear-cut relation between lesion location (left anterior in acute stroke, with proximity to the left frontal pole correlating with depression severity) and incidence or severity of PSD. Lesion volume was considered as well. Results were not always replicated, and in the late 1980s Primeau wrote that “association of PSD and left hemisphere is far from being widely accepted and has been consistently challenged in recent years”. In 1998, Singh and others stated that “based on systematic review of the present literature, any definitive statements regarding stroke lesions location and risk for depression are not substantiated.” In 1999, Shimoda and Robinson concluded that “anatomical correlates of PSD change over time and may explain interstudy differences in the association of lesion locations with PSD”.
Within the last decade, cerebral blood flow (CBF) studies have been used to examine PSD. They review whether reduced blood flow in regions of the brain contributes to PSD. Other biological aspects explored have been cortisol levels and serotoninergic mechanisms. In 1997, Alexopoulos wrote about a “vascular depression hypothesis” that could be applied to PSD: “Cerebrovascular disease may predispose, precipitate, or perpetuate some geriatric depressive syndromes. The hypothesis is supported by the comorbidity of depression, vascular disease, and vascular risk factors and the association of ischemic lesions to distinctive behavioral symptoms. Disruption of prefrontal systems or their modulating pathways by single lesions or by an accumulation of lesions exceeding a threshold are hypothesized to be central mechanisms in vascular depression.”. This hypothesis has its opponents and proponents.
To our knowledge, no systematic review of psychosocial risk factors exists in the literature. This is the aim of the present review.
Selection of Articles
First, using the key word “poststroke depression,” Medline was searched for all relevant articles published between January 1, 1966, and June 30, 2000. Using the key words “cerebrovascular disease” and “depression,” another Medline search was conducted for all relevant articles published between June 1, 1996, and June 30, 2000. Second, additional references were identified from the bibliographies of relevant articles.
The first author screened all the articles to ensure that they represented original research in either French or English;
1. they were either prospective, case-control, or cross-sectional studies;
2. they assessed PSD within the first 6 months of the stroke;
3. they defined depression acceptably, using DSM or ICD criteria, or a scale employing them;
4. they defined stroke acceptably;
5. they studied at least 1 psychosocial risk factor.
Finally, the second author tested the screening process results for interrater reliability.
As in Elie's 1998 study, each selected article was systematically reviewed according to a checklist to abstract the following information: study year, country, study type, patient type, PSD criteria, stroke criteria, inclusion and exclusion criteria, mean age and age range, number of patients enrolled and followed, and psychosocial risk factors studied.
Study Quality Assessment
We used the scale to Assess Quality of Risk Factor Studies developed by Elie and others to evaluate the study quality. The first author rated the selected articles, and the second author tested interrater reliability.
Risk Factor Analysis
As in Elie's study, all selected articles were screened to identify the risk factors studied. The statistical analysis varied among studies. Risk factors were tabulated to report their association (positive, negative, or not applicable) with the development of depression. This comparison provided a qualitative analysis.
The 2 Medline searches (20 articles) and the corollary articles found through references in bibliographies (6 articles) yielded a total of 26 articles. One article could not be located, for a final total of 25 articles. The first author selected 9 articles because they met the 6 inclusion criteria. The second author, blind to the foregoing choices, selected separately the same articles for an interrater agreement of 100%. The 9 studies included were prospective (that is, patients were diagnosed with stroke and prospectively followed to determine who among them developed depression). Patients were included from the Helsinki Stroke Aging Memory Study, from hospitals and outpatient departments (OPDs) in Denmark, from a rehabilitation centre, and from hospitals in Sweden, the US, Australia, and Finland. All these studies considered both major depressive disorder and minor depressive episode, except one, which did not include minor depressive episode in its analysis. Stroke criteria varied and included those of the National Stroke Data Bank, the WHO, the Institute of Neurological Disorder and Stroke, CT lesions, and “stroke patients” as defined in 1 study. Age ranged from 25 to 100 years. Five studies had a mean age of approximately 70 years, 2 of approximately 60 years, and 1 did not specify age range or mean. Total patients assessed could vary within a study due to loss in follow-up (for example, due to death, migration, or refusal to be reevaluated).
The first author ranked each article using the scale to Assess Quality of Risk Factor Studies. Blind to this ranking, the second author completed the scale separately. Over the 8 criteria across the 9 studies (yielding 72 ratings), interrater agreement was 79%. Questions not answered in almost all the studies referred to comparability of cases and controls and the sample size needed for adequate power. For these 2 criteria, interrater reliability was 83%. The total results varied between 7 and 12 over a possible maximum total of 16.
Qualitative Psychosocial Risk Factor Analysis
Eighteen psychosocial risk factors were identified across the articles. Three different risk factors were studied in 7, 6, and 5 articles respectively. Two risk factors were studied in 4 articles. Three risk factors were addressed in 3 articles. Five risk factors were part of 2 articles each, and 5 risk factors were also considered in one article each.
Demographic Risk Factors. Marital status was studied in one article and was found not to be a risk factor for PSD. One study found lower socioeconomic status (SES) to be a risk factor, whereas another study stated the contrary. The criteria to define SES were different across the studies. Age was not associated with PSD in 3 studies, but 3 others found an association: 1 observed that younger age was significant (for men only), 1 stated that older age was significant for major depression, and 1 stated that older age was significant for both major and minor depression. Sex was addressed in 7 studies. Of these, 2 studies shared the same patients, and the results were therefore considered only once. Three studies (found no association between PSD and sex, 3 found female sex to be a risk factor, and 1 (specifying that men in their study differed from women in terms of age, marital status, and attachment figure) found that male sex was a risk factor for minor depression.
Social Risk Factors. Prior social distress was positively associated with PSD in one study, when defined in terras of “social loss” in the 6 months prior to stroke. When measured according to a 12-item list of threatening negative experiences, no association was found. Two studies used the Social Activities Index to address poststroke social isolation at repeated intervals up to 1 year poststroke and found it to be significant when measured in terms of few social activities and decreased social activities. A questionnaire investigating social network found a significant association with PSD at 3 months, 1 year, 2 years, and 3 years. Perceived adequacy of social support assessed in 1 study, using the Interview Schedule for Social Interaction, was positively associated with PSD. Living arrangements were addressed in 3 different studies. Two found that living alone was significantly associated with PSD (after stroke, prior to stroke). The other study did not replicate the association with PSD and living arrangements. The study that reported living alone poststroke to be a risk factor stipulated that, without further definition, housing was not significantly associated.
Psychiatric History Risk Factors. Familial psychiatric history, personal anxiety history, and personality neuroticism assessed using the Eysenck Personality Inventory were all studied in 1 article each and were all positively associated with PSD. Dementia was assessed using DSM-III-R criteria in 1 study and with the Mattis Dementia Rating Scale in another. Both studies reported no association with PSD.
Three studies addressed past personal psychiatric history, and 2 found positive association. One specified that past psychiatric history was significant for major depression, as was past personal, familial, affective, and anxiety history. Affective psychiatric history was used to identify past depression (not bipolarity) and was positively associated with PSD in 4 different studies.
Impairment Risk Factors. Dependency assessed using the scale for ADL was significantly associated with PSD in 1 study. In another study, it was not associated with PSD having onset in hospital. This latter study reported that later in the disease it would affect depression severity. A study that used the Motricity Index at 1 month poststroke, as well as the Barthel Index (within 7 days and at 1, 6, and 12 months), to assess functional impairments did not find a correlation with PSD. Other studies using the Barthel Index to define functional impairments found significant association, and 1 specified an association for minor depression but not for major depression. Cognitive impairments assessed in several studies using the Mini-Mental Status Examination (MMSE) concluded that there was no association (1 study assessed these impairments at 3-year follow-up). Conversely, another study, which assessed cognitive impairments using the BriefCognitive Rating Scale, reported a significant association with PSD. Communicative impairments were assessed in 3 studies. Two stated that dysphasia was significantly associated with PSD. One did not define the assessment method, whereas the other used the Western Aphasia Battery. The last of these 3 studies reported no association with aphasia, measured with the Acute Aphasia Screening Protocol (AASP).
-Ouimet, M.A.; Primeau, F.; Cole, M.G.. Canadian Journal of Psychiatry, Nov2001, Vol. 46 Issue 9
Reflection Exercise #9
The preceding section contained information about psychosocial risk factors in poststroke depression. Write three case study examples
regarding how you might use the content of this section in your practice.
According to Ouimet, a patient with major or minor depression after stroke is how much more likely to die, compared with patients without depression? Record the letter of the correct answer the