This study has several limitations. First, the searches were limited to Medline, and other key words could have selected a different set of articles. Nevertheless, to avoid missing important articles in the field, we systematically reviewed the bibliographies of all potential articles obtained through the searches. Second, other inclusion criteria might have excluded fewer studies and perhaps addressed more risk factors. Third, the analysis included different types of patient populations (for example, hospital, outpatient, and rehabilitation centre), which did not allow population-specific psychosocial risk factors to emerge (for example, factors specific to people in nursing homes). Fourth, risk factors like prior social distress, poststroke social isolation, and dementia were assessed in 2 different studies each but using different scales or assessment methods, which can further decrease the strength of comparison for a factor across studies. Fifth, the analysis was qualitative and not quantitative. Sixth, the total number of patients (considering only those patients who completed the studies) was small, at not more than 1000 subjects. Seventh, the interraters' study quality assessments were all in the middle range (7 to 12 points out of a maximum of 16 for the scale), indicating that articles were of similar quality. Was this because the inclusion criteria selected studies of the same quality, or because the scale was not able to discriminate between articles, or because the state of the research is such that they all share the same quality? Eighth, interrater reliability tests would have been beneficial at all steps of the research. Ninth, our review included studies with assessments within the first 6 months, and 3 had repeated measures up to 1 and 3 years. Thus, we could not quantitatively confirm or invalidate the relative importance of psychosocial risk factors over time.
Aware of all these limitations, we can still mention that certain psychosocial risk factors are more consistently found to be associated with PSD. Of the studies that considered a specific risk factor, those that found positive associations are as follows: past history of depression (4/4 articles), past personal psychiatric history (2/3 articles), functional impairment (3/4 articles), poststroke social isolation (2/2 articles), living alone (2/3 articles), and dysphasia (2/2 articles). For risk factors more consistently found not to be associated with PSD, the findings were as follows: dementia (2/2 articles) and cognitive impairments (4/5 articles).
Age led to inconsistent results: 3 studies found a positive correlation, whereas 3 studies did not. Lower SES was a positive risk factor in 1 study while 1 study, without specifying how SES was assessed, did not find it to be a risk factor. Prior social distress gave rise to the same type of inconsistent results (once positive and once not) in studies using different instruments to assess this risk factor. Dependency evaluated with ADL scales was once positive and once negative.
Sex was studied within 7 different groups of patients: female sex was positively correlated in 3 studies, male sex in 1 study. In this study, men differed from women on certain demographic factors. Three studies found no association with sex. These results are different from the sex prevalence of depression in the general population, where the female: male ratio is 2:1. Could this indicate that brain injury affects men and women differently due to sex-based differences in brain organization and that greater biological predisposition in women is opposed to nonbiological factors associated with depression in poststroke men (for example, issues of physical impairment and social support), which could decrease the prevalence gap? Sex prevalence should be clarified in further studies.
A psychiatric history of familial disorder, of anxiety disorder, and of personal neuroticism were all associated with a positive relation to PSD in the only study in which they were addressed. Perceived adequate social support was also positively correlated, but aphasia was not. Nevertheless, it should be noted that these 5 results were obtained from only a single article each and therefore need replication before any conclusion can be drawn.
A brief review of the literature identified the following psychosocial risk factors for depression: age, sex, personality development, neuroticism, degree of disability, lack of social support, disruption of social roles, dependence on others for ADLs, negative life events, personal and familial history of affective and anxiety disorders, nonfluent aphasia, and cognitive impairment. Of these factors, age and sex were not generally documented as important (not mentioned were education level and marital status).
This reinforces Aström's observation that "the contributions of physical impairments and psychosocial risk factors to depression after stroke are less well known and probably change over time." It also points to the difficulty in determining whether certain risk factors are a result or a cause of depression—or only interact with depression to increase its severity or the severity of the outcome. In agreement with earlier work of Primeau, Spencer and others report in their extensive paper that PSD research displays methodological limitations which may contribute to conflicting outcomes and conclusions.
In addition, it has been suggested that the heterogeneous distribution of risk factors across studies reflects the varying nature of poststroke depression—that it is more biologically determined in the first months and more reactive (implying more impact on the part of psychosocial risk factors) later in the illness course. Investigating that hypothesis, however, Gainotti concluded that a psychological model of PSD was consistent over time and that the evidence did not support a biological basis in the acute stage.
Despite the fact that PSD is prevalent and has been demonstrated to increase morbidity and mortality after stroke, the last 3 decades have given rise to a small number (approximately 25) of studies investigating strictly psychosocial risk factors and not psychosocial risk factor predictors of severity. As Ramasubbu reports: "poststroke depression has been criticized on the premise that stroke lesions are judged to be causally linked with depression to the exclusion of psychosocial risk factors." To our knowledge, our review of psychosocial risk factors for PSD is the first. No study addressed the issue of ethnicity. Our selected articles included subjects of European descent from Australia, Denmark, Finland, and the US, as well as Chinese and African-American subjects.
The results point to a need for teamwork in the case of patients with stroke because past psychiatric history, social aspects, and impairments are often identified as risk factors to the development of PSD. Efforts should aim to prevent social deterioration and impairments and to identify patients with past psychiatric history who deserve an early psychiatric consultation. Knowledge of risk factors is important because some are easy to identify (for example, psychiatric history and depression history) and some are amenable to interventions (for example, functional impairment, social isolation, and living arrangements).
- The methodology of studies of the psychosocial risk factors for poststroke depression (PSD) should be improved to allow quantitative analysis.
- Screening for PSD psychosocial risk factors should be considered.
- Teamwork to identify past psychiatric history and to prevent social deterioration and impairment would be beneficial.
-Ouimet, M.A.; Primeau, F.; Cole, M.G. Canadian Journal of Psychiatry, Nov2001, Vol. 46 Issue 9
Reflection Exercise #10
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.
What may the lack of association between PSD and the sex of the patient indicate, considering the sex prevalence of depression in the general population has a female: male ratio of 2:1? Record the letter of the correct answer the