Post-stroke depression: Prevalence and relationship with disability in chronic stroke survivors

Objectives: To evaluate (1) the prevalence of operationally defined depressive disorder (ICD-10) in chronic stroke subjects and (2) the relationship of post-stroke depression (PSD) with disability. Design: Cross-sectional, descriptive study. Setting: Neurological rehabilitation unit of a tertiary care university research center. Materials and Methods: Participants were those with first episode of supratentorial stroke of more than 3 months' duration with impaired balance and gait who had been referred for rehabilitation. Data were collected on demographic data, stroke data (side and type of lesion and post-stroke duration), cognition (mini mental state examination), depressive ideation (Hamilton Depression Rating Scale - HRDS), impairment (Scandinavian Stroke Scale), balance (Berg Balance Scale), ambulatory status (Functional Ambulation Category), walking ability (speed), and independence in activities of daily living (Barthel Index). Statistical analysis was done using SPSS 13.0. We carried out the chi-square test for ordinal variables and the independent t test for continuous variables. Results: Fifty-one patients (M:F: 41:10) of mean age 46.06 ± 11.19 years and mean post-stroke duration of 467.33 ± 436.39 days) were included in the study. Eighteen of the 51 participants (35.29%) met the criteria for depression. Demographic variables like male gender, being married, living in a nuclear family, urban background, and higher HRDS score were significantly correlated with PSD (P < 0.05). Depression was related to functional disability after stroke but to a statistically insignificant level (P > 0.05) and was unrelated to lesion-related parameters. Conclusion: Depression occurs in one-third of chronic stroke survivors and is prevalent in subjects referred for rehabilitation. PSD is related primarily to demographic variables and only to a lesser extent to functional disability following stroke.


Introduction
Stroke is a major public health problem.Traditionally, epidemiological stroke studies have focused on mortality and recurrence [1,2] and not on the long-term morbidity.The prevalence of disability among stroke survivors is between 24-54%. [3]The progressive decrease in stroke mortality observed in the last few decades, and the subsequent increase of survivors with residual impairments and disabilities, have been accompanied by a growing interest in the factors that could interfere with functional outcome and quality of life. [4]Depression is considered as the strongest predictor of poor quality of life among stroke survivors. [5]st-stroke depression (PSD) is one of the common emotional disorders affl icting stroke survivors.Previous studies have reported prevalence rates that have ranged from 18% to 61%, depending upon patient selection and criteria used. [6,7]iagnosis of PSD is challenging; therefore, it oft en remains unrecognized and/or undertreated.PSD is associated with stroke of more than 3 months' post-stroke duration, within the age range of 16-65 years, able to follow three-step commands, having impaired balance and gait but with ability to walk with or without support were included in the study.Patients with recurrent strokes, receptive aphasia, signifi cant cognitive defi cits aff ecting participation, and those with past history of depression were excluded from the study.The Institute Ethics Committee approved the protocol and written informed consent from all participants was taken prior to data collection.

Design
A cross-sectional, descriptive study design was used.The rehabilitation physician managing the patient did the screening of the patients for the study.The patients fulfi lling the inclusion criteria were evaluated on all outcome variables and then referred to the in-house psychiatrist for diagnosis of the operationally defi ned depressive disorder (ICD-10). [11]ociodemographic and clinical information was obtained in a pre-designed format by a single evaluator.

Outcome variables
Demographic data (age, gender, marital and residential status, and type of family), stroke data (side and type of lesion and post-stroke duration), cognition (Mini Mental State Examination), [12] depressive ideation (Hamilton Depression Rating Scale), impairments (Scandinavian Stroke Scale), balance (Berg Balance Scale), ambulatory status (functional ambulation category), walking ability (speed), and independence in activities of daily living (ADL -Barthel Index) were noted, and their relationship with PSD was evaluated.

Statistical analysis
Continuous variables were expressed as mean ± standard deviation and were evaluated by the Student's t test.Similarly, categorical variables were expressed as percentage of the total and were evaluated by the chi-square test.The two-tailed probability value was set at P < 0.05 for statistical signifi cance.

Hamilton Rating Scale for Depression
Depressive ideation was assessed on the 24-item Hamilton Rating Scale for Depression (HRDS).It consists of queries on mood, sleep, somatic symptoms, anxiety, and engagement in pleasure activities on graded score of 0-4.This has been validated for assessment of depression in stroke survivors. [13]

Scandinavian Stroke Scale
Impairment aft er stroke was assessed on the Scandinavian Stroke Scale (SSS).It consists of eight parameters, including consciousness, cognitive defi cits, motor power, and walking ability, on a graded score of 0-12.The maximum score is 58; higher scores indicate bett er outcome. [14]rgman Balance Scale It evaluates 14 sitt ing and standing activities, each on a 5-point scale.The maximum score is 56.Higher scores indicate bett er balance.It has been tested on patients with stroke and has shown good inter-rater and intra-rater reliability of 0.98 and 0.99, respectively. [15]nctional Ambulation Category It distinguishes six levels of support required during gait without taking into consideration any aid used.It is based on a walking distance of 15 meters.This has been validated for classifying qualitative walking handicap aft er stroke. [16]The test is performed with a cane but without orthosis.Levels are defi ned as follows: • Level 0: the subject cannot walk at all or requires the help of two or more people • Level I: the subject needs continuous support from one person who helps to carry the patient's weight and helps with balance • Level II: the subject is dependent on the continuous or intermitt ent support of one person to help with balance and coordination • Level III: the subject needs only verbal supervision • Level IV: the subject requires help for stairs and uneven surfaces • Level V: the subject can walk independently anywhere

Walking speed
Walking speed was measured in meters per second as the subject walked across a 10-meter walkway.Subjects were allowed to use the walking aids they required and necessary assistance to compensate for any loss of balance was provided.This method has been validated and is widely used for quantitative classifi cation of walking handicap following stroke. [17]rthel Index Independence in ADL was assessed with the Barthel Index.It can be subdivided to measure discrete functions of self-care and mobility, with an overall score of 100.This index has been widely used and for stroke subjects and has high reliability and validity and moderate responsiveness to changes in functional ability over time.[18]

Results
One hundred and eighty-fi ve stroke patients were screened during the study period and 73 of them met the inclusion criteria.Of those fulfi lling the inclusion criteria, 51 patients gave informed consent and were included in the study.The clinical and demographic profi le of these patients is given in Table 1.The mean age of the study population was 46.06 ± 11.19 years, mean post-stroke duration was 467.33 ± 436.39 days [Figure 1], and mean Barthel Index (BI) score was 74.33 ± 13.38 [Figure 2].
Eighteen patients (35.29%) were diagnosed to have operationally defi ned depressive disorder as per ICD-10 criteria.Surprisingly, none of these patients were on any specifi c treatment for depression.When patients who were depressed (n = 18) were compared to those who were not depressed (n = 33), it was noted that depressed patients were mostly male (P < 0.05), married (P < 0.03), living in nuclear families (P < 0.03), from urban backgrounds (P < 0.03), and had higher scores on HRDS (P < 0.003).
Depression was not related to parameters like age, type and side of lesion, and post-stroke duration.However, when diffi cult to defi ne the true prevalence rate of PSD.This variability between studies arises not only from the methodological problems of the investigations but also from the complexity in recognizing, assessing, and diagnosing depression. [8]While most of the studies have based their diagnosis on diff erent rating scales, others have used a structured interview format and the diagnostic standards defined by International Classifi cation of Diseases or the Diagnostic and Statistical Manual of Mental Disorders. [8]It is important to recognize that rating scales were designed to measure depression severity in primary depressive illness rather than to diagnose depression itself, especially depression occurring as a comorbidity.The correct att ribution of somatic symptoms to either PSD or stroke is a very relevant issue, because such symptoms may aff ect rating scales.This may be compounded by the inadequacies of physicians who do not have proper psychiatric training. [19]he present study was designed taking into account all these problems.The diagnosis of PSD in this study was done by a psychiatrist on operationally defi ned norms (ICD-10).The Hamilton Rating Scale for Depression was used only as an adjunct and it correlated well with the diagnosis of PSD.
prevalence was calculated on yearly basis, there was signifi cant reduction in prevalence in those with stroke of more than 2 years' duration [Table 2].
Depression was related to functional disability, i.e., cognition, balance, ambulatory status, walking ability, and independence in ADL but the relationship did not reach statistically significance (P < 0.05).However, when evaluated as per different grades of independence in ADL (BI score), PSD showed high prevalence in moderately (BI score < 80) to severely (BI score < 60) dependent patients as compared to those who had mild dependency (BI score > 80) [Table 3].

Discussion
At present, despite the abundant available literature, it is still     at 33%, even with relevant diff erences across studies. [20]In particular, the pooled estimate from the population-based studies was equal in the acute and medium-term phases (33%), with a slight increase to 34% in the long-term phase of recovery aft er stroke.Moreover, there were only slight diff erences in the pooled frequencies in the hospital-based (acute 36%, mediumterm 32%, and long-term 34%) and rehabilitation-based studies (acute 30%, medium-term 36%, and long-term 34%) over time [ Table 4].
The prevalence of PSD in this study was 35.29%, which approximates the pooled estimates of 34-36% for medium-term to long-term rehabilitation-based studies. [21]The diff erence in prevalence rate as compared to other studies [7,[21][22][23] might be due to the sample characteristics, since this study was confi ned to patients who were referred for rehabilitation, that too in the late phase following stroke.These were likely to be patients who were preselected on the basis of their rehabilitation potential, and in such a sample patients with mild and very severe stroke are likely to be excluded.The relatively younger mean age of 46.06 years in our study is probably refl ective of this bias.The exclusion of patients with aphasia, because of the expected diffi culty in evaluating depressive symptoms, was also an important confounding variable. [24]major concern is that none of the depressed patients in our study were on any antidepressant treatment.PSD is associated with increased disability and worse rehabilitation outcome [9,10] and therefore the need to routinely screen chronic stroke survivors for depression and institute appropriate treatment should be emphasized.Absence of PSD in young adults is a signifi cant predictor of the ability to return to work.Improvement of depressive symptoms has been associated with bett er functional recovery. [25]23] Similarly, in our study, the most important determinants of depression were demographic variables like male gender, marital status, and living situation.This refl ects the sociodemographic profi le of a developing country, where there is lower per capita expenditure on health and few, if any, benefi ts provided to the disabled population.The majority of the earlier studies [21][22][23] had reported female gender as being an important risk factor for PSD.However, in contrast, male gender was an important determinant in our study.This might be due to the low proportion of females in our study population.
PSD in this study was not related to stroke lesion-related parameters like age, type and side of lesion, and post-stroke duration.[23] However, one study each had reported site of the lesion, [26] younger age, [21] and older age [27] as being among the important risk factors for PSD.Although, post-stroke duration was not signifi cantly correlated to PSD in our study, when yearly prevalence was calculated, there was signifi cant reduction in the prevalence of PSD aft er 2 years.This is because longer post-stroke duration does have a positive eff ect on emotional responses to disability. [7]rrelation between depression and impairment in ADL peaked at 6 months and thereafter fell, but it remained significant at 1 and 2 years post stroke. [28]The persisting signifi cant association of impairment in ADL with depression may refl ect the eff ect of severe depression in sustaining-and possibly even retarding recovery from-physical impairment.
Depression was related to functional disability in our study, as patients with PSD have lower mean scores for all functional parameters (cognition, balance, ambulatory status, walking ability, and independence in ADL) as compared to those without depression; however, this did not reach statistically signifi cant levels (P < 0.05).This is one among few studies where multiple dimensions of functional disability, such as post-stroke impairment, cognitive status, balance, walking ability, and independence in ADL are studied together. [29,30]One earlier study reported that PSD was unrelated to functional disability. [22]However, the majority of the previous studies have reported a signifi cant relationship between disability and PSD. [23,27,28]This diff erence between our study and most of the earlier studies is due to two major reasons: 1) Our study population was less disabled due to a preselection bias, as patients with greater rehabilitation potential were included, and 2) the longer post-stroke duration allowed more time for the patients to come to terms with their disability. [7,28]is study has several limitations because of the nature of the study population: for example, the small sample size, unequal gender distribution, exclusion of patients with aphasia, and longer post-stroke duration.The major limitation was the preselected nature of the study population, with only those with bett er rehabilitation potential being included; patients with mild or severe strokes are not likely to have been included, and the study population is therefore not representative of the general stroke population.The sample in this study was screened by the rehabilitation physicians providing care, but the diagnosis of PSD was made by a psychiatrist and that too on operationally defi ned guidelines; this emphasizes the need for multidisciplinary teams in rehabilitation.This is among very few studies that have evaluated the relationship of demographic factors, lesion-related parameters, and multiple functional disability parameters simultaneously in chronic stroke survivors.Such an approach aids bett er understanding of the origin and eff ects of depression following stroke.Knowledge of PSD and its relationship with disability may be of use during the planning, provision, and allocation of health services for the care of stroke survivors, especially in this era of health care cost management.

Conclusion
The results of this study indicate that depression is underrecognized following stroke.The prevalence of PSD is high, with about one-third of stroke survivors suff ering from it.Demographic variables are important determinants of PSD.Depression is related to functional disability following stroke but the relationship does not reach statistical signifi cance in our study.The association of disability with PSD may refl ect its eff ect on sustaining-and possibly retarding recovery fromphysical impairment.We suggest that chronic stroke survivors be routinely screened for depression, given its high prevalence and its negative impact on further recovery and quality of life.
Srivastava, et al.: Post stroke depression

Figure 1 :Figure 2 :
Figure 1: Post-stroke duration (years) of the stroke population Figure 2: Independence in activities of daily living (Barthel Index scores) of the study population

Table 2 : Relationship of post-stroke duration with depression Post-stroke duration Number of patients
Figures in parentheses are in percentage

Table 3 : Relationship of functional disability (Barthel index scores) with PSD
Figures in parentheses are in percentage