Discriminative factors for post-stroke depression

Discriminative factors for post-stroke depression

Journal Pre-proof Discriminative Factors for Post-stroke Depression Amin Ghaffari, Malahat Akbarfahimi, Hamid Reza Rostami PII: S1876-2018(19)30944-...

3MB Sizes 0 Downloads 73 Views

Journal Pre-proof Discriminative Factors for Post-stroke Depression Amin Ghaffari, Malahat Akbarfahimi, Hamid Reza Rostami

PII:

S1876-2018(19)30944-X

DOI:

https://doi.org/10.1016/j.ajp.2019.101863

Reference:

AJP 101863

To appear in:

Asian Journal of Psychiatry

Received Date:

20 September 2019

Revised Date:

26 October 2019

Accepted Date:

31 October 2019

Please cite this article as: Ghaffari A, Akbarfahimi M, Rostami HR, Discriminative Factors for Post-stroke Depression, Asian Journal of Psychiatry (2019), doi: https://doi.org/10.1016/j.ajp.2019.101863

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2019 Published by Elsevier.

Discriminative Factors for Post-stroke Depression Amin Ghaffari (PhD Candidate) a, Malahat Akbarfahimi* (PhD) b, Hamid Reza Rostami (PhD)c

a

Department of Occupational therapy, School of Rehabilitation Sciences, Iran University of

Medical Sciences, Tehran, Iran. [email protected] Associated Professor, Department of Occupational therapy, School of Rehabilitation Sciences,

of

b

Iran University of Medical Sciences, Tehran, Iran. [email protected]

Assistant Professor, Department of Occupational therapy, Musculoskeletal Research Center,

ro

c

School of Rehabilitation Sciences, Isfahan University of Medical Sciences, Isfahan, Iran.

re

-p

[email protected]

* Corresponding author: Malahat Akbarfahimi, Mirdamad blv, Madar sq, Shahnazari st, School

lP

of Rehabilitation Sciences, Department of Occupational therapy, Iran University of Medical Sciences, Tehran, Iran. Tel: 09123362925, Email:[email protected]

ur na

Conflict of Interest: There was not any conflict of interest. Running Title: Post-stroke depression Discriminators Word count for the abstract: 224

Jo

Complete manuscript word count: 2281 Number of references: 32 Number of figures/tables: 1/3

Highlights

1

Trunk control and basic activities of daily living are discriminators of post stroke depression The high educated people with Stroke are more at risk of depression

-p

ro

of

Post-stroke depression is more common in patients with cardiac diseases

Abstract

re

Background: Depression is the most common mood disorder following stroke. It can negatively affects different domains of patient’s life. The present study

lP

aimed to explore demographic and clinical predictors of post stroke depression and determine discriminative cognitive, motor, and functional factors in stroke

ur na

patients with and without depression. Methods: In a cross sectional study, 100 patients with stroke were investigated. Measurements consisted of Beck Depression Inventory-II, Trail Making Test A

Jo

& B, Digit Span Subtest of Wechsler Memory Scale, Motricity Index (arm and leg motor), Trunk Control Test, Barthel Index, and Lawton Instrumental Activities of Daily Living. Demographics and clinical data including educational level, marital status, limb affected, cigarette smoking habits, diabetes mellitus, cardiac diseases, and blood pressure were also collected.

2

Results: Multivariate logistic regression revealed that college level of education (OR=8.78, 95% CI: 2.65-29.11, P<0.001) and cardiac diseases (OR=3.11, 95% CI: 1.19 -8.13, P<0.001) were significant demographic and clinical predictors of post stroke depression. Using stepwise discriminant function analysis, basic activities of daily living and trunk control with 88.0% classification accuracy,

of

81.1% sensitivity, and 95.7% specificity were as the best discriminators of post stroke depression.

ro

Conclusions: Rehabilitation experts working with patients with stroke should pay special attention to trunk control and basic activities of daily living for

-p

preventing consequences of PSD particularly in those with higher educational

re

level and cardiac diseases.

PSD1

ur na

ADL2

lP

Abbreviations:

BDI-II3 TMT4

WMS-R5

Jo

TCT6

1

Post-stroke depression Activities of Daily Living 3 Beck Depression Inventory-II 4 Trail Making Test 5 Wechsler memory scale 6 Trunk Control Test 2

3

OR7 CI8 DFA9

Keywords: Post-Stroke Depression; Cognition; Activities of Daily Living;

lP

1. Introduction:

re

-p

ro

of

Trunk control

Depression is the most common mood disorder following stroke that can

ur na

increase mortality rate up to ten times (Esparrago Llorca et al., 2015). Nearly 33 percent of stroke survivors suffer from post-stroke depression (PSD) (Das & G, 2018), while this range is up to 47 percent in Iran (Dalvand et al., 2018). Investigation of predictive and risk factors of PSD seems crucial considering

Jo

its negative effects on rehabilitation outcomes such as cognitive and physical recovery (Robinson & Jorge, 2016), social participation (Micaela Silva et al., 2016), quality of life and activities of daily living (ADL) (Haghgoo et al.,

7

Odd Ratios Confidence Interval 9 Discriminant Function Analysis 8

4

2013); and for designing a preventive or early intervention (van de Port et al., 2007). The risk factors/predictors of PSD are investigated in various systematic review and meta-analysis studies such as: history of mental illness (Shi et al., 2017), female sex, age <70 years, neuroticism (Shi et al., 2017), family

of

history (Mitchell et al., 2017; Shi et al., 2017), stroke severity (Ayerbe et al., 2013; Hackett & Anderson, 2005; Kutlubaev & Hackett, 2014; Shi et al.,

ro

2017), physical disability (Ayerbe et al., 2013; Hackett & Anderson, 2005;

Kutlubaev & Hackett, 2014; Shi et al., 2017), aphasia (Mitchell et al., 2017),

-p

dominant hemisphere lesions (Mitchell et al., 2017), pre-stroke depression

re

(Ayerbe et al., 2013; Kutlubaev & Hackett, 2014; Mitchell et al., 2017), cognitive impairment (Ayerbe et al., 2013; Hackett & Anderson, 2005), lack

lP

of social or family support and anxiety (Ayerbe et al., 2013). Meanwhile, there are studies reporting no relationship between depression and

ur na

demographic or clinical characteristics in patients with stroke (Kutlubaev & Hackett, 2014).

Despite a large number of studies on the risk factors of PSD, the lack of consensus about the PSD risk factors is obvious (Dalvand et al., 2018;

Jo

Haghgoo et al., 2013). This non-consensus and controversy may be due to dynamic and multifactorial nature of PSD, which makes predictive factors variable depending on different stages of stroke, selected factors, etc. (De Ryck et al., 2014). Another main issue in determining PSD risk factors is cultural aspect of depression, which makes cultural-based investigation of

5

PSD and implementation of rehabilitation programs based on cultural background an important issue (Hinojosa et al., 2011; Kaadan & Larson, 2017; Sarfo et al., 2017). Therefore, present study was aimed to: 1) determine discriminative factors in patients with and without PSD among cognitive, motor, and functional factors; 2) explore the demographic and clinical

of

predictors between patients with and without PSD. 2. Methods

ro

2.1. Participants

In this cross sectional study, 100 patients with stroke (with PSD: n= 53,

-p

without PSD: n=47) were participated from five occupational therapy centers.

re

Centers were selected based on cluster randomization method. A convenient sample of 18 to 30 people with stroke were enrolled from each clinic.

lP

Participants in two groups were matched by age, gender, and time since stroke. Inclusion criteria were: first-ever stroke diagnosed by two

ur na

neurologists, right hand dominance (because of laterality effects on emotional centers in the brain), age 30-80 years, successful formal education ≥ 9 years, and normal or corrected-to-normal visual acuity and hearing. Depression diagnosis was confirmed by a psychiatrist. Patients with a history of transient

Jo

ischemic attack, other acute or chronic neurologic and psychiatric disorders, unilateral neglect, aphasia, pre-stroke depression history or taking antidepressant medications prior to stroke were excluded. The study protocol was approved by local ethics committee (IR.IUMS.REC1395.9411355005).

6

After signing informed consent, demographic and clinical data including age, gender, educational level, marital status, affected limb, cigarette smoking habits, diabetes mellitus, cardiac diseases and blood pressure were collected. Data collection was conducted in a quiet room between 9: 00 Am to 12:00. Following gathering demographic data, psychological, cognitive, motor, and

of

activities of daily living domains were measured using different assessment tools, which were administered in a random order by an occupational therapist

ro

blinded to the study’s aims. 2.2. Measures

-p

The severity of depression was assessed using the Beck Depression Inventory-

re

II (BDI-II). BDI-II consists of 21 items concerning people’s feeling in different situations during the last week. Each item scores from 0-3. Cut-off

al., 2005).

lP

point for depression in Iranian population is calculated 15 (Ghassemzadeh et

ur na

Cognitive domain included Trail Making Test (TMT) and Wechsler memory scale (WMS-R). TMT was used to measure mental flexibility, visual attention, and speed with a motor component. TMT consists of two parts including TMT-A and TMT-B. In TMT-A, participant should match the numbered

Jo

circles orderly as fast as possible. In TMT-B participant is asked to alternatively match the circled numbers and letters. Time required to complete each task is recorded in seconds (Spreen & Strauss, 1988). Raw scores of TMT (B-A) were transformed to Z-score for data analysis. In addition, digit span subtest of WMS-R was used for measuring verbal memory in the form of

7

forward and backward numbers. Scores in this test range from 3 to 8 for forward and 3 to 7 for backward numbers and total score is sum of both parts (Wechsler, 1945) Motricity index was used to evaluate arm motor ability (pinch grip, elbow flexion, shoulder abduction) and leg motor ability (ankle dorsiflexion, knee

of

extension, hip flexion) (Demeurisse et al., 1980). The Trunk Control Test (TCT) examines the ability of participants in four

ro

simple aspects of trunk movement including roll to the weak side from supine position, roll to the strong side, sit up from lying down, and sit in a balanced

-p

position on the edge of the bed with the feet off the ground for a minimum of

re

30 seconds. Its scores are zero for unable to perform movement without assistance, 12 for able to perform movement but in an abnormal style, 25 for

lP

able to complete movement normally (Wade & Hewer, 1987). Barthel Index was used to measure performance of participants in basic

ur na

activities of daily living such as bowel and bladder function, grooming, toilet use, feeding, transfer, mobility, dressing, steps, and bathing. Its scores range from 0 - 100 which means complete dependency to completely independency (Oveisgharan et al., 2006).

Jo

Lawton instrumental activities of daily living scale measures performance in activities such as using telephone, shopping, food preparation, housekeeping, laundry, transportation, responsibility for own medication, and ability to handle finances(Chong, 1995). Its scores range from 0 (low function, dependent) to 8 (high function, independent).

8

2.3. Statistical analysis Data were analyzed using SPSS software version 17.0 by a biostatistician blinded to the study. The chi square test and Fisher exact test were used to compare categorical variables between groups and independent T-test for continuous variables. Univariate logistic regression analysis was used to

of

examine clinical and demographic risk factors for PSD. Variables with P < 0.10 were introduced into multivariate regression analyses with forward

ro

elimination likelihood ratio (LR) method (Kleinbaum et al., 1997). The odd ratios (OR) and 95% confidence interval(CI) were estimated. In order to

-p

discriminate patients with and without PSD, step-wise Discriminant Function

re

Analysis (DFA) was performed on continuous data. After calculating sensitivity and specificity of DFA model, results higher than 0.80 (80%)

lP

revealed a good model, 0.50 - 0 .80 (50% to 80%) showed an acceptable model, and lower than 0.50 were considered as poor model (Marôco, 2011).

ur na

Significant level was set at P values <0.05. 3. Results

Selection and allocation of participants at each stage of data collection is shown in a Figure 1. As shown in Table 1, there was not any significant

Jo

difference between groups regarding to age, time since stroke, Z score (TMT B-A), sex, marital status, affected limb, cigarette smoking habits, diabetes mellitus, blood pressure, and cardiac diseases. The logistic regression analysis was used to determine PSD predictors among demographic and clinical variables. As shown in Table 2, participants with

9

college level education were 6.41 times (CI 95%: 2.10-19.60) more likely to have PSD. In the next step, the variables which predicted the PSD in univariate analysis (P<0.10), were entered into multivariate logistic regression analysis by indicating forward elimination likelihood ration (LR) method. These variables were education level (p<0.001), smoking habits (p<0.06), and

of

cardiac disease (p<0.09). The college educational level and suffering from cardiac diseases were left in the model and identified as significant

ro

demographic and clinical risk factors for post stroke depression (Table 2).

Discriminant Function Analysis was used to conduct a multivariate analysis of

-p

variance test of our hypothesis that patients with and without PSD differ

re

significantly in a linear combination of 7 variables including Barthel index, Lawton instrumental activities of daily living, arm and leg motor function of

lP

Motricity index, Trunk control test, digit span subtest of Wechsler memory scale and TMT (B-A). DFA results using enter method revealed a significant

ur na

association between all variables in with and without PSD groups (Eigenvalue= 1.624, Wilks λ = 0.381, Chi-square = 91.151, df = 7, Canonical correlation = 0.787, p <. 001). Stepwise DFA was developed for all variables to determine the most parsimonious way to distinguish between stroke

Jo

patients with and without PSD. The overall Chi-square test was significant (Eigenvalue= 1.421, Wilks λ = 0.413, Chi-square = 85.784, df = 2, Canonical correlation = 0.766, p < 001). According only one function, during 2 steps two variables including Barthel index score (step 1), and Barthel index and trunk control test scores (step 2) were found as discriminators of patients with from

10

those without PSD. As Table 3 presents, the discriminant function revealed a significant association between groups and two predictors, accounting for 58.67 % between groups variability. Our findings provide linear combinations of these two predictors as a group membership in the discriminator model. The predictive model is: "DF = 0.082 ×Barthel index score - 0.041 ×Trunk

of

control test score -2.032". The predictive accuracy of the model for the analyzed sample was 88.0 percent. Although stroke patients without PSD

ro

were classified more precise (95.7 %) than those with PSD (81.1 %), both values represent good model accuracy. The cross validation classification

-p

showed that overall 88.0 percent was correctly classified.

re

4. Discussion

Results of the present study revealed basic activities of daily living and trunk

lP

motor control as two discriminative factors for distinguishing patients with PSD from those without PSD with 88.0% classification accuracy, 81.1%

ur na

sensitivity, and 95.7% specificity.

Following stroke, the majority of survivors complain about some degrees of restriction in activities of daily living and a variety of predictors have been reported for activities of daily (Robinson & Jorge, 2016). In our study, basic

Jo

activities of daily living was the strongest discriminator of PSD. It seems presence of depression makes patients more dependent in basic ADL because of negative effects of depression on volition and motivation of patients for participation in daily life and also, the vicious circle of disability, dependency to others, and depression.

11

Trunk stability is an important prerequisite for limbs’ movements, balance and postural control (Verheyden et al., 2007). Our results showed that trunk motor control in patients with PSD is weaker than those without PSD. Perhaps this outcome is due to psychomotor complications such as trunk and postural control deficits caused by depression which are reported in patients with

of

major depression disorders (Canales et al., 2010). The involved limbs motor function was not recognized as a discriminative

ro

factor for PSD in the present study, despite previous studies which have been found strong correlation between PSD and physical disability (Ayerbe et al.,

-p

2013; Eriksen et al., 2016; Hackett & Anderson, 2005; Kutlubaev & Hackett,

re

2014; Shi et al., 2017). This finding may support the key role of trunk control in recovery of motor function of involved limbs, which is influenced by the

lP

mechanism of depression.

Despite cognitive impairments in our participants as measured by TMT and

ur na

digit span test, but no one of them was found as a discriminator for PSD in our study. The affected cognitive impairments related to PSD are reported as executive functions, attention, visuospatial ability, verbal fluency, Language, and orientation (Robinson & Jorge, 2016). Memory decline is not reported as

Jo

a discriminator for PSD in some systematic reviews in line with our study (Cumming et al., 2013), but some points in the present study such as complexity of assessment tools, limited considered cognitive areas, and sample characteristics (excluding stroke patients with aphasia and dementia)

12

makes precisely evaluating different areas of cognition necessary in future studies. Multivariate logistic regression showed that among demographic and clinical characteristics, college educational level (OR=8.78) and cardiac diseases (OR= 3.11) were significant predictors of PSD. Increasing the risk of PSD

of

with higher educational level is in contrast to the some meta-analysis and systematic reviews (Backhouse et al., 2018; Shi et al., 2017; Ayasrah et al.,

ro

2018). In one study in China, the education was not related to PSD (Zhang et al., 2013). The point worth noting in these studies is the classification of

-p

education levels. In these studies the education was considered low (between

re

illiterate to 8 classes) and high (more than 9 classes); while in our study, the minimum educational level of 9 classes was an inclusion criteria for

lP

participants. Therefore, in comparison to other studies, our participants were high educated. Although expectedly higher education is correlated with faster

ur na

and more successful coping with new situations, may be factors such as life style or complexity of roles and duties influenced by culture could make depression more prevalent in higher educated patients with stroke. This issue needs more research to clarify facilitators and barriers for better adjusting to

Jo

the disease in patients with stroke. Cardiac diseases were correlated with increasing the risk of PSD in our study; in line with Robinson and Jorge’s study (Robinson & Jorge, 2016). However, when interpreting this issue, special medication regime of stroke patients with

13

cardiac diseases and functional restrictions due to cardiac disease should be considered. We think main limitation in our study was lack of control over body function and structures before stroke. In a study by Duan et al., it is reported that patients with PSD have higher infarct volumes and increased stroke

of

severity (Duan et al., 2019). Therefore, considering stroke severity and location of brain lesions as well conditions of body functions and structures of

ro

patients before stroke are issues, which need to be cleared more in future studies. In addition, available evidence show the relationship between

-p

biological factors with PSD and cardiac diseases (Duan et al., 2019;

re

Donnellana et al., 2019). Since cardiac diseases were found as a predictive factor for PSD in the present study, careful clarification of association among

lP

biological factors and cardiac diseases and PSD would be a main research

ur na

question for future studies.

Funding

Jo

This work was supported by the Iran University of Medical Sciences [grant numbers IR.IUMS.REC1395.9411355005].

Conflict of interest All authors assert that they have no conflicts of interest.

Acknowledgment 14

Our special thanks to the patients with stroke who consented to participate in this study. This research did not receive any specific grant from public agencies.

References Ayasrah, S. M., et al. (2018). Post-Stroke Depression in Jordan: Prevalence

doi: 10.1016/j.jstrokecerebrovasdis.2017.11.027

of

Correlates and Predictors. J Stroke Cerebrovasc Dis, 27(5), 1134-1142.

ro

Ayerbe, L., et al. (2013). The natural history of depression up to 15 years

-p

after stroke: the South London Stroke Register. Stroke, 44(4), 11051110. doi: 10.1161/STROKEAHA.111.679340

re

Backhouse, E. V., et al. (2018). Cognitive ability, education and

lP

socioeconomic status in childhood and risk of post-stroke depression in later life: A systematic review and meta-analysis. PLoS One, 13(7), e0200525. doi: 10.1371/journal.pone.0200525

ur na

Canales, J. Z., et al. (2010). Posture and body image in individuals with major depressive disorder: a controlled study. Rev Bras Psiquiatr, 32(4), 375380.

Jo

Chong, D. K.-H. J. S. (1995). Measurement of instrumental activities of daily living in stroke. 26(6), 1119-1122.

Cumming, T. B., et al. (2013). Stroke, cognitive deficits, and rehabilitation: still an incomplete picture. Int J Stroke, 8(1), 38-45. doi: 10.1111/j.1747-4949.2012.00972.x

15

Dalvand, S., et al. (2018). Prevalence of poststroke depression in Iranian patients: a systematic review and meta-analysis. Neuropsychiatr Dis Treat, 14, 3073-3080. doi: 10.2147/NDT.S184905 Das, J., & G, K. R. (2018). Post stroke depression: The sequelae of cerebral stroke. Neurosci Biobehav Rev, 90, 104-114. doi:

of

10.1016/j.neubiorev.2018.04.005 De Ryck, A., et al. (2014). Poststroke depression and its multifactorial nature:

159-166. doi: 10.1016/j.jns.2014.09.038

ro

results from a prospective longitudinal study. J Neurol Sci, 347(1-2),

re

Neurol, 19(6), 382-389.

-p

Demeurisse, G., et al. (1980). Motor evaluation in vascular hemiplegia. Eur

Donnellana, C., et al., (2019). Associations between ApoE gene and

lP

psychological consequences post stroke in a Bahraini cohort. Asian Journal of Psychiatry, 39, 135–142. doi: 10.1016/j.ajp.2018.12.019

ur na

Duan, Z., et al. (2019). Association between serum retinoic acid levels and risk of post-stroke depression in patients with ischemic stroke. Asian Journal of Psychiatry, 46, 87-91. doi: 10.1016/j.ajp.2019.09.038

Jo

Eriksen, S., et al. (2016). Acute phase factors associated with the course of depression during the first 18 months after first-ever stroke. Disabil Rehabil, 38(1), 30-35. doi: 10.3109/09638288.2015.1009181

Esparrago Llorca, G., et al. (2015). Post-stroke depression: an update. Neurologia, 30(1), 23-31. doi: 10.1016/j.nrl.2012.06.008

16

Ghassemzadeh, H., et al. (2005). Psychometric properties of a Persian‐ language version of the Beck Depression Inventory‐Second edition: BDI‐II‐PERSIAN. Depression and anxiety, 21(4), 185-192. Hackett, M. L., & Anderson, C. S. (2005). Predictors of depression after stroke: a systematic review of observational studies. Stroke, 36(10),

of

2296-2301. doi: 10.1161/01.STR.0000183622.75135.a4 Haghgoo, H. A., et al. (2013). Depression, activities of daily living and

ro

quality of life in patients with stroke. J Neurol Sci, 328(1-2), 87-91. doi: 10.1016/j.jns.2013.02.027

-p

Hinojosa, R., et al. (2011). Social isolation poststroke: relationship between

re

race/ethnicity, depression, and functional independence. Top Stroke Rehabil, 18(1), 79-86. doi: 10.1310/tsr1801-79

lP

Kaadan, M. I., & Larson, M. J. (2017). Management of post-stroke depression in the Middle East and North Africa: Too little is known. J Neurol Sci,

ur na

378, 220-224. doi: 10.1016/j.jns.2017.05.026 Kleinbaum, D. G., et al. (1997). Applied Regression Analysis and Other Multivariable Methods 3rd Edition: Duxbury Press.

Jo

Kutlubaev, M. A., & Hackett, M. L. (2014). Part II: predictors of depression after stroke and impact of depression on stroke outcome: an updated systematic review of observational studies. Int J Stroke, 9(8), 10261036. doi: 10.1111/ijs.12356

Marôco, J. (2011). Análise Estatística com o SPSS Statistics (5ª ed.): Lisboa:Report Number.

17

Micaela Silva, S., et al. (2016). Impact of depression following a stroke on the participation component of the International Classification of Functioning, Disability and Health. Disabil Rehabil, 38(18), 18301835. doi: 10.3109/09638288.2015.1107774 Mitchell, A. J., et al. (2017). Prevalence and predictors of post-stroke mood

of

disorders: A meta-analysis and meta-regression of depression, anxiety and adjustment disorder. Gen Hosp Psychiatry, 47, 48-60. doi:

ro

10.1016/j.genhosppsych.2017.04.001

Oveisgharan, S., et al. (2006). Barthel index in a Middle-East country:

-p

translation, validity and reliability. Cerebrovascular Diseases, 22(5-6),

re

350-354.

Robinson, R. G., & Jorge, R. E. (2016). Post-Stroke Depression: A Review.

lP

Am J Psychiatry, 173(3), 221-231. doi: 10.1176/appi.ajp.2015.15030363

ur na

Sarfo, F. S., et al. (2017). Post-stroke depression in Ghana: Characteristics and correlates. J Neurol Sci, 379, 261-265. doi: 10.1016/j.jns.2017.06.032

Jo

Shi, Y., et al. (2017). Risk Factors for Post-stroke Depression: A Metaanalysis. Front Aging Neurosci, 9, 218. doi: 10.3389/fnagi.2017.00218

Spreen, O., & Strauss, E. A. (1988). Compendium of Neuropsychological Tests. New York: Oxford University Press.

18

van de Port, I. G., et al. (2007). Determinants of depression in chronic stroke: a prospective cohort study. Disabil Rehabil, 29(5), 353-358. doi: 10.1080/09638280600787047 Verheyden, G., et al. (2007). Trunk performance after stroke: an eye catching predictor of functional outcome. J Neurol Neurosurg Psychiatry, 78(7),

of

694-698. doi: 10.1136/jnnp.2006.101642 Wade, D. T., & Hewer, R. L. (1987). Motor loss and swallowing difficulty

ro

after stroke: frequency, recovery, and prognosis. Acta Neurol Scand, 76(1), 50-54. doi: 10.1111/j.1600-0404.1987.tb03543.x

-p

Wechsler, D. (1945). A standardized memory scale for clinical use. The

re

Journal of Psychology, 19(1), 87-95.

Zhang, W. N., et al. (2013). A prospective study of the incidence and

lP

correlated factors of post-stroke depression in China. PLoS One, 8(11),

Jo

ur na

e78981. doi: 10.1371/journal.pone.0078981

19

of ro -p re lP ur na

Jo

Searching documentaries of 5 cluster randomized clinics (n=154) Excluded (n= 50) Lack of inclusion criteria (n= 15) Having exclusion criteria (n= 19) Declined to participate (n=16)

Allocation to groups (n=104)

With PSD (n= 54)

Without PSD (n= 50)

20

of ro -p

Lack of cooperation during the study implementation (n= 4)

lP

re

Final Analysis

(n= 53)

Without PSD (n= 47)

ur na

With PSD

Jo

Figure 1: Flow chart of the selection and allocation of participants (N=100)

21

22

re

lP

ur na

Jo -p

ro

of

of ro

Jo

ur na

lP

re

-p

Table 1: Distribution of demographic and clinical characteristics and descriptive analysis results

23

Without PSD n=47

P

25(53.2) 22(41.5)

0.24*

26(70.3) 20(54.1) 7(26.9)

11(29.7) 17(45.9) 19(73.1)

0.003*

11(68.8) 42(50.0)

5(31.2) 42(50.0)

0.17*

22(53.7) 31(52.5)

19(46.3) 28(47.5)

0.54*

No (n (%)) Yes (n (%))

38(48.1) 15(71.4)

41(51.9) 6(28.6)

0.06*

No (n (%)) Yes (n (%))

39(52.7) 14(53.8)

35(47.3) 12(46.2)

0.92*

No (n (%)) Yes (n (%))

27(51.9) 26(54.2)

25(48.1) 22(45.8)

0.82*

No (n (%)) Yes (n (%))

40(58.8) 13(40.6) 57.75(14.40)

28(41.2) 19(59.4) 53.51(13.69)

0.09*

Gender

re

lP

Educational level 9-11 years (n (%)) 12 years (n (%)) Above 12 years (n (%)) Marital status Single (n (%)) Married (n (%))

22(46.8) 31(58.5)

-p

Male (n (%)) Female (n (%))

Affected limb

of

With PSD n=53

Right (n (%)) Left (n (%))

ro

Variables

ur na

Cigarette smoking habits Diabetes mellitus

Jo

Blood pressure

Cardiac disease Age (years) mean(SD)

24

0.14**

of ro

12(3-24)

0.10***

66.00(15.95) 68.98(14.03) 99.45(2.65) 9.32(1.56) -0.105(0.65) 89.89(6.56) 5.85(1.65)

<0.001** <0.001** <0.001** <0.001** <0.001** <0.001** <0.001**

-p

6(1-16.5)

re

47.83(20.75) 52.52(18.28) 83.67(23.05) 7.47(2.14) 0.093(1.22) 53.02(23.08) 2.70(1.99)

Jo

ur na

lP

Time since stroke(months) median(Q1-Q3) Motricity index Arm mean(SD) Leg mean(SD) Trunk control mean(SD) Digit span mean(SD) Z score(TMT B-A) mean(SD) Barthel index mean(SD) Lawton ADL mean(SD)

25

of ro **

-p

Fisher Exact

***

Independent T-test

U Mann Witney

PSD= Post-stroke depression

ur na

ADL=activities of daily living

lP

re

*

Jo

TMT=trail making test

26

of ro -p re

Jo

ur na

lP

Table2: Univariate and multivariate logistic regression on demographic and clinical characteristics

27

0.49 0.57

0.15 0.001

0.58

0.18

1 2.01 6.41

0.28-1.38

0.77-5.23 2.10-19.60

1 2.20

0.002

0.89 2.17

0.52 0.61

0.09 >0.001

2.41 8.78

0.88-6.63 2.64-29.11

1.11

0.49

0.02

3.11

1.19-8.13

0.70-6.89

0.41

0.91

0.96 1

0.43-2.13

0.53

0.06

1 0.37

0.13-1.05

0.46

0.92

1 0.96

0.39-2.34

-0.09

0.40

0.82

1 0.92

0.42-2.01

0.73

0.44

0.09

1 2.09

0.89-4.91

-0.05

-p

1 0.63

Multivariate logistic regression SE P OR 95% CI

re

0.24

B

ro

of Right Left

0.40

lP

Affected limb

Univariate logistic regression SE P OR 95% CI

ur na

Variable B Gender Male -0.47 Female Educational level(years) 9-11 years 12 years 0.70 Above 12 years 1.86 Marital status Single 0.79 Married

Cigarette smoking habits No -0.99 Yes

Diabetes mellitus

0.05

Jo

No Yes

Blood pressure

No Yes

Cardiac disease

No Yes

28

of ro

-1.38

0.45

0.002

OR odd ratio

Jo

ur na

CI Confidence Interval

lP

re

-p

Constant

29

0.25

30

re

lP

ur na

Jo -p

ro

of

of ro -p re lP

Jo

ur na

Table 3: Summary of interpretive measures for Stepwise Discriminant Function Analysis

31

of ro

-p

Discriminant Univariate F ratio loading(rank) 111.796** 21.720**

0.896 0.395

Without PSD 1.253

Jo

ur na

lP

re

Unstandardized Standardized Discriminant Discriminant Discriminants Function Function Coefficient Coefficient Barthel Index 0.082 1.419 Trunk control Test -0.041 -0.686 Constant -2.032 Group centroid : With PSD -1.111

32

33

re

lP

ur na

Jo -p

ro

of