Reliability and validity of a new post-stroke depr...
Background：Nowadays there is still a lack of effective method to evaluate post-stroke depression.To distinguish patients with and without depression after stroke reliably，this study proposes a new PostStroke Depression Scale(PSDS).
Methods：PSDS was developed based on various depression scales and clinician experiences.158 stroke patients who were able to finish PSDS and Hamilton Depression Rating Scale(HDRS)were recruited.Cronbachα，Spearman rank coefficient and Kruskal-Wallis test were respectively used to examine reliability，internal consistency and discriminate validity.Then the Receiver Operating Characteristic(ROC)curve was used to determine the ability of scale and categorized scales to the range of depression.Finally，the factors of the PSDS were classified by average clustering analysis.
Results：The Cronbachαof PSDS was 0.797(95%CI)indicted a good reliability.The Spearman correlation coefficient between PSDS and HDRS was 0.822(P＜0.001)showed an excellent congruent validity.The discriminate validity displayed significant difference between patients with and without depression(P＜0.001).6/24 was set to be the cut-off value by ROC analysis.Moreover，the different severity was distinguished by the value 6/24，15/24 and 17/24.
Limitations：The small sample size maybe the main limitation，the larger sample used in different fields according sex，age and side-lesion was needed to verity the results.The cut off value calculated by ROC curve maybe react the severity of the disease to some extent，but it is not absolute.
Conclusions：PSDS is a valid，reliable and specific tool for evaluating post-stroke depression patients and can be conveniently utilized.
Key words：stroke；Depression；Assessment；Post-Stroke Depression Scale(PSDS)
中文摘要
卒中后抑郁评估量表在中国人群中的信效度研究
岳莹莹 刘 瑞 陆 建 王晓菁
张石宁 吴爱勤 王 桥 袁勇贵
目的：当前对于卒中后抑郁（post stroke depression，PSD）患者，仍缺乏有效的评估量表进行早期筛查，为了能够有效识别PSD患者，我们制定了一个新的卒中后抑郁评估量表（post stroke depression scale，PSDS）。方法：该量表根据已有的抑郁评估量表和临床医生经验制定。本研究入组158例脑卒中患者，采用汉密尔顿抑郁量表（Hamilton Depression Rating Scale，HDRS）和PSDS对抑郁症状进行评估，通过Cronbachα系数、Spearman秩相关、Kruskal-Wallis检验对该量表的信度、内部一致性和区分效度进行检测；之后通过受试者工作特征曲线（receiver operating characteristic，ROC）和约登指数（Youden index，YI）对量表的有效性进行测定和界值划分；最后通过聚类分析方法寻找PSD的特异症状。
结果：PSDS的Cronbachα为0.797，表明该量表有较好的信度。PSDS和HDRS的相关系数为0.822（P＜0.001），PSD和卒中非抑郁（Non-PSD）患者的PSD-S评分存在显著差异（P＜0.001）表示其有较好的聚合效度和区分效度。ROC曲线和YI指数示6/24，15/24，分别为可疑/轻度抑郁，中重度抑郁的界值。
局限：本研究样本量较小，未来可以在大样本中通过对年龄、性别和卒中部位进行分层研究，界值划分在一定程度上反应了抑郁严重程度。
结论：PSDS是简单易行的自评量表，在卒中人群中具有较好的信度和效度，可以用来对卒中幸存者进行广泛筛查，实现PSD的早期发现。
关键词：卒中；抑郁；评估；卒中后抑郁评估量表
1 Introduction
Post-stroke depression is a frequent complication that worsens rehabilitation outcomes，quality of life and confers substantial risk for suicide(de Man-van Ginkel et al.，2012；Pompili et al.，2012).A pooled estimate indicates that depressive symptoms are present in one third of all stroke survivors at any time during the follow-up(de Man-van Ginkel et al.，2013；Zhang et al.，2012).Increasing evidence shows that regularly antidepressant treatment will lead to decreased depression symptoms and improved functional status(Loubinoux et al.，2012).Therefore，the early detection of poststroke depression is essential to optimize the recovery of stroke patients and avoid unfortunate incidents.
The diagnosis criteria of post-stroke depression are virtually unclear because there is no accurate description of this disease in the three international diagnosis systems，including Diagnostic and Statistical Manual of Mental Disorder，Fourth Edition(DSM-IV)by American Psychiatry Association，the International Classification of Disease，Tenth Edition of the World Health Organization and Chinese Classification of Mental Disorders，Third Version.Among them，DSM-IV is one of the most common diagnostic criteria at present(Kang et al.，2013)，but it just recommends that post-stroke depression is distinguished from a“Mood Disorder due to a General Medical Condition”(§293.83)when the mood disturbance is judged to be a direct consequence of a medical condition such as stroke.It was described as“depressive disorder due to another medical condition”in the DSM-5 published in May，2013.The etiology is nearly unknown and confused by many factors，so the diagnostic nosology is focused on clinical manifestations，duration of disorders and interference with social function.
The evaluation of depression in the stroke patients is performed generally using the scales developed for the psychiatric population(Gabaldon et al.，2007)，which include the Hamilton Depression Rating Scale(HDRS)(Hamilton，1960)，Beck Depression Inventory(Beck et al.，1961)，Montgomery Asberg Depression Rating Scale(Montgomery and Asberg，1979)，Zung Self-rating Depression Scale(Zung，1965)and Geriatric Depression Screening Scale(Yesavage et al.，1983).HDRS has been one of the most widely used in the study of post-stroke depression and has been shown to be a reliable instrument(Bjerg et al.，1997；Aben et al.，2001；Aben et al.，2002).Originally，it consists of 17 items that evaluate the symptoms of depression，in which 3 items are sleep-related.However，there have been some concerns raised about their low specificity(Salter et al.，2007).For example，the use of verbal skills to diagnose and measure depression is a matter for debate since there is no instrument designed specifically for patients who have suffered from stroke that usually caused physical and cognitive impairments(Kang et al.，2013).On the other hand，a state of agitation or psychomotor retardation，insomnia，or significant loss of weight that may be directly related to the somatic disorder accounted for a large part percentage of the total score biased to depression.Furthermore，motor aphasia patients may even make it difficult to communicate with patients.Several scales have been proposed for specifically evaluation for aphasic patients，including the Visual Analog Mood Scales(Arruda et al.，1999)，Stroke Aphasic Depression Questionnaire(Lincoln et al.，2000)and the Aphasic Depression Rating Scale(Benaim et al.，2004).However Stroke Aphasic Depression Questionnaire is commonly used by caregiver or rehabilitators through the observation of the patients，but few study has been used by nurses'or rehabilitation observation and proxy ratings in assessing depression after stroke patients.Then a test named post-stroke depression rating scale specifically devised for post-stroke patients was consisted of 10 sections including catastrophic reaction as well as difficulty in emotional control and each section has different subtype(Gainotti et al.，1997).Maybe some items that have not been commonly accepted limit their widespread application by clinical and scientific research of post-stroke depression patients(Colasanti et al.，2010；Spalletta and Robinson，2010).
Because of no specific self-rating assessment of post-stroke depression patients，this study is aimed to develop a new PostStroke Depression Scale(PSDS)in Chinese population and then verify the reliability and validity of the scale.
2 Materials and methods
2.1 The development of the PSDS
The selection of items was as followed：first，55 items were collected after being analyzed，arranged and merged from HDRS，Zung Self-rating Depression Scale，Beck Depression Inventory，Montgomery Asberg Depression Rating Scale，Aphasic Depression Rating Scale，Stroke Aphasic Depression Questionnaire，Visual Analog Mood Scales，Post-Stroke Depression Rating Scale，Center for Epidemiologic Studies Depression Scale(Radloff，1977)，the Hospital Anxiety and Depression Scale(Zigmond and Snaith，1983)and the Patient Health Questionnaire Depression Scale-9(Williams et al.，2005).Second，17 items were selected by 10 senior psychiatrists or neurologists according to their clinical experiences in our research group.Then，these items were emailed to the national experts to choose the common symptoms in depressive patients after stroke according to their clinical experiences.Replies were received from 65 chief doctors consisting of 39 psychiatrists and 26 neurologists.The PSDS consisted of 8 items[decreased speech(do not want to speak)，easy fatigability，easy to cry，insomnia(waking up too early)，feeling of decreased capability，suicidal ideation，feeling of difficult to recover，more irritable than usual]which were most selected by more than half-experts after statistics(see Supplementary material).PSDS is a self-rating scale，the subjects were asked to read each of the 8 items and carefully decided how often the statement describes according their feeling during the last 7 days in the following four quantitative terms：absent，some of the time，part of the time，or most of the time.A value of 0，1，2，and 3 is assigned to a response depending upon whether the item was positive or negative.The PSDS is constructed so that the more depressed subjects and his complaint will have a higher score on the scale.Add up the score of each item for a total score and the highest possible score is 24.
2.2 Participant selection
This study was approved by the Medical Ethics Committee for Clinical Research of Zhongda Hospital Affiliated to Southeast University.A prospective study was conducted in 2 cooperation hospitals：Nanjing Ruihaibo Rehabilitation Hospital and Nanjing Brain Hospital in 2013.A total of 158 patients were recruited and they were all given written informed consent.To be enrolled in our study，participants are required to meet the following criteria：(1)all participants with ischemic stroke and intracerebral hemorrhage determined by Computed Tomography(CT)or Magnetic Resonance Imaging(MRI)data；(2)each participant evaluated HDRS and PSDS；(3)participants were antidepressant-naïve，and the age of onset was under 80 years；(4)participants were free of other major psychiatric disorders，including schizophrenia，bipolar disorder，substance abuse(caffeine，nicotine and alcohol)(American Psychiatric Association，1994)，neurodegenerative illness，severe physical illnesses and other medical illnesses；(5)participants were free of anosognosia，neglect，hemianopia，cortical blindness，amnesia，aphasia，dementia and other symptoms hindered assessment.
Diagnostic evaluations of post-stroke depression were carefully conducted on all participants who fulfilled the following diagnostic criteria combined with the previous literature by two trained senior psychiatrists.Five aspects were included in the diagnostic criteria：(1)had stroke before，or stroke occurs earlier than depressive symptoms；(2)met at least two depressive symptoms except core criterion symptoms of depressed mood and loss of interest or pleasure in nine symptoms of major depressive disorder in DSM-IV；(3)impairment to fit personal and work functioning(not induced by somatic disorder)，the motor deficits since stroke were evaluated by Barth Index(Mahoney and Barthel，1965)；(4)depressive symptoms lasting more than 1 week；(5)free of other major psychiatric disorders，including schizophrenia，bipolar disorder，substance abuse(caffeine，nicotine and alcohol)(see Supplementary material).
2.3 Data analysis
SPSS 18.0 software(SPSS，Inc，Chicago，IL)was used to calculate the reliability and validity of PSDS and MATLAB(R 2012b)was to clustering analysis and multiple linear regression.We would illustrate the PSDS from the following five aspects.
First，in order to ensure reliability of the results，PASS 13 was used to calculate the power whileαwas set to 0.05.
Second，the reliability of PSDS and HDRS was assessed by Cronbachα which evaluated the internal consistency of each items.In the exploratory analysis，the value of Cronbachα0.6—0.9 indicates good consistency reliability(Duncan et al.，1999).
Third，the validity of PSDS was assessed by Spearman rank correlation coefficient and Kruskal-Wallis test(Rosner，2004).Validity is used to show the accuracy，usefulness of assessment which included criterion validity，content validity and construct validity.Content validity contains congruent validity and discriminated validity.The congruent validity was calculated by the Spearman rank correlation coefficient and discriminate validity was estimated by Kruskal-Wallis test.In order to assess the criterion validity of PSDS，Spearman rank correlation coefficient was used to indicate the relationship between each item and the total score.A high correlation during-1 to-0.5 and 0.5 to 1 indicates good consistency and shows that PSDS could be used to evaluate the depression(McCrate et al.，2011).The items which Spearman rank correlation coefficients were larger than 0.6 were considered to be the specific characteristics of post-stroke depression patients.In order to test whether the suspected specific characteristics can explain the depression，the correlation between the total score of four suspected characteristics of post-stroke depression and HDRS total score was calculated.
Fourth，the diagnose cut-off score was developed by Youden Index combing sensitivity value and specificity value(Fluss et al.，2005)，which resulted from ROC curve.The areas of ROC curve were rational index of the whole diagnostic precision of the test.Youden Index was calculated by the formula：
Youden Index=sensitive value+specificity value-1 (1)
The maximum of Youden Index is the best cut-off score.To verify the accuracy and effectivity of this cut-off value，we calculated the accordance rate compared with HDRS and used multiple linear regression to observe the comparison directly.
Finally，the main factors of the PSDS were classified by an average clustering analysis method.
3 Results
From all participating patients，data were collected regarding potential risk factors for post-stroke depression(Hackett and Anderson，2005).These included socio-demographic and strokerelated factors medical history concerning vascular risk factors and vascular diseases(van Swieten et al.，1988).Details are described in Table 1.Table 2 describes the relative stage of stroke patients.For a variable to be considered candidate predictor of post-stroke depression，it had to be easily collected in a clinical setting with a view to future applicability.Consequently，the data collected were a proportion of the data normally collected in hospital care.
There were 101 stroke patients without depression(63.92%)and 57 post-stroke depression patients(36.08%)which contained 43 males and 14 females in this study.
Table 1 Characteristics of study sample
续表
Table 2 The duration distribution of the study sample.
3.1 Power of test
βis the probability of accepting a false null hypothesis which should be small whenαis 0.05.Conventionally a test with a power greater than 0.8(orβ≤0.2)is considered statistically powerful(Park，2008).Through the analysis of PASS 13，when the size of sample is 130，βis 0.118，so its power is 88%.If the size of sample is 170，βis 0.05，so its power is 95%.In this study，the size is 158，so the power between 88%and 95%.The value of power is enough to show patients with and without depression had significant difference and this study is effective.
3.2 Reliability
Cronbachα was used to estimate the reliability of scales.The Cronbachαof PSDS and HDRSis 0.797(95%CI)and 0.643(95%CI)respectively，which shows that PSDS has a better reliability than HDRS.
3.3 Validity
All data of subjects were used to calculate the Spearman rank correlation of PSDS and HDRS scores.The correlation coefficient was 0.822(P＜0.001)which showed a good congruent validity.In hence it can be used to evaluate depression.Then，we used Kruskal-Wallis Test to calculate the discriminate validity of PSDS and the result showed PSDS had significant difference between patients with and without depression after stroke.The result demonstrated a significant effect on estimating depression in stroke patients(P＜0.001).The boxes plot of the group of poststroke patients and stroke patients without depression display that the two median lines have a far distance between two groups[see Fig.1(a)].Therefore，PSDS has a good discriminate validity that could be used to distinguish the depression patients after stroke.
The Spearman's rank correlation coefficients between every item and total score of PSDS is listed in Table 3.All the items correlated with the total score significantly(r＜0.49；P＜0.001).The value shows that PSDS has good internal consistency.The Spearman rank correlation coefficients of decreased Speech，easy fatigability，feeling of decreased capability and feeling of difficult to recover were larger than 0.6 with the total score，so these four items were thought to be the specific characteristics of poststroke depression patients.Easy to cry，insomnia，waking up too early，suicidal ideation and more irritable than usual were the general characteristics of post-stroke depression patients.Furthermore，we calculated the relevant between the score of characteristic items and HDRS to verify our hypothesis.The Spearman rank coefficient indicated a significant relevant of 0.718(P＜0.001).From Fig.1(b)，we can see that the correlation is relevant directly and has higher correlation in the range of minor to serve.
Table 3 The result of spearman rank correlation of PSDS
3.4 Cut-off value
The cut-off value of PSDS was 6 by the ROC curves.While，the cut-off values of minor，middle and severe degree were 6，15，and 17.Fig.1(c)demonstrates the ROC curves of PSDS and HDRS.Short dotted line is the curves of PSDS and long dotted line is the curves of HDRS.From Fig.1(c)，we can seethat HDRS has a larger area than the PSDS，but the line of PSDS is in close proximity to the line of HDRS.The PSDS and HDRS areas under the line were 0.894(P＜0.001；95% CI，0.845～0.943)and 0.944(P＜0.001；95%CI，0.911～0.977).The result shows that PSDS has a good diagnostic accurate.
Fig.1 (a)The box plots of patients with and without depression after stroke，the result of Kruskal-Wallis test showed a significant difference between two group(PSD and non-PSD，P＜0.001).(b)The Spearman rank coefficient between specific characteristics of post-stroke depression patients(the score of decreased speech，easy fatigability，feeling of decreased capability and feeling of difficult to recover)and HDRS indicated a significant relevant of 0.718(P＜0.001).(c)ROC curve for PSDSand HDRS which areas under the line were 0.894(P＜0.001；95%CI，0.845-0.943)and 0.944(P＜0.001；95%CI，0.911-0.977).(d)The PSDS was divided into three parts by clustering analysis.PSD：post-stroke depression，non-PSD：stroke patients without depression，PSDS：Post-Stroke Depression Scale，HDRS：Hamilton Depression Rating Scale，ROC：Receiver Operating Characteristic.
In the process of setting the cut-off value of PSDS，we used the one-toone correspondence and the average method.As shown in Table 4，HDRS's cut-off value 7.5 corresponded to the maximum Youden Index 0.772；6.5 corresponded to the second Youden Index 0.747；7 was the average of 7.5 and 6.5.We calculated the cut-off value of PSDS in the light of the same method.On the other hand，the values of 15 and 17 indicate the level of middle and severe PSD according to ROC curve.Based on the cut-off values of PSDS，there were 49 patients with minor depression，4 patients with middle depression and 1 patient with severe depression.Similarly，based on the cut-off value of HDRS，the number of patients was 51，5 and 1 with minor to severe(see Table 5).
Table 4 Cut-off value of PSDS and HDRS
续表
Table 5 The accordance rate of PSDS and HDRS
(https://www.daowen.com)
To verify the effective of cut-off value，we fitted the HDRS and the PSDS by the multiple linear regression method.We can see that the 57 depressive patients'scores were all above 7，but 3 patients is less than 6 score，and PSDS accordance rate is 94.7%(Table 5).Furthermore，the PSDS cut-off value is used to 101 patients without depression，the accordance rate is 89%.PSDS plays a good auxiliary diagnostic role and distinguishes the symptom of depression excellently.
3.5 The main factors of PSDS
Fig.1(d)shows the result of clustering analysis extracting the factors of post-stroke depression.8 items were divided into three parts.Part 1 was insomnia，waking up too early；Part 2 was more irritable than usual：Part 3 included 6 factors that is feeling of difficult to recover，easy fatigability，feeling of decreased capability，decreased speech，easy to cry and suicidal ideation.Therefore，Part 1 was named as sleeping factors；Part 2 was named as mood factors；Part 3 was named as regressive factors.
4 Discussion
The aim of our study presents two folds：(1)draft a new specific self-rating scale screening for post-stroke depression patients；(2)verified the reliability and validity and obtained the cut-off value of the new PSDS scale.
The power of test(1-β)calculated by PASS demonstrated between 88%and 98% while sample is 158 which showed a significant difference as well as an effective result.
Although DSM-IV criteria for mood disorders have been often used to diagnose post-stroke depression(Robinson，2003)，this approach may have been misleading because some symptoms considered as diagnostic of depression can frequently occur also in patients without depression.Psychomotor retardation，fatigue，sleep，and appetite disturbance，may be a consequence of the stroke event and not necessarily indicative of depression.In particularly，previous study showed that reports of somatic complaints were not useful indicators of depression in elderly stroke patients(Gordon and Hibbard，1991).However，fewer study attempts to revise the diagnosed criteria.In the present study，it was modified based on DSM-IV that at least two depressive symptoms excepting one core criterion symptoms of depressed mood either loss of interest or pleasure and duration of at least 1 week compared with five symptoms besides one core criterion during the same 2-week period.
Furthermore，the cognitive impairment of post-stroke depression patients may limit a patient's ability to describe or pour their anguish(Tilanus and Timmerman，2005).In the previous study we found that memory，orientation，language and attention are the most likely impaired cognitive domains after a stroke(Tatemichi et al.，1994).So the evaluation of the cognitive efficacy of the participants is important，and it also guarantees greater strength of the relation between questionnaire answers and the effective psychological outcome of the patient.In order to avoid the influence of cognitive impairment，the patients enrolled in our study had no complain of cognitive impairment，in addition we also interviewed with the family dependents of patients aimed to know the state of their emotional expression.Then we evaluate the cognitive function by Montreal Cognitive assessment(MoCA)and the result showed that MoCA had less influence on PSDS than HDRS(see Supplementary material).
Results of the present study are relative comprehensive from several aspects.It is also specific scale similar to post-stroke depression rating scale which is an observer-rating scale in previous preliminary report，as to the nature of the major form of poststroke depression(Gainotti et al.，1995).First，the reliability and validity showed a good inter-consistency which is security of the diagnosis，Cronbachαand spearman rank correlation could illustrate the effective of PSDS.Second，the cut-off value was estimated.The area of PSDS was enough to show its rational index of the whole diagnostic precision of the test.We also calculated the number of minor，middle and severe level of PSDS and compared the rate with HDRS.The HDRS appears similar to self-rating scales in terms of the accurate screening for the presence of depression after stroke.Three terms of insomnia focused on the somatic aspects significantly inflated scores of HSRS.Moreover，in terms of post-stroke depression individuals，a cut-off score of 10 of this scale has been recommended as indicative of presence of depression rather than the usual 7 because of high proportion of some items(Agrell and Dehlin，1989).However，its usefulness as a screening tool is limited by long time consuming，more factors construct and the level of examiner expertise required.Results obtained by HDRS are linked closely to the expertise of the interviewer(Mc Dowell and Newell，1996).In hence，the self-rating scale PSDS may be widely used to screen of the stroke patients especially the motor aphasia patients.It may benefit the patient's functional recovery，improvement the life quality and reintegration into the society ultimately by screening the depression and timely treatment.
In addition，the items of scale were clustered and fitted to find the characteristics of post-stroke depression patients.The PSDS was divided into three parts，which were named as sleeping factor，mood factors and regressive factors.Regressive characteristics contain 6 factors reflecting the decreased ability of competent and social adapt function.Differential anatomical substrates of two different subtypes of post-stroke depression：“apathetic”and“affective”one(Hama et al.，2007).This approach has been widely criticized mainly because of the controversies in assimilating poststroke depression to endogenous mood disorders(Gordon and Hibbard，1991).Decreased speech，easy fatigability，feeling of decreased capability and feeling of difficult to recover were listed to the suspected characteristics of post-stroke depression according to the result of Spearman rank coefficient and clustering analysis.In our opinion，even if our results uncertainty at all，however，it could lead to an updated model for the pathogenesis of post-stroke depression.
The present work was an exploratory study，technical and biological limitations inevitably exist.First，this was a crosssectional study with a relatively small sample size，the larger sample were needed to verity the results.Second，the PSDS should be used in different fields according to sex，age and side-lesion.Thirdly，the cut off value calculated by ROC curve maybe react the severity of the disease to some extent，but it is not absolute.
Overall，post-stroke depression is a common sequel of stroke that is associated with reduced functional recovery，social outcomes，quality of life and increased mortality.Early screening and assessment can lead to a significant，positive impact on the individual recovery.Reconsidering the diagnosis and finding a reliable and validity scale specifically devised to assess post-stroke depression patients is the main purpose of this study.We are optimistic about the utility of this new scale.However，as required in the development of any scale，an ongoing work to evaluate this scale and to explore the generalizability of the results in one population to other stroke populations is needed.We hope that this scale may be helpful for screening depression patients after stroke then promote early discovery and timely treatment.
5 Contributors
The order of authorship is Yingying Yue，Rui Liu，Jian Lu，Xiaojing Wang，Shining Zhang，Aiqin Wu，Qiao Wang，Yonggui Yuan.We thank to all individuals who participate in this study.
5.1 Role of funding source
The funding source had no further role in study design，in the collection，analysis and interpretation of data，in the writing of the report，and in the decision to submit the paper for publication.
5.2 Conflict of interest
No conflict declared.
5.3 Acknowledgments
We wish to thank all the participants in this study.This research was funded by Jiangsu Provincial Special Program of Medical Science(BL2012025，Yonggui Yuan).
6 Appendix A.Supplementary material
Supplementary data associated with this article can be found in the online version at http：//dx.doi.org/10.1016/j.jad.2014.11.031.
References
[1] Aben，I.，Verhey，F.，Honig，A.，Looder，J.，Lousberg，R.，Maes，M.，2001.Research into the specificity of depression after stroke：a review on an unresolved issue.Prog.Neuro-Psychopharmacol.Biol.Psychiatry 18，671-689.
[2] Aben，I.，Verhey，F.，Lousberg，R.，Lodder，J.，Honing，A.，2002.Validity of the Beck Depression Inventory，Hospital Anxiety and Depression Scale，SCL-90，Hamilton Depression Rating Scales screening instruments for depression in stroke patients.Psychosomatics 43，386-393.
[3] Agrell，B.，Dehlin，O.，1989.Comparison of six depression rating scales in geriatric stroke patients.Stroke 20，1190-1194.
[4] American Psychiatric Association，1994.Diagnostic and Statistical Manual of Mental Disorders—DSM-IV.American Psychiatric Press，Washington，DC.Arruda，J.E.，Stern，R.A.，Somerville，J.A.，1999.Measurement of mood status in stroke patients：validation of the visual analogue mood scales.Arch.Phys.Med.Rehabil.80，676-680.
[5] Beck，A.T.，Ward，C.H.，Mendelson，M.，Mock，J.，Erbaugh，J.，1961.An inventory for measuring depression.Arch.Gen.Psychiatry 4，561-571.
[6] Benaim，C.，Cailly，B.，Perennou，D.，Pelissier，J.，2004.Validation of the Aphasic Depression Rating Scale.Stroke 35，1692-1696.
[7] Bjerg，B.B.，Bjerg，B.E.，Lauritzen，L.，Vilmar，T.，Bech，P.，1997.Poststroke patients in rehabilitation：the relationship between biological impairment(CT scanning)，physical disability and clinical depression.Eur.Psychiatry 12，399-404.
[8] Colasanti，V.，Marianetti，M.，Micacchi，F.，Amabile，G.A.，Mina，C.，2010.Tests for the evaluation of depression in the elderly：a systematic review.Arch.Gerontol.Geriatr.50，227-230.
[9] de Man-van Ginkel，J.M.，Hafsteinsdottir，T.，Lindeman，E.，Burger，H.，Grobbee，D.，Schuurmans，M.，2012.An efficient way to detect poststroke depression by subsequent administration of a 9-item and a 2-item Patient Health Questionnaire.Stroke 43，854-856.
[10] de Man-van Ginkel，J.M.，Hafsteinsdottir，T.B.，Lindeman，E.，Ettema，R.G.，Grobbee，D.E.，Schuurmans，M.J.，2013.In-hospital risk prediction for post-stroke depression development and validation of the Post-stroke Depression Prediction Scale.Stroke 44，2441-2445.
[11] Duncan，P.W.，Wallace，D.，Lai，S.M.，Johnson，D.，Embretson，S.，Laster，L.J.，1999.The stroke impact scale version 2.0.Evaluation of reliability，validity，and sensitivity to change.Stroke 30，2131-2140.
[12] Fluss，R.，Faraggi，D.，Reiser，B.，2005.Estimation of the Youden Index and its associated cutoff point.Biom.J.47，458-472.
[13] Gabaldon，L.，Fuentes，B.，Frank-Garcia，A.，Diez-Tejedor，E.，2007.Post stroke depression：importance of its detection and treatment.Cerebrovasc.Dis.24，181-188.
[14] Gainotti，G.，Azzoni，A.，Lanzillotta，M.，Marra，C.，Razzano，C.，1995.Some preliminary findings concerning a new scale for the assessment of depression and related symptoms in stroke patients.Ital.J.Neurol.Sci.16，439-451.
[15] Gainotti，G.，Azzoni，A.，Razzano，C.，Lanzillotta，M.，Marra，C.，Gasparini，F.，1997.The Post-Stroke Depression Rating Scale：a test specifically devised to investigate affective disorders of stroke patients.J.Clin.Exp.Neuropsychol.19，340-356.
[16] Gordon，W.A.，Hibbard，M.R.，1991.Issues in the diagnosis of post-stroke depression.Rehabil.Psychol.36，71-85.
[17] Hackett，M.L.，Anderson，C.S.，2005.Predictors of depression after stroke：a systematic review of observational studies.Stroke 36，2296-2301.
[18] Hama，S.，Yamashita，H.，Shigenobu，M.，Watanabe，A.，Kurisu，K.，Yamawaki，S.，Kitaoka，T.，2007.Post-stroke affective or apathetic depression and lesion location：left frontal lobe and bilateral basal ganglia.Eur.Arch.Psychiatry Clin.Neurosci.257，149-152.
[19] Hamilton，M.，1960.A rating scale for depression.J.Neurol.Neurosurg.Psychiatry 23，56-62.
[20] Kang，H.J.，Stewart，R.，Kim，J.M.，Jang，J.E.，Kim，S.Y.，Bae，K.Y.，Kim，S.W.，Shin，I.S.，Park，M.S.，Cho，K.H.，Yoon，J.S.，2013.Comparative validity of depression assessment scales for screening post-stroke depression.J.Affect.Disord.147，186-191.
[21] Lincoln，N.B.，Sutcliffe，L.M.，Unsworth，G.，2000.Validation of the Stroke Aphasic Depression Questionnaire(SADQ)for use with patients in hospital.Clin.Neuropsychol.Assess.1，88-96.
[22] Loubinoux，I.，Kronenberg，G.，Endres，M.，Schumann-Bard，P.，Freret，T.，Filipkowski，R.K.，Kaczmarek，L.，Popa-Wagner，A.，2012.Post-stroke depression：mechanisms，translation and therapy.J.Cell.Mol.Med.16，1961-1969.
[23] Mahoney，F.I.，Barthel，D.W.，1965.Functional evaluation：the Barthel Index.Md.State Med.J.14，61-65.
[24] McCrate，R.R.，Kurtz，J.E.，Yamagata，S.，Terracciano，A.，2011.Iternal consistency，retest reliability，and their implications for personality scale validity.Personal.Soc.Psychol.Rev.15，28-50.
[25] McDowell，I.，Newell，C.，1996.Measuring Health.A Guide to Rating Scales and Questionnaires，New York，2nd ed.Oxford University Press.Montgomery，S.A.，Asberg，M.，1979.A new depression scale designed to be sensitive to change.Br.J.Psychiatry 134，382-389.
[26] Park，H.M.，2008.Hypothesis Testing and Statistical Power of a Test.Working Paper.The University Information Technology Services(UITS)Center for Statistical and Mathematical Computing，Indiana University.
[27] Pompili，M.，Venturini，P.，Campi，S.，Seretti，M.E.，Montebovi，F.，Lamis，D.A.，Serafini，G.，Amore，M.，Girardi，P.，2012.Do stroke patients have an increased risk of developing suicidal ideation or dying by suicide?An overview of the current literature.CNS Neurosci.Ther.18，711-721.
[28] Rosner，B.，2004.Fundamentals of Biostatistics，Seventh Edition Harvard University，Canada.
[29] Radloff，L.S.，1977.The CES-D scale：a self-report depression scale for research in the general population.Appl.Psychol.Meas.1，385-401.
[30] Robinson，R.G.，2003.Poststroke depression：prevalence，diagnosis，treatment，and disease progression.Biol.Psychiatry 54，376-387.
[31] Salter，K.，Bhogal，S.K.，Foley，N.，Jutai，J.，Teasell，R.，2007.The assessment of poststroke depression.Top.Stroke Rehabil.14，1-24.
[32] Spalletta，G.，Robinson，R.G.，2010.How should depression be diagnosed in patients with stroke?Acta Psychiatr.Scand.121，401-403.
[33] Tatemichi，T.K.，Desmond，D.W.，Stern，Y.，Paik，M.，Sano，M.，Bagiella，E.，1994.Cognitive impairment after stroke：frequency，patterns，and relationship to functional abilities.J.Neurol.Neurosurg.Psychiatry 57，202-207.
[34] Tilanus，J.，Timmerman，L.，2005.Poststroke depression.Rev.Clin.Gerontol.14，37-43.
[35] van Swieten，J.C.，Koudstaal，P.J.，Visser，M.C.，Schouten，H.J.，van Gijn，J.，1988.Inter observer agreement for the assessment of handicap in stroke patients.Stroke 19，604-607.
[36] Williams，L.S.，Brizendine，E.J.，Plue，L.，Bakas，T.，Tu，W.，Hendrie，H.，Kroenke，K.，2005.Performance of the PHQ-9 as a screening tool for depression after stroke.Stroke 36，635-638.
[37] Yesavage，J.A.，Brink，T.L.，Rose，T.L.，Lum，O.，Huang，V.，Adey，M.，Leirer，V.O.，1983.Development and validation of a geriatric depression screening scale：a preliminary report.J.Psychiatr.Res.17，37-49.
[38] Zhang，N.，Wang，C.X.，Wang，A.X.，Bai，Y.，Zhou，Y.，Wang，Y.L.，Zhang，T.，Zhou，J.，Yu，X.，Sun，X.Y.，Liu，Z.R.，Zhao，X.Q.，Wang，Y.J.，Prospective Cohort study on Incidence and Outcome of Patients with Poststroke Depression in China(PRIOD)Investigators，2012.Time course of depression and one-year prognosis of patients with stroke in mainland China.CNS Neurosci.Ther.18，475-481.
[39] Zigmond，A.S.，Snaith，R.P.，1983.The Hospital Anxiety and Depression Scale.Acta Psychiatr.Scand.67，361-370.
[40] Zung，A.，1965.A self-rating depression scale.Arch.Gen.Psychiatry 12，63-70.
[作者及发表刊物：
Yingying Yue，Rui Liu，Jian Lu，Xiaojing Wang，Shining Zhang，Aiqin Wu，Qiao Wang，Yonggui Yuan.Reliability and Validity of a new poststroke depression scale in Chinese population[J].Journal of Affective Disorders，2015，174：317-323.]