Quality of life of stroke survivors

Quality of life of stroke survivors

Peter Kim1, Sharon Warren1, Helen Madill1 & Margaret Hadley21Department of Occupational Therapy, Faculty of Rehabilitation Medicine, University of Alberta, Edmonton, Canada;2Stroke and Brain Injury Units, Glenrose Rehabilitation Hospital, Edmonton, Alberta, Canada

Received 20 April 1998; accepted in revised form 17 November 1998

Abstract. Adaptation to stroke requires complex,long-term change in stroke survivors' lives. Thisstudy aimed at identifying those factors that in¯uencequality of life (QOL) of geriatric stroke survivors 1±3years post-discharge. The objectives were: to describethe overall quality of life of stroke survivors; to ex-amine the relationships between sociodemographicvariables, neurological variables, functional status,social support, perceived health status, depression,and overall QOL; and to determine the best predic-tors of QOL. Data were collected on 50 stroke sur-vivors using a cross-sectional design and standardizedquestionnaires, including the Quality of Life Index,

the Functional Independence Measure, the SocialSupport Inventory for Stroke Survivors and theCentre for Epidemiologic Studies Depression Scale.The overall quality of life of the study participantswas low. The most important predictors of QOL weredepression, marital status, quality of social support,and functional status. Depression was the strongestpredictor of QOL. By employing a multi-dimensionalperspective, this study con®rmed that adaptation tostroke involves much more than physical function.Thus, rehabilitation programs for this group wouldbe more e�ective if they are based upon a holisticapproach.

Key words: Quality of Life, Rehabilitation, Social Support, Stroke

Introduction

Cerebrovascular accident (CVA), or stroke, is amajor cause of disability among many elderlyCanadians, and therefore presents a major socialchallenge [1]. In the last four decades, advances inmedicine and technology, improved quality of care,and increased emphasis on health promotion havecontributed to a reduction in the incidence of stroke[2]. However, despite this decline, improved survivalrates and an aging population contribute to agrowing number of elderly stroke survivors living inthe community [2].

The impact of stroke on an individual is usuallyunanticipated, and its e�ects can be devastating [3].Depending on the severity and type, a stroke canleave a person with residual impairment of physical,psychological and social functions. The severity ofdisability depends upon the level of mobility, com-munication skills, or the degree to which behavioraland personality changes or cognitive de®cit a�ecteach individual. The occurrence of a stroke oftenhas several negative consequences for the strokesurvivor. Increased dependency on others for activ-ities of daily living, alteration in mood, anddisruption in social interaction with family membershave all been reported [4, 5]. Such changes have apotentially far-reaching in¯uence on the person'squality of life.

Even though previous studies have shown that re-habilitation can help a stroke survivor to regain andmaintain physical function [6, 7], therapeutic inter-ventions that focus on physical restoration in isola-tion from the context of a person's psychosocialenvironments are insu�cient to improve quality oflife following disability [8, 9]. This narrow approachtoward stroke rehabilitation usually identi®es dis-ability as a characteristic which resides in the indi-vidual rather than the environment [10].Consequently, even though the ®eld of rehabilitationhas recently placed a growing emphasis on improvingquality of life, functional status or health status arefrequently the only criteria used to measure quality oflife [8]. This approach tends to neglect the psycho-social and environmental in¯uences which a�ect howindividuals cope with stroke-related disabilities.Studies have shown that a signi®cant number ofstroke survivors with little or no physical problemsstill experience a deterioration in their quality of lifeover an extended period of time [6, 7, 11]. Improvedfunctional status may be of little consequence if in-dividuals are unable to resume meaningful and pro-ductive social roles within their environment. Ifcurrent approaches to rehabilitation do not lead to animprovement in the client's perceived quality of life,these should be reevaluated and further researchshould be conducted to determine what other factorsbesides functional status need to be considered.

Quality of Life Research 8: 293±301, 1999.Ó 1999 Kluwer Academic Publishers. Printed in the Netherlands.

Literature review

Using quality of life as an outcome criterion has beena controversial issue in the ®eld of health care [12].Since quality of life is a multidimensional concept,existing de®nitions lack clarity [13±15]. Lack of con-sensus regarding a satisfactory de®nition of quality oflife has complicated the operationalization of thisconcept as an outcome criterion. Nevertheless, qual-ity of life is often de®ned by a broad range of lifedomains which include psychosocial, physical, andsocial well-being [12, 16]. The most commonly useddimensions are physical function and psychosocialwell-being and include ``life satisfaction'' [12, 14, 17].Many authors believe that individual perception isintegral to the concept of quality of life [14, 16, 18].The underlying assumption is that, as people's valuesdi�er, so will the impact of functional limitations onvarious aspects of their lives [14]. Assumptions suchas this one focus on the subjective experiences of in-dividuals rather than on objective conditions (e.g.,income) of their lives. With a growing emphasis onhealth promotion and individual empowerment, aconceptual framework which emphasizes the patient'sperception of quality of life is gaining wider accep-tance among the health-related professions [19].

Functional status, referred to as the degree of in-dependence in performing Activities of Daily Living(ADL), is of particular signi®cance. ADL is the mostcommonly used measure of quality of life in the lit-erature on the impact of stroke. Studies of strokesurvivors have generally focused on the physical as-pects of stroke recovery and have shown that a ma-jority of survivors report depression anddeterioration in quality of life despite their level ofphysical recovery [7, 8, 11, 20±22]. For instance, in alongitudinal study on the quality of life of patientsfollowing stroke, AhlsioÈ et al. [11] used a visual an-alogue scale to measure change in quality of life andfound that even though lower quality of life was as-sociated with greater disability, perceived quality oflife failed to improve over time even when ADLfunction increased. To validate this observation,Niemi et al. [6] used a questionnaire concerning fourdomains of life (work conditions, activities at home,family relationships, and leisure time activities) toassess the quality of life for 46 stroke survivors underthe age of 65 years for 4 years after their ®rst stroke.Niemi and his associates [6], Nydevik and Hulter-AÊ sberg [7], and Viitanen et al. [22] reported that mostparticipants rated their quality of life as low despiteadequate functional recovery three to four years afteronset of stroke. Independence in functional status,thus, does not mean absence of handicap which isclosely related to the quality of a patient's life in thecommunity [23].

The social environment, speci®cally informal socialsupport, may also in¯uence the quality of life ofstroke survivors in the community. The individual

who returns to the community after a stroke en-counters problems beyond their level of independencein activities of daily living. These individuals andtheir families are in need of support in the struggle toimprove the quality of their lives. There is a rapidlyexpanding body of literature on the e�ects of infor-mal and formal social support on mental and physicalhealth [24, 25]. Social supports are known to delay orreduce the risk of institutionalization, and generallyimprove well-being and the degree of life satisfactionof elderly persons in the community [26]. Empiricalresearch on social support and quality of life amongstroke survivors speci®cally has produced inconsis-tent ®ndings. Biegel et al. [4] report that psychologi-cal and social factors may be as important as physicalfunction in determining the patient's quality of lifeand social reintegration. Further, in a cross-sectionalstudy examining the relationship of social support toquality of life of individuals 1 to 3 years after onset ofstroke, King [27] found that perceived social support,depression, and functional status contributed to 38%variance in quality of life. On the contrary, Lin,Simeone, Ensel, and Kuo [28] reported a very weakassociation between social support and quality of life.Although social support can have positive as well asnegative e�ects, in general, a growing number ofstudies indicate that during times of stress, the patientneeds to perceive that others are supportive or will bethere if necessary to help cope with demands posed bythe environment [24]. Considering the stressful natureof adapting to the residual impairments and disabil-ities of stroke, one would expect that a person'squality of life would be related to perceived socialsupport, as social support is known to mediate thee�ect of stressful experiences on quality of life [25].

Perceived health status, de®ned within a broadercontext of health determinants such as social envi-ronments, personal behaviors, and lifestyles [29], isthought to be another factor in QOL. Norland [30]found that elderly Canadians rated their own healthas good, very good, or excellent despite the increasingnumber of chronic conditions and disabilities, al-though positive responses declined with age. How-ever, stroke survivors are apparently more likely torate their health as poor than age-matched nondis-abled seniors living in the community [7, 31], or thanpersons without a history of stroke but at elevatedrisk [32]. Further, studies have found that both astroke patient's perceptions of speci®c impairmentsand general health status are appropriate outcomemeasures in clinical trials [33, 34] and that perceivedhealth status is closely related to quality of life [35].

With regard to a relationship between post-strokedepression and quality of life, study ®ndings againseem to be inconsistent. In a prospective study, Niemiet al. [6] found that the incidence of post-stroke de-pression among patients was 48% , and a tendency todepression was one of the strong predictors of qualityof life. Similarly, King [36] reported that quality of

294

life was negatively a�ected by depression, and it wasthe strongest predictor of overall quality of life in 86post-discharge participants with stroke. Further,AhlsioÈ et al. [11] found that depression and anxietya�ect quality of life as much as physical disability.Other investigators also observed problems of socialfunctioning in depressed patients even two years afterstroke [37]. However, some authors have not foundany association between social dysfunction and de-pression [38] or between depression and quality of life[39].

Other health and socio-demographic factors havebeen studied for their in¯uence on quality of life. Asfor the type of stroke and the localization of lesions,in a four year follow-up study, Niemi et al. [6] foundthat patients with subarachnoid hemorrhage experi-enced less deterioration in their quality of life whencompared to those with a brain infarction. Further,patients with either right- or left-hemisphere lesionswere reported to have more severely deterioratedquality of life compared to patients with no localiz-able or brain stem lesions. On the other hand, deHaan, Limburg, Van der Meulen, Jacobs, and Aa-ronson [40] found that survivors of hemorrhagicstrokes did not evidence lower quality of life thansurvivors of ischemic strokes.

The in¯uence of age and gender factors have alsobeen studied with varying results. Niemi et al. [6]found that older (51±64 yr.) patients experiencedmore severe deterioration in quality of life thanyounger (17±50 yr.) patients, and men had a slightlymore deteriorated quality of life. On the contrary,AhlsioÈ et al. [11] found no signi®cant age and genderdi�erences when ADL capacity was kept constant. Ingeneral, good health, marital status, socioeconomicstatus, and better education were found to be posi-tively associated with quality of life [41]. However,Labi et al. [21] found that individuals with highereducation and income levels had a lower quality oflife after stroke. The authors speculated that the de-terioration in quality of life was related to decreasedparticipation in social structures which resulted fromperceived stigmatization associated with disability.

In spite of extensive research on stroke, little at-tention has been focused on multi-dimensional as-pects of quality of life in stroke outcome research.Rather, physical dysfunction has long been the pri-mary focus in stroke research and, in rehabilitation,quality of life has been commonly assessed in terms offunctional status [8]. The presence of psychosocialdysfunction has always been acknowledged but hasbeen emphasized less; its potential in¯uence onquality of life has not yet been fully explored.

Purpose

The primary objectives of this study were (1) to des-cribe the overall quality of life of community-based

stroke survivors 1 to 3 years after discharge from alocal rehabilitation hospital, (2) to examine the rela-tionship between Quality of Life (QOL) and possiblepredictors including functional status in terms ofBasic Activities of Daily Living (BADL) and Instru-mental Activities of Daily Living (IADL), perceivedsocial support, perceived health status, and post-stroke depression, (3) to examine the in¯uence ofeight possible intervening socio-demographic andstroke-related variables on quality of life: age, gender,marital status, education, income, living arrange-ments, stroke type, and localization of lesions, and (4)to determine which combination of all the variablesmeasured best predicts quality of life for long-termsurvivors of stroke.

Method

Sample

A cross-sectional design was used to determine therelative impact of functional impairment, perceivedsocial support, perceived health status, and post-stroke depression on quality of life of stroke survi-vors. Participants were recruited from 433 geriatricstroke survivors discharged from a large rehabilita-tion hospital in the province of Alberta, Canada be-tween 1993 and 1995 with their ®rst diagnosis ofstroke. Out of 433, 120 survivors met the inclusionand exclusion criteria. Out of these 120, initially 55consented to participate in the study. Of these ®fty-®ve, 50 were able to fully participate in the datacollection process. To be included in the study, par-ticipants must have sustained their ®rst stroke as anadult between the ages of 60 and 85 years and beendischarged for 12 to 36 months from the rehabilita-tion facility. Although research has shown that familycaregivers [42] or signi®cant others [43] can provideuseful information on QOL among non-communi-cative stroke survivors, it was decided to limit thestudy to survivors who could communicate forthemselves. Because the data collection process re-quired a high level of cognition and communicability,stroke survivors with severe cognitive impairmentand receptive or expressive aphasia were excludedfrom the study. All participants were screened forsevere cognitive impairment using the Mini-MentalState Examination (MMSE) [44, 45] at the initialinterview. Study participants' scores on the MMSEranged from 24 to 30 (Mean � 27:3, SD � 2:10).Four participants who scored below 24 on theMMSE, which indicated severe cognitive de®cits,were not included. Stroke survivors were also ex-cluded on the basis of a history of psychiatric illnessbecause of its possible confounding in¯uence onquality of life. Likewise, stroke survivors with otherdegenerative neurologic diseases that could interferewith BADL or IADL function were excluded. How-

295

ever, study participants were not excluded on thebasis of localization of brain lesions or type of strokebecause the in¯uence of these factors has not beenfully established.

Data collection

This research utilized face to face data collection witha number of standardized instruments. Most of thedata was collected cross-sectionally during partici-pant interviews. However, supplementary data suchas stroke survivors' neurological status was retrievedretrospectively from the hospital's medical recordsand veri®ed at the time of data collection. All ques-tionnaires were administered in the homes of partic-ipants. All interviews were carried out by theprincipal investigator, lasted approximately twohours, and were completed in a maximum of twohome visits.

Instruments

Quality of life was measured using the Ferrans &Powers Quality of Life Index (QLI-Stroke Version).The QLI is a 72-item measure composed of two parts:the ®rst part measures satisfaction with various do-mains of life, and the second part measures the im-portance of the same domains to the participants [46].It covers four major underlying domains: health andfunctioning, socioeconomic, psychological and spiri-tual, and family. Participants are required to respondto the items on a six-point Likert scale in which thecategories range from ``very satis®ed'' to ``very dis-satis®ed,'' and ``very important'' to ``very unimpor-tant'' [47]. The range possible for the overall score is 0to 30, with higher scores indicating higher quality oflife [47]. Test±retest correlations of 0.87 with a two-week interval and 0.81 with a month interval dem-onstrate the stability of this measure [36, 45, 47].High internal consistency of the QLI is supported byCronbach's a coe�cient of 0.91 for the entire in-strument [36], and its content, construct, and criteri-on validity are satisfactory [14, 47, 48].

Functional status was measured using the 18 itemsof the Functional Independence Measure (FIM),which is a comprehensive scale that includes a widerange of BADL items concerning mobility, socialbehavior, communication, and cognitive impair-ment. Measurement of each item ranges from com-plete dependence (1) to complete independence (7)and the possible total scores range from a minimumof 18 to a maximum of 126 [49]. High internalconsistency of the FIM was supported by aCronbach's a of 0.92 reported for 86 stroke survi-vors [36]. Further, convergent validity was supportedby a su�cient correlation (Pearson correlationcoe�cient of 0.80) between the FIM and theFrenchay Activities Index (FAI) for thirty-eightstroke survivors [50].

The FIM, although an adequate measure ofBADL, does not assess the patient's ability to per-form complex activities (IADL) such as housekeep-ing, recreation, hobbies, and social interaction.Because limitations in IADL may a�ect quality of lifeconsiderably, this variable was further measuredusing the FAI [51]. The FAI was chosen because thisinstrument was developed speci®cally for use withstroke patients. The FAI measures lifestyle in termsof more complex physical activities and social func-tioning. It rates the frequency with which respondentsperform 15 activities (e.g., gardening, washing dishes)that have been content-validated for application tothe stroke population. The FAI is reported as ade-quate for internal consistency, Cronbach's a coe�-cients of 0.78±0.87 [52]. Convergent validity issupported by a substantial correlation between theFAI and disability scores of the Barthel Index [52]and between the physical activities subscales of theShort Form 36 Health Survey questionnaire and theAdelaide Activities Pro®le which is based on the FAI[53].

Perceived Social Support was assessed using theSocial Support Inventory for Stroke Survivors (SSI-PAD). The SSIPAD, a self-report inventory in theform of a structured interview, was developed spe-ci®cally for use with stroke patients [54]. The SSI-PAD was constructed to measure the supportperceived from ®ve sources (personal ties, friends,community, group, and professional ties). Each of the®ve source sections includes 15 questions that cor-respond to dimensions of quality (reciprocity, de-pendability, and closeness), quantity (frequency andintensity), and satisfaction (satisfaction with bothquality and quantity). Scores are calculated in each ofthe ®ve source categories by summing quality,quantity, and satisfaction questions within the cate-gory. Total scores are generated by summing thethree informal support categories (personal, familyand close friends, and community individuals). Theinventory also addresses concerns unique to thestroke population, such as change in available sup-port since the stroke, and the need for both instru-mental and emotional support. The advantages ofthis instrument include high content validity and in-ternal consistency [54], and multi-dimensional scalingthat is consistent with theoretical frameworks forsocial support.

Perceived health status (PHS) was measured using asingle age-referenced health question answered on a5-point continuum, ``excellent, good, fair, poor, bad''(with 5 being excellent). Self-reported health is deemeda strong and valid predictor of quality of life in theelderly population residing in the community [55].

Post-stroke depression was measured using theCenter for Epidemiologic Studies Depression Scale(CES-D) [56]. The CES-D Scale is a self-rating scaledesigned for use in community surveys to measure theseverity of depressive symptomatology. The CES-D

296

Scale is a 20-item questionnaire investigating per-ceived mood and level of functioning within the past7 days. Scores range from 0±60, with higher scoresindicating increasing severity of depression. Scores of16 or higher are considered indicative of depression.Shinar et al. [57] have examined the inter-observerreliability and validity of the CES-D Scale as ameasure of depressive symptomatology in strokepatients and have reported that the CES-D Scale hasan inter-rater reliability of 0.76, p < 0:001. Further,the CES-D scale scores are signi®cantly correlatedwith the other depression measures such as HamiltonDepression Scale, Zung Depression Scale and PresentState examination (r � 0:57, r � 0:65, and r � 0:72respectively, p < 0:002). Thus, the CES-D Scale wasfound to be reliable and valid as a screening tool forassessing depression in stroke patients.

Data was also collected on eight sociodemographicand stroke-related variables that might intervene toin¯uence QOL including: age, gender, marital status,education, income, living arrangements, stroke typeand localization of lesions.

Statistical analysis

Means and standard deviations, or percentages, wereused to describe QOL and possible predictor or in-tervening variables. Pearson's r was used to describethe relationship between QOL and both possiblepredictors and intervening variables and the t-testwas used to examine whether any of these relation-ships were statistically signi®cant. Finally stepwisemultiple regression analysis was used to determinethe amount of variance in QOL explained by thepossible predictor and intervening variables. Thevariable entry and removal criteria were set atthe conventional alpha level p < 0:05.

Results

Sample characteristics

The average age of the ®nal sample at the time of thedata collection was 75 years ( SD � 6:08 years, Range= 63±86 years). The majority of participants weremale (58%), married (68%) and lived with theirspouse or another family member (66%). The ma-jority of participants (74%) had incomes between$20,000 to $40,000 per year or more, and the majority(56%) had graduated from high school or had somepost secondary education.

The majority of the participants (55%) had expe-rienced a right brain lesion. The relatively highnumber of right brain lesions was not unexpected,because the exclusion criteria related to aphasiaeliminated many participants with left brain lesions.Regarding stroke type, the majority of the survivorsin this sample had su�ered from infarcts (74%).

Quality of Life and its predictors

Descriptive statistics on the QOL and possible pre-dictor variables are provided in Table 1. The meanoverall QOL score as measured by the QLI was 18.5,indicating moderate quality of life, and was lowerthan those of other stroke studies [36, 48]. The meanscore for the health and functioning subscale of theQLI was the lowest of the four subscales (17.8);whereas, the family subscale had the highest mean(24.9). Slightly higher but similar ®gures were re-cently reported by King [27] who found that the meansubscale scores of the QLI ranged from 21.5 (healthand functioning) to 26.9 (family).

The mean FIM score of 104.7 indicated that onaverage people in this study were quite independentin their basic activities of daily living. The frequencydistribution for the FIM values was negativelyskewed, since almost two thirds of the sample had ascore above 90/126. FAI (IADL) scores averaged31.2, indicating at least moderate independence ininstrumental activities of daily living.

As for the SSIPAD, total scores for quality of so-cial support averaged 55.2 and total scores forquantity of social support 14.2, both indicatingmoderate support. The most frequently reportedsources of informal support were from children(32%), spouse (25%), and friends (22%). The mostfrequently identi®ed types of support were emotionalsupport (52%) followed by instrumental support(41%) and informational support (7%).

Mean PHS reported was 3.3, with 42% of partici-pants reporting ``good'' or ``excellent,'' but more thanhalf (58%) of the sample reporting ``fair,'' ``poor,'' or``very poor'' health.

Table 1. Descriptive statistics on QOL possible predictor

variable measures

Instrument Mean Standarddeviation

Range

Overall Quality of LifeIndex

18.5 2.5 12.1±22.6

Health and Functioningsubscale

17.8 6.1 4.5±28.9

Socioeconomic subscale 23.2 4.4 8.4±30.0

Psychological/Spiritualsubscale

21.5 5.9 2.5±30.0

Family subscale 24.9 6.4 3.0±30.0

Functional IndependenceMeasure

104.7 17.2 63±126

Frenchay Activities Index 31.2 10.0 9±56

SSIPADTotquant 14.2 3.2 6±20Totqual 55.2 9.9 33±69

Perceived Health Status 3.3 0.86 1±5CES-D Scale 12.2 7.9 0±37

297

The average score on the CES-D was 12.2. Scoresranged from 0 to 37; hence, the frequency distributionof the CES-D scale values showed a notable disper-sion. Sixteen participants scored at or above 16,which indicates that the prevalence of post-strokeclinical depression among people in this study was32% . This rate is similar to those of other stokestudies in the literature [6, 36] but is higher than thatof the general population [56].

Table 2 provides correlation coe�cients for possi-ble predictors that were signi®cantly related to theoverall QOL score. The FIM, FAI, Total qualitySSIPAD, PHS and CES-D scores were all positivelycorrelated with QOL. Only the Total quantity SSI-PAD score was not associated with QOL.

Marital status and Income were two interveningsocio-demographic variables that were signi®cantlyassociated with overall QOL score. Being marriedwas positively related to the overall QOL (r � 0:29,p < 0:05), as was income (r � 0:30, p < 0:05).Age, gender, education, living arrangements, stroketype, and localization of lesions did not correlatesigni®cantly with QOL.

Study variables showing an individual statisticallysigni®cant correlation with QOL were simultaneouslyconsidered for selection into the stepwise multipleregression analysis. As a group, depression, maritalstatus, quality of social support, and functional status(IADL) were highly signi®cant predictors of QOL.Depression was entered into the equation ®rst andaccounted for 32% of the variance in quality of life.When all remaining variables were entered and thee�ect of depression removed, marital status ac-counted for 15% , social support 9% and functionalstatus (IADL) 6% respectively. Together, these fourpredictors explained 60.1% of the variance in qualityof life.

Limitations

The sample used for this study was a group of strokesurvivors whose medical status, cognitive and com-munication abilities permitted participation in theinterview methods employed. The generalizability ofthe study results from this small, relatively homoge-nous sample is reduced by the entry criteria em-ployed. Further, the validity of the study results isa�ected by the refusals of large number of eligible

stroke survivors to participate in the study. Com-parisons of socio-demographic and neurologicalstatus variables between participants and non-par-ticipants at discharge could have helped to detect anybias. However, demographic and neurological infor-mation on non-participants could not be retrievedfrom the hospital records due to con®dentiality reg-ulations. Because of the cross-sectional approach todata collection the causal nature of some relation-ships observed in this study is not clear. In particular,poor quality of life may cause depression as opposedto the reverse.

Discussion

This study aimed at identifying those factors thata�ect QOL following stroke. Based on the ®ndings,rehabilitation professionals should consider the fol-lowing observations when o�ering an intervention inrehabilitation settings.

Community-based stroke survivors like the ones inthis study report a relatively low overall QOL, this isnot surprising given the impact of stroke on an in-dividual's performance. These ®ndings are consistentwith reports by other investigators [11, 22, 58], al-though they contrast with those of King [36] andFerrans [48]. The health and functioning subscale ofthe QLI generated the lowest mean subscale scores,this was not unexpected because close to two thirds ofstroke survivors su�er some degree of physiologicaldi�culties following stroke. In addition, these survi-vors 1±3 years post-stroke reported low satisfactionwith, but placed high importance on, speci®c qualityof life items related ``ability to get around'', ``abilityto do things'', ``not having a job'', ``usefulness toothers'', and ``stress''. Low reported QOL and em-phasis on health imply that long-term psychosocialand physical management of stroke would be bene-®cial to survivors during post-discharge periods.Further, the vast majority of the participants hadretired, and a subsample of 42% reported dissatis-faction or slight satisfaction with leisure time activi-ties which is congruent with other stroke studieswhere a decline in leisure activities was also reported[6, 21, 22, 59]. E�ective rehabilitation and follow-upintervention should focus on identifying those envi-ronmental factors that either foster or impede strokesurvivors' participation in their leisure activities. Thiskind of focus on the contributions of leisure toquality of life would allow stroke survivors to rede-®ne the meaning of their optimal productivity and toexplore new avenues.

There is evidence that even 1±3 years after dis-charge from the hospital, stroke survivors continue tosu�er from depression and may bene®t from a long-term intervention to counter such psychological dis-tress. In this study, participants who reported apoorer quality of life experienced greater depression.

Table 2. Possible predictor variables related to QOL

FIM(BADL)

FAI(IADL)

SSIPADTotal

quality

PHSHealth

CES-DDepre-

ssion

Correlationcoe�cients

0.46b 0.41b 0.33a 0.48c )0.50c

a p <0.05; b p <0.01; c p<0.001.

298

Further, depression was the single best predictor ofquality of life, and it contributed as much as 32% ofthe variance in QOL. This ®nding on depression isnotably similar to the report by King [36], who founddepression to be the strongest predictor of quality oflife, and by other stroke studies which reported theimportance of depression to quality of life [6, 11, 60,61]. In light of this relationship, rehabilitation pro-fessionals should incorporate psychological assess-ment and counseling into their practice when treatingstroke patients. Evaluating patients' emotional re-sponses to a stressful event such as a stroke is criticalin understanding its potential e�ects on their lives[62]. Further, a complete adjustment to stroke willprobably not happen during the brief period of strokerehabilitation. Therefore, when providing psycho-logical intervention in rehabilitation, emphasisshould be placed on long-term intervention that in-cludes pre/post-discharge counseling. Long-term fol-low-up is advisable as it may help prevent emotionalbreakdowns and subsequently decreased quality oflife.

Quality of social support from a close personalrelationship accounted for 9% of the variance inquality of life among stroke survivors in this study.This con®rms the ®ndings of other stroke studies,indicating that persons who report greater socialsupport report higher quality of life. It should also benoted that a higher quantity of social support was notrelated to higher quality of life, which is also consis-tent with reports by other investigators [61, 63±66]. Inthis study, family caregivers provided a large portionof social support, and participants'perception of ad-equate support from a key personal relationship(usually with a spouse) was related to higher qualityof life. Additionally, the vast majority of participantsreported that they were not capable of reciprocatingthe support they received from their family members.Consequently, in order to help families continue in asupportive role, early development and implementa-tion of family support intervention in rehabilitation isof paramount importance to assist family caregiversin developing new coping strategies or strengtheningcoping skills. Such programs are valuable becausethey give the family members the opportunity to takecare of their relatives at home without becomingphysically and emotionally exhausted. In particularin this study, being married accounted for nearly16% of the variance in quality of life. The ®nding ofthe signi®cant relationship between marital statusand quality of life is consistent with that of Osberg,DeJong, Haley, Seward, McFinnis, and Germaine[67] and supports the current view that, in general,stable marital status is positively associated withquality of life. The ®ndings of this current study in-dicate that functional status is not as strong a pre-dictor of quality of life as psychosocial variables suchas depression and social support. Only IADL ex-plained about 6% of the variance in quality of life.

The lack of a relationship between BADL and QOLis consistent with observations made by other in-vestigators who have reported that depression oranxiety decrease QOL as-much-as or more thanphysical disability [11, 36]. In other words improve-ment in functional status is important, but is not asu�cient condition for successful rehabilitation.Thus, to more e�ectively facilitate improved qualityof life, rehabilitation programs need to place moreemphasis on psychosocial issues that a�ect strokesurvivors' life as a whole.

The two strongest predictors of quality of life(depression & social support) in this study representstroke survivors' perceptions of their life situations.This ®nding indicates that individual perception isintegral to the concept of quality of life. Thus, inrehabilitation, the subjective experiences of strokesurvivors, rather than the objective aspects of theirlives, should become a judge of satisfactory quality oflife following stroke. Therefore, in designing an ef-fective post-stroke rehabilitation program, rehabili-tation professionals need to recognize that strokesurvivors must be involved as partners in the processand be allowed to identify their own problems andplan their solutions.

The socio-ecological concept of health emphasizesthat barriers to occupational function are the resultof disabling environments. Rehabilitation personnelhave been challenged to integrate psychosocialconcerns into clinical practice. By integrating thesocio-ecological concept of health into their practice,rehabilitation therapists can further contribute to theoccupational health of stroke survivors. These chan-ges in the ®eld of rehabilitation will not happensmoothly or without con¯ict. Such changes do,however, represent an alternative and positive visionfor the future, a future in which the individuals withpermanent disabilities caused by stroke, along withtheir families, can sustain healthy lifestyles as con-tributing members of our communities.

Acknowledgement

This research was supported by a grant from theGlenrose Rehabilitation Hospital, Edmonton,Canada.

References

1. Petrasovits A, Nair C. Epidemiology of stroke in

Canada. Health Reports 1994; 6(1): 39±44.2. Duncan PW. Stroke disability. Physical Therapy, 1994;

74(5): 399±407.

3. Farzan DT. Reintegration for stroke survivors: Homeand community considerations. Nursing Clinics ofNorth America 1991; 26(4): 1037.

4. Biegel D, Sales E, Schulz R, Rau MT. Caregiving in

stroke. In: Biegel D, Sales E, Schulz R, eds, Family

299

Caregiving in Chronic Illness. London: Sage Publica-tions Inc, 1991.

5. Schultz SK, Castillo CS, Kosier JT, Robinson RG.Generalized anxiety and depression. Assessment over 2years after stroke. American Journal of Geriatic Soci-ety 1997; 5(3): 229±237.

6. Niemi M, Laaksonen R, Kotila M, Waltimo O. Qualityof life 4 years after stroke. Stroke 1988; 19: 1101±1107.

7. Nydevik I, Hulter-AÊ sberg K. Sickness Impact Pro®le:

A three year follow-up. Scandinavian Journal of Pri-mary Health Care 1988; 10: 284±289.

8. Renwick R, Brown I, Nagler M. Quality of Life in

Health Promotion and Rehabilitation: ConceptualApproaches, Issues, and Applications. ThousandsOaks, CA: Sage Publications Inc, 1996.

9. Law M. 1991 Muriel Driver Lecture. The environment:A focus for occupational therapy. Canadian Journal ofOccupational Therapy 1991; 58(4): 171±179.

10. Jongbloed L, Crichton A. A new de®nition of disabil-

ity: Implications for rehabilitation practice and socialpolicy. Canadian Journal of Occupational Therapy1990; 57(1): 32±37.

11. AhlsioÈ B, Britton M, Murray V, Theorell T. Disable-ment and quality of life after stroke. Stroke 1984; 15:886±890.

12. Grant M, Padilla GV, Ferrell BR, Rhiner M. Assess-ment of quality of life with a single instrument. Semi-nars in Oncology Nursing 1990; 6(4): 260±270.

13. Fallow®eld L. The Quality of Life: The Missing Mea-

surement in Health Care. London: Souvenir Press,1990.

14. Ferrans, CE. Quality of life: Conceptual issues. Semi-

nars in Oncology Nursing 1990; 6(4): 248±254.15. Goodinson SM, Singleton J. Quality of life: A critical

review of current concepts, measures and their clinical

implications. International Journal of Nursing Studies1989; 26(4): 327±341.

16. Calman, KC. De®nitions and dimensions of quality of

life. In: Aaronson NK, Beckmann J, eds, The Qualityof Life of Cancer Patients. New York: Raven Press,1989.

17. Padilla GV, Present C, Grant MG, Metter G, Lipsett J,

Heide F. Quality of Life Index for patients with cancer.Research in Nursing and Health 1983; 6: 117±126.

18. Oleson M. Subjectively perceived quality of life. Jour-

nal of Nursing Scholarships 1990; 22(3): 187±190.19. Schipper H, Clinch J, Powell V. De®nitions and con-

ceptual issues. In: Spiler B. Ed, Quality of Life Assess-

ments in Clinical Trials. New York: Raven Press, 1990.20. Fiebel JH, Springer CJ. Depression and failure to re-

sume social activities after stroke. Archives of PhysicalMedicine and Rehabilitation 1982; 63: 276±278.

21. Labi MLC, Phillips TF, Gresham GE. Psychosocialdisability in physically restored long-term stroke sur-vivors. Archives Physical Medicine Rehabilitation

1980; 61: 561±565.22. Viitanen M, Fugl-Meyer KS, BernspaÈ ng B, Fugl-Meyer

AR. Life satisfaction in long-term survivors after

stroke. Scandinavian Journal of Rehabilitation Medi-cine 1988; 20: 17±24.

23. Wood-Dauphinee S, KuÈ chler T. Quality of life as re-

habilitation outcome: Are we missing the boat? Cana-dian Journal of Rehabilitation 1992; 6(1): 3±12.

24. Chappell NL. Social Support and Aging. Toronto:Butterworths Canada Ltd, 1992.

25. Cohen J, Cohen P. Applied Multiple Regression/Cor-relational Analysis for the Behavioral Sciences. Hills-dale, NJ: Lawrence Earlbaum Associates, 1983.

26. Sauer WJ, Coward RT. Social Support Networks and

the Care of the Elderly. New York: Springer PublishingCo, 1985.

27. King RB. Quality of life after stroke. Stroke 1995;

27(9): 1467±1472.28. Lin RS, Simeone WM, Ensel WM, Kuo W. Social

support, stressful life events, and illness: A model and

empirical test. Journal of Health and Social Behavior1979; 20: 108±119.

29. Siler-Wells H. Directing Change and Changing Direc-

tion: A New Health Policy Agenda for Canada. Otta-wa: Canadian Public Health Association, 1988.

30. Norland J. Pro®le of Canada's seniors: Focus onCanada. Ottawa: Statistics Canada & Prentice Hall

Canada Inc, 1994.31. Becker G. Continuity after a stroke: Implications of

life-course disruption in old age. Gerontology 1993; 33:

148±158.32. Duncan PW, Samsa GP, Weinberger M, Goldstein LB,

Bonito A, Witter DM, Enarson C, Matchar D. Health

Status of individuals with mild stroke. Stroke 1997;28(4): 740±745.

33. Spilker B. Quality of Life in Clinical Trials. New York:Raven Press, 1990.

34. Whiteneck GG. The 44th John Stanley Coulter Lec-ture. Measuring what matters: Key rehabilitation out-comes. Archives of Physical Medicine and

Rehabilitation 1994; 75: 1073±1076.35. de Haan R, Aaronson N, Limburg M, Hewer RL, van

Crevel H. Measuring quality of life in stroke. Stroke

1992; 24: 320±327.36. King RB. Quality of Life after Stroke. Unpublished

doctoral dissertation, University of Illinois, Chicago,

1990.37. Parikh R, Lipsey J, Robinson R, Price TR. A two-year

longitudinal study of post-stroke mood disorders:Findings during the initial evaluation. Stroke 1987; 14:

736±741.38. Stern R, Bachman D. Depressive symptoms following

stroke. American Journal of Psychiatry 1991; 148: 351±

356.39. Parikh R, Robinson R, Lipsey J. The impact of post-

stroke depression on recovery in activities of daily liv-

ing over a 2-year follow-up. Archives of Neurology1990; 47: 785±789.

40. de Haan RJ, Limburg M, Van der Meulen JHP, JacobsHM, Aaronson, NK. Quality of life after stroke: Im-

pact of stroke type and lesion location. Stroke 1995;26(3): 402±408.

41. Palmore E, Kivett V. Change in life satisfaction: A

longitudinal study of persons aged 46-70. Journal ofGerontology 1977; 32: 311±316.

42. Mathias SD, Bates MM, Pasta DJ, Cisternas MG,

Feeny D, Patrick DL. Use of the Health Utilities Indexwith stroke patients and their caregivers. Stroke 1997;28(10): 1888±1894.

43. Sneeuv KG, Aaronson NK, de Haan RJ, Limburg M.Assessing quality of life after stroke. The Value and

300

limitations of proxy ratings. Stroke 1997; 28(8): 1541±1549.

44. Folstein MF, Folstein SE, McHugh PR. ``Mini-MentalState'': A practical method for grading the cognitivestate of patients for the clinician. Journal of PsychiatricResearch 1975; 12: 189±198.

45. Evelyn LT, Chui HC. The Modi®ed Mini-Mental StateExamination. Journal of Clinical Psychiatry 1987;48(8): 314±317.

46. Ferrans CE, Powers MJ. Quality of life index: Devel-opment and psychometric properties. Advances inNursing Science 1985; 8(1): 15±24.

47. Ferrans CE, Powers MJ. Psychometric assessment ofthe Quality of Life Index. Research in Nursing &Health 1992; 15: 29±38.

48. Ferrans CE. Development of a Quality of Life Indexfor patients with cancer. Oncology Nursing Forum1990b; 17(3): 15±19.

49. Granger CV, Hamilton BB. The uniform data system

for medical rehabilitation report of ®rst admissions for1991. American Journal of Physical Medicine and Re-habilitation 1992; 72(1): 33±38.

50. Segal ME, Schall RR. Determining functional/healthstatus and its relation to disability in stroke survivors.Stroke 1994; 25(12): 2391±2397.

51. Holbrook M, Skilbeck CE. An activities index for usewith stroke patients. Age and Ageing 1983; 12: 166±170.

52. Schuling J, de Haan R, Limburg M, Groenier KH. TheFrenchay Activities Index: Assessment of functional

status in stroke patients. Stroke 1993; 24: 1173±1177.53. Anderson C, Laubscher S, Burns R. Validation of the

Short Form 36 (SF-36) health survey questionnaire

among stroke patients. Stroke 1996; 27(10): 1812±1816.54. McColl MA, Friedland J. Development of a multidi-

mensional index for assessing social support in reha-

bilitation. The Occupational Therapy Journal ofResearch 1989; 9(4): 219±235.

55. Mossey J, Schapiro E. Self-rated health: A predictor of

mortality among the elderly. American Journal ofPublic Health 1982; 72: 800±808.

56. Radlo� LS, Locke BZ. The community mental healthassessment survey and the CES-D Scale. In: Weissman

MM, Myers JK, Ross CG, eds, Community Surveys ofPsychiatric Disorders. New Brunswick, NJ: RutgersUniversity Press, 1985.

57. Schinar D, Gross CR, Price TR, Banko M, Bolduc PL,Robinson RG. Screening for depression in stroke pa-tients: The reliability and validity of the Center for

Epidemiological Studies Depression Scale. Stroke 1985;17(2): 241±245.

58. Kelly-Hayes M, Paige C. Assessment and psychologicfactors in stroke rehabilitation. Neurology 1995; 45(1):29±32.

59. Wide n-Holmqvist L, Pedro-Cuesta J, Holm M,

SandstroÈ m B, Hellblom A, Stawiarz L, Bach-y-Rita P.Stroke rehabilitation in Stockholm: Basis for late in-tervention in patients living at home. Scandinavian

Journal of Rehabilitation Medicine 1993; 25: 173±181.60. Kameyama M, Udaka F, Mezak T, et al. Quality of

life after stroke. Therapeutic Research, 1990; 8(11):

26±32.61. Kaun YY. Subjective Well-being of Stroke Survivors.

Unpublished doctoral dissertation, Vanderbilt Univer-

sity, TN, 1990.62. Popovich JM. Commentary on ``poststroke family

function: An evaluation of the family's role in reha-bilitation.'' Rehabilitation Nursing 1992; 17(3): 131±

132.63. Fuhrer MJ, Rintala DH, Hart, KA, Clearman R,

Young ME. Relationship of life satisfaction to im-

pairment, disability, and handicap among persons withspinal cord injury living in the community. Archives ofPhysical Medicine and Rehabilitation 1992; 73: 552±

557.64. Goodenow C, Resine ST, Grady KE. Quality of social

support and associated social and psychological func-tioning in women with rheumatoid arthritis. Health

Psychology 1990; 9: 266±284.65. Schulz R, Decker S. Long-term adjustment to physical

disability: The role of social support, perceived control,

and self-blame. Journal of Personality and Social Psy-chology 1985; 48: 1162±1172.

66. Smith JL. Quality of Life after Traumatic Brain Injury.

Unpublished master dissertation. University of Alber-ta; Edmonton, AB, 1995.

67. Osberg JS, DeJong G, Haley S, Seward ML, McGinnis

MP, Germaine J. Predicting long-term outcome amongpost-rehabilitation stroke patients. American Journalof Physical Medicine and Rehabilitation 1988; 67: 94±103.

Address for correspondence: Dr. S. Warren, Faculty of Re-

habilitation Medicine, Rm 3-48 Corbett Hall, University ofAlberta, Edmonton, Alberta, Canada T6G 2G4Fax: (780) 492 1626

301