Self-reported changes in everyday life and health of significant others of people with aphasia: a quantitative approach

ABSTRACT

Background: Changes in the everyday life and health of the significant others (SOs) of people living with stroke sequalae have been widely investigated. However, information regarding the frequency of the changes in daily life and the variables most associated with these perceived life changes is limited.

Aims: To examine the extent to which SOs of persons with aphasia (PWAs) experience changes in everyday life and health after the stroke event and how they evaluate these changes and to identify which variables are associated with these perceived life changes.

Methods & Procedures: A study-specific questionnaire about perceived changes in everyday life and health was completed by 173 SOs of PWAs living in Sweden (response rate 67.8%). The items concerned the SO’s working and financial situation, leisure time and social life, relationship with the PWA, household work and responsibility, and health and quality of life.

Outcomes & Results: The everyday life and health of the SOs were greatly affected by the stroke event. The changes the SOs experienced were mainly appraised negatively. The relationships with the PWA and immediate family were least affected by the stroke. The perceived existence and severity of physical, cognitive, and language impairments of the PWA were the variables most strongly associated with the everyday life situation of the participants. The sex and age of the participants and the nature of the relationship with the PWA were only marginally associated with the experience of the situation.

Conclusions: SOs’ perceptions of the PWA’s stroke-related disabilities and need for assistance may be a key factor in identifying SOs who may require support and guidance to help them cope with their new life situation.

KEYWORDS:

• Aphasia
• significant others
• everyday life
• health
• quality of life

Introduction

Significant others’ (SOs’) well-being and knowledge about stroke are known to affect the rehabilitation outcomes of persons living with post-stroke sequalae. Tsouna-Hadjis et al. (2000) reported that family support significantly improved the functional status of patients during post-stroke recovery, and Barskova and Wilz (2007) found that both the cognitive and emotional recovery of stroke patients were positively influenced by their partners’ attitudes and health. However, due to their often challenging situation, SOs may require additional support beyond information about stroke to effectively support the patient after stroke (Barskova & Wilz, 2007; Camak, 2015).

It has been established in the literature that the SOs of persons with post-stroke sequalae experience reduced psychological health, such as increased prevalence of stress (Carod-Artal et al., 2009; Rigby et al., 2009), anxiety, depression (Forsberg-Wärleby et al., 2004; Kruithof et al., 2016), and reduced quality of life (Jönsson et al., 2005; Rombough et al., 2007; Schlote et al., 2006; White et al., 2004). Furthermore, increased marital problems (Visser-Meily et al., 2009), a reduced social network, and time for leisure activities (Glozman, 2004; Greenwood & Mackenzie, 2010b) have all been reported. In contrast, some evidence suggests that SOs perceive providing care as rewarding and feel appreciated and closer to the person after the stroke (Mackenzie & Greenwood, 2012).

In a review of quantitative studies, Greenwood et al. (2008) found that several factors were associated with the well-being of SOs of people with stroke sequalae, including the psychological characteristics of SOs (such as coping strategies and self-esteem) and their physical health. The disability and dependency of stroke survivors also influenced the situation of the SOs (Franzen-Dahlin et al., 2008). Greenwood et al. (2008) suggested other possible factors that might be of importance, including the impact of role changes or cohabitating. These findings have been supported in more recent research (Jessup et al., 2015). Greenwood et al. (2008) also called for analyses of subgroups to determine how demographic factors (such as age and sex), as well as impairments secondary to stroke (such as aphasia), may affect the well-being of SOs. Moreover, they emphasised that previous research had focused on the objective assessment of disability, rather than disability as perceived by the SOs. However, the perception of the SOs seems to be important. According to Franzen-Dahlin et al. (2008), SOs’ perception of the person with stroke sequalae’s need for assistance with daily activities, rather than the functional disability per se, predicted the mental health of the SOs.

Although not conclusive, research findings have indicated that the experiences of male and female SOs differ. In general, female SOs are reported to perceive their situation as more burdensome (Jessup et al., 2015), reporting more anxiety (Carod-Artal et al., 2009) and more depressive symptoms (Jessup et al., 2015) than male SOs. Women have also been reported to perceive their partners who had had a stroke more negatively than did men (Croteau & Le Dorze, 2001). Greenwood et al. (2008) found that younger partners of persons with stroke sequalae perceived the situation as more taxing than did older partners, while McCullagh et al. (2005) reported that caregivers’ quality of life was negatively correlated with caregivers’ age. It has also been suggested that the experiences of SOs may differ depending on the mood of the person post-stroke and changes in personality or behaviour secondary to stroke (Forsberg-Wärleby et al., 2004; Persson et al., 2015).

Although not explicitly stated, it is to be presumed that the literature regarding SOs of people with post-stroke sequalae includes data from SOs of individuals with aphasia secondary to stroke. Research focusing on this specific subgroup of caregivers, SOs of persons with aphasia (PWAs), has mainly been investigated via qualitative studies (e.g., Blom Johansson, 2006; Le Dorze & Brassard, 1995; Grawburg et al., 2013; Kitzmüller et al., 2012; Nyström, 2011; Sjöqvist Nätterlund, 2010; Winkler et al., 2014). In these studies, the SOs describe their situations as involving increased responsibilities and workloads due to role changes. Moreover, communication difficulties and a reduced social and recreational life are characteristic of their situations. Furthermore, SOs report feelings of being confined, living under difficult circumstances, and that everyday life necessitates careful planning. When comparing the situation of SOs of persons with stroke, with and without resultant aphasia, the situation appears to be more challenging for the SOs of PWAs (Bakas et al., 2006; Draper & Brocklehurst, 2007). This suggests that the communication difficulties caused by aphasia compound an already difficult situation. In addition, there are some indications that the situation of SOs of PWAs does not improve with time to the same extent as has been reported for SOs of stroke survivors without aphasia (Blom Johansson et al., 2012).

The body of previous qualitative research allows for a reasonably good understanding of how the everyday life and health of the SOs of PWAs changes. However, it is still not well known how prevalent the life and health changes are; that is, what percentage of SOs of PWAs perceive these changes. Furthermore, there is still limited or unclear knowledge regarding how these changes are influenced by demographic factors of the SOs and the stroke-induced impairments of the stroke victim, as called for by Greenwood et al. (2008). This information is important when designing and providing rehabilitation services for PWAs and their SOs (McGurk & Kneebone, 2013). The primary aim of this study was thus to investigate the extent to which the SOs of PWAs perceive changes in everyday life and health after the stroke event, as well as how they evaluate these changes. A secondary aim was to investigate the relationships between these perceptions and demographic factors, time post-stroke, perceived disabilities, and the PWA’s need for assistance.

Materials and methods

Participants

The participants were SOs of PWAs, preferably cohabitating partners. If no such partner was available, the SO was defined as a person who communicates with the PWA on a regular basis (at least once a week) and helps the PWA with societal contacts. Participants were recruited from across Sweden by local speech-language pathologists (SLPs) or by a contact person at a local aphasia association. The participants were included if they were SOs of a person with aphasia secondary to stroke who was at least three months post-stroke onset, was aged 18 or older, was living at home or expected to return home after rehabilitation, and had no other language difficulties except for aphasia.

The participants were excluded if they or their related PWA had diagnosed dementia or any other known significant cognitive impairment, substantial non-corrected hearing or vision problems or known alcohol or drug abuse (as determined by the recruiting person).

Data collection

Data collection procedure

Heads of SLPs in all county councils of Sweden (n = 21) were approached by mail for permission to ask SLPs working with PWAs in their respective geographic areas to distribute questionnaires to SOs of PWAs. Heads of 15 county councils approved and 41 SLPs distributed 205 questionnaires to SOs of PWAs between May 2010 and April 2011. To reach participants from all across Sweden, the researchers contacted the local aphasia associations in the six county councils in which the SLP heads had declined to participate. Each aphasia association was asked to distribute ten questionnaires to SOs of PWAs to limit their extra work and to achieve an even representation from these county councils. Five of the six associations agreed to participate, resulting in the distribution of 50 additional questionnaires. After having checked that inclusion criteria were fulfilled, the regular SLP or the contact person at the aphasia association informed the PWA about the study using study-specific pictographic material to facilitate understanding. When the PWA gave his or her oral informed consent, written information about the study and a questionnaire were sent to the SO selected by the PWA. The SO mailed the completed questionnaire to the principal author in a pre-stamped envelope. It was emphasised that the SLPs should approach their patients with aphasia in a consecutive order (i.e. in the order the PWA was scheduled for an SLP session) instead of choosing which patients to ask. More information about the data collection procedure can be found in Blom Johansson et al. (2012).

Measures

Data were collected via a comprehensive questionnaire designed for a larger project that aimed to elucidate different aspects of the living conditions of SOs of PWAs, covering all domains of the International Classification of Functioning, Disability and Health (ICF, World Health Organization, 2001). The items in the questionnaire concerned everyday communication, support, and services from SLPs (Blom Johansson et al., 2012); changes in everyday life and health (present paper); perceptions of support provided by healthcare services and the community; and perceived need for additional support (forthcoming paper). The questionnaire was constructed based on the first author’s 15 years of clinical experience as an SLP providing SLP services to PWAs and their SOs and previous research on the lived experience of SOs of PWAs and their needs (e.g., Blom Johansson, 2006; Le Dorze & Brassard, 1995; Franzen-Dahlin et al., 2008), with the specific intention of capturing changes in the participants’ life situation and health, as well as their evaluation of these changes and their perceptions of their related PWA’s disabilities. In order to satisfy the requirement of covering a broad range of items without being too exhaustive, in addition to having a similar design for the items and scales within in each section, no existing questionnaires or scales (such as the Life Satisfaction Questionnaire [Carlsson & Hamrin, 2002] and the Montgomery Caregiver Burden [Montgomery et al., 2000]) were used. Instead, these instruments served as a basis for the content and wording of the items and the scales. The scales were chosen with regard to the requested information in each section. Eight SOs of PWAs piloted the questionnaire. Their responses were not included in the study. Response time varied between 30 and 80 min (mean: 59 min), excluding encouraged breaks. The questionnaire was considered to have a high face validity by the pilot participants, summarised as “A very well-composed survey. Clear questions and answer options. Gives an impression that the person who formulated the questions has met with many patients [with aphasia] and relatives.” A few minor changes, such as changing a word or word order to clarify, were made in response to their comments.

The questionnaire consisted of eight parts: sections A (demographics, 10 items), B1 (stroke-related data, 13 items), B2 (PWA’s need for assistance, 3 items), C (communication with PWA, 22 items), D (SLP services, 36 items), E (everyday life, 29 or 32 items), F (support provided by healthcare and community services, 22 items), and G (current need of support, 11 items), for a total of 146 or 149 items.

Data relevant for the present study are found in sections A, B1, B2, and E. A short description of these sections is given below. The specific items, translated into English, are presented in Appendix A. Scores on ordinal scales are indicated as 1–5.

Items in section A requested demographic data such as age, sex, education, working situation, and nature and length of relationship with the PWA.

Items in section B1 assessed the participant’s perception of the stroke-related impairments of the PWA. No existing measure of the SO’s own judgement of the severity of the stroke-induced impairments suitable for the study was found. Instead, the participant was asked to mark if they perceived the severity of aphasia as mild, moderate, or severe and if they perceived leg and arm mobility and the cognitive functions of memory, attention, initiative, executive function, time perception, and environmental orientation as impaired, and if so, the perceived severity of the impairment (mild, moderate, or severe). The items about cognitive functions were selected because they pertained to common impairments following a stroke (Helm-Estabrooks, 2002) that may affect the caregiver’s everyday life and health (Blom Johansson, 2006; Carod-Artal et al., 2009) and were presented in terms that were thought to be comprehensible to laypeople. Perceived depressive mood (1 item) and personality change (1 item) were assessed by a yes–no question. The examples of personality changes were based on impairments that were frequently mentioned in previous qualitative research (e.g., Blom Johansson, 2006) and that affect the everyday life and health of caregivers.

Items in section B2 were partly assessed via an abbreviated version of a scale constructed by Åberg-Wistedt et al. (1995) that was originally developed to evaluate the need for assistance in patients with schizophrenia but that has also been used in stroke patients (e.g., Franzen-Dahlin et al., 2008). The original version consisted of 15 suggested activities with which the patient may need help. The eight-item modified version consisted of six of these 15 activities which are often affected in persons with stroke: help with indoor and outdoor transportation, toilet visits, dressing, mealtimes, shopping and running bank errands. Two communication-demanding activities were also added: talking with friends and acquaintances and contact with healthcare professionals and public authorities. The items were answered using a yes/no format. The total score (number of “yes” responses) ranged from 1 to 8 (the higher the score, the greater the perceived need for assistance). If there was no perceived need for assistance with these activities, the score was 0.

The items in section E concerned the working situation (4 items), the financial situation (2 items), leisure time (4 items), one’s own time and social life (7 items), relationship with the PWA (5 [partner] or 2 [non-partner] items), household work (4 items), and health and quality of life (6 items). Twenty-five items had fixed-response alternatives and seven were open-ended questions. In general, each domain begins with an item for which the SOs mark the perceived change from before the stroke event on a five-point ordinal scale from 1 (substantially worsened/decreased) to 5 (substantially improved/increased). The SOs then marked how they perceived this change from 1 (very negative) to 5 (very positive) on a five-point ordinal scale. Finally, the SOs were asked to describe what they perceived as the cause of the change in their own words. They were also asked to mark whether they associated any changes in their own health and quality of life with the stroke event.

For those SOs recruited by an SLP, the SLP performed the same rating of aphasia severity as the SOs did, based on the SLP’s own clinical judgement, in order to compare the aphasia severity rankings between SLPs and SOs. This ranking was not performed by the contact persons at the aphasia associations since they were laypeople and did not have the professional competence to do so.

Data analysis

Section E and parts of sections B1 and B2 were tested for internal structure by conducting principal component analyses with direct oblimin rotation. For sections B1 and B2, Kaiser-Meyer-Olkin’s measure of sampling adequacy was 0.851 and Bartlett’s test of sphericity was significant (p <.001), indicating good factorability. The analysis resulted in two components explaining 67.8% of the variance: 1) Cognitive function and 2) Physical function and need for assistance (Appendix B). When tested for internal consistency (Cronbach’s alpha) the results were α = .874 (component 1) and α = .856 (component 2). Component scores were used as independent variables in subsequent analyses.

For section E, Kaiser-Meyer-Olkin’s measure of sampling adequacy was 0.882 and Bartlett’s test of sphericity was significant (p < .001). The analysis resulted in three components explaining 60.0% of the variance: 1) Time, 2) Health and quality of life, and 3) Social life (Appendix C). When tested for internal consistency (Cronbach’s alpha) the results were α = .873 (component 1), α = .824 (component 2), and α = .808 (component 3). The three components were used as dependent variables in the statistical analyses. For the factor analyses, missing data were replaced with the mean.

To investigate the relationship between independent and dependent variables, appropriate statistical tests were selected according to the data type. For interval data (age, time post-stroke onset, cognitive function, and physical function and need for assistance), Pearson’s product-moment correlation was used. For ordinal data (perceived aphasia severity), Spearman’s rank correlation was used. For nominal data (sex, cohabitating with PWA or not, and the two emotional variables [depression and personality change]), Pearson’s chi-squared test was used. For chi-squared tests, each dependent variable (Time, Health and quality of life, and Social life) was divided into two categories: above and below zero. Cohen’s (1988) guidelines for the interpretation of correlation coefficients were used. The alpha level was set at .05. Statistical tests used are indicated in the tables of results. Correlation analyses were calculated for the valid data for each item. Missing data varied from 1 to 12 per item (mean 3.8; 2.2%). The mean percentages of missing data for each questionnaire section were: A (demographic data, mean 0.9%), B1 (stroke-related issues, mean 3.0%), B2 (perceived need of assistance, mean 1.7%), and E (perceived changes in everyday life, mean 2.5%). The item with the greatest amount of missing data (6.9%) concerned the participant’s relationship with the PWA (item E:20).

IBM SPSS Statistics version 26 was used for statistical analyses.

The answers to the open-ended questions were analysed according to the following systematic procedure of qualitative content analysis described by Graneheim and Lundman (2004). First, the answers to a specific question (the unit of analysis) were read through several times to provide an overall sense of the content. Each answer was considered as a condensed meaning unit, which was labelled with a code describing the content. The codes were then sorted into categories and subcategories based on differences and similarities. The number of participants who provided an answer to a specific open-ended question is indicated by (n = [no of participants]) in the Results section. During the analysis process, great importance was attached to awareness and reflexivity in order to secure rigor and ensure that data were interpreted according to the intent of the participants. The first author conducted the analysis and the results were then reviewed by co-authors. The few uncertainties were discussed until consensus was reached.

Ethical considerations

The study followed ethical regulations and guidelines according to Swedish Law 2003:460 (Codex) and was reviewed by the Regional Ethical Review Board in Uppsala, Sweden (Dnr 2010/071). The PWAs gave their permission to send the questionnaire to SOs of their choice. The SOs received written information about the study and were informed that participation was voluntary and that their answers would be anonymous. Informed consent was considered to have been given by the SOs when they returned the completed questionnaire.

Results

A total of 173 completed questionnaires were received (response rate 67.8%). Of the participants (SOs), 66.9% were female and 33.1% were male. The mean age was 64.2 years (SD: 10.9; range 33–87 years). Most participants (85.5%) were a cohabitating partner of the PWA, and the mean length of the relationship was 40.3 years (SD: 14.9; range 3–70 years). Of the PWAs, 38.4% were female and 61.6% were male (mean age 67.1 years; SD: 11.1; range 38–97 years). The mean time since stroke onset was 41.3 months (SD: 54.9; range 3–312 months). See Table 1 for descriptions of the participants and the PWAs.

Table 1. Characteristics of participants (significant others; n = 173) and persons with aphasia (n = 173)

	Participants		Persons with aphasia
	N	%	N	%
Age
< 65	80	46.5	61	35.5
≥ 65	92	53.5	111	64.5
Missing value	1		1
Sex
Female	115	66.9	66	38.4
Male	57	33.1	106	61.6
Missing value	1		1
Education (years)
≤ 9	47	27.6	59	34.9
10–12	60	35.3	61	36.1
> 12	63	37.1	49	29.0
Missing value	3		4
Occupational situation
Full-time work	39	22.7	2	1.2
Part-time work	26	15.1	4	2.3
On sick leave	2	1.2	34	19.8
Retired	99	57.6	128	74.4
Job-seeking	6	3.4	4	2.3
Missing value	1		1
Relationship with PWA
Cohabitating partner	147	85.5
Non-cohabitating partner	4	2.3
Father/mother	1/5	0.6/2.9
Son/daughter	2/11	1.1/6.4
Brother/sister	1	0.6
Close friend	1	0.6
Missing value	1
Time since aphasia onset (months)
3–6			33	19.3
7–12			37	21.6
13–24			35	20.5
25–60			28	16.4
> 60			38	22.2
Missing value			2
Perceived severity of aphasia (SO/SLP^a)
Mild			23/32	14.0/23.4
Moderate			68/57	41.5/41.6
Severe			73/48	44.5/35.0
Missing value			9/2
Perceived physical impairment (0–6)
None (0)			25	14.9
Mild (1–2)			27	16.1
Moderate (3–4)			38	22.6
Severe (5–6)			78	46.4
Missing value			5
Perceived cognitive impairment (0–18)
None (0)			4	2.4
Mild (1–6)			73	43.5
Moderate (7–12)			61	36.3
Severe (13–18)			30	17.9
Missing value			5
Perceived depressed mood
Yes			72	43.9
No			92	56.1
Missing value			9
Perceived personality change
Yes			134	81.7
No			30	18.3
Missing value			9
Perceived need for assistance (0–7)
None (0)			19	11.2
Low (1–2)			35	20.6
Moderate (3–5)			77	45.3
High (6–7)			39	22.9
Missing value			3

Notes:

^aOnly for the SLP sample (n = 139)

PWA = person with aphasia, SO = significant other, SLP = Speech Language Pathologist

Where N does not add up to 173, there are missing values (participants, n = 1–3, PWA n = 1–9).

Participants recruited from the SLP services (SLP sample, n = 139) and local aphasia associations (AA sample, n = 34) were compared in terms of age, sex, perceived aphasia severity, nature of relationship with the PWA, and time since stroke onset. One statistically significant difference was found: Mean time post-stroke onset was longer in the AA sample (103.5 months, SD: 77.9; range 4–312 months) than in the SLP sample (25.9 months, SD: 33.0; range 3–192 months; t(169) = −8.9, p < .001).

The two ratings of aphasia severity of the SLP sample (one by the participant and one by the SLP) were positively correlated (r = .695, p < .001). In general, the participants rated the aphasia as being as severe as or more severe than did the SLPs.

Perceived changes in everyday life

For almost half of the participants, their financial situation had changed for the worse due to decreased income (due to the participant or the PWA not being able to work) or increased expenses (such as medical consultations, medications, transportation services, or home help services). Six participants reported improved finances due to reduced expenses (for example, the PWA had stopped smoking or driving; Table 2).

Table 2. Number of participants (in percentages) perceiving changes (or not) on questionnaire items related to everyday life and health (E:5 to E:30) (n = 173)

Item (Section E)	Decreased or worsened			Unchanged	Increased or improved
	Substantially (%)	Somewhat (%)	Total (%)	(%)	Somewhat (%)	Substantially (%)	Total (%)
E:5 Financial situation	14	32	46	50	3	1	4
E:7 Own leisure activities	44	32	76	23	-	1	1
E:8 Leisure activities together with PWA^a	55	21	76	19	3	2	5
E:11 Time for yourself	44	35	79	20	1	-	1
E:13 Time with own friends	22	34	56	43	-	1	1
E:14 Time with family	7	17	24	57	15	4	19
E:15 Time with others (relatives, friends, neighbours)	21	32	53	40	6	1	7
E:18 Relationship with PWA	8	25	33	54	12	1	13
E:22 Household work^b	1	4	5	18	27	50	77
E:23 Responsibility at home^b	1	2	3	18	29	50	79
E:26 Physical health	12	34	46	51	3	-	3
E:28 Psychological health	12	42	54	44	2	-	2
E:30 Quality of life	23	44	67	28	5	-	5

Notes:

^aPWA = person with aphasia

^bNote that Household work and Responsibility at home are scored in the opposite way to the other items, regarding how a change was perceived; that is, increased is perceived as negative by most participants, contrary to the other items.

Of the participants who were working outside the home at the time of the stroke event (n = 89), 42 (47%) had a changed working situation. Of these participants, 33 (79%) had either given up work or reduced their working hours and 16 (38%) had changed their workplace or tasks (8 participants reported both). For those who reported a changed working situation, 41% perceived this change as negative and 23% perceived it as positive (Table 3).

Table 3. Number of participants (in percentages) who perceived changes in everyday life and health as negative, neutral or positive (n = 173)

Item (Section E)	Negative			Neutral	Positive
	Very (%)	Somewhat (%)	Total (%)	(%)	Somewhat (%)	Very (%)	Total (%)
Perception of change in …
E:3 working situation^a	15	26	41	36	10	13	23
E:9 leisure time	29	47	76	22	1	1	2
E:16 social life	21	41	62	31	4	3	7
E:24 role at home	16	34	50	41	8	1	9

Notes:

^aQuestions about the working situation were answered only by those participants who were working at the time of the stroke event (n = 89).

According to the qualitative analysis (n = 30), the change in working situation for some participants was due to reorganisation at the workplace, retirement, or their own illness. However, for the majority (n = 24, 80%), the change was due to the stroke event: The participants had to or wanted to stay at home to take care of the PWA, wanted to work near the home, or could not manage a pre-stroke level workload in addition to the increased responsibility at home. One participant had increased her working hours for financial reasons.

For most of the participants, time for leisure activities, either on their own or with the PWA, had decreased after the stroke event (Table 2). The changes in leisure time were mainly perceived as negative but were perceived as positive by 2% (Table 3). Almost 80% of the participants reported having less time for themselves (Table 2): 7% reported having no time at all for themselves; 70% reporting having some time, but not enough according to their perceived needs; and 23% felt that they had the time they needed (not shown in the table). Time spent with their own friends decreased for 56% of the participants. For just as many (57%), time spent with the family was unchanged, while almost one-fifth (19%) perceived that time with family had increased. Time spent with others (relatives, friends, neighbours) decreased for more than half of the participants (Table 2). Most participants (62%) perceived the changes in social life as negative (Table 3).

The qualitative analysis revealed that the reasons for the decreased time for leisure activities (n = 110) and social life (n = 107) were the result of the stroke and its consequences for all participants. The participants chose or felt obligated to stay at home with the PWA to provide help and support. Physical or language impairments limited the activities that were possible to achieve or in which the PWAs could participate. In addition, fatigue (of the PWA or the participant) or reduced time became barriers to social participation. In addition, family members and friends of the PWA sometimes chose to terminate contact due to communication difficulties with the PWA (as reported by the participants).

While 54% of the participants perceived the relationship with the PWA to be unchanged, 33% reported that the relationship had deteriorated and 13% reported that the relationship had improved since the stroke onset (Table 2). Approximately half of the participants who were partners of the PWA (52%) characterised the relationship as a wife–husband relationship, whereas the remaining partner participants who described the relationship as having changed described a relationship characterised predominantly as caregiver–receiver (23%), adult–child (11%), brother–sister (4%), or as a combination of these alternatives (10%). Most partner participants (81%) claimed never to have considered divorce (or separation). The remainder reported having considered divorce occasionally (16%) or frequently (3%).

According to the qualitative analysis (n = 74), the main reason for a deteriorated relationship was that the PWA had changed in a negative direction after the stroke in terms of mood (irritated, mean, depressed, egoistic, or ungrateful) and reduced physical and cognitive abilities. In addition, the communication difficulties, perceived caregiver burden, and increased responsibility negatively affected the relationship. Those who reported an improved relationship attributed this to spending more time with each other, showing a great appreciation of the partner and of life itself, and the PWA having changed in a positive direction (such as being more understanding and happier).

For most participants, responsibility increased at home and in household work (Table 2). These changes were perceived as negative by half of the participants and positive by 9% (Table 3). Reported reasons (n = 101) for increased responsibility at home and in household work were all linked to the stroke. Due to aphasia, physical and cognitive impairments, fatigue, etc., the PWAs were no longer able to share responsibility at home and take part in household chores. In addition to the participants’ usual tasks, new tasks were introduced, such as providing the PWA support with personal care. A few participants reported that their household work had actually decreased after the PWA’s stroke because the PWA was at home and was able to carry out some household chores.

Forty-six percent of the participants reported that their physical health had deteriorated after the PWA’s stroke onset (Table 2). While most participants linked their physical deterioration to the PWA’s stroke, answering that it was definitely (30%) or partly (58%) due to the stroke, others reported that it was not at all (12%) linked. Likewise, more than half reported deteriorated psychological health and that their psychological deterioration was definitely (55%) or partly (41%) associated with the stroke, with only 4.0% answering that it was not at all linked. Two-thirds reported that their quality of life had decreased (Table 2), and the vast majority associated this deterioration with the stroke: 58% answered that it was definitely associated, 39% answered that it was partly associated, and only 3% answered that it was not at all associated with the stroke.

Relationship between independent variables (demographic factors, time post-stroke, cognitive function, and physical function and need for assistance) and dependent variables (time, health and quality of life, and social life)

In general, there were few and weak relationships between demographic factors and the dependent variables, while there were more and stronger relationships between the dependent variables and the domains of perceived impairments of and need for assistance by the PWAs. The exception was being a cohabitant of the PWA, which was significantly related to the Time component (i.e. time restrictions were more commonly reported by cohabitating SOs). There was also a tendency for female participants to report a worse health and quality of life than males (Table 4).

Table 4. Relations (r) between independent variables (age, sex of the significant other, cohabitant with PWA (or not), length of relationship, time post-stroke, perceived aphasia severity, perceived depressed, perceived personality changed, Cognitive function and Physical function and need for assistance), and dependent variables (Time, Health and Quality of life, and Social life)

Independent variables	Dependent variables
	Time	Health and Quality of life	Social life
Age (significant other)^a	r = −.078 p =.307	r =.010 p =.896	r = −.008 p =.921
Sex (significant other)^c	p =.554	p =.057	p =.301
Nature of relationship to PWA (cohabitant or not)^c	p =.005	p =.883	p =.131
Time post-stroke^a	r =.108 p =.158	r =.054 p =.479	r = −.231 p =.002
Perceived aphasia severity^b	r = −.408 p <.001	r = −.289 p <.001	r = −.235 p =.002
PWA perceived as depressed^c	p =.740	p =.001	p =.006
PWA perceived as personality changed^c	p =.007	p =.040	p =.114
Perceived cognitive function of PWA^a	r = −.324 p <.001	r = −.297 p <.001	r = −.300 p <.001
Perceived physical function and need for assistance of PWA^a	r = −.614 p <.001	r = −.283 p <.001	r = −.187 p =.014

Notes:

^a= Pearson’s product-moment correlation – statistically significant (p <.05) moderate and high correlations (.30 ≤ r ≤ 1.0) (Cohen, 1988) are in bold.

^b= Spearman rank correlation – statistically significant (p <.05) moderate and high correlations (.30 ≤ r ≤ 1.0) (Cohen, 1988) are in bold.

^c= Pearson’s Chi-2 test – statistically significant correlations are in bold (p <.05)

PWA =person with aphasia.

Regarding the stroke-related independent variables, the perceived severity of aphasia and cognitive and physical impairment were related to all dependent variables: the more severe the impairments, the more time restriction, worse health, reduced quality of life, and restricted social life were reported. In addition, a perceived personality change of the PWA was correlated with time restriction, whereas perceived depression in the PWA was not. Instead, perceived depression in the PWA was related to a decreased social life, as was time post-stroke. The health and quality of life domain was related to all stroke-related items except time post-stroke.

Discussion

The results of this study confirm that SOs of PWAs experience considerable changes in everyday life and health after a stroke event. The results are summarised and discussed below with regard to the research questions.

Perceived changes in everyday life

Almost half of the participants who were working at the time of the stroke terminated their work or changed their working hours, workplace, or work tasks. Approximately 75% of the participants experienced a reduced recreational and social life and had less time for themselves. These changes were predominantly perceived as negative. Almost 80% took on increased responsibility for home and family and one-third felt that the relationship with the PWA had deteriorated. These results are in accordance with several previous qualitative studies (e.g., Kitzmüller et al., 2012; Nyström, 2011; Sjöqvist Nätterlund, 2010).

The extent of the changes was pronounced. The vast majority of the participants reported considerable changes to everyday life, mostly pertaining to social and recreational life, lacking time of their own, and increased responsibility/workload at home. Similar results were found by Jessup et al. (2015). It is notable that many participants in the present study perceived these changes as negative, which may have affected their ability to manage the situation and to support the PWA in a constructive manner.

Life areas that seemed more stable, for which more than half of the participants reported an unchanged situation, were time with family and the relationship with the PWA. In fact, almost one-fifth of the participants reported that time with family had increased, a finding consistent with the literature suggesting that the family becomes the main provider of socialisation after the onset of aphasia (Glozman, 2004). The relationship between the PWA and the participant did not seem as strongly affected as were other aspects of everyday life. When changes were noted, it was predominantly perceived changes in mood or personality and cognitive impairments of the PWA that seemed to affect the participant’s experience of the relationship. Emotional or cognitive impairments of the PWA have previously been reported to negatively affect SOs’ perceptions of their life situations and their ability to cope with their changed life situation (Kitzmüller et al., 2012). Although the data did not show explicit evidence of increased marital problems (Visser-Meily et al., 2009) or negative attitudes toward the PWA (Croteau & Le Dorze, 2001), the finding that some participants perceived the marital relationship as an adult–child or brother–sister relationship may indicate that such problems and attitudes existed to some extent.

Perceived changes in health and quality of life

A majority of the participants perceived that their health (physical and psychological) and/or quality of life had deteriorated and that the change was, to a large extent, associated with the stroke event. This finding is in accordance with the results of the systematic review by Rigby et al. (2009), which concluded that caregiving was associated with high rates of cardiovascular disease, general ill-health, and mortality. In addition, research into the psychological health of caregivers of stroke survivors identified that 30%–40% of the caregivers were depressed (Han & Haley, 1999), and an even higher prevalence of anxiety (approximately 50%) has been reported (Greenwood & Mackenzie, 2010a). Anxiety itself seems to be a predictor of depression (Han & Haley, 1999). Another recognised predictor of caregiver’s health status is caregiver burden (Carod-Artal et al., 2009), defined as “the weight or load carried by caregivers as a result of adopting the caregiving role” (Rigby et al., 2009, p. 285). Other risk factors for deteriorating health are fewer social contacts, the caregiver’s own physical illness, or the person who had a stroke being perceived to have cognitive impairments or exhibit abnormal behaviour (Greenwood et al., 2008; Han & Haley, 1999). All of these risk factors were reported frequently in the present study.

In general, the quality of life of caregivers of people post-stroke is reported to decrease after the stroke event (McCullagh et al., 2005; White et al., 2004). According to Schlote et al. (2006), caregivers’ health-related quality of life was far below normative values during the sub-acute phase after the stroke, and even one year after admission from hospital. Jönsson et al. (2005) found that the quality of life of caregivers of stroke patients was lower than the quality of life of the stroke patients themselves 16 months post-stroke. This finding is supported by Rombough et al. (2007), who concluded that the quality of life of caregivers of stroke survivors seems to decrease over time rather than increase. However, the trajectory of the quality of life of caregivers of persons post-stroke is not conclusive. According to a comprehensive review by Gaugler (2010), several studies have reported a negative development over time concerning the emotional, social, psychological, and physical well-being of the caregivers of persons with stroke sequalae, whereas others describe the successful adaptation of caregivers, with increased knowledge, reduced stress, decreased prevalence of depression, and increased quality of life. The present study did not identify any associations between time post-stroke onset and the health and quality of life of the participants.

The questionnaire used was constructed to cover all domains of the framework of ICF. The way a health condition of a relative affects the everyday life of SOs in the ICF domains Body Functions, Activity Limitations, and Participation Restrictions is also called “third-party disability” (World Health Organization, 2001). Grawburg et al. (2013) suggested that “third-party functioning” should be used when the consequences of the relative’s health condition are neutral or positive. The results of the present study, describing the prevalence of changes in Body Functions, Activity Limitations, and Participation Restrictions, confirm that the vast majority of SOs of PWAs can be considered to experience a third-party disability according to the definition of the WHO (World Health Organization, 2001), while it would be more appropriate to use the term “third-party functioning” for only a minority of SOs of PWAs. Regardless of the label used, the results show the need for and importance of providing clinical support to the SOs of PWAs.

Associations between independent and dependent variables

The perceived existence and severity of physical, cognitive, and language impairments of the PWA were the variables most strongly associated with the everyday life situation of the participants. The sex and age of the participants and the nature of the relationship with the PWA were only marginally associated with the experience of the situation.

No significant differences were identified between male and female participants in the present study. However, there was a tendency for female participants to report lower health and quality of life than male participants. Greenwood and Mackenzie (2010b) commented in their review that sex differences among caregivers were seldom reported. One possible explanation for this was that sex differences were less common than might be expected (Greenwood & Mackenzie, 2010b), which is supported by the results of the present study.

One of the few relationships that were identified between demographics and outcome factors was cohabitating with the PWA. Those living together are more likely to influence each other’s lives (such as role changes) than those living apart. However, in Rigby et al. (2009), evidence was mixed regarding the importance of the caregiver’s relationship with the patient. For example, McCullagh et al. (2005) found no correlation between caregiver burden and the nature of relationship with the patient, whereas Franzen-Dahlin et al. (2008) concluded that living with the PWA was a predictor of the life situation experienced by the SOs. Overall, concurrent with the findings of the present study, demographic variables of the SOs (or the person with post-stroke sequalae), including the nature of relationship with the stroke survivor, are seldom found to be major predictors of the situation or health of SOs (Jessup et al., 2015; McCullagh et al., 2005).

As could be hypothesised, participants’ everyday life and health were more affected by aphasia that was perceived as severe than aphasia that was perceived as less severe. In addition, perceived physical and cognitive impairments were associated with the experiences of the participants. In particular, people’s cognitive impairments secondary to stroke have been reported to affect the everyday life of SOs (Carod-Artal et al., 2009; Greenwood et al., 2008). Physical or functional disability produces a more conflicting picture, with some studies finding a correlation with perceived caregiver burden or quality of life, while other studies did not (Rigby et al., 2009).

Included within the factor of physical function was the perceived need for assistance. Franzen-Dahlin et al. (2008) found that the best predictor of the perceived life situation of caregivers of stroke patients with aphasia or depression was the need to assist the person with stroke. Particularly noteworthy is that the patients in their study had only minor functional impairments (as measured by the Barthel Index). The authors suggested that it is the SO’s appraisal, or, again, perception, of the needs of the stroke survivors that influences their life situations, not the factual, objectively measured need, a suggestion that is supported by the present study. Although a broad scale was used in the present study, the results revealed that the participants tended to consider the severity of the aphasia as at least as high as or even higher than the SLPs did, which may have influenced their perceptions of their life situations. The SOs’ perceptions of the need for assistance of the PWA has also been found to be associated with the prevalence of depression in SOs (Molloy et al., 2005).

Greenwood and Mackenzie (2010b) emphasised the ability of SOs of people with stroke sequalae to eventually adapt to, accept, and deal with their situation in a constructive way. However, only a small proportion of the SOs participating in our study spontaneously reported positive experiences of being caregivers. Taken together, the results suggest the need for further development and implementation of interventions aimed at supporting the SOs of PWAs. Based on our results and the importance of caregivers’ support in stroke rehabilitation, it is encouraging to see that the health and actual life situations of caregivers have received increasing attention in stroke and aphasia rehabilitation. In clinical work, instruments such as the Carer communication outcome after stroke (Carer COAST; Long et al., 2009) and Family Aphasia Measure of Life Impact (FAMLI; Grawburg et al., 2019) can be used to investigate how the caregiver’s life situation and quality of life are affected by the communication difficulties of the PWA in addition to the caregiver’s perception of the PWA’s ability to communicate and the effectiveness of their communication. While the Carer COAST could be seen as a screening instrument, the FAMLI is a more comprehensive instrument covering a lot of the issues presented in the present study.

Methodological considerations

There are several limitations of the study that must be acknowledged. A major limitation of the study is that only study-specific measures with limited psychometric properties were used. Although the validity of the questionnaire was strengthened by conducting a pilot study and undertaking the factor analyses, the results must be interpreted with caution. The length of the questionnaire is another possible threat to validity. If the participants chose to complete the questionnaire in one sitting, they might have been tired at the end of the questionnaire and therefore responded to the items with less care. They were therefore encouraged to take breaks during the completion of the questionnaire. In addition, different scales were used between and within sections, particularly in section B. This could have been confusing for the participants, although no pilot informant or participant commented on this or seemed to have misunderstood the response format. Furthermore, it might have been unclear how to respond to an item with the response format of, “If there has been a change – - -, how do you find that change?” if the respondent had not perceived any change. These issues with scales and phrasing should be addressed if the questionnaire is to be used in future research. A strength of the questionnaire is that a neutral scale was used regarding life changes and the participants’ evaluation of the changes, which made positive or negative answers equally likely to be chosen. A further strength is that the data were based on caregivers’ perceptions of the PWA’s disability, as was called for in Greenwood et al. (2008).

The number of participants (n = 173) constitutes a limited sample of the population, which restricts the generalisability of the results. Furthermore, we have no demographic data about those who declined to participate, and it appears that more women than men responded. In addition, the participants recruited by aphasia associations (n = 34; 19%) were not randomly selected. Thus, the possibility of selection bias must be considered. However, there were no differences between the AA sample and the SLP sample except time since stroke onset, and the AA sample thus provided important information from participants with longer experiences of being informal caregivers. The AA sample also ensured that participants from all over Sweden were included, which is a strength of the study. In addition, the response rate was relatively high.

Despite the limitations, the results are believed to provide new insights into how everyday life changes (based on the qualitative data), the prevalence of these changes, and how such changes are perceived and explained by the SOs of PWAs. The results also identify some factors that may affect SOs’ perceptions; however, this requires further investigation, preferably in a longitudinal study using psychometrically sound measures. The present study does not provide new information about whether there are any differences between SOs of stroke survivors with or without resultant aphasia. It would be of interest to use the items in section E of the questionnaire with SOs of stroke patients with no communication disorder to better understand how aphasia may contribute to the life situation of SOs.

Conclusion

The everyday life and health of the participating SOs were substantially affected by the stroke event. These everyday changes were mainly perceived negatively by the participants. The findings of this study support findings from previous research that have demonstrated the need to provide support for the SOs of PWAs. The SOs’ perceptions of the PWA’s stroke-related impairments and need for assistance may be a key factor in identifying SOs who may require support and guidance to help them to cope with the challenges they face.

Acknowledgments

This study was undertaken with the financial support of the Department of Public Health and Caring Sciences, Uppsala University, and the Department of Neuroscience, Uppsala University.

Disclosure statement

The authors report no conflicts of interest.

Supplementary material

Supplemental data for this article can be accessed here.