Stroke Caregiver Depression: Qualitative Comparison of Treatment Responders and Non-Responders at One-Year

Abstract

BACKGROUND:

Mixed method studies can shed light on the mechanisms through which interventions effectively reduce caregiver depression.

OBJECTIVE:

The objective of this qualitative study was to understand factors contributing to reduced depressive symptoms in stroke caregiver treatment responders compared with non-responders.

METHODS:

A mixed methods explanatory sequential design was used to assess cognitive behavioral coping use and helpfulness in reducing depression among experimental study treatment responders and non-responders. Using purposive sampling at one-year caregiving, we selected responders who were not depressed or demonstrated reduction to probable depression, and non-responders whose scores indicated worsened or new depression. Participants responded to semi-structured interviews on life changes and the use and helpfulness of intervention coping strategies. Thematic analysis identified codes and themes.

RESULTS:

Twelve of 25 participants were responders. We identified two overarching outcome themes: “It was about me, changing to make life better” (responders) and “Lamenting our changed lives” (non-responders). Five subthemes related to coping were identified. Whereas responders coped actively and were less dependent on the counseling relationship, non-responders were passive in coping, and more dependent on the initial counseling relationship. Anger or prior psychologic treatment were found in over a quarter of non-responders and in no responder. Negative life change was found in 9 of 13 non-responders compared with one responder.

CONCLUSION.

Findings support the helpfulness of cognitive behavioral coping. For many caregivers engagement Interviews and a more intense, longer intervention, such as stepped-care and anger management is indicated to mitigate treatment resistance and depressive symptoms, respectively.

Introduction

The strain of providing care for stroke survivors presenting a variety of physical, cognitive, and emotional morbidities has been associated with caregiver depression prevalence as high as 40% or greater.^1–3 Although studies to reduce caregiver depression are more common in recent years, outcomes have varied. Reviews have suggested that combined skill building/psycho-educational or psychosocial interventions were most successful in reducing caregiver depression.^4–7 Other researchers suggested that specific psychological strategies, such as cognitive-behavioral therapy (CBT), problem-solving, and coping skill training were most effective in depression reduction.⁸ Several studies found improvement in depressive symptoms at one year.^9,10,11 The interventions tested in these studies ranged from caregiver skill training in activities of daily living and stroke complications¹⁰ to psycho-educational strategies,⁹ and problem-solving.¹¹ However, little is known about the factors that differentiated outcomes among caregivers receiving the same treatment. The objective of this qualitative study was to increase understanding of factors contributing to reduced depressive symptoms in stroke caregiver treatment responders (Rs) compared with non-responders (NRs). We assessed treatment-related determinants of one-year depression outcomes for caregiver participants in a problem-solving intervention.

Researchers have recommended qualitative methods to aid in understanding outcomes of clinical trials, particularly those testing complex interventions.^{12, 13,14} We found no report of a qualitative assessment of variations in experimental treatment outcomes in stroke caregiver trials. However, a recent qualitative study of dementia caregivers assessed treatment non-responders to identify ways to improve interventions.¹⁵ Lewin et al.,¹⁶ in a review of mixed-methods trials, found only three studies (all non-caregiver) that assessed the reasons for study findings.

Methods

The participants in this study were purposively sampled from stroke caregivers enrolled in the treatment arm of a randomized clinical trial that tested the efficacy of a Caregiver Problem-Solving Intervention (CPSI) on caregiver outcomes after one year of caregiving. The Moos & Schaefer Model of Coping guided intervention development. The model emphasizes the role of mediators (cognitive appraisal, coping strategies) on adaptation outcomes.¹⁷ The mixed methods explanatory sequential design (Quant → qual) enabled assessment of coping strategy use and helpfulness among Rs and NRs.¹⁸ A publication from the quantitative study provides study details,¹⁹ which are described briefly here. We recruited caregivers during the stroke survivor’s acute rehabilitation inpatient program from two rehabilitation hospitals and two hospital rehabilitation units. The Institutional Review Boards of each hospital provided initial and annual study approval. Caregivers provided informed consent at recruitment.

After screening for a cut-off of ≥10 (at risk for depression) on the Center for Epidemiologic Studies-Depression (CES-D),²⁰ caregivers were assigned randomly to the CPSI treatment or a wait-list control group and followed for one year. 360 caregivers consented, 84 were ineligible; 145 caregivers (CPSI treatment, 70; wait-list , 75) completed one year quantitative assessments. CPSI caregivers participated in a 10-session treatment that included structured information, guided problem-solving, and coping with stress and emotional responses. Emotional responses to problems and concerns were the treatment focus. A major component of the intervention was social problem-solving therapy, which is a CBT treatment that combines cognitive and coping strategies to manage emotional responses by altering problem orientation and improving problem-solving.²¹ A cognitive-behavioral treatment paradigm²² (ABC model, cognitive reframing, thought-stopping) was incorporated to strengthen the intervention and aid in improving emotional responses to the stress of caregiving. The ABC (i.e., Activating Event, Belief, Consequence) model is a basic paradigm of CBT indicating that cognitions, emotions, and behaviors are linked and provide an opportunity to challenge dysfunctional beliefs.^22,23 Problem-solving coaching included using the acronym, STAR, as a reminder to Stop, Think, Act, and Review when faced with problems or tense situations. All sessions were structured and accompanied by a specifically designed workbook focusing on information and resources about stroke and management of its effects, caregiver stress, and positive coping with it, including recognizing the positive effects of caregiving. The CPSI treatment participants had one to two in-person meetings with a counselor to establish rapport and the remainder of the sessions were conducted by phone for the convenience of the caregiver. Telephone sessions were conducted weekly through session seven and bi-weekly through session 10, ending approximately 3 months post-discharge. The wait-list control group received telephone well-being checks at two and five months and at six months began five weekly telephone sessions consisting of supportive listening. Both groups received standard services provided by their rehabilitation team. Assessments of depressive symptoms were collected for CPSI treatment and wait-list groups pre-intervention (T1), post-intervention (T2; 3 months), and 6 (T3), and 12 months (T4). Wait-list control participants were not included in this qualitative analysis of CPSI treatment responders and non-responders. Most participants in the parent study were female (77%), White (61%), and caring for a spouse (64%). We found no significant difference in depressive symptoms between treatment and control groups at one year.¹⁹ With the exception of one participant, who missed one session, participants in this qualitative analysis completed 10 CPSI treatment sessions.

Qualitative methods

We used criterion sampling based on 12-month depression scores to purposely select CPSI treatment participants. Two investigators and two research assistants, who were not interventionists or data collectors, conducted the interviews. Rs had a 12-month CES-D score <16 or improved to probable depression from moderate-severe from baseline. Non-responder’s CES-D score either increased from < 16 at baseline to ≥ 16 or remained ≥ 16 with no improvement. We asked participants to respond to semi-structured questions to gain insight into why the intervention was effective for some caregivers and not others.

Qualitative Analysis

Interviews were conducted by telephone, audio-recorded, and transcribed verbatim. The research team qualitative expert guided the development of the interview guide. Questions were based primarily on caregivers’ perceptions of stroke-related life changes, and the use of and perceptions of the value of the intervention strategies. We analyzed interviews using thematic analysis.²⁴ The transcribed interviews were imported into Atlas Ti (ATLAS.ti Scientific Software Development GmbH Berlin, Germany) to facilitate analysis. Following the development of the initial codes by the qualitative expert, the principal investigator (PI) read the transcriptions repeatedly, and further developed and revised codes. The co-authors then coded the interviews independently; coding disagreements were discussed to reach consensus. Case summaries aided in the development of themes. Separate matrices of code categories were developed to aid in the comparison of R and NR responses on CPSI treatment strategy use and helpfulness. Strategies included both cognitive-behavioral, such as problem-solving, thought-stopping, and others, such as relaxation techniques and self-care options.

Results

Twenty-five caregivers participated in the one-year qualitative interviews. Table 1 presents characteristics of the total sample and responders and non-responders. Their mean age was 54.67 [SD =14.40]; eighteen were women (72%), and the majority were White. No significant difference was found in group demographics, caregiver health, survivor functional status, or caregiving hours. However, five of 13 NRs were men, compared to two of 12 Rs. Also, 10 NR caregivers were in a spousal relationship compared to five Rs. Five NRs compared with two Rs were taking psychotropic medication at one year. A non-significant difference was found for three NRs and no R reporting previous psychological treatment, Fisher exact test X ² = 3.15, P= .220. Differences were found in the one-year Profile of Moods²⁵ (POMS) Anger scale, t₂₃ = 3.84, P=.003 with greater anger found in NRs.

Table 1.

Caregiver Characteristics for total group, responders, and non-responders

	Total (N = 25)	Responder (n=12)	Non-responder (n = 13)	X 2	P value
Mean (SD) or n (%)
Characteristic
Sex (female)	18 (72)	10 (83)	8 (62)	1.47	.378
Race (White)	18 (72)	10 (83)	8 (62)	1.47	.378
Spousal caregiver (Yes)	15 (60)	5 (42)	10 (77)	3.23	.111
History psychological treatment (Yes)	3 (12)	0 (0)	3 (23)	3.15	.220
Health:
Poor/Fair	8 (32)	3 (25)	5 (38)	.52	.673
Good/Excellent	17 (68)	9 (75)	8 (62)
				t value
Age	54.67 (14.4)	50.21 (15.8)	58.79 (12.2)	−1.53	.140
Education (Years)	13.4 (3.6)	12.5 (4.3)	14.23 (2.7)	−1.22	.237
Hours caregiving (daily)	6.98 (6.7)	6.42 (5.2)	7.50 (8.0)	−.40	.696
Profile of Moods Anger	5.08 (2.9)	3.33 (2.9)	6.69 (1.9)	−3.84	.003

Note. SD = Standard Deviation.

Of the 12 R participants, one scored 11 on the CES-D at baseline. The remaining Rs’ baseline scores ranged from 16 to 35 with seven caregivers scoring ≥ 30 (moderate depression). Baseline CES-D scores for 13 NRs ranged from 11 to 43, with three caregivers scoring ≥ 30. One-year depression scores (0 to 11) remitted in all but one R whose score decreased from moderate to probable depression (16). Among NRs, symptoms increased to ≥ 16 for three with baseline scores under 16. Among NRs whose baseline score was ≥16; one-year scores ranged from 21 to 43.

Themes and Subthemes

Table 2 compares two overarching outcome themes and five subthemes. The outcome theme representing Rs was, “It was about me, changing to make life better”. “Lamenting our changed lives” represented the NR outcome. Responders tended to talk about how life improved from the stroke onset and how to continue improving. Several mentioned they did not have the time or inclination to focus on loss. An older R wife stated, “I don’t cry, I don’t think of the past. I think I have to be healthy to help him. The program helped me do this.” In contrast, a common thread among NRs was comparing the present with pre-stroke life, focusing on what was lost. These caregivers used few, if any, study strategies. A NR wife said:

I had a wonderful life, I could do what I wanted, when I wanted, and all of a sudden I am confronted with all these things and I was mad, I still am at times, and maybe that made it harder for me

Table 2.

Comparison of themes between responders and non-responders

SUB-THEME	Responders’ outcome theme It was about me; Changing to make life better	Non-Responders’ outcome theme Lamenting our changed lives
View of life change	PRIMARILY AMBIVALENT	PRIMARILY NEGATIVE
Finding other coping options	PROBLEM-SOLVING	Problem-solving
	CHANGING PERSPECTIVE/ABC	CHANGING PERSPECTIVE/ABC
	Empower the survivor	Empower the survivor
	Use manual	Use manual
	Black humor
	Journaling
Learning to care for myself	Self-care	Self-care
	RELAXATION, Pleasant events	Relaxation, Pleasant events
	Pray	Pray
	RESPITE	Respite
I’m not open to it		No time
		Can’t think positively
		Not open to it
		Did not digest the program
Making a difference for me	PROBLEM-SOLVING	Problem-solving
	Relationship/Interaction with counselor plus other coping options	RELATIONSHIP/INTERACTION WITH COUNSELOR
		Self-care advice/relaxation

Note. Bold uppercase indicates over 50% of group endorsed the item. Lowercase indicates less than 50% endorsed the item.

Subthemes

Four subthemes observed in both groups were View of life change, Finding other coping options, Learning to care for myself, and Making a difference for me. One theme, I am not open to it, was found only in the NR group.

View of life change.

Life changes were coded as negative, ambivalent, or positive view of the stroke experience. Negative views represented unpleasant responses to the stroke caregiving experience. These caregivers expressed few, if any, neutral or positive statements. Negative statements indicated difficulty, hardship, frustration, loss of freedom, resentment, no control of life, loss of companionship, poor support, emotional upheaval, and personal loss. Ambivalent views included negative statements that were softened or countered with neutral or positive observations, such as feeling grateful for support, finding meaning in the experience, feeling that life goes on albeit in a different way, focusing on the survivor’s needs, or finding satisfaction with the caregiver role or outcomes. Positive views focused on positive responses to the situation, such as positive aspects about current life with little emphasis on problems or personal loss, and a focus on the survivor’s needs. In contrast to NRs, most Rs talked about life changes with an attitude of being able to manage them and to restructure life to move forward. Compared to 9 NRs, one R was classified as having a negative view. Three of the NRs with negative views expressed anger. Three Rs and one NR expressed positive life change/experience. Ambivalent views were expressed by 8 Rs compared with 3 NRs holding this perspective.

Negative View examples.

A 49-year-old female spouse R related multiple challenges and frustrations caring for her husband, her mother, and in-laws. She worked full time and was resourceful in procuring respite. Her greatest changes were “loss of freedom …and you don’t have that companion you had before.” She said, “it’s hard, but I can handle it.” … “You have something negative, but there’s nothing you can do to change it, you have to change the way you’re feeling about the situation.” A 57-year-old NR husband and father of teenagers said: “my life turned upside down and I have almost lost complete control over everything …I am not sure it will ever get back in my hands.”

Ambivalent example.

A 42-year-old R wife said, “It never gets easy… It’s still hard, everything is my responsibility.” However, she found meaning in their situation, experimented with ways to improve her husband’s independence, and accepted their altered lifestyle resulting from her husband’s severe disability:

I told him, maybe you were supposed to have this stroke so we can spend more time together and get to travel. You’re getting disability and it’s not the greatest pay, but I didn’t ever think we were going to be rich.

Positive examples.

A responder (63-year-old female spouse) with a positive view of life reported, “Life is different, but it’s good, we can deal with it.” An outlier among NRs, a 57-year-old wife and mother of teens focused on positive life changes. She said:

Let’s look at positives first. First, was taking time to be with my husband; the attention changed my priorities in that he was first rather than second, and to learn to appreciate that time… what was really positive was that it allowed time to completely reorganize our life.

Finding other coping strategies

A variety of coping strategies were reported, with NRs using fewer strategies.

Problem-Solving.

The use of problem-solving methods was much more common among Rs (10 of 12) compared to 6 of 13 NRs. Those using problem-solving used more coping strategies than participants who did not use problem-solving. A 79-year-old sibling responder, speaking of problem-solving, said, “it’s very helpful, especially the Stop & Think because I’m the type that would start screaming. I don’t do that anymore.” A responder daughter said, “They were the best. They gave me cues to think about other options.” A NR wife reported, “It helped me to identify and figure out issues … to look at something before making a judgment about it.”

Perspective change strategies.

Rs used perspective change more often than NRs (10 vs 8). Two of these NRs used them only rarely when very stressed and several could not give examples of use. NRs who did not use thought stopping or relaxation presented a negative view of life change. Whereas 3 NRs reported angry feelings/no Rs did. Two Rs spontaneously mentioned journaling as a helpful technique and one of these also identified dark humor as helpful.

A small number of caregivers in each group explored ways to increase the survivor’s decision-making and independence. Approximately 35% in each group continued to use the study manual.

Learning to care for myself.

Caregivers in both groups practiced relaxation and a variety of other self-care activities including exercise, pleasant events, praying, and respite care. A female R, caring for her severely disabled adult daughter and other children while working full time, used relaxation regularly and said:

“I thought it was ridiculous… taking 20 minutes to take breaths. I learned you could. I would go 7 minutes, then 10, then I said, I can spend 30 minutes on the phone, don’t tell me I can’t take 20 minutes for myself.”

In contrast, a NR husband indicated, “That didn’t happen; I’m supposed to stop and take 30 minutes to do relaxation things? What I would like is to take 30 minutes for a walk… I can’t even find 30 minutes for that.” A responder daughter, talking about self-care indicated, “taking care of yourself is really important and learning to take that step back was extremely important.” A NR husband in his 60s said: “She was teaching me relaxation things and things to do because you get overwhelmed and even depressed because you’re thinking my whole life has changed.” Several caregivers used pleasant events to relieve stress and relax. A 64-year-old NR spouse found needlework and lunch with friends helped to focus on other things. A 49-year-old wife R said, “I also get 40 hours a month respite time for myself. That helps… I found just doing little things for myself.” Whereas over half of Rs used respite care, only two NRs used respite. Prayer was used by a minority in both groups. A 46-year-old NR daughter said that she “turned to God for support when the counseling sessions ended”. When asked what helped her when she was upset by her mother’s mood swings, a 45-year-old NR, indicated “I pray and I go through the book (treatment manual) to see what could be causing that effect.”

I’m not open to it.

This subtheme was found among four NRs, who used little or none of the recommended coping options. They felt they could not change their ways of coping or had no time to use strategies. A middle-aged NR wife stated, “They were excellent ideas, but sometimes it’s hard because of my personality to integrate that into what I was doing, it is just not the way I was made.” A 79 year old wife said: “It was helpful, but I was too mixed up in my mind and not open to it… Maybe if it started later.”

Making a difference for me

Many participants, particularly Rs, mentioned more than one strategy or aspect of the intervention when asked what part of the program made the biggest difference for them. Both groups highly valued the counselor relationship and the benefit of talking with someone outside their circle of family and friends. Several indicated they did not know what they would have done without the counselor. The majority of NR caregivers (9/13) mentioned the relationship with the counselor with vague descriptions, such as getting support or encouragement. Only three NRs also mentioned a coping strategy. In contrast, one-half of Rs identified the counselor relationship as most helpful, describing the importance of normalizing feelings and someone listening to and guiding them in problem-solving. All of these responders added one or more strategies with 8 also identifying problem-solving as most helpful. As one R female spouse said, “The biggest difference is talking to you – explaining things, but keeping positive helps and the program helped that. Also, problem-solving and talking about how I felt and how I think were helpful.” Two Rs could identify nothing in the intervention that made a difference; however, both used coping strategies.

Discussion

The groups differed in several ways. Rs used active coping, including respite, and were less dependent on the counseling relationship. They may have come to the study more resilient. In contrast, NR caregivers seemed more vulnerable and needy (angrier, more with psychological treatment history, less open to change, and more passive in coping). Passive coping has been correlated with depression after one year of stroke caregiving.²⁶ Nevertheless, NRs reported benefitting from the study, particularly from support and attention from the counselor during the first months as a caregiver. We could not assess the endurance of coping strategy use because the research team had no contact with caregivers after their six-month assessment. Interim checks on coping strategy use may provide important information but could act as an intervention.

The adaptations of these caregivers in managing life after a stroke were diverse, ranging from challenging, but successful, to feeling overwhelmed, hopeless, and angry. In our comparison of Rs and NRs we found marked contrasts in caregivers’ coping and views of life post-stroke. The findings suggest that Rs whose depression remitted or decreased markedly took an active role in achieving improvement, whereas NRs were less engaged in stabilizing their emotional response. Most of the non-responder caregivers focused on how supportive it was to talk with the counselor regularly. Despite this apparent benefit, few NRs used coping strategies taught by the counselors, such as problem-solving or the ABC method to change negative perspectives to decrease depressive symptoms. This finding is similar to a study of depression in later-life.²⁷ A marked contrast is noted in Views of life change with ambivalent and positive views more common in Rs compared with NRs. The use of coping strategies and the remittance of depression may have contributed to this difference. Our finding of a high prevalence of ambivalent experience in the non-depressed Rs differed from findings reported in prior research indicating greater depression in ambivalent dementia caregivers.²⁸ Methodological differences may contribute to the contrasting findings as the dementia caregiver study was quantitative, using a brief 6-item measure compared to our use of a semi-structured interview guide, which allowed caregivers to expand on a range of emotions. Similar to our findings, a report on spousal dementia caregivers indicated that positive caregiving experiences were found in those who focused on their spouses’ needs rather than their own, found meaning in caregiving, and accepted life changes.²⁹ Coping strategies, such as identifying positive aspects of caregiving, using relaxation exercises, the ABC method and effective problem-solving to analyze stressful situations likely contributed to a more positive, or perhaps ambivalent perspective. In contrast, the NRs who utilized fewer coping strategies reported more anger and tended to focus on loss while the Rs who used more coping strategies focused on changes that could make life better for the survivor and themselves.

The findings support the helpfulness of aspects of the intervention. All, but two participants, who were responders, indicated the program helped or made a difference in their lives. Responders used problem-solving, relaxation, and respite care much more than NRs. Several NRs used no coping strategies. Their statements that they could not “think that way” may indicate they could not deal with problems contributing to negative emotions, or they may be denying their effect on themselves and family life. It may be that these caregivers did not believe the strategies could help their emotions and relied on familiar ways of coping. R caregivers may have been more open to tackling the problems encountered in their new role. Considering the frequency of anger expressed by NRs (23%), and the greater POMS anger score in NRs, future stroke caregiver studies should consider including anger assessments.

Assessments of anger and interventions to aid in anger regulation may promote positive coping strategies to reduce depressive ideation and negative life change perceptions.³⁰

We found no stroke caregiver study documenting prior treatment for psychological issues. Such information is likely important in tailoring interventions as the three NRs with a history of psychological treatment were depressed at one year. A stepped care paradigm that provides more intense treatment based on response post-intervention and booster sessions may be indicated for such caregivers. In their study of Alzheimer caregivers, Nehrig and Chen (2019) also concluded that treatment non-responders may have benefitted from a stepped-care approach and longer therapy.¹⁵ Stepped care models have been recommended for depression in caregivers of older adults; they have been effective in preventing and treating depression in older adults, and in reducing major depression.^31–33

Limitations

A limitation is that more men and spouses were in the NR group. Cameron et al., 2011³⁴ found increased depression in spouses after one and two years of caregiving. However, our larger study indicated no significant relationship between dyadic relationship and gender and one-year depression. Increased depression in women compared to men caregivers has been reported.^34,35 Future explanatory mixed-methods studies may benefit from selecting equal proportions of sex and spouses in responder and non-responder caregiver participants. Another limitation is the reliance on self-report of coping strategies since counselor contact ended approximately 9 to 10 months before the qualitative interviews were completed.

Conclusion

Given that active coping strategies appeared to aid in reducing one-year depression in this small sample, the question of how to promote such behaviors remains. Engagement interviews (a combination of motivational interviewing and ethnographic interviewing) prior to initiating treatment have been used to promote the participation of depressed individuals in treatment³⁶. This process aids understanding of barriers to treatment and highlights ambivalence about treatment participation. Since it appears treatment non-responders derived only selective, temporary benefit from the intervention, a stepwise strategy might offer increased intensity of treatment to increase the chance of enduring benefit. Documenting a history of psychological treatment should be included in future studies as it can aid in identifying caregivers at increased risk for sustained depression and in need of a stepped-care, more intense approach.