Quality of life among caregivers of a vulnerable population living with HIV: Caregiving and relationship factors

Abstract

Literature on health-related quality of life (HRQOL) has focused on people living with chronic conditions, with less attention given to HRQOL among informal caregivers. We used cross-sectional dyadic data from both care recipients (CR) living with HIV and the person they identified as their primary informal (unpaid) caregiver (CG) to identify psychosocial and caregiving relationship factors (including, CG role ambivalence and caregiving-related stress) associated with CG HRQOL. We conducted confirmatory factor analysis and structural equation modeling testing. The results highlight interdependent effects of the CG-CR relationship and reveal pathways whereby relationship interactions positively and negatively impact CGs’ HRQOL. Affiliative stigma, CG-CR communication, CRs’ reciprocity of support and other psychosocial factors indirectly and differentially affected physical and mental HRQOL through effects on secondary stress and role ambivalence. Dyad-focused intervention on interpersonal communication and support exchange may improve HRQOL and resilience of CGs of vulnerable people living with HIV.

INTRODUCTION

Literature on health-related quality of life (HRQOL) has focused on people living with chronic conditions and given less attention to family and friends caring for them. The health and well-being of caregivers of vulnerable people and their relationship with their care recipients likely play a key role in patients’ HRQOL and continuity of caregiving. Informal caregiving is often defined as unpaid assistance from family and friends and has been studied primarily within the context of care recipient impairments (e.g., dementia) and need for care [1,2]. Little research has examined caregiving within the context of disadvantaged, African American, potentially stigmatized populations. Research on informal caregiving for people living with HIV (PLHIV) is particularly important given research revealing these caregivers’ role in promoting and sustaining care recipients’ adherence to HIV treatments [3]. HIV caregiving is increasingly important as effective treatments allow PLHIV to live longer and therefore experience greater rates of comorbidities and associated impairments [4,5]. While HRQOL of PLHIV has been studied [6–8], less is known about the HRQOL of their caregivers. We sought to identify aspects of caregiving and caregiving relationships associated with HRQOL of caregivers of vulnerable PLHIV using data from both caregivers and care recipients.

Informal caregiving

As with other seriously ill populations, the majority of PLHIV receive unpaid instrumental or emotional support from partners, family or friends (as informal caregiving is often operationalized), and informal caregivers provide most of their care [4,9]. In the US, low income African American women (compared to other economic-race groups) are most likely to provide informal HIV care and to provide the most intensive forms of such care [10]. Caregiving is culturally normative and may provide positive benefits to caregivers [11]. Caregiving may also be highly stressful. Among vulnerable populations, caregiving may strain already limited physical, mental and social resources [12,13]. Indeed, being a main caregiver has been found to predict social isolation, morbidity and mortality [14].

Research suggests that immediate stressors from providing instrumental support (primary stressors) and the subsequent (secondary) stressors that arise from these caregiving tasks are likely draining on the caregiver’s physical and mental health. The stress process of caregiving has been found to include the direct burden of caregiving, often quantified as the number of hours spent caregiving [15], and burdens related to the health conditions, behavior and care needs of the care recipient. The specific tasks associated with HIV care may differ from tasks associated with other illnesses and may differ from person to person. Previous research has found that caregiver stress is more highly associated with secondary effects of caregiving, namely caregiving-related interferences in engagement in other valued roles such as work, family, leisure, and self-care activities [16,17]. Such role strains have been found to be associated with caregiver depressive symptoms and caregiver role overload, the state of feeling overwhelmed by caregiving [18]. It is likely that caregivers’ appraisals of their relationship with their care recipient affects caregivers’ perceptions of burden and overload, which may impact HRQOL [19]. Caregiving may adversely impact caregivers’ own needed social support, which has been found to predict cessation of caregiving. Thus, it is important to assess the quality of interpersonal relationships with caregivers’ other support network members, in addition to their care recipient [20].

In the context of caregiving in a vulnerable population, other under-studied psychosocial factors may also be important. Caregiver role ambivalence, or perceiving a need for respite from care, may affect HIV caregivers’ stress, though has received little attention in the literature. In the context of HIV caregiving, affiliative stigma as a result of caring for someone with stigmatizing conditions may affect caregiver stress. Conversely, having a positive relationship characterized by reciprocity (defined as the care recipient showing affection or appreciation for the caregiver) and good communication, have been shown to have positive health effects and enhance caregivers’ HRQOL [21]. This study explored how these caregiver-recipient relationships factors impact primary HIV caregivers’ HRQOL.

Health-Related Quality of Life

HRQOL has been conceptualized in the literature as comprising multiple domains of well-being and has been operationalized as overall health status, physical functioning, social functioning, mental health, and health effects on social role performance [22–25]. Measures emphasize physical health and mental health dimensions of HRQOL [7,26–28]. Mental HRQOL (MHRQOL) is often measured by depressive symptoms and health-related interference in normal activities [29,30]. Physical HRQOL (PHRQOL) is often measured by overall health ratings, physical limitations in performing activities of daily living, and health interference in social activities [29–31]. HRQOL is a robust construct for assessing effects of caregiving on caregivers [32,33].

Primary and Secondary Care-Related Stressors

According to Pearlin and colleagues’ [34] stress proliferation theory, primary caregiving-related stressors include care recipients’ actions or attitudes that make caregiving difficult, while secondary stressors are challenges that occur as a result of caregiving. Primary stressors include the care recipient’s problem behaviors (e.g., erratic or aggressive behaviors), which may be especially prevalent and stressful for caregivers of persons who use illicit drugs and violate prosocial norms while using drugs. Secondary stressors may include financial strains, due to loss of income opportunities or increased expenditures, and psychosocial strain related to the care recipient’s illness and its adverse impact on the caregiver-recipient relationship or family, such as disagreement or conflict [35–39]. Within the context of caring for vulnerable PLHIV, social conflict may be exacerbated by financial strains and stigma related to the care recipient’s illness or behavior while using drugs [4,40]. Primary and secondary stresses can contribute to losses in health-related quality of life for both caregivers and PLHIV directly, by making it more difficult to adhere to HIV related treatment and thereby lessen the burden of disease, or indirectly, by increasing stress and anxiety within the relationship. The proliferation of these stresses may also contribute to role ambivalence.

Role Ambivalence

Caregiver role ambivalence is defined as caregiver perception of frustration with caregiving or wanting time away from caregiving or their care recipient (though not to the extent of perceiving their caregiving-related stress as inescapable) [37]. Ambivalence may indicate stress above and beyond caregivers’ expectations of stress associated with caregiving, though ambivalent caregivers may also experience personal satisfaction with helping a loved one [41,42]. Ambivalence toward the caregiving role is thought to result from shifting expectations of relationships between the caregiver, care recipient, and other family members, who may be or feel neglected as a result of the caregiver’s focus and attention on the care recipient [43].

Affiliative Stigma

Affiliative stigma is defined as perceived stigma due to association with a stigmatized person. Affiliative stigma may affect HRQOL among caregivers of PLHIV due to the stigma associated with HIV and may be especially acute if the PLHIV is a current or former user of illicit drugs [44,45]. Affiliative stigma may limit the availability of care and support to PLHIV. HIV caregiver’s perceived affiliative stigma may alienate them from their family and friends, further impeding potential support to the care recipient as well as social support important for the caregiver’s own well-being [45].

Affiliative stigma can also adversely impact the caregiving relationship, generating feelings of shame [46]. Research indicates that affiliative stigma is associated with depressive symptoms among caregivers of PLHIV with a history of drug use [47]. Caregivers that report greater affiliative stigma also tend to report worse HRQOL [48].

Positive Caregiver-Care Recipient Relationship

Positive aspects of the caregiving relationship may be protective of caregivers’ HRQOL. Good communication and mutual exchange of support in the caregiver-care recipient relationship are indicators of quality relationships. Reciprocity of social support is associated with positive appraisals of support, better caregiving relationships, and predictive of long term physical and mental health outcomes for care recipients and caregivers [4,33,49]. Caregivers’ ability to count on their care recipient for emotional support, socializing and other support may affect their HRQOL. Distinct from reciprocity of support, Pearlin and colleagues proposed that caregiving-related stress may also be affected by reciprocity of caring, defined as caregiver’s perception that their care recipient cares about their own health and well-being and would do the same for them if possible [50,51]. Indeed, relationship qualities and expectations (or care recipients’ ability) to reciprocate may differ in normative caregiving of senior family members as compared to caring for PLHIV who use drugs [51,52].

Study Objective

The purpose of this study was to examine associations among interpersonal caregiving constructs and HIV caregivers’ HRQOL. Interpersonal factors included in the analyses modeled not only chronic caregiving-related stressors and the outcomes of these stressors, but also aspects of social support and coping, which can buffer the effects of stress on mental health [53–55]. Data from both caregivers and care recipients were used to model how relationship dynamics affect caregiver HRQOL. Specifically, in the present study we tested a standard error of the mean (SEM) to examine how characteristics of the quality of the caregiver-receiver relationship and caregiving demands related to HRQOL among HIV caregivers (see Figure 1). In doing so, we sought to not only improve theoretical understandings of caregiving relationships, but also inform the development and testing of potential interventions to improve the HRQOL of caregivers of the population. Such intervention may have potential implications for vulnerable care recipients’ HRQOL as well.

Figure 1.

The initial conceptualization of relationships among caregiving-related psychosocial constructs and quality of life.

METHODS

Procedure

Data for this secondary analysis were from the baseline survey of the BEACON (Being Active and Connected) study (2006–2012) in Baltimore, Maryland, USA [56]. This study examined social relationship factors associated with health outcomes and well-being among disadvantaged PLHIV and their informal caregivers. PLHIV care recipients were recruited from community venues and an HIV clinic affiliated with an academic hospital. Selection criteria included: HIV seropositive adult on antiretroviral treatment (ART); Baltimore City resident; history of or current injection drug use; and willing to invite a main supportive person or caregiver to participate in the study. Main caregivers were selected based on criteria of PLHIV care recipient report of persons who provided the recipient with emotional support or instrumental assistance (e.g., help with errands or household chores) or health-related assistance (e.g., assistance with medications or attending medical visits) in the prior six months. The care recipient authorized the caregiver’s recruitment to the study prior to caregiver recruitment. Caregiver exclusion criteria included providing care to the recipient in a professional (paid) capacity.

Measures

Health-related Quality of Life. Indicators of HRQOL used in this study were based on measures from Ware and colleagues’ short forms for assessing quality of life (SF-12) [19,30,35,57,58]. Specific measures were self-rated health (from 1=poor to 5=excellent); a sum of measures of physical functioning limitations (e.g., bending, walking uphill) with each item rated from 0=not at all to 2=a lot [17,20,59]; two items measuring the degrees to which mental health and physical health interfered with the caregiver’s time spent with family or friends [60]; and depressive symptoms using cut-points of none/low (0 to 8), mild (9 to 15), and moderate/high (16+) [58,61].

Care provision.

Care provision in the past year was measured using a series of yes/no questions about whether the caregiver provided aid to the recipient by going with the recipient to a doctor’s appointment or to the emergency department, helping the recipient manage money or bills, get transportation, run errands, visiting the recipient in the hospital, helping with personal care (activities of daily living, i.e., grooming, eating, toileting), and assisting with medications [9].

Caregiving-related primary stress.

Ten questions, rated from 0=never to 2=a lot, assessed the degree to which certain problems made it difficult for the caregiver to help the recipient. These included caregivers’ financial limitations, time limitations, caretaking of other people; care recipients’ problem behaviors (e.g. aggressive behavior) and amount of needs; recipients’ not telling their caregivers when they are sick and not wanting others to know their HIV status; and care recipients’ family conflict, namely, the degree to which caregiving affected conflict with recipients’ friends and family, and conflict with recipients over the care of their children [34].

Affiliative HIV stigma.

Affiliative stigma was measured by four items, comprising the question stem, “thinking about [care recipient’s] HIV,” and items pertaining to the caregiver’s fears that s/he might lose friends due to affiliation with the recipient, that people might avoid or feel uncomfortable around him/her, or that people might blame him/her for the recipient’s having HIV [62]. All four items were rated from 0=not at all to 3=very much.

Recipient’s reciprocity of support.

Caregivers reported on specific kinds of support they had received from their care recipients in the past year, including help around the house; gifts; money or something valuable; caring for the caregiver’s children, family or friends; time spent with the caregiver; and care recipient’s participation in the caregiver’s family activities (rated from 0=never to 2=a lot) [31,63].

Caregiving-related secondary stress.

Problems caregivers incurred from providing caregiving to their recipients, rated from 1=not at all to 4=a great deal, included: experiencing financial problems, losing or missing work, having less time with friends or family, having reduced ability to care for children, experiencing interference with caring for their own health needs, undergoing family arguments, having interference with church or religious practices, and suffering physical strain [34].

Communication.

Caregiver-recipient communication was measured using three items assessing caregivers’ feelings that they could count on their recipients to listen when they needed to talk to someone, could state their feelings about their relationship without recipients getting defensive, and could enjoy spending time with their care recipients (rated from 1=strongly disagree to 4=strongly agree) [64].

Reciprocity of caring.

Perceived reciprocity was measured with four questions, rated from 1=strongly disagree to 4=strongly agree, assessing the degree to which caregivers perceived their care recipients were concerned about their health and appreciated their help [65], that their care recipient would help them if needed and that their care recipient helped out as much as possible [66].

Role ambivalence.

Role ambivalence was measured with three items assessing how often caregivers wished they did not have to help out their care recipients as much, wished they felt less obligated to help (0=rarely or none of the time to 3=most or all of the time), or felt that they needed time away from their care recipient (0=never to 2=most of the time) [37].

Covariates.

Four variables, including caregiver sex, age, HIV serostatus, and substance use status were control variables in the final SEM. HIV serostatus was coded 0 for HIV seronegative and 1 for HIV seropositive. Current substance use was coded 0 for no use and 1 for use of illicit drugs in the previous 6 months (heroin, cocaine, stimulants, barbiturates, opiates, or tranquilizers) or hazardous alcohol use including binge drinking (5 or more drinks per day) at least once per week in the previous month or daily drinking in the past month.

Data Analyses

Factor analyses and SEM were conducted using Mplus version 7.1 [63]. Exploratory factory analyses were performed on each group of items to determine if the items were uni- or multi-dimensional after using oblique rotation. Next, a confirmatory factor analysis model was fit with all the latent factors, where items were constrained to load only on their relevant factors. This measurement model was fit and a correlation matrix of all the latent variables was generated.

Using the latent factors from the previous step, we conducted SEM based on the conceptual model in Figure 1. The model was fit step by step, starting with having the independent or exogenous factors on the far left of the model being directly associated with the next level to the right in the diagram, which were labeled tier 1 dependent or endogenous factors. Subsequently, we added endogenous factors at higher levels to the right, one tier at a time, using theoretical relationships defined in Figure 1 and the latent variable correlation matrix as guides.

In these analyses, we considered potential confounding by age, sex, current drug use and/or hazardous alcohol use, and HIV status of the caregiver at baseline. In the context of SEM, we examined whether each of these variables had a potential confounding effect on any of the structural effects in the model, i.e., for each structural path arrow, whether the potential confounder was correlated with both the predictor and outcome factors. If so, the potential confounder was then included in the SEM as a predictor of the outcome factor, and as a predictor of the exogenous factor in cases when the latter was also an endogenous factor

RESULTS

Sample description

Caregivers were mostly female (59.7%), African American (88.4%), and had monthly incomes of $999 or less (63.0%) (Table I). Approximately half had a high school education or less (52.3%). Slightly less than half were HIV seropositive (43.4%) and reported current substance use (33.7%). Role relations between caregivers and care recipients comprised kin (45.2%), partner (31.6%) or friend (17.2%). The mean age of caregivers was 47 (SD = 11 years) and for care recipients, mean age was 48 (SD = 6 years). Most care recipients were male (61.4%), African American (83.8%), had a high school education or less (83.8%), earned $999 or less per month (79.9%), and three-fifths reported substance use in the past six months (58.2%).

Table I.

Demographic characteristics of caregivers and care recipients (BEACON study; N = 258)

Variable	Care Recipient N (%) or Mean (SD)	Caregiver N (%) or Mean (SD)
Age	48 (6)	47 (11)
Sex (male)	148 (57.4)	105 (40.7)
Race/ethnicity (African American/Black)	218 (84.5)	228 (88.4)
Education (High School/GED or less)	205 (83.3)	201 (52.3)
Income (< 1,000 per month)	210 (81.4)	160 (63.0)
Depression (16+ on CES-D)	101 (39.1)	84 (32.7)
Substance use (Illicit drug use or	151 (58.5)	87 (33.7)
hazardous alcohol use)
HIV seropositive	258 (100.0)	112 (43.4)
Viral load (Undetectable)	173 (71.5)	--
Main supporter relationship:
Kin	110 (42.6)	--
Partner	99 (38.4)	--
Friend	46 (17.8)	--

Overall, most caregivers reported recent care provision to the recipient, care recipient’s reciprocity of support and caring to the caregiver, and positive communication, while a minority reported secondary caregiving-related stress and affiliative HIV stigma (Table II). Forty-two percent reported providing the recipient assistance with personal care (activities of daily living, e.g., bathing, feeding, toileting), 57% assisted with medications, and 68% accompanied care recipients to medical appointments or the emergency department in the prior year. Approximately a third of caregivers reported that caregiving to the recipient caused them financial problems (36%) or conflict with recipients’ families (32%). Approximately one-quarter of caregivers perceived affiliative HIV stigma, i.e., that people might feel uncomfortable around them due to the care recipients’ HIV status (26%), and 32% of caregivers wished they were less obligated to help their care recipients at least some of the time.

Table II.

Summary of variables used in analysis (BEACON study; N=258).

Care provision: n (%)	No	Yes
Doctor/ER accompaniment	82 (32)	176 (68)
Money/bills	79 (31)	179 (69)
Transportation	63 (24)	195 (76)
Errands	73 (28)	185 (72)
Hospital visit	101 (39)	150 (58)
Personal care	148 (57)	109 (42)
Help with medications	112 (43)	146 (57)
Care-related primary stress (make caregiving difficult): n (%)	Never	Some	A lot
Recipient’s behavior	114 (44)	109 (42)	35 (14)
Recipient’s amount of needs	133 (52)	100 (39)	25 (10)
Financial limitations	90 (35)	132 (51)	36 (14)
Time limitations	130 (50)	108 (42)	20 (08)
Care of other people	139 (54)	98 (38)	21 (08)
Recipient not telling when sick	113 (44)	101 (39)	44 (17)
Recipient keep HIV secret	177 (69)	45 (17)	36 (14)
Conflict with recipient’s friends	176 (68)	52 (20)	30 (12)
Conflict with recipient’s family	191 (74)	48 (19)	19 (07)
Conflict over recipient’s children	191 (74)	46 (18)	14 (05)
Affiliative HIV stigma: n (%)	Not at all	Just a little	Somewhat	Very much
People uncomfortable being around	187 (72)	34 (13)	27 (10)	9 (03)
People avoid	226 (88)	18 (07)	10 (04)	2 (01)
People blame for HIV	248 (96)	6 (02)	2 (01)
Fear losing friends	219 (85)	15 (06)	16 (06)	8 (03)
Care-related secondary stress (problems as a result of caregiving): n (%)	Not at all	A little	Some	A great deal
Financial problems	165 (64)	54 (21)	25 (10)	14 (05)
Losing/missing work	223 (86)	17 (07)	15 (06)	2 (01)
Less able to care for family	170 (66)	53 (21)	27 (10)	8 (03)
Less able to care for children	224 (87)	21 (08)	10 (04)	3 (01)
Less able to care for own health	204 (79)	34 (13)	14 (05)	6 (02)
Family arguments	183 (71)	42 (16)	22 (09)	12 (05)
Physical strain	183 (71)	43 (17)	23 (09)	9 (03)
Recipient’s reciprocity of support to caregiver: n (%)	Never	Some	A lot
R helped around the house	54 (21)	92 (36)	112 (43)
R lent money	39 (15)	132 (51)	86 (33)
R gave gifts	28 (11)	124 (84)	106 (41)
R helped care for C’s children	59 (23)	82 (32)	117 (45)
R spent time with C	4 (02)	73 (28)	181 (70)
R joined C’s family activities	63 (24)	96 (37)	99 (38)
Recipient’s reciprocity of caring: n (%)	Strongly disagree	Disagree	Agree	Strongly agree
R would help if needed		7 (03)	80 (31)	171 (66)
R cares about C’s health	1 (00)	9 (03)	91 (35)	157 (61)
R helps out as much as can	4 (02)	14 (05)	98 (38)	141 (55)
R appreciates C	3 (01)	10 (04)	84 (33)	161 (62)
Caregiver-recipient communication: n (%)	Strongly disagree	Disagree	Agree	Strongly agree
CG counts on CR to listen	4 (02)	19 (07)	99 (38)	136 (53)
CG can express feelings to CR	10 (04)	45 (17)	108 (42)	95 (37)
CG enjoys spending time w/ CR	1 (00)	7 (03)	139 (54)	111 (43)
Caregiver role ambivalence: n (%)	Rarely or none of the time	Some or little of the time	Occasionally or moderate amount of time	Most or all of the time
Wish not having to help so much	176 (68)	54 (21)	20 (08)	8 (03)
Wish less obligated	178 (69)	61 (24)	13 (05)	6 (02)
Need time away from CR	100 (43)	126 (49)	22 (09)
Health-related quality of life: n (%)	Excellent	Very good	Good	Fair	Poor
General health rating	58 (22)	68 (26)	74 (29)	44 (17)	14 (05)
	Not at all	A bit	A lot
Mental health interferes with social activities (social function-mental)	187 (72)	47 (18)	24 (09)
Physical health interferes with social activities (social function-physical)	158 (61)	81 (31)	19 (07)
	Low	Mild	Moderate	High
Depressive symptoms	126 (49)	48 (19)	26 (10)	58 (22)
	Minimum	Maximum	Mean	SD
Physical functioning limitations: value	0	10	3.55	3.21

Care provision
Variable	No		Yes
	n	%	n	%
Doctor/ ER accompaniment	82	32	176	68
Money/ bills	79	31	179	69
Transportation	63	24	195	76
Errands	73	28	185	72
Hospital visit	101	39	150	58
Personal care	148	57	109	42
Help with medications	112	43	146	57

Care-related primary stress (making caregiving difficult)
Variable	Never		Some		A lot
	n	%	n	%	n	%
Recipienťs behavior	114	44	109	42	35	14
Recipienťs amount of needs	133	52	100	39	25	10
Financial limitations	90	35	132	51	36	14
Time limitations	130	50	108	42	20	8
Care of other people	139	54	98	38	21	8
Recipient not telling when sick	113	44	101	39	44	17
Recipient keep HIV secret	177	69	45	17	36	14
Conflict with recipienťs friends	176	68	52	20	30	12
Conflict with recipienťs family	191	74	48	19	19	7
Conflict over recipienťs children	191	74	46	18	14	5

Care-related secondary stress (problems as a result of caregiving)
Variable	Not at all		A little		Some		A great deal
	n	%	n	%	n	%	n	%
Financial problems	165	64	54	21	25	10	14	5
Losing/ missing work	223	86	17	7	15	6	2	1
Less able to care for family	170	66	53	21	27	10	8	3
Less able to care for children	224	87	21	8	10	4	3	1
Less able to care for own health	204	79	34	13	14	5	6	2
Family arguments	183	71	42	16	22	9	12	5
Physical strain	183	71	43	17	23	9	9	3

Recipienťs reciprocity of support to caregiver
Variable	Never		Some		A lot
	n	%	n	%	n	%
R helped around the house	54	21	92	36	112	43
R lent money	39	15	132	51	86	33
R gave gifts	28	11	124	84	106	41
R helped care for C's children	59	23	82	32	117	45
R spent time with C	4	2	73	28	181	70
R joined C's family activities	63	24	96	37	99	38

Recipient reciprocity of caring
Variable	Strongly disagree		Disagree		Agree		Strongly agree
	n	%	n	%	n	%	n	%
R would help if needed	0	0	7	3	80	31	171	66
R cares about C's health	1	0	9	3	91	35	157	61
R helps as much as they can	4	2	14	5	98	38	141	55
R appreciates C	3	1	10	4	84	33	161	62

Caregiver-recipient communication
Variable	Strongly disagree		Disagree		Agree		Strongly agree
	n	%	n	%	n	%	n	%
C counts on R to listen	4	2	19	7	99	38	136	53
C can express feelings to R	10	4	45	17	108	42	95	37
C enjoys spending time with R	1	0	7	3	139	54	111	43

Caregiver role ambivalence
Variable	Rarely or none		Some or little		casionally or moder		Most or all of the time
	n	%	n	%	n	%	n	%
Wish not having to help so much	176	68	54	21	20	8	8	3
Wish less obligated	178	69	61	24	13	5	6	2
Need time away from R	100	43	126	49	22	9	0	0

Affiliative HIV stigma
Variable	Not at all		Just a little		Somewhat		Very much
	n	%	n	%	n	%	n	%
People uncomfortable being around	187	72	34	13	27	10	9	3
People avoid	226	88	18	7	10	4	2	1
People blame for HIV	248	96	6	2	2	1	0	0
Fear losing friends	219	85	15	6	16	6	8	3

Health-related quality of life
Variable	Excellent		Very good		Good		Fair		Poor
	n	%	n	%	n	%	n	%	n	%
General health rating	58	22	68	26	74	29	44	17	14	5
Variable	Not at all		A bit		A lot
	n	%	n	%	n	%
Mental health interferes	187	72	47	18	24	9
Physical health interferes	158	61	81	31	19	7
Variable	Low		Mild		Moderate		High
	n	%	n	%	n	%	n	%
Depressive symptoms	126	49	48	19	26	10	58	22
Variable	Minimum		Maximum		Mean		SD
Physical functioning limitations	0		10		3.55		3.21

Abbreviations: R = recipient. C = caregiver. w/ = with. SD = standard deviation.

Factor Analyses

Factor analyses revealed several of the latent constructs were unidimensional, including affiliative stigma, recipient’s reciprocity of support, recipient’s reciprocity of caring, communication, caregiving-related secondary stress, and role ambivalence (Table III). Care provision items loaded on two highly correlated factors (travel companion/aid and personal care/finances, r = .81, p < .05), with four items each. Care-related primary stress also loaded on two highly correlated factors (caregiving demands and caregiving conflict, r = .78, p < .05), with five items each. Consistent with standard HRQOL scales [53,54,66], HRQOL items loaded on two factors reflecting mental health related quality of life (MHRQOL) and physical health related quality of life (PHRQOL), with a correlation of .52 (p < .05). Using confirmatory factor analysis, the measurement model that included all of these factors evidenced good fit (Comparative Fit Index [CFI] = .97, Tucker-Lewis Index [TLI] = .96, and Root Mean Squared Error of Approximation [RMSEA] = 0.03, with 90% CI=0.02, 0.03) [67].

Table III.

Confirmatory Factor Analysis (CFA) loadings for indicator variables on their respective latent factors (BEACON study; N = 258).

	Loading (SE)		Loading (SE)
Affiliative HIV stigma		Secondary stress
People uncomfortable	0.75 (0.07)	Financial problems	0.70 (0.05)
People avoid	0.80 (0.08)	Losing/missing work	0.52 (0.08)
People blame	0.82 (0.15)	Less able to care for family	0.76 (0.04)
Fear losing friends	0.76 (0.07)	Less able to care for children	0.75 (0.06)
		Less able to care for own health	0.80 (0.04)
		Family arguments	0.75 (0.04)
		Physical strain	0.72 (0.05)
Perceived reciprocity of caring		Recipient’s reciprocity of support to caregiver
R appreciates C	0.72 (0.05)	R helped around the house	0.71 (0.05)
R would help if needed	0.81 (0.03)	R lent money	0.76 (0.04)
R cares about C’s health	0.88 (0.03)	R gave gifts	0.88 (0.03)
R helps out as much as can	0.91 (0.03)	R helped care for C’s children	0.66 (0.05)
		R spent time with C	0.75 (0.05)
		R joined C’s family activities	0.68 (0.05)
Caregiver/Recipient Communication		Caregiver role ambivalence
C counts on R to listen	0.89 (0.03)	Wish not having to help so much	0.80 (0.04)
C can express feelings to R	0.79 (0.04)	Wish less obligated	0.89 (0.03)
C enjoys spending time w/ R	0.78 (0.05)	Need time away from R	0.76 (0.05)
Travel companion/aid		Personal Care/finances
Doctor/ER accompaniment	0.84 (0.05)	Money/bills	0.87 (0.05)
Food/cleaning	0.79 (0.08)	Transportation	0.64 (0.08)
Errands	0.81 (0.05)	Personal care	0.80 (0.06)
Hospital visit	0.74 (0.05)	Help with medications	0.81 (0.05)
Caregiving demands		Caregiving conflict
Recipient’s behavior	0.74 (0.05)	Recipient not telling when sick	0.69 (0.05)
Recipient’s amount of needs	0.85 (0.04)	Recipient keep HIV secret	0.80 (0.06)
Financial limitations	0.57 (0.06)	Conflict with recipient’s friends	0.64 (0.06)
Time limitations	0.70 (0.05)	Conflict with recipient’s family	0.64 (0.07)
Care of other people	0.47 (0.07)	Conflict over recipient’s children	0.67 (0.07)
Mental Health-Related Quality of Life		Physical Health-Related Quality of Life
Depressive symptoms	0.86 (0.05)	Self-rated health	0.66 (0.08)
Mental health interfering w/ time spend w/ family/friends	0.86 (0.04)	Physical health interfering with time spent with family/friends	0.42 (0.10)
Physical health interfering w/ time spent w/ family/friends	0.56 (0.09)	Physical functioning limitations	0.67 (0.08)

Abbreviations: R = recipient. C = caregiver. SE = Standard error. MHRQOL = Mental health-related quality of life. PHRQQOL = Physical health-related quality of life.

Correlations of Latent Variables

Most path coefficients suggested in the structural equation model specified in Figure 1 were supported by the bivariate analyses (Table IV). For example, strong bivariate relationships were found between affiliative HIV stigma and care-related secondary stress (r = .67, p < .05), as well as between caregiving demands (r = .74, p < .05) and caregiving conflict (r = .69, p < .05) with secondary stress.

Table IV.

Correlation matrix of latent variables (BEACON study; N = 258).

	CD	CC	TC/A	PC/F	AS	SS	RR	RS	Comm	RA	MHQOL	PHQOL
Caregiving demands	1.00
Caregiving conflict	0.81	1.00
Travel companion/aid	−0.07	−0.03	1.00
Personal care/finances	0.00	0.03	0.78	1.00
Affiliative HIV stigma	0.06	0.09	0.46	0.50	1.00
Care-related secondary stress	0.32	0.22	0.74	0.69	0.67	1.00
Recipient’s reciprocity of caring	0.29	0.19	−0.44	−0.24	−0.14	−0.22	1.00
Recipient’s support to CG	0.64	0.58	−0.29	−0.12	0.15	0.02	0.57	1.00
CR/CG Communication	0.17	0.15	−0.56	−0.35	−0.24	−0.40	0.82	0.52	1.00
CG Role ambivalence	0.24	0.10	0.76	0.72	0.51	0.80	−0.37	−0.09	−0.45	1.00
MHQOL	−0.01	−0.13	0.57	0.59	0.57	0.60	−0.17	−0.07	−0.12	0.61	1.00
PHQOL	0.07	−0.02	0.13	0.19	0.16	0.24	0.07	−0.06	0.02	0.13	0.52	1.00

Boldface denotes statistically significant correlations between two latent variables (p < .05).

CR = Care recipient; CG = Caregiver; CD = caregiving demands; CC = caregiving conflict; TC/A = travel companion/aid; PC/F = personal care/finances; AS = affiliative HIV stigma; SS = care-related secondary stress; RR = recipient’s reciprocity of caring; RS = recipient’s support to caregiver; Comm = CR/CG communication; RA = CG role ambivalence; MHQOL = mental health quality of life; PHQOL = physical health quality of life.

Structural Equation Modeling

In the structural equation model, several independent (affiliative stigma, recipient’s reciprocity of support) and dependent (caregiving-related secondary stress, role ambivalence, MHRQOL and PHRQOL) variables were represented by single-level factors. In addition, we used three second-order factors. Care provision was defined by the sub-factors, travel companion/aid and personal care/finances, while caregiving-related primary stress was defined by caregiving demands and caregiving conflict. Perceived reciprocity of caring and communication were found to be highly correlated (r = .82), suggesting overlap in the constructs; therefore, they were combined as indicators of a second-order factor that was named positive relationship. All exogenous factors were then allowed to be associated with both care-related secondary stress and positive relationship.

Caregiver’s age, sex and HIV status were not confounders at any stage of the structural equation model. Age was statistically significantly correlated with the two HRQOL variables but not with any of their predictors; sex was not associated with any variables in the structural model; and HIV status was associated only with positive relationship but not with its predictor or outcomes. Baseline substance use status, on the other hand, was marginally associated with care-related primary stress (r = .15), care-related secondary stress (r=.18), role ambivalence (r = .21) and MHQOL (r = .21). Substance use was therefore included initially in the SEM with paths predicting care-related secondary stress, role ambivalence and MH-related QoL. Non-significant paths were removed one at a time (starting from the one pointing to the most distant outcome), leaving one path adjusting for drug use’s confounding of the association between care-related primary stress and care-related secondary stress.

In the final model, care provision, caregiving-related primary stress and affiliative HIV stigma were positively associated with caregiving-related secondary stress (Figure 2). Positive relationship was positively associated with recipient’s reciprocity of support and negatively associated with caregiving-related primary stress. Caregiver role ambivalence was strongly associated with caregiving-related secondary stress and was negatively associated with positive relationship. Role ambivalence was strongly associated with lower MHRQOL, while caregiving-related secondary stress and positive relationship were indirectly associated with MHRQOL through role ambivalence. Caregiving-related secondary stress was directly and negatively associated with PHRQOL but was not associated with role ambivalence or positive relationship. The final model fit the data well, with RMSEA = 0.03 (90% CI = (0.024, 0.035), suggesting excellent fit), CFI = .95 and TLI = .95.

Figure 2.

Revised structural equation model based on results from the correlation matrix of latent variables (BEACON study; N=258).

Note: CD = caregiving demands; CC = caregiving conflict; TC/A = travel companion/aid; PC/F = personal care/finances. All coefficients are XY-standardized. The model adjusts for baseline substance use (see details in the text). To improve visual clarity, the measurement models for all first-level latent factors are omitted from this figure. The indicators are the same, and factor loadings are very similar, to those in the CFA model in Table II. *< .05, **< .01, ***< .001

PHRQOL R2 = 065; MHRQOL R2 = .65.

DISCUSSION

The study is, to our knowledge, one of the first to comprehensively conceptualize and test a model of HRQOL among caregivers of a vulnerable, stigmatized care recipient population. The study tested a model of stress proliferation proposed by Pearlin and colleagues [60] and expanded on it by identifying caregiving resiliencies. The findings contribute to theoretical understandings of effects of the caregiving role and context-specific stressors and resiliencies in caregivers’ health and well-being. As in a prior study with PLHIV, we found separate dimensions of HIV caregivers’ HRQOL across domains of physical and mental health [67] and identified factors that differentially affect their physical and mental HRQOL. Specifically, affiliative stigma, caregiver-recipient communication, care recipients’ reciprocity of support and other psychosocial factors, under-studied in the HRQOL or caregiving literatures, appeared to indirectly affect physical and mental HRQOL through their effects on secondary stress and role ambivalence.

Caregiver demographics

Caregivers in our study were similar in demographic characteristics (e.g. race/ethnicity, age) to their care recipient and the majority were kin. Similar proportions of caregivers as care recipients reported probable depression (one-third of caregivers versus two-fifths of care recipients), and nearly half of caregivers reported living with HIV themselves. While caregivers of older persons in prior studies usually have greater resources and stability compared to their care recipients, the caregivers in the present study appeared to be only somewhat better off than their care recipient, and overall to be a highly disadvantaged population of caregivers. Overall, findings suggested that caregiving-related social and interpersonal factors appeared to have more importance to caregivers’ mental HRQOL than their physical HRQOL. Future studies are merited to examine effects of social and interpersonal factors on their physical HRQOL via effects on caregivers’ self-care behaviors.

Care provision and primary stress

The study findings support prior research theories of caregiving as a stress process. We found that care provision and care-related primary stress were associated with secondary stress, which was in turn associated with physical and mental HRQOL. These findings support Pearlin’s conceptualization of caregiving as an interpersonal process of stress proliferation, whereby caregivers’ appraisals of caregiving-related stress, as well as care recipients’ coping behaviors, impact caregivers’ health and well-being [35,54,60]. In our model, we included both objective stressors, as measured by care provision, and the stress stemming from the care activities themselves, as represented by primary stress. Consistent with the stress buffering hypothesis of social support, we observed that positive caregiver-recipient relationships, indicated by better caregiver-recipient communication and caregiver perceived reciprocity, buffered associations between primary stress and health outcomes of mental and physical HRQOL [68]. This adds to the dearth of evidence on caregivers’ resiliencies.

Direct demands of caregiving (primary stress) were associated with subjective perceptions of care (secondary stress), as theorized and supported with empirical evidence [30]. Primary stressors, defined as hardships in caregiving due to, for example, the care recipient’s amount of care needs, along with care provision, were found to be associated with secondary stressors, defined as stress related to caregiving interferences in valued roles and strains on resources consequent to caregiving activities. Compared to the more subjective reports of care-related stress, care provision refers to the instrumental activities and objective demands on the caregiver’s time. The findings suggest that both these objective and subjective aspects of care provision impeded HIV caregivers’ ability to care for loved ones as well as tend to their own health needs.

Care-related secondary stress

Consistent with growing objective evidence, results revealed that subjective care-related secondary stress was strongly correlated of physical HRQOL, suggesting that caregiving activities impinged upon the caregivers’ physical health. It is plausible that instrumental assistance affected physical strains on caregivers directly, or that mental stress related to care provision or to other secondary stresses such as conflict with the care recipient’s family and friends adversely affect caregivers’ somatic symptoms and physical health. Further study is needed on the impact of caregivers’ physical health via engagement in healthcare, adherence to medical regimens and other self-care and coping behaviors.

Care-related secondary stress was also associated with perceived affiliative HIV stigma. While considerable research attention has been given to effects of HIV stigma on PLHIVs’ health-related behaviors and outcomes, our study suggests the role of affiliative stigma on caregivers’ own health and well-being. HIV affiliative stigma may amplify stigma perceptions based on other aspects of care recipients, such as drug use behaviors, poverty or disability. Caregiver’s and recipient’s perceived stigma may exacerbate each other’s stress and strain their relationships, threatening relationship viability. Interestingly, though, affiliative stigma was not directly associated with caregiver role ambivalence, rather, it affected role ambivalence through its adverse effects on family relationships and other stress secondary to caregiving.

Stigma and ambivalence

Study findings on affiliative stigma and role ambivalence extend the literature on caregiver stress. Affiliative stigma did not have a direct effect on mental or physical HRQOL, as was found in previous studies [20,69], but rather, an indirect effect through caregiver role ambivalence. Role ambivalence is an understudied construct in the caregiving literature and is related to role captivity (perceived inability to escape caregiving-related stress) [37]. In contrast to prior observations on the role of HIV caregiver role captivity in health, HIV caregiver role ambivalence was found to play a pivotal role in HIV caregiver HRQOL in our study. It is plausible that this reflects more tenuous main caring relationships in the present study population.

Ambivalence about caregiving, such as wishing one was less obligated to help or wanting time away from the care recipient, was strongly negatively associated with caregivers’ mental HRQOL. As in other populations, HIV caregiving in this population may represent a role that is tacitly defined and not consciously assumed, but rather evolves over time as illness progresses and support needs increase [37,55]. Ambivalence about assisting the care recipient may reflect the quality of the relationship; and has been considered to contribute to negative caregiver outcomes, such as perceived caregiving burden, caregiver role overload or entrapment, which have implications for care quality and cessation of caregiving [70]. The finding suggests that caregiver role ambivalence merits further attention in the caregiving literature, especially in studies with vulnerable populations with serious illness and their caregivers.

Positive caregiving relationship

Having a positive caregiving relationship, defined by better communication and greater reciprocity of support to the caregiver, was negatively associated with caregivers’ care-related primary stress, such as care recipients’ greater amount of needs, problem behavior and/or conflict with the recipients’ family or friends. It was also negatively associated with role ambivalence and thereby indirectly positively associated with MHRQOL. Caregiving relationship quality was not directly associated with secondary care-related stress. This may be due to already strained caregiving relationships due to stigma, drug use, disability and poverty of the care recipient and/or caregiver. There was, however, a negative indirect effect due to the positive association between primary and secondary stress. Likewise, there was a weak negative indirect effect between positive caregiving relationship and care provision. This makes intuitive sense since we would tend to expect greater care provision to be associated with greater disability and greater disability can impair communication and the ability to reciprocate.

Our study results suggest the importance of community support mobilization and caregiver-recipient dyad training to improve quality support exchange in vulnerable populations and access to respite services for caregivers. Observed interdependencies among caregivers and their care recipients suggests that identifying and improving the quality and effective functioning of these relationships interpersonal coping with serious chronic conditions, may be beneficial for HIV caregivers’ HRQOL. Caregivers who reported poor communication with and low reciprocity of support from their care recipients may benefit from dyad-focused interventions emphasizing context-specific social and emotional interpersonal skills building.

Limitations

We cannot determine causal directions of associations only direct and indirect associations between the factors because the data were cross-sectional. Generalizability may be limited due to the sampling of caregivers of PLHIV, all of whom were on antiretroviral therapy and most virally suppressed. Moreover, there was some overlap in HIV status between caregivers and care recipients. This, along with observed interdependence of the relationships, call into question to what extent the designation of caregiver for these dyads was arbitrary. Additionally, most of the sample comprised low-income African American women, a caregiver population that provides higher levels and more intensive forms of caregiving compared to caregivers of other incomes or race. These results may not generalize to HIV caregivers of other income-race groups or to low-income, African American caregivers of persons with other serious conditions. It is also possible that generalizability to other populations is affected by temporal and population-specific factors of African Americans’ disproportionate burden of serious health conditions and more accelerated aging with HIV compared to other racial/ethnic groups of PLHIV. This may affect these care recipients’ greater need for or reliance on caregivers and challenges accessing health resources, as well as additional stigma and stress. Finally, although there was significant variance explained by the mental health dimension of HRQOL, the R-squared value for the physical HRQOL was relatively low, suggesting that the model was omitting important measured variables and latent factors that may explain additional variance. Alternatively, it may be that mental HRQOL and physical HRQOL are distinct constructs with separate pathways.

Conclusions

The study identified aspects of caregiving and the caregiving relationship directly or indirectly associated with HRQOL among caregivers of a population with health disparities, stigma and vulnerability to alienation. Specifically, the study examined dyadic relationship variables, such as reciprocity of support and care provision, as well as communication and affiliative stigma. Results suggest that positive relationship and caregiving-related secondary stress indirectly affect MHRQOL through caregiver role ambivalence. Caregiving-related secondary stress was negatively associated with PHRQOL, however there was no evidence of a significant effect of either positive relationship quality or role ambivalence. The findings contribute to theories and methods of examining effects of caregiving on the mental and physical health and well-being of caregivers and contribute to an understanding of caregiving as an interpersonal stress process. With population aging and growing needs for community caregiving of persons living with HIV, drug use disorders and other serious chronic conditions, our findings suggest the need for identification of their caregivers and dyad-focused intervention for promoting quality of caregiving relationships and mitigating caregiving effects on their HRQOL. Intervention is especially indicated for dyads with poor relationship quality and caregivers reporting greater caregiving-related stress and ambivalence about their caregiving role within disadvantaged African American and low-income communities.