Taking a Break: Daily Respite Effects of Adult Day Services as Objective and Subjective Time Away From Caregiving

Abstract

Background and Objectives

Adult day services (ADS) can provide emotional and physical relief for caregivers of persons with dementia. While prior studies conceptualized ADS use at the aggregate level as a dichotomous construct, little is known about objective and subjective respite as distinct constructs. This study investigated how objective and subjective breaks from caregiving were associated with caregivers’ daily emotional well-being.

Research Design and Methods

Family caregivers (N = 173) whose relatives were using ADS at least twice a week participated in daily interviews over 8 consecutive days (day N = 1,359). Participants provided information on daily respite hours and daily affect. They also reported perceived frequency of breaks from caregiving responsibilities and primary and secondary caregiving stressors (i.e., overload and work conflict). Multilevel models were used to examine the research questions.

Results

On average, caregivers reported 7.12 respite hours on ADS days and 1.74 respite hours on non-ADS days. Having more objective respite was associated with higher positive affect, whereas more subjective respite was associated with lower negative affect, after controlling for ADS use and other covariates. Further, caregivers with greater work conflict experienced more benefits to their positive affect as a result of objective respite.

Discussion and Implications

Objective and subjective respite are unique aspects of caregiving that may have varying impact on caregivers. Respite may be especially beneficial for caregivers experiencing conflict between work and caregiving.

Adult day services (ADS) provide out-of-home care and supervision to individuals with various functional and cognitive impairments. These services are designed to meet the needs of persons with dementia (PWDs), as well as provide respite to caregivers through planned, temporary relief from the accumulation of primary (care-related) and secondary stressors (Fields et al., 2014). Numerous studies have tested the effectiveness of ADS use in attenuating negative caregiver outcomes, with some evidence suggesting ADS’s contribution to reducing stress, burden, and worry (Måvall & Thorslund, 2007; Zarit et al., 2011).

Most studies have relied on dichotomous constructs at the aggregate level such as ADS use or nonuse to explain disparities in caregiver well-being outcomes (Gramain & Malavolti, 2004; Schacke & Zank, 2006). In doing so, these studies assume that ADS use reduces care hours and increases the frequency of getting a break without measuring actual time away (objective respite) or perceived breaks from care responsibilities (subjective respite). Further, although caregivers with higher levels of stress, as measured by the primary subjective stressor of role overload and secondary stressor of work conflict, may be most in need of respite services, few have examined these contextual factors as moderators of respite and caregiver well-being (Pearlin et al., 1990). Taken together, the current study aimed to examine how objective and subjective respite affected daily emotional well-being of dementia family caregivers who were using ADS. We also examined the moderating effects of role overload and work conflict on these associations between respite and caregivers’ daily emotional well-being.

Caregiver Respite and Well-Being

Positive and negative affect are not mere bipolar opposites; the experience of emotion is context-dependent. According to the Dynamic Model of Affect (DMA), in situations of acute stress, a person’s attention narrows to the threat at hand, producing a bias toward processing negative information over positive in order to react quickly and adaptively. Under increasing levels of stress, the relationship between positive and negative affect becomes more strongly, inversely correlated. However, in situations of low stress, persons are able to simultaneously process both positive and negative characteristics of the environment (Zautra et al., 2005). In these contexts of reduced stress, which may include respite, positive and negative affect are largely independent of one another. Therefore, it is important to treat positive and negative affect as distinct constructs. Most prior studies have attended to the reduction of negative outcomes, such as stress, depression, and burden; yet, it is unclear if taking time away can boost positive affect (Fields et al., 2014). Positive affect may offer unique benefits to the caregiver such as creativity in problem-solving and decreased rumination (Folkman & Moskowitz, 2000).

It is likely that during respite, caregivers experience lower negative affect and higher positive affect than they did under the stress of dementia caregiving. It is possible, however, that positive affect depends more on the status of the PWD and the PWD–caregiver relationship. In one qualitative diary study of 57 Chinese primary caregivers of a family member with Alzheimer’s disease, caregivers experienced positive gains when witnessing the PWDs’ condition stabilize or improve, or when bonding with the PWD (Cheng et al., 2016). Therefore, increasing positive affect may require direct positive changes in the care situation. Respite may increase a caregiver’s patience for the role and thereby buffer negative emotions, but the challenges of caregiving could remain overwhelming.

Objective Respite: Actual Respite Hours Away From Caregiving

Although rarely examined, actual respite hours away from caregiving can be a meaningful predictor of well-being associated with ADS use. ADS can reduce the number of hours that a caregiver spends on care-related tasks, and this reduction in care hours has been related to decreased burden and increased well-being (Måvall & Thorslund, 2007). According to Hobfoll’s (1989) Conservation of Resources theory, time is a valuable resource that individuals aim to gain and preserve. The constraints of a caregiving role can lead to the depletion of one’s resources (e.g., objective time), which then results in stress; actual respite hours, therefore, can replenish one’s resources. Gaugler et al. (2003) indicated that hours of ADS use predicted significant reductions in role overload among caregivers using ADS for 3 months. Further, Zarit and colleagues (2011) found from a daily diary study of 121 caregivers in New Jersey that ADS days were associated with reduced exposure to stressors caused by dementia-related behavior problems, which was thought to operate through reduced time with the PWD. Although objective respite offers caregivers a physical break from direct care of the PWD, this time away may still be spent on care-related responsibilities. Therefore, the relationship between actual hours away and caregivers’ mood should be thoughtfully explored.

Subjective Respite: Perceived Breaks From Caregiving

Just as objective respite is thought to be an important resource for caregivers, subjective respite can be equally valuable. One psychological facet of time resources is perceived time adequacy, or the extent to which individuals feel they have sufficient time to accomplish desired activities (Lee et al., 2015). According to the Conservation of Resources theory, variations in stressors and resources in multiple roles can predict one’s perceived time adequacy. Employed caregivers, for example, may feel they have less time for family caregiving than other groups of caregivers if they experience more stressors and fewer resources in work and family contexts. Given the demands of the dementia caregiving role, which typically extend for many years, dementia caregivers may feel that they lack a balance between time spent providing care and time spent on other responsibilities, such as employment.

Lee et al. (2015) have argued that perceived time is distinct from actual time: psychological assessments of time are context-dependent. For example, caregivers may feel that they have personal and free time resources during care hours. From in-depth interviews with 243 informal caregivers, Chappell and colleagues (2001) identified alternative definitions of respite; “stolen moments,” for instance, refer to activities or situations that temporarily relieve a caregiver within the constraints of ongoing caregiving responsibilities, such as watching the evening news, drinking coffee, or going on a vacation, all with the PWD present. Respite, too, may simply refer to instances where the PWD is comfortable or “angst-free” (Chappell et al., 2001). Taken together, caregivers may experience emotional and physical breaks even when with the PWD. Respite, therefore, is not restricted to actual hours away from the PWD, and subjective evaluations of time away should be considered.

Primary and Secondary Stressors as Moderators: Role Overload and Work Conflict

Although many caregivers may experience similar care-related stressors such as problematic behaviors and memory issues, they may not respond to these common stressors in the same way. In line with the Conservation of Resources theory, feelings of role overload can arise when caregivers perceive that they do not have adequate resources, such as time, to fulfill the demands of multiple roles (DePasquale et al., 2018).

Role overload, a subjective primary stressor, refers to the burnout or fatigue associated with care-related activities (Pearlin et al., 1990). Highly stressed caregivers likely have lower perceptions of time adequacy for activities outside of caregiving, such as caring for their own emotional well-being. One study indicated that, compared to caregivers for older adults with less severe disabilities, those caring for older adults with moderate physical and mental disabilities experienced greater reductions in burden as a result of ADS use (Gramain & Malavolti, 2004). Thus, it is possible that caregivers with greater care demands and potentially higher levels of stress may benefit more from respite.

Given that many family caregivers are employed either full- or part-time, another important caregiving context is work conflict—the way in which caregivers experience the task of balancing the caregiving role with professional work. A caregiver’s multiple roles are thought to compete for a finite supply of resources, thus causing difficulty in meeting role demands (DePasquale et al., 2018). Although evidence suggests that ADS use can help attenuate stress associated with work conflict (Schacke & Zank, 2006), few studies examined the intensity of work conflict as a moderator of the association between caregiving respite and well-being.

Those with high levels of work conflict would likely benefit most from respite, as the intense demands of paid work are thought to increase strain for caregivers juggling multiple roles (Utz et al., 2012). Caregivers with more work conflict are likely highly motivated to use free time wisely, whether for leisure or for accomplishing necessary tasks. On the other hand, it is also possible that even with significant amounts of respite time, employed caregivers who are experiencing high levels of work conflict use most of their free time worrying about their job.

Advantages of Daily Diaries

Much of the literature concerning ADS has been cross-sectional or longitudinal with follow-up measurements at two or three time points. Cross-sectional designs may confound between-person differences and within-person variations in outcomes. Additionally, most longitudinal respite studies ask caregivers to average their emotional responses over a period of a week or longer, conflating the effect of both ADS and non-ADS days (e.g., Måvall & Thorslund, 2007). Because ADS use may bring more day-to-day variations in time away from caregiving and thus contribute to daily variations in well-being, it is valuable to use a within-person perspective with a closer time proximity for self-reports to improve recall accuracy (Zarit et al., 2011).

The Current Study

The present study aims to expand upon previous literature regarding the impact of ADS use on caregiver well-being in four ways. First, we considered positive affect in addition to negative affect as an outcome variable. Second, we examined objective respite and subjective respite as distinct constructs, and examined how each related to caregivers’ daily affect. Third, we also examined how primary and secondary stressors of caregiving moderated the associations between objective and subjective respite and well-being. Last, we used daily diaries over eight consecutive days to examine day-to-day variations in objective respite hours. The intensive longitudinal method had the advantage of examining within-person associations between respite use and emotional well-being with fewer retrospective report errors. We tested the following hypotheses:

Hypothesis 1: Caregivers would have higher positive affect and lower negative affect on days when they had more respite hours (objective respite), compared to days when they had fewer respite hours (1a); caregivers who perceived more breaks away from caregiving (subjective respite) at baseline would have higher positive affect and lower negative affect across days, compared to their counterparts who perceived less breaks (1b).

Hypothesis 2: Caregivers with higher levels of role overload would experience a greater increase in positive affect and a greater decrease in negative affect on days when they had more objective respite (2a) and when they had more subjective respite at baseline (2b).

Hypothesis 3: Caregivers with higher levels of work conflict would experience a greater increase in positive affect and a greater decrease in negative affect on days when they had more objective respite (3a) and when they had more subjective respite at baseline (3b).

Method

Participants and Procedure

Data for this study were collected from the Daily Stress and Health (DaSH) study (Zarit et al., 2014a). Eligible participants were primary family caregivers residing with individuals with dementia who had been using an ADS program at least 2 days per week for a minimum of 1 month at baseline. ADS programs were identified through state associations in New Jersey, Pennsylvania, Virginia, and Colorado, and caregivers from a total of 57 programs contacted study personnel for eligibility screening. Of the 241 caregivers interested in participating, 41 (17%) did not meet eligibility criteria, and 16 (8%) of the remaining 200 chose not to enroll. Of the 184 caregivers who did participate, two caregivers (1%) did not complete daily interviews, and nine (4.5%) were excluded from the analysis because their interviews did not capture data from both days when the PWD attended ADS and did not attend ADS. Thus, the final analytic sample included 173 caregivers (86.5% of eligible participants). Caregivers were offered a small remuneration for their participation.

In-person home interviews were conducted to collect baseline data and sociodemographic information. Daily phone interviews took place over the course of 8 days (day N = 1,359), during which PWDs either attended ADS (52% of the eligible days) or remained in the home (48% of the eligible days).

Measures

Daily affect

Measures of affect were collected on each of the 8 days via daily phone interviews. Daily affect was measured using an adapted inventory from the Non-Specific Psychological Distress Scale (22 items; Kessler et al., 2002). Two items (i.e., interested, attentive) from the Positive and Negative Affect Schedule (Watson et al., 1988) were added to capture the breadth of positive emotions. Caregivers rated the frequency of each emotion occurring on a day using a 5-point scale (1 = none of the day to 5 = all day) so that higher scores indicated higher levels of the respective affect. To confirm the structure of positive and negative domains relevant to caregiver affect, we performed a factor analysis using the 24 items. We found that four items did not load well. For the current study, we used nine items for positive affect (mean scores; α = .92) and 11 items for negative affect (mean scores; α = .88).

Objective respite

On days when the PWD used ADS, caregivers were asked, “What time did your relative leave for and return from day care?” On days when the PWD remained at home, caregivers were asked, “What time were you away from your relative today?” Using the answers, hours caregivers spent away from the primary caregiving role were calculated for both ADS and non-ADS days (military hours).

Subjective respite

Overall perceived breaks were measured during in-person interviews at baseline. Participants were asked, “In a typical week, how often are you able to get a break from caregiving responsibilities?” Responses were coded on a scale from 1 (never) to 4 (often) such that higher scores indicate more perceived breaks from caregiving.

Role overload

Caregivers’ perceptions of the time and effort they invest in providing care were measured using a 3-item scale adapted from Pearlin and colleagues (1990) during baseline interviews. The scale asks participants to report how often they have found each statement to be true in the past week: (a) “You are exhausted when you go to bed at night”; (b) “You have more things to do than you can handle”; and (c) “You have time just for yourself” (reverse-coded). Responses were rated from 1 (none of the time) to 4 (all of the time) such that higher scores indicated greater feelings of role overload (mean scores; α = .63).

Work conflict

Caregivers’ perceived stressors at work were measured using a 9-item scale from Aneshensel and colleagues (1995) during baseline interviews. The scale asks employed participants to report how often they have found each statement of negative and positive experiences associated with work to be true in the past week (e.g., I have had too little energy for work; I have worried about my relative while I am at work; I have felt successful at my job). Responses were rated from 1 (none of the time) to 4 (all of the time) and positive experiences were reverse-coded such that higher scores indicated more work conflict while providing care (mean scores; α = .74).

Covariates

We included covariates at both person and daily levels that have been commonly associated with caregivers’ time use and well-being in prior studies (see Supplementary Table 1 for correlations among variables). As between-person covariates, we considered caregivers’ sociodemographic characteristics, including gender (1 = female, 0 = male), relation to PWD (1 = child, 0 = others), employment status (1 = employed, 0 = not-employed), education level (1 = less than high school to 6 = postcollege degree), race (1 = White, 0 = others), self-rated health (1 = poor to 5 = excellent), and neuroticism. Neuroticism was included given previous findings that individuals with greater neuroticism display higher emotional reactivity to events that induce negative mood (Larsen & Ketelaar, 1989). Neuroticism was assessed with four items that asked participants to what extent each adjective described them (i.e., moody, worrying, nervous, and calm [reverse-coded]; Lachman & Weaver, 1997). Responses were rated from 1 (not at all) to 5 (a great deal); higher scores indicated greater neuroticism (mean scores; α = .72). We also included caregiving-related variables measured at baseline: duration of care (in months), relationship quality with PWD, PWD’s activities of daily living (ADL) impairments and memory problems, and number of ADS days. Relationship quality was assessed with two items (i.e., how close do you feel and how well do you get along together). Response options ranged from 1 (not at all) to 4 (very); higher scores indicated better relationship quality (α = .70). ADL impairments were assessed using 13 items (Katz et al., 1963; Lawton & Brody, 1969). Responses were rated from 1 (does not need help) to 4 (cannot do without help); higher scores indicated greater impairment in PWD (mean scores; α = .83). We also assessed the frequency of PWD’s memory symptoms in the past week using seven items from the Revised Memory and Behavior Problems Checklist (Teri et al., 1992). Responses were rated from 1 (did not occur) to 5 (often/everyday); higher scores indicated worse cognitive functions (α = .70).

For daily characteristics (i.e., within-person covariates), we controlled for ADS use status (1 = ADS used, 0 = no ADS), care-related stressors, noncare-related stressors, and day of the week (1 = weekday, 0 = weekend). Daily care-related stressors were evaluated with the Daily Record of Behavior, a 19-item scale assessing the daily frequency of behavioral and psychological symptoms of dementia (Femia et al., 2007). Caregivers were asked to report the occurrence of PWD behavior problems (e.g., ADL resistance, restless behaviors, disruptive behaviors, depressive behaviors, and reality problems) during three time periods each day. The total number of problems each day were calculated (range = 0–57). Noncare-related stressors, including those related to caregivers’ social networks, health, finances, work, or any other incidents, were measured using the 8-item Daily Inventory of Stressful Events (Almeida et al., 2002). The number of events were similarly calculated (range = 0–8).

Analytic Strategy

In our preliminary analysis, we first examined the descriptive statistics of caregivers’ sociodemographic characteristics and day characteristics (such as the use of ADS and the experience of daily stressors and affect). Further, we looked at whether daily objective respite differed by daily ADS use and employment status.

We estimated 2-level multilevel models to examine caregivers’ daily affect as the outcome (i.e., days nested within persons). To address the first hypothesis regarding the association between respite and daily affect, we tested the main effects of objective respite (actual time away centered at person-mean to represent the within-person effect; H1a) and subjective respite (perceived breaks; H1b) on daily affect outcomes (i.e., positive and negative affect, separately):

$${\displaystyle \begin{array}{ll}{\displaystyle \mathrm{A}\mathrm{f}\mathrm{f}\mathrm{e}\mathrm{c}{\mathrm{t}}_{ti}=}& {\gamma }_{00}+{\gamma }_{10}(\mathrm{O}\mathrm{b}\mathrm{j}\mathrm{e}\mathrm{c}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{e}\mathrm{s}\mathrm{p}\mathrm{i}\mathrm{t}{\mathrm{e}}_{ti})\\ & +{\gamma }_{01}(\mathrm{S}\mathrm{u}\mathrm{b}\mathrm{j}\mathrm{e}\mathrm{c}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{e}\mathrm{s}\mathrm{p}\mathrm{i}\mathrm{t}{\mathrm{e}}_i)+e_{ti}\end{array}}$$

For the second and third hypotheses, we tested moderation effects of stress appraisals to address whether the association between respite and daily affect varied by levels of role overload (H2a & H2b) and work conflict (H3a & H3b). Interaction terms were computed between these stress appraisals (i.e., role overload and work conflict) and respite (i.e., daily actual hours and perceived breaks), respectively:

$${\displaystyle \begin{array}{l}{\displaystyle \begin{array}{ll}{\displaystyle \mathrm{A}\mathrm{f}\mathrm{f}\mathrm{e}\mathrm{c}{\mathrm{t}}_{ti}=}& {\gamma }_{00}+{\gamma }_{10}(\mathrm{O}\mathrm{b}\mathrm{j}\mathrm{e}\mathrm{c}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{e}\mathrm{s}\mathrm{p}\mathrm{i}\mathrm{t}{\mathrm{e}}_{ti})+{\gamma }_{01}(\mathrm{S}\mathrm{u}\mathrm{b}\mathrm{j}\mathrm{e}\mathrm{c}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{e}\mathrm{s}\mathrm{p}\mathrm{i}\mathrm{t}{\mathrm{e}}_i)\\ & +{\gamma }_{02}(\mathrm{S}\mathrm{t}\mathrm{r}\mathrm{e}\mathrm{s}\mathrm{s}\mathrm{a}\mathrm{p}\mathrm{p}\mathrm{r}\mathrm{a}\mathrm{i}\mathrm{s}\mathrm{a}{\mathrm{l}}_i){\displaystyle +{\gamma }_{11}(\mathrm{O}\mathrm{b}\mathrm{j}\mathrm{e}\mathrm{c}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{e}\mathrm{s}\mathrm{p}\mathrm{i}\mathrm{t}{\mathrm{e}}_{ti})}\\ & \times (\mathrm{S}\mathrm{t}\mathrm{r}\mathrm{e}\mathrm{s}\mathrm{s}\mathrm{a}\mathrm{p}\mathrm{p}\mathrm{r}\mathrm{a}\mathrm{i}\mathrm{s}\mathrm{a}{\mathrm{l}}_i){\displaystyle +{\gamma }_{03}(\mathrm{S}\mathrm{u}\mathrm{b}\mathrm{j}\mathrm{e}\mathrm{c}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{e}\mathrm{s}\mathrm{p}\mathrm{i}\mathrm{t}{\mathrm{e}}_i)}\\ & \times (\mathrm{S}\mathrm{t}\mathrm{r}\mathrm{e}\mathrm{s}\mathrm{s}\mathrm{a}\mathrm{p}\mathrm{p}\mathrm{r}\mathrm{a}\mathrm{i}\mathrm{s}\mathrm{a}{\mathrm{l}}_i)+e_{ti}\end{array}}\end{array}}$$

In all models, we controlled for within- and between-person covariates which were known to be associated with caregiver daily affect—including daily ADS use, daily care and noncare stressors, weekday, caregiver and PWD characteristics as well as the person-mean levels of the three daily variables (i.e., time away, care stressors, and noncare stressors). Missing data in daily affect outcomes were handled using restricted maximum likelihood (SAS PROC MIXED).

Results

Descriptive statistics of our sample are shown in Table 1. Caregivers were, on average, 62 years old, and the majority of them were female (87%). Among them, 58% were caring for a parent and 42% were employed. On average, participants used ADS for 4 days during the 8-day interview period. Participants reported that in a typical week they were able to get a break from caregiving responsibilities rarely to some of the time (M = 2.91, SD = 0.77).

Table 1.

Summary of Key Variables

Variable	M     (SD)	Range
Person characteristics
Age	61.97 (10.66)	39–89
Female, %	87
Relation to PWD, %
Adult child	58
Spouse	38
Other	4
Educationa	4.46 (1.20)	1–6
White, %	72
Employed, %	42
Duration of care (month)	61.12 (45.55)	3–264
Self-rated healthb	3.24 (0.99)	1–5
Neuroticismc	2.53 (0.80)	1–4.75
Relationship quality with PWDd	3.57 (0.66)	1–4
PWD ADL impairmente	3.06 (0.49)	2–4
PWD memory problemsf	3.11 (0.91)	1–5
Total number of ADS days	4.09 (1.46)	1–6
Perceived breaks from caregivingg	2.91 (0.77)	1–4
Stress appraisals
Role overloadh	2.82 (0.65)	1.33–4
Work conflicti	1.81 (0.50)	1–4
Day characteristics
ADS day, %	52
Care stressorj	4.31 (5.74)	0–46
Noncare stressork	1.16 (1.26)	0–6
Weekday, %	73
Daily affect
Positive affectl	3.02 (0.94)	1–5
Negative affectm	1.45 (0.52)	1–4.64

Notes: Caregiver N = 173; caregiver-day N = 1,359. ADL = activities of daily living; ADS = adult day services; PWD = person with dementia.

^a1 = less than high school to 6 = postcollege degree. ^b1 = poor to 5 = excellent. ^cMean scores of four items (1 = not at all to 5 = a great deal). ^dMean scores of two items (1 = not at all to 4 = very). ^eMean scores of 13 items (1 = independent to 4 = cannot do without help). ^fMean scores of seven items (1 = did not occur to 5 = often/everyday). ^g1 = never to 4 = often. ^hMean scores of three items (1 = none of the time to 4 = all of the time). ⁱMean scores of nine items (1 = none of the time to 4 = all of the time); asked only for employed caregivers (n = 72). ^jSum of 57 behavioral and psychological symptoms of dementia. ^kSum of eight stressful events not directly related to the caregiving role. ^lMean scores of nine items (1 = none of the day to 5 = all day). ^mMean scores of 11 items (1 = none of the day to 5 = all day).

Table 2 showed objective respite by ADS use. On average, caregivers reported 7.12 hr away from the PWD on ADS days and 1.74 hr on non-ADS days. The difference in objective respite by ADS use was more pronounced for nonemployed caregivers (6.73 vs 0.90 hr) than employed caregivers (7.62 vs 3.11 hr). We confirmed that ADS use as well as employment status were significant predictors (p < .001) of daily objective respite, after controlling for other caregiver and day characteristics (see Supplementary Table 2).

Table 2.

Daily Actual Hours Away From Caregiving by ADS Use and Employment Status

	ADS day (n = 707)		Non-ADS day (n = 652)
Sample	M  (SD)	Range	M  (SD)	Range	t
Total sample	7.12 (1.47)	2–13	1.74 (2.85)	0–16	42.94***
Employed (n = 72)	7.62 (1.55)	2.25–11.83	3.11 (3.76)	0–16	17.61***
Nonemployed (n = 101)	6.73 (1.27)	1.50–13	0.90 (1.63)	0–13.25	56.06***

Notes: Caregiver N = 173; caregiver-day N = 1,359. Number of hours shown in military time format. ADS = adult day services.

***p < .001.

To address the first hypothesis, we examined effects of objective respite (H1a) and subjective respite (H1b) on caregiver daily affect (see Table 3). Controlling for both within- and between-person characteristics, we found that having more daily objective respite was associated with higher levels of positive affect (B = 0.03, p < .001), but not with negative affect on that day. In contrast, more subjective respite was associated with lower levels of negative affect (B = −0.08, p = .013), but not with positive affect.

Table 3.

Multilevel Models for Associations Between Respite Time and Daily Affect

	Positive affect	Negative affect
Variable	B    (SE)	B    (SE)
Fixed effects
Intercept	2.07**    (0.71)	1.41*** (0.33)
Objective respitea	0.03*** (0.01)	0.00         (0.00)
Subjective respiteb	0.12         (0.07)	−0.08*     (0.03)
Within-person covariates
ADS day	−0.04         (0.05)	−0.05         (0.03)
Care stressorc	−0.01**    (0.00)	0.03*** (0.00)
Noncare stressorc	−0.08*** (0.01)	0.10*** (0.01)
Weekday	−0.11**    (0.04)	0.06**    (0.02)
Between-person covariates
Female	0.17         (0.16)	−0.03         (0.07)
Adult child	0.15         (0.12)	−0.07         (0.06)
Educationd	−0.07         (0.05)	−0.03         (0.02)
White	−0.27         (0.14)	0.11         (0.06)
Employed	−0.01         (0.14)	−0.11         (0.06)
Duration of care (month)	0.00         (0.00)	0.00         (0.00)
Self-rated healthe	0.19**    (0.06)	−0.04         (0.03)
Neuroticismf	−0.36*** (0.08)	0.20*** (0.04)
Relationship qualityg	0.22*        (0.09)	0.02         (0.04)
PWD ADL impairmenth	−0.06         (0.11)	−0.08         (0.05)
PWD memory problemsi	0.11         (0.07)	−0.02         (0.03)
Total number of ADS days	0.01         (0.05)	0.02         (0.02)
Objective respitej	0.07         (0.05)	0.00         (0.02)
Care stressorj	−0.01         (0.01)	0.03*** (0.01)
Noncare stressorj	−0.03         (0.07)	0.11**    (0.03)
Random effects
Intercept variance	0.44*** (0.05)	0.09*** (0.01)
Residual variance	0.22*** (0.01)	0.08*** (0.00)
−2 log likelihood	2,328.6	859.3

Notes: Caregiver N = 173; caregiver-day N = 1,359. ADL = activities of daily living; ADS = adult day services; PWD = person with dementia.

^aDaily actual time away (hour); person-mean centered (i.e., time-varying). ^bPerceived breaks from caregiving; 1 = never to 4 = often (i.e., time-invariant). ^cPerson-mean centered (i.e., time-varying). ^d1 = less than high school to 6 = postcollege degree. ^e1 = poor to 5 = excellent. ^fMean scores of four items (1 = not at all to 5 = a great deal). ^gMean scores of two items (1 = not at all to 4 = very). ^hMean scores of 13 items (1 = independent to 4 = cannot do without help). ⁱMean scores of seven items (1 = did not occur to 5 = often/everyday). ^jPerson-mean across days (i.e., time-invariant).

*p < .05. **p < .01. ***p < .001.

Next, for the second and third hypotheses, we examined whether role overload (H2a & H2b) and work conflict (H3a & H3b) moderated the association between respite and caregiver daily affect (see Table 4). The results showed that work conflict moderated the association between objective respite and positive affect (B = 0.03, p = .029), such that it was stronger for those with higher levels of work conflict. However, work conflict did not moderate the relationship between subjective respite and daily affect, whether positive or negative. Role overload did not have any moderating effects. Thus, the association between respite time (both objective and subjective) and daily affect did not vary by levels of role overload.

Table 4.

Multilevel Models for Moderation Effects of Stress Appraisals on the Associations Between Respite Time and Daily Affect

	Positive affect		Negative affect
	Main effects	Moderation effects	Main effects	Moderation effects
Variable	B    (SE)	B   (SE)	B            (SE)	B                 (SE)
Moderated by role overload
Objective respitea	0.03*** (0.01)	0.00    (0.02)	0.00    (0.00)	0.00 (0.01)
× Role overloadb	—	0.01    (0.01)	—	0.00 (0.00)
Subjective respitec	0.12            (0.07)	0.07    (0.32)	−0.08* (0.03)	0.10 (0.15)
× Role overloadb	—	0.02    (0.11)	—	−0.06 (0.05)
Role overloadb	−0.09            (0.10)	−0.14    (0.34)	0.01    (0.05)	0.20 (0.16)
−2 log likelihood	2,330.5	2,338.9	863.5	875.0
Moderated by work conflict
Objective respitea	0.03**   (0.01)	0.07    (0.02)	0.00    (0.00)	0.00 (0.01)
× Work conflictd	—	0.03* (0.01)	—	0.00 (0.01)
Subjective respitec	0.03            (0.12)	0.18    (0.39)	−0.01    (0.04)	−0.09 (0.14)
× Work conflictd	—	−0.08    (0.20)	—	0.04 (0.07)
Work conflictd	−0.18                (0.20)	0.06    (0.65)	0.04    (0.07)	−0.09 (0.23)
−2 log likelihood	917.2	920.7	222.8	234.1

Notes: Caregiver N = 173; caregiver-day N = 1,359 for role overload moderation. Employed caregiver n = 72; caregiver-day n = 563 for work conflict moderation. Models are fully adjusted for the same within-person and between-person covariates as the previous models.

^aDaily actual time away (hours); person-mean centered (i.e., time-varying). ^bMean scores of three items (1 = none of the time to 4 = all of the time). ^cPerceived breaks from caregiving; 1 = never to 4 = often (i.e., time-invariant). ^dMean scores of nine items (1 = none of the time to 4 = all of the time); asked only for employed caregivers.

*p < .05. **p < .01. ***p < .001.

Discussion

The current study makes a contribution to the caregiving literature by exploring how ADS can help (in terms of positive and negative affect) and for whom they best help (those with stressors in specific domains). By using daily diary data, we examined caregivers’ mood at a more granular level and not as an aggregate estimated retrospectively over the past week or longer.

Objective and Subjective Respite Are Uniquely Related to Daily Affect

Rather than examining ADS use as a descriptive variable (whether or not it was delivered), we explored how two distinct constructs of respite related to caregiver daily well-being. Results partially supported our first hypothesis. Though we expected both objective and subjective respite to be associated with higher positive affect and lower negative affect, we found each to affect mood uniquely. In line with the DMA, positive and negative affect were not found to be opposite ends of a spectrum.

Daily objective respite was only associated with greater daily positive affect, but not reduced negative affect. Actual physical breaks, which are provided by respite, are an objective time resource that creates opportunities for mental, physical, and social stimulation that could boost positive mood (Chappell et al., 2001). Prior research using the DaSH data set found that family caregivers had more positive experiences on ADS days, reporting more positive events at work and more time “sharing a laugh” with someone (Zarit et al., 2014a). Daily ADS use was also associated with an adaptive physiological response to stress, which in turn was linked to increased positive mood but did not decrease depressive symptoms (Zarit et al., 2014b). Therefore, our findings align with those of previous studies suggesting that objective respite acts more as a resource to boost positive emotions and less as a buffer of negative emotions. These findings support the DMA; caregivers are better able to process positive aspects of the environment during actual breaks from the care role, but can simultaneously attend to negative aspects. Caregivers may engage in mood-boosting activities during time away, but the perceived stressors of the care role might still persist.

We found a different pattern of results for subjective respite. Perceived caregiving breaks were related to lower levels of negative mood, but not to any differences in positive mood. Previous research on family caregivers who also doubled as professional certified nursing assistants found that perceived family time adequacy (the feeling of having sufficient time with one’s family) acted as a psychological resource that buffered the negative effects of work stress on emotional exhaustion (DePasquale et al., 2018). Our finding that subjective respite is associated with negative affect is consistent with cognitive models of depression that emphasize the central role of subjective appraisals of events and resources (e.g., Beck et al., 1979). Caregivers’ subjective appraisals of stress predicted depression better than objective measures of physical and cognitive impairment, behavior problems, or duration of care. Taken together, the findings suggest that subjective and objective time resources affect caregivers differently. Our findings support the claim that the experience of positive and negative affect is context-dependent and affective reactions during respite cannot collapse into one bipolar construct (Zautra et al., 2005).

Role Overload and Work Conflict as Moderators

Regarding the moderating effects of primary and secondary stressors of caregiving, we found partial support for work conflict. Caregivers who experienced higher levels of work conflict while providing care benefited more from respite, but only in terms of their positive mood. Hobfoll’s (1989) Conservation of Resources theory posits that individuals try to make up for lost resources in one domain when relieved of stressors in another. It may be, then, that when caregivers experience a break from the care role, they try to compensate in their professional work. In this way, objective respite may facilitate more time for work, which can lead to an increase in positive feedback and positive affect in this domain. It is also possible that caregivers who are having higher levels of stressors at work, when given the opportunity, will take full advantage of their free time, engaging in recreational and leisure activities that could boost positive mood. A qualitative study that examined time-log diaries of caregivers showed that employed primary caregivers using ADS for their relative were more satisfied with respite time than nonemployed caregivers (Utz et al., 2012). The authors attributed this finding to higher consistency between desired and actual time use in employed caregivers; in other words, employed caregivers actually did what they intended to do. This time-use consistency may be more pronounced in those with a greater need for respite (i.e., those with greater work conflict); the intensity of a high-stress work environment may lead one to select the most enjoyable activities during respite. At the same time, caregivers experiencing higher levels of work stress are likely to bring residual work stressors home with them, allowing negative affect to still permeate free time.

We did not find any support for the moderating effect of role overload. Those with higher role overload might not benefit uniquely from respite time because they may be preoccupied with the caregiving role even during time away. ADS may not offer a sufficient amount of respite for those feeling especially burned out by caregiving. Because overload specifically describes one’s feelings about the care role, caregivers with high stress levels in domains other than caregiving may be promising targets for ADS interventions.

A few limitations of the present study should be discussed. First, there was a relatively small number of caregivers who were employed in the sample, which may affect the generalizability of the current findings on work conflict. Second, we did not have information on how caregivers used time away from caregiving responsibilities. Additionally, inter- and intrapersonal characteristics may affect one’s time use and emotional well-being, such as attachment style, relationship quality, and personality traits. Qualitative research may be best poised to understand the deeper, subjective experiences of caregivers, their family members, and the intricate relationships between them.

Contributions and Future Research Directions

The current study supports previous findings that ADS use promotes caregiver well-being, and carries further implications for respite program development and evaluation. How much time a caregiver spends away from the PWD and how they perceive breaks from caregiving are distinct constructs of respite. Employed caregivers with high levels of work conflict may benefit from ADS, whereas those already highly fatigued from care-related stressors may need a more specialized intervention than taking a break from caregiving.

Future research could address the exact mechanisms by which time away relates to affect. For example, during actual respite time, socializing with others could drive boosts in positive mood. Time away and affect may also differ by other caregiving characteristics, such as relationship conflict or duration of care. Those with greater financial strains, too, may experience respite differently from others. In addition to addressing the psychological outcomes associated with respite programs, future research should also consider the accessibility and affordability of such programs.

Funding

This research was supported by a grant R01 AG031758 “Daily Stress and Health Study” (PI: Steven H. Zarit, PhD) from the National Institute on Aging (NIA).

Conflict of Interest

None declared.