555319 research-article2014

JAHXXX10.1177/0898264314555319Journal of Aging and HealthLiu et al.

Article

Health Trajectories of Family Caregivers: Associations With Care Transitions and Adult Day Service Use

Journal of Aging and Health 2015, Vol. 27(4) 686­–710 © The Author(s) 2014 Reprints and permissions: sagepub.com/journalsPermissions.nav DOI: 10.1177/0898264314555319 jah.sagepub.com

Yin Liu, MS1, Kyungmin Kim, PhD2, and Steven H. Zarit, PhD1

Abstract Objective: The study examines family caregivers’ health changes over 1 year on four health dimensions and explores the association of differential health trajectories with adult day service (ADS) use and caregiving transitions. Method: The participants were 153 primary caregivers of individuals with dementia (IWDs) who provided information on care situations and their own health at baseline, 6-month, and 12-month interviews. Results: Caregivers showed increasing functional limitations and decreasing bodily pain over time, whereas role limitation and general health perception remained stable. Furthermore, caregivers’ trajectories of functional limitation were associated with their extent of ADS use at baseline and their relatives’ placement. Discussion: Health is multidimensional; all dimensions of caregiver health do not change in a uniform manner. The findings underscore the importance of the association of caregiving transitions and caregiver health and the potential health benefits of ADS use for family caregivers.

1The 2The

Pennsylvania State University, University Park, USA University of Texas at Austin, USA

Corresponding Author: Yin Liu, Department of Human Development and Family Studies, The Pennsylvania State University, 315-A Health and Human Development East, University Park, PA 16802, USA. Email: [email protected]

Liu et al.

687

Keywords caregivers, dementia, health, adult day service, caregiver transition

Family caregiving to individuals with dementia (IWDs) is highly stressful and physically demanding. Caregivers, in general, report more health problems than noncaregivers, and dementia caregivers have poorer physical health compared with caregivers of other disease types (Pinquart & Sörensen, 2003; Vitaliano, Zhang, & Scanlan, 2003). Furthermore, caregiving, especially when accompanied by emotional strain, is an independent risk factor for mortality (Perkins et al., 2013). Among studies on caregiver health, most are cross-sectional; longitudinal analyses examining both within-person health trajectories and between-person differences on these individual trajectories and associations are relatively rare. It remains unclear what the typical trajectory of health is for family caregivers of IWDs over time and whether respite use such as adult day service (ADS) and caregiving transitions are associated with individual differences in typical health trajectories. In this study, we examine these questions.

Caregiving and Health Many of the prior enquiries on caregiver health and well-being have been guided by the stress process model (Pearlin, Mullan, Semple, & Skaff, 1990). Assisting dementia patients differs from many other caregiving relationships in that IWDs’ increasing dependency as a consequence of progressive memory loss and chronic impairment restructures caregiving relationships to be dominant and nonreciprocal. From caregivers’ perspective, such relationships may become less encompassing and mutual compared with the dyad’s prior relationship. The stress process model denotes that this imbalanced caregiving relationship creates a rich context for proliferation of the effects of chronic stressors, which may compromise caregivers’ physical health and lower their psychosocial well-being in the long run (Aneshensel, Pearlin, Mullan, Zarit, & Whitlatch, 1995; Pearlin et al., 1990). Providing care for IWDs has been found to be more stressful than caring for older adults who are physically impaired and dependent. Dementia caregiving involves higher levels of physical and emotional demands, which in turn are more likely to take a toll on caregivers’ physical health. The chronic stress of dementia caregiving has been found to affect several dimensions of physical health, including general health perceptions, physical symptoms, illness, medication use, health service use, indicators of sympathetic arousal,

688

Journal of Aging and Health 27(4)

and cardiovascular reactivity. Baumgarten and colleagues (1992) found that caregivers had poorer self-rated health perceptions and greater use of prescription medications. Furthermore, caregivers were twice as likely to experience physical symptoms like persistent cough and fatigue, chest pain, headache, and morning fatigue compared with noncaregivers. These results were robust even after controlling for demographic characteristics, chronic health conditions, recent life events, coping, and social support. Son and colleagues (2007) showed that stressors associated with dementia caregiving had negative impact on caregivers’ health outcomes, specifically poorer selfreported health, more negative health behaviors, and greater use of health care services. Furthermore, Mausbach and colleagues (2010) suggested that the chronic stress of caregiving was associated with compromised endothelial function, which may contribute to the increased risk of cardiovascular diseases in older adult caregivers. Two meta-analyses depict a dismal picture of caregivers’ health and wellbeing. Pinquart and Sörensen (2003) found that caregivers, especially dementia caregivers, had significantly poorer physical health as measured by a single item on general health perceptions. Another meta-analysis by Vitaliano and colleagues (2003) showed that for self-reported health measures, such as general health perceptions and chronic illnesses, caregivers reported more health problems than did noncaregivers. Vitaliano and colleagues also found that caregivers had greater potential illness risks than did noncaregivers as indicated by levels of antibodies and stress hormones. Furthermore, women caregivers reported poorer global health than did female noncaregivers. Female caregivers also reported poorer global health than did male caregivers, who in turn reported worse global health than did male noncaregivers. Because of the negative effects of dementia caregiving on caregiver health found in most of the cross-sectional studies, interventions have been designed focusing on either caregiver self-care or physical activity promotion to enhance caregiver health and well-being. The Resources for Enhancing Alzheimer’s Caregiver Health (REACH) II caregiver intervention (Belle et al., 2006) was effective in ameliorating depression and emotional distress at 6-month follow-up. Caregivers in the intervention group also reported better global health and physical health (Elliott, Burgio, & DeCoster, 2010). King, Baumann, O’Sullivan, Wilcox, and Castro (2002) conducted a randomized controlled trial (RCT) to test effects of physical activity intervention versus nutrition education (control) on health and quality of life in a sample of 100 women dementia caregivers, whose age ranged from middle to old age. Although both conditions improved caregiving distress, physical activity

Liu et al.

689

additionally benefited caregivers’ total energy expenditure and stress-induced blood pressure reactivity. Longitudinal studies have shown different findings on caregiving and physical health decline. In a small longitudinal study examining changes in caregivers’ mental and physical health with the transition into caregiving roles, Lawton, Moss, Hoffman, and Perkinson (2000) did not find any significant changes in new caregivers’ quality of life or any detrimental effects on health as measured by health conditions and general health perceptions over 1 year. Focusing on physical health symptoms, Goode, Haley, Roth, and Ford (1998) found caregivers of dementia patients had stable physical health over time. It may be that health effects are found over longer durations or that there are individual differences in that some caregivers may be at greater risk of declining health.

Factors Associated With Caregivers’ Physical Health Caregivers’ physical health is associated with a number of individual and contextual factors. In general, poor physical health was more strongly associated with subjective distress (e.g., feelings of burden) than with objective stressors (e.g., IWD’s behavior problems or cognitive impairment). Caregivers in poor health were more likely to be older, married, less educated, unemployed, and assisting their relatives with basic personal activities of daily living (PADL) and instrumental activities of daily living (IADL) needs like bathing, dressing, and managing finances (Navaie-Waliser et al., 2002). Furthermore, worse physical health of caregivers was associated with longer caregiving duration, coresidence with care recipients, and fewer caregiving tasks (Pinquart & Sörensen, 2007). Caregivers who had fair to poor health or who had a serious health condition were likely to report their health suffered as a result of caregiving, to have unmet needs in providing care, and to have difficulties in managing daily caregiving responsibilities. Caregivers who continue providing high intensity care for a prolonged period of time can exacerbate their existing vulnerability, making caregiving unsafe and stressful for both themselves and IWDs (Navaie-Waliser et al., 2002). Pinquart and Sörensen (2007) indicated that performing more caregiving tasks was associated with better caregiver physical health although it was also related to higher burden. This “suppressor” effect shows that after activities of daily living (ADL) dependency of IWDs is accounted for, caregivers with better health are able to meet higher demands of caregiving responsibility and provide higher levels of support.

690

Journal of Aging and Health 27(4)

Caregivers’ Declining Health as a Contributing Factor to Caregiving Transitions In the caregiving literature, “transition” in the caregiver role is often defined by patient institutionalization and death (Aneshensel et al., 1995). Transition from caregiving to patient bereavement versus institutionalization has been found to have different health consequences among caregivers (Bass & Bowman, 1990). Moreover, while caregiver health may not be the antecedent of patient death, the increasing likelihood of placing IWDs into an institutional setting could be one consequence of caregivers’ declining health or health problems (Pinquart & Sörensen, 2007). The most common type of placement is for caregivers to place an IWD in a nursing home or other institutional settings, but caregivers may also turn care over to another family member. In this study, we focus on these types of placement as caregiving transitions. Caregivers need good physical health and functioning to provide ongoing care, and thus deteriorating health may be a precipitant of placement and caregiving relinquishment. Among studies of nursing home placement, caregivers’ own health as measured by comorbidities and general health perceptions has been consistently found to be an important factor. Using a single-item measure for general health perceptions, Gaugler and colleagues (2000; 2003) found that better health of caregivers was associated with decreased likelihood of nursing home placement, whereas caregivers who self-rated their health as poor institutionalized IWDs sooner. Baumgarten and colleagues (1994) found in their longitudinal study that caregivers who had placed their relatives into institutional care showed worsening health trajectories, compared with those who did not experience the transition. McCann and colleagues (2004) conducted initial interviews on both family caregivers and noncaregivers, and then followed them up at 3 years. The study showed the importance of physical health, specifically, that initial physical fitness was an important predictor of beginning and continuing the caregiving role. In another longitudinal study over a 3-year period, Buhr, Kuchibhatla, and Clipp (2006) examined female spouse caregivers’ primary reasons for placement. Over the study period, 26% of caregivers institutionalized their IWDs. Nearly half of these caregivers endorsed their own health problems as being the primary reason.

ADS Use and the “Health Preserving” Hypothesis Interventions that lower caregivers’ stressor exposures may reduce the risk of having poor health, particularly for caregivers of IWDs who often provide extensive help and ongoing supervision. Community-based long-term care

Liu et al.

691

services such as ADS programs provide a predictable amount of time away from IWDs, reducing exposure to care-related stressors and creating opportunities for caregivers to engage in self-care activities (Zarit et al., 2011). In a carefully controlled evaluation study, Zarit, Stephens, Townsend, and Greene (1998) showed that regular ADS use for at least 3 months lowered caregivers’ feelings of overload, worry, and strain, and also led to lower depressive symptoms and anger, compared with caregivers in the control group not using ADS. These benefits were maintained by ADS users at a 1-year follow-up. Other studies have found that as a result of ADS use, caregivers reported better self-efficacy in behavior management and enhanced general well-being, which in turn enabled them to provide more efficient home care (Gitlin, Reever, Dennis, Mathieu, & Hauck, 2006). In addition, regular ADS use of 2 to 3 days per week was found to be associated with a substantial lower risk of mortality among community-dwelling frail elders (Kuzuya et al., 2006). One of the outcomes most often associated with ADS use is the timing of institutionalization. In particular, ADS use was found to delay placement of IWDs for adult daughter caregivers, though not spouses (Cho, Zarit, & Chiriboga, 2009; Gaugler & Zarit, 2001). Based on the positive effects of ADS found in prior studies on caregiver well-being, ADS use may also have a health-preserving effect at both between-person and within-person levels. In particular, more frequent and sustained ADS use may have positive associations with some dimensions of caregiver health over time.

The Present Study The aims of this study are to describe longitudinal health change trajectories among dementia caregivers who utilized ADS service and to examine the associations between caregiver health change, ADS use, and caregiving transitions. The present study builds on and extends prior studies on caregiving and health in the following ways. A longitudinal design was used to depict within-person health trajectories and between-person differences in health changes over a 1-year period for caregivers who had been providing active care at baseline. Guided by the stress process model (Aneshensel et al., 1995), these differential health trajectories over time were related to baseline caregiving characteristics, between- and within-person ADS use, and caregiving transitions during the course of study. To examine the effects of ADS use among caregivers who experienced transitions versus who did not, the interaction between amount of ADS use at baseline and caregiving transitions on health change were modeled. Finally, instead of relying on a single measure of health, caregiver health was conceptualized as a multidimensional system with various dimensions.

692

Journal of Aging and Health 27(4)

There is evidence that the effect of stressors on caregiver health is best examined using a multidimensional approach to health (Aneshensel et al., 1995; Pearlin et al., 1990). In contrast to the single health measure approach adopted by much of the existing literature on caregiver physical health, four dimensions of health over a 1-year period were examined. The general trajectories of health changes over a 1-year period in each of the health dimensions were examined first. Then the associations between these health trajectories and caregiving transitions, specifically, relinquishing the caregiving role by placing IWDs in the nursing home or turning care over to another family member were examined. In addition to caregiving transitions, other factors that have been linked with caregivers’ physical health changes were controlled (i.e., ADL dependency of IWDs, caregiver age, gender, marital status, duration of care, role overload, role captivity, and caregiver depression). Finally, both baseline and longitudinal ADS utilization were considered. Both the extent of ADS use at baseline and the effect of longitudinal ADS use throughout the study on caregiver health change over time were examined. Specifically, controlling for caregiving characteristics, we expect caregivers who utilized more ADS services at baseline to demonstrate less health decline (betweenperson associations); we also expect sustained ADS use throughout the study period to be associated with less health decline (within-person associations). The hypotheses were the following. First, there will be linear deteriorations in all four dimensions of caregiver health—increasing functional limitation, increasing bodily pain, poorer subjective health, and increasing role limitation. Second, caregivers’ health trajectories will vary systematically as a function of (a) the extent of baseline (time-invariant) ADS use as well as sustained ADS use over time, (b) caregiving transition status (i.e., continuing care at home, institutionalization of IWD), and (c) the interactions between baseline ADS use and caregiving transition status, controlling for baseline caregiving characteristics known to be associated with health dynamics. Specifically, greater extent of baseline ADS use is hypothesized to be associated with less health deterioration; ADS use is also hypothesized to have positive within-person associations with health changes. However, caregiving transitions (i.e., no longer providing primary care for IWDs) will be associated with a steeper decline in health trajectory over time.

Method Participants, Procedures, and the Sample Data for this study were drawn from the Daily Stress and Health (DaSH) study (Zarit, Kim, Femia, Almeida, & Klein, 2014). The sample consisted of family caregivers, who were providing primary care to IWDs and using ADS

693

Liu et al. 241

241 Referred 41 Did not meet eligibility criteria 200 Screened 16 Not enrolled

184 Participated 31 Excluded: - 17 had IWDs deceased at 6-month interview - 14 had IWDs deceased at 12-month interview

153 Analyzed 122 Did not experience transitions, and were active caregivers 31 Experienced transitions, and were no longer primary caregivers for the living IWDs: - 22 Nursing home or assisted living placement - 2 Another family member became the primary caregiver - 7 Other types of transition: - 3 Alzheimer’s special care - 2 Rehabilitation unit - 2 Senior apartment

Figure 1.  Flow of participants.

Note. IWD = individual with dementia.

programs for a minimum of 2 days per week at baseline. To be eligible for participation, IWDs needed to be clinically diagnosed with a type of dementia and to live with their primary caregivers in the same household. We identified ADS programs through their state associations in New Jersey, Pennsylvania, Virginia, and Colorado. Programs that agreed to participate were contacted by the study’s research coordinator who provided them with detailed information about the study and brochures for potential participants. Caregivers from 57 programs subsequently contacted the research coordinator for more information about the study and were screened for eligibility. As shown in Figure 1, 41 of the 241 referrals did not meet eligibility for the study, and 16 (8%) of the remaining 200 decided subsequently not to enroll. Initial face-to-face interviews were conducted at caregivers’ homes with 184 caregivers, during which they signed consent forms and provided sociodemographic information and baseline data. We interviewed these same caregivers by telephone at 6 months, and again at 12 months to obtain repeated measures of their emotional well-being, physical health, ADS use, and other caregiving

694

Journal of Aging and Health 27(4)

situations. Family caregivers were paid a small honorarium for the baseline and longitudinal follow-up interviews at 6 months and 12 months. Because one of our main research questions was whether caregivers’ health trajectories were associated with caregiving transitions, we operationalized caregiving transition to include family caregivers (a) who were no longer providing the primary care at the time of interview and (b) whose living relatives were cared for either formally in institutional settings (n = 29) or informally by another family caregiver (n = 2). Types of institutional settings included nursing homes, assisted living facilities, Alzheimer’s special care, rehabilitation units, and senior apartments. We excluded caregivers (n = 31) whose IWDs were deceased at 6-month or 12-month interview as bereavement and placement were likely to be different types of caregiving transitions and to have distinctive associations with caregiver health. The final working sample of family caregivers in this study (N = 153) included those (a) whose relatives were alive between time of entry into the study and at 1-year followup and (b) who were either providing primary care or had one of the aforementioned types of caregiving transition (Figure 1).

Measures Physical health.  We took a multidimensional approach to health, using four scales derived from the Medical Outcomes Studies (MOS) 36-item short form health survey (SF-36; Ware & Sherbourne, 1992) that addressed functional health and global subjective health. We did not include items that assessed emotional health and well-being as they are included in other study measures. The measure has been widely used in medical outcome studies with younger and older adults and has good subscale reliabilities and validity (Hays & Stewart, 1992; Stewart, Ron, & Ware, 1988). The four scales were as follows. General Health Perception. This single-item subscale assessed caregivers’ general health perceptions: “In general, rate your health” (5-point scale; coded 1 = poor to 5 = excellent). Higher scores indicated better global subjective health. Functional Limitation.  This subscale (nine-item, α = .82) assessed degrees of limitation in physical activities: moderate activities such as pushing a vacuum, moving a table; carrying groceries; climbing stairs; bending, kneeling, or stooping; walking; bathing and dressing (Stewart et al., 1989). Caregivers were asked about how much physical limitation (3-point scale; coded 1 = not limited to 3 = limited a lot) they had in carrying out these activities, and higher scores indicated greater functional limitation.

Liu et al.

695

Bodily Pain.  This single-item subscale asked caregivers how much bodily pain have they generally had during the past month (6-point scale; coded 1 = none to 6 = very severe). Higher scores indicated more severe sensation of bodily pain. Role and Activity Limitation.  This subscale (four-item, α = .83) asked caregivers how much of the time they have felt (5-point scale; coded 1 = none of the time to 5 = all of the time) that they accomplished less than they would like, they were limited in the kind of activities that they could do because of their physical health, it was difficult to perform their work or other activities, and their physical health or emotional problems interfered with their normal social activities with family, friends, neighbors, or groups (Ware & Sherbourne, 1992). Higher scores indicated greater role limitation. Key predictors and other covariates Caregiving transition.  Caregiving transition (1 = experienced a transition and 0 = did not experience a transition) was coded to indicate any types of transition reported at either 6- or 12-month interview, such that the caregiver was no longer the primary caregiver of their living IWDs. Not experiencing a transition indicated that the caregiver was still providing the primary care. ADS use.  ADS use at the between-person level (time-invariant) was coded as self-reported total number of days per week at baseline (M = 3.89 days, SD = 1.15 days). ADS use at the within-person level (time-varying) was dummycoded as 1 (use) and 0 (nonuse) throughout the study period at baseline, 6 months, and 12 months. Covariates. We considered additional variables from the baseline interview that are often associated with caregiver physical health as covariates: caregiver age, relation type (1 = spouse and 0 = other kin relationship), duration of caregiving (in months), depressive symptoms (depression subscale from the Nonspecific Psychological Distress Scale; seven items on a 5-point scale ranging from 1 = none of the day to 5 = all day; coded so that higher scores indicated more depressive symptoms; α = .81; Kessler et al., 2002), caregiving role overload (three items on a 4-point scale ranging from 1 = none of the time to 4 = all of the time coded so that higher scores indicated more subjective experiences of being worn out and overloaded; α = .63; Pearlin et al., 1990), caregiving role captivity (three items on a 4-point scale ranging from 1 = none of the time to 4 = all of the time coded so that higher scores indicated that the caregiving role was unwanted; α = .83; Aneshensel et al., 1995; Pearlin et al., 1990), and the IWDs’ ADL dependency (mean of

696

Journal of Aging and Health 27(4)

13 items on a 4-point scale ranging from 1 = does not need help to 4 = cannot do without help coded so that higher scores indicated greater ADL needs and dependency; α = .83; Katz, Ford, Moskowitz, Jackson, & Jaffe, 1963; Lawton & Brody, 1969).

Analysis The main research interest was to examine the time trend in caregivers’ health longitudinally. To test Hypothesis 1, four unconditional growth curve models were fit with random intercepts (Model 1) to each of the repeatedly measured health outcomes, following standard model fitting and trimming procedures (Singer & Willett, 2003). In the Level-1 within-person model, caregiver health was specified as Health ti = π0i + π1i ( Timeti ) + εti (1) where health for caregiver i at time t was a function of an intercept (π0i, baseline health status), linear time (π1i, within-person association between time and health), and the within-person residual, εti , whose variance was σε2 and assumed to be homogeneous across persons. In the Level-2 model, the individual specific intercepts and slopes were specified as

π 0i = β00 + υ0i (2) π1i = β10 where βs are population-level parameters, representing the mean intercept ( β00 ) and mean slope ( β10 ) of the health trajectory pooling over all caregivers in the sample. Furthermore, υ0i are unexplained, between-person differences in the intercept with a variance, σ2υ , representing the degree of individual variability around the mean intercept. The second research interest was to relate caregivers’ longitudinal health trajectories to the extent of ADS use at baseline, sustained ADS use over time, caregiving transition status, and the interaction between extent of baseline ADS use and caregiving transition. To test Hypothesis 2, three linear growth curve models were fit sequentially (Models 2-4). The Level-1 equation was specified as Health ti = π0i + π1i ( Timeti ) + π2i ( ADSti ) + εti (3) where caregiver health was modeled as a function of time-varying ADS use throughout the study (π2i, within-person association between ADS use and

697

Liu et al.

health) in addition to the intercept, the linear time, and the within-person residual term as described for Equation 1. Level-2 equations modeled the effects of extent of baseline ADS use (Model 2), caregiving transition (Model 3), and the interaction between the extent of baseline ADS use and caregiving transition (Model 4) on caregiver health trajectories, while controlling for caregivers’ characteristics and IWDs’ level of dependency. Level-2 equations in Model 2 were specified as π0i = β00 + β01 ( IWDs' ADL dependencyi ) + β02 ( Caregiver agei ) + β03 ( Spousse caregiveri ) + β04 ( Duration of caregivingi ) + β05 ( Role overload i ) + β06 ( Role captivityi ) + β07 ( Depressive symptomsi )

(4)

+ β08 ( ADS days per week i ) + υ0i π1i = β10 + β11 ( ADS days per week i ) π2i = β20 where βs are population-level parameters; the person-specific intercept, π0i, from Equation 3 is now modeled as a function of sum of ADS days per week at baseline, while controlling for other caregiving characteristics; the personspecific linear time slope, π1i, is also modeled as a function of sum of ADS days per week at baseline. Furthermore, υ0i is between-person differences in the intercept with a variance, σ2υ . Next, to test any differentiating effect of caregiving transition status on health trajectories, Level-2 equations in Model 3 were specified as π0i = β00 + β01 ( IWDs' ADL dependencyi ) + β02 ( Caregiver agei ) + β03 ( Spouse caregiveri ) + β04 ( Duration of caregiving i ) + β05 ( Role overload i ) + β06 ( Role captivityi ) + β07 ( Depressive symptomsi ) + β08 ( ADS dayys per week i ) + β09 ( Transition statusi ) + υ0i π1i = β10 + β11 ( Transition statusi ) π2i = β20

(5)

698

Journal of Aging and Health 27(4)

where both the person-specific intercept (π0i) and linear time slope (π1i) are modeled as functions of caregiving transition status. Last, to test the effect of interaction between extent of baseline ADS use and caregiving transition on longitudinal health, Level-2 equations in Model 4 were specified as π0i = β00 + β01 ( IWDs' ADL dependencyi ) + β02 ( Caregiver agei ) + β03 ( Spousse caregiveri ) + β04 ( Duration of caregiving i ) + β05 ( Role overload i ) + β06 ( Role captivityi ) + β07 ( Depressive symptomsi ) + β08 ( ADS dayys per week i ) (6) + β09 ( Transition statusi ) + υ0i π1i = β10 + β11 ( Transition statusi ) + β12 ( ADS days per week i ) + β13 ( ADS days per week i × Transition statusi ) π2i = β20 where the person-specific intercept (π0i) and linear time slope (π1i) are modeled as functions of sum of ADS days per week at baseline, caregiving transition status, and the interaction between sum of ADS days per week at baseline and transition status. To facilitate interpretation, the nature of any significant cross-level interactions were further explored with simple slope tests (Curran, Bauer, & Willoughby, 2006; Preacher, Curran, & Bauer, 2006). All four models were fit to each of the four repeatedly measured health outcomes, and all betweenperson covariates on demographics and caregiving characteristics were grand-mean centered.

Results As a preliminary analysis, descriptives on caregiver characteristics were run, which are presented in Table 1. Next, unconditional growth curve models (Model 1) were fit to test Hypothesis 1. Nonsignificant quadratic slopes were trimmed, and health changes presented in Table 2 are all linear. Over time, caregivers had increasing functional limitation (β = 0.028, p = .026) and decreasing bodily pain (β = −0.023, p = .007). Role limitation showed a linear increase, but the effect was at a borderline significance level (β = 0.043, p = .069). General health perception stayed stable without any significant time trend (β = −0.008, p = .418). However, caregivers showed significant individual differences in the levels of all four health dimensions.

699

Liu et al. Table 1.  Caregivers’ Characteristics at Baseline (N = 153). M or frequency Age Female Spouse Child Educationa Married with a partner Depressive symptomsb Role overloadc Role captivityc Duration of cared Number of ADS days per week IWD’s ADL dependencye

SD or percent

60.65 133  50  94 4.38  99 1.55 2.82 2.10 62.67 3.89 3.00

Minimum

Maximum

39

89

10.84 86.93 32.68 61.44 1.18 64.71 0.65 0.66 0.79 46.27 1.15 0.50

      6   5 4 4 216 5 4

1 1 1 1 2 2 2

Note. ADS = adult day service; IWD = individual with dementia; ADL = activities of daily living. aMeasured on a 6-point scale: 1 (less than high school) to 6 (post college degree). bMeasured as the mean of seven items on a 5-point scale: 1 (none of the day) to 5 (all day). cMeasured as the mean of three items on a 4-point scale: 1 (none of the time) to 4 (all of the time). dMeasured in months. eMeasured as the mean of 13 ADL items on a 4-point scale: 1 (does not need help) to 4 (cannot do without help).

Table 2.  Unconditional Model of Longitudinal Health Change (Model 1).

  Fixed effects  Intercept   Linear time Random effects   Intercept VAR   Residual VAR −2 log likelihood AIC, BIC

Functional limitation

Bodily pain

Health perception

Role limitation

β (SE)

β (SE)

β (SE)

β (SE)

8.113 (0.224)*** 0.028 (0.012)*

3.381 (0.103)*** −0.023 (0.008)**

7.084 (0.152)*** −0.008 (0.010)

7.769 (0.306)*** 0.043 (0.023)†

6.345 (0.796)*** 1.608 (0.134)*** 1,854.3 1,858.3, 1,864.4

1.010 (0.147)*** 0.731 (0.061)*** 1,366.8 1,370.8, 1,376.9

2.647 (0.348)*** 1.043 (0.088)*** 1,587.3 1,591.3, 1,597.3

9.597 (1.339)*** 5.662 (0.472)*** 2,295.5 2,299.5, 2,305.6

Note. VAR = variance; AIC = Akaike information criterion; BIC = Bayesian information criterion. < .10. *p < .05. **p < .01. ***p < .001.

†p

Next, while controlling for within-person ADS use over time, sequential models were fit to test Hypothesis 2 on caregiver health trajectories over time and associations with extent of baseline ADS use (Model 2), caregiving

700

Journal of Aging and Health 27(4)

transition (Model 3), and the interaction between extent of baseline ADS use and caregiving transition (Model 4). Models with significant time effects are presented in Table 3. Model 2 shows that sum of ADS days at baseline was associated with changes in functional limitation (β = −0.025, p = .011), but not the other dimensions of health. Simple slope tests showed further that caregivers who were below-the-average ADS users at baseline (1 SD below the average days) had significant deterioration in functioning over time (β = 0.059, p = .001), whereas above-the-average ADS users at baseline (1 SD above the average days) remained stable in functional limitation over time (β = −0.004, p = .834). Model 3 shows that caregiving transitions were significantly associated with both trajectories of functional limitation (β = 0.096, p = .005) and bodily pain (β = 0.063, p = .009). However, caregiving transition was not associated with trajectories of health perceptions or role and activity limitation. Specifically, simple slope tests showed that caregivers who experienced caregiving transitions had increasing functional limitation (β = 0.106, p = .001) and reported stable bodily pain over time (β = .030, p = .166; Figure 2), whereas caregivers who did not experience a caregiving transition had stable functioning (β = 0.011, p = .423) and decreasing pain over time (β = −0.032, p = .001; Figure 2). Model 4 shows a significant three-way interaction between sum of ADS days at baseline, caregiving transition, and caregivers’ functional changes over time (β = −0.080, p = .007). We conducted simple slope tests to help interpret functional trajectories among four profiles of caregivers: aboveversus below-the-average ADS users among active caregivers who had not experienced any caregiving transitions (Figure 3), and above- versus belowthe-average ADS users among caregivers who had experienced caregiving transitions (Figure 4). Specifically, simple slope tests showed that the only significant functional limitation trajectory over time was for caregivers who were below-the-average ADS users at baseline, and who had experienced transitions (β = 0.213, p = .000). That is, this group of caregivers showed significant deterioration in functionality over time, whereas the other caregivers remained stable on the functional limitation scale. For all the models presented in Table 3, the same set of covariates was modeled as controls. Turning to the hypothesis on ADS use over time and caregiver health trajectory, within-person ADS use had significant positive association with general health perception only (not shown in Table 3) in Model 2 (β = .894, p = .001), Model 3 (β = .700, p = .007) and Model 4 (β = .738, p = .005). That is, caregivers who had been using ADS throughout the study period also had better general health perception. At the between-person level, baseline depressive symptoms, caregiver age, role overload, and captivity

701

Liu et al.

Table 3.  Models for Caregivers’ Health Change and Associations With Baseline ADS Use and Caregiving Transitions. Model 2

Model 3

Model 4

 

Functional limitation

Functional limitation

Bodily pain

Functional limitation

 

β (SE)

β (SE)

β (SE)

β (SE)

Fixed effects  Intercept  Timea   ADS daysb   Transition statusc  Timea × ADS daysb  Timea × Transition statusc  Timea × ADS daysb × Transition statusc   Time-varying ADS used Covariates   IWD’s ADL dependency  Age  Spouse   Duration of care   Role overload   Role captivity  Depressive symptoms Random effects   Intercept VAR   Residual VAR −2 log likelihood AIC, BIC

8.512 (0.379)*** 0.023 (0.013)† −0.014 (0.184) n/a −0.025 (0.011)* n/a

8.536 (0.387)*** 0.006 (0.014) −0.201 (0.178) −0.525 (0.594) n/a

3.413 (0.237)*** −0.032 (0.010)*** 0.026 (0.080) −0.312 (0.287) n/a

8.538 (0.382)*** 0.006 (0.013) −0.125 (0.179) −0.511 (0.592) n/a

0.096 (0.034)**

0.063 (0.024)**

0.090 (0.033)**

n/a

n/a

n/a

−0.093 (0.027)***

−0.400 (0.319)

−0.364 (0.311)

0.028 (0.210)

−0.366 (0.306)

0.381 (0.408)

0.418 (0.417)

0.283 (0.187)

0.367 (0.417)

0.066 (0.024)** −0.743 (0.538) −0.000 (0.004) 0.611 (0.320) −0.570 (0.272)* 1.534 (0.320)***

0.058 (0.025)* −0.725 (0.551) −0.000 (0.004) 0.515 (0.329) −0.487 (0.286) 1.535 (0.366)***

0.005 (0.011) −0.085 (0.246) 0.001 (0.002) 0.404 (0.147)** −0.304 (0.128)* 0.570 (0.163)***

0.058 (0.025)* −0.689 (0.551) −0.001 (0.004) 0.543 (0.329) −0.494 (0.286)† 1.486 (0.366)***

5.030 (0.665)*** 1.587 (0.133)*** 1,834.2 1,838.2, 1,844.3

4.939 (0.675)*** 1.486 (0.127)*** 1,712.7 1,716.7, 1,722.6

0.840 (0.136)*** 0.729 (0.062)*** 1,302.1 1,306.1, 1,312.0

4.952 (0.674)*** 1.428 (0.122)*** 1,706.0 1,710.0, 1,715.9

Note. Models 2 to 4 were fit to each of the repeatedly measured health outcome variables and nonsignificant models are not presented here; Nonsignificant interaction terms with time were trimmed. ADS = adult day service; IWD = individual with dementia; ADL = activities of daily living; VAR = variance; AIC = Akaike information criterion; BIC = Bayesian information criterion. aTime effect is linear. bMeasured as the total number of ADS days per week at baseline. cCoded 1 (experienced a caregiving transition) and 0 (did not experience a caregiving transition). dCoded 1 (use) and 0 (nonuse). †p < .10. *p < .05. **p < .01. ***p < .001.

were significantly associated with levels of health in almost all dimensions at baseline. On average, older caregivers reported greater functional limitation. Caregivers who reported more depressive symptoms also reported greater

702

Journal of Aging and Health 27(4) 2.6

Bodily Pain

2.5 2.4

Trasition No transition*

2.3 2.2 2.1 Baseline

6 month

12 month

Figure 2.  Caregivers’ bodily pain and caregiving transition. Note. *p = .001.

8.6

Functional Limitation

8.5 8.4 8.3

ADS high

8.2

ADS low

8.1 8 7.9 Baseline

6 month

12 month

Figure 3.  Functional limitation and ADS use among caregivers providing active assistance. Note. ADS = adult day service.

functional limitation and more severe bodily pain. In addition, caregivers who reported more role captivity reported less functional limitation and less bodily pain, whereas the ones who reported greater role overload also reported more severe bodily pain.

703

Liu et al. 10.5

Functional Limitation

10 9.5 9 ADS high 8.5

ADS low*

8 7.5 7 Baseline

6 month

12 month

Figure 4.  Functional limitation and ADS use among caregivers who experienced transitions.

Note. ADS = adult day service. *This group of caregivers showed significant deterioration in functioning over time (p = .000).

Discussion This study contributes to the existing literature regarding effects of dementia caregiving on family caregivers’ health. Although prior studies have suggested an overall detrimental health effect associated with caring for IWDs, the findings from this study provide a different and more nuanced view. Family caregivers had significant between-person variations in their general levels of health, and the rates of health change differed for health dimensions. On the one hand, the extent of ADS use at baseline and their relatives’ nursing home placement were associated with caregivers’ functional limitation trajectories; on the other hand, sustained ADS use over time had positive association with caregivers’ general health perception at the within-person level. This study provides insights into dementia caregivers’ health dynamics in the context of caregiving transitions, and ADS use at both within- and between-person levels.

The Good and the Bad: A Mixed Picture of Health The descriptive aspect of the study depicted a mixed picture of caregiver health in that all health dimensions do not change in a uniform manner. The multiple indicators of health have delineated the subtle nuances in different

704

Journal of Aging and Health 27(4)

aspects of caregiver health and well-being. In contrast to most cross-sectional findings that reported poor perceived health (Pinquart & Sörensen, 2007), caregivers in our study had stable general health perception, and they reported experiencing less bodily pain over time. Meanwhile, they had greater functional limitation over time and perceived greater role limitation due to health problems. These observations may be a result of using specific measures of functional limitations, which have previously been found sensitive to detect changes in health, whereas the more general measure of health perceptions tend to show stability as indicated by relatively high test–retest reliability (McHorney, Ware, & Raczek, 1993). The dimension of bodily pain among dementia caregivers was rarely examined in prior studies of caregivers. The finding on decreasing pain trajectories was somewhat unexpected. It is known that more chronic pain is associated with aging, stress, environmental, and affective factors (Loeser & Melzack, 1999). The current study showed that depression and caregiving burden as measured by role overload and captivity were associated with caregivers’ pain severity. The decrease in pain over time may indicate more positive coping or adaptation to the caregiving role (Zarit, Todd, & Zarit, 1986). It may also reflect that care-related stressors, such as behavior problems, reach their peak during the middle phases of dementia, and then begin to decrease, thereby lowering one of the major sources of subjective burden on caregivers (McCarty et al., 2000). Both mechanisms may be at work to bring down reports of pain over time.

Contextual Factors and Caregiver Health We observed significant individual variation in caregivers’ general levels of health. Contextual factors that were associated with these differential levels of caregiver health were primarily consistent with prior studies. Caregivers who were older and more depressed had greater functional limitation; being a spouse, duration of caregiving, and ADL dependency of the IWD were, however, not associated with levels of functioning or bodily pain (Pinquart & Sörensen, 2003, 2007; Son et al., 2007). Regarding the hypothesis on associations between within-person ADS use and health changes, sustained ADS use over time had positive association with only caregivers’ general health perception. On the one hand, the lack of significant association may be due to little variation in longitudinal patterns of ADS use. A trend that can be found in the current study was that sustained ADS use was associated with continued home care, and ADS nonuse almost always happened with caregiving transitions. On the other hand, there could be a selection bias in that caregivers who selfperceived to be healthier tend to use ADS more and continued providing care

Liu et al.

705

at home. Thus, although the study showed some benefits of ADS use on caregivers’ physical health, to what extent sustained ADS use has the potential to better preserve caregiver physical health needs to be further explored. The findings showed that health trajectories, specifically how functional limitation change in the long run was associated with IWDs’ placement. Specifically, caregivers who showed greater health decline also experienced caregiving transitions. Furthermore, the extent of baseline ADS use was related to changes in caregiver health, and this extends the literature on psychological benefits of ADS use for caregivers (Zarit et al., 1998). It is possible that the threat of an illness, an actual occurrence of a physical health event, or anticipation of functional decline in the near future may have caused caregivers to give up their caregiving role to protect themselves from further decline. Caregivers who reported increasing functional limitation may also be experiencing more difficulty related to carrying out caregiving responsibilities, leading to their turning care of the IWD over to an institution or another family member. However, among caregivers who had experienced transitions, only the group of below-the-average ADS users at baseline (who had an average of 2.74 ADS days per week) showed a significant functional decline over time. This three-way interaction suggests that caregivers’ physical functioning is at the core of being able to sustain care provision, which is consistent with prior literature (Baumgarten et al., 1994). In addition, as with any type of intervention, the “dosage” or extent of ADS use is important to show its health benefits of minimizing caregivers’ functional deterioration (Zarit et al., 1998). ADS and other forms of caregiving respite may be particularly useful in preventing health decline because they provide predictable relief from exposure to care-related stressors (Zarit et al., 2011).

Limitations and Implications The study is limited by potential biases in the relatively small sample of family caregivers who self-selected to use ADS, and at varying amounts. Another limiting factor is our sample may have included a greater proportion of caregivers who were adult daughters or daughters-in-law than in the general population, a group who may benefit more from ADS use (Cho et al., 2009). We did, however, control for kin relationship and factors such as caregiver age in the analysis. Furthermore, biomarkers of health could also be incorporated to provide improved estimates of changes in caregiver health. One more consideration is the pattern of associations between health trajectories and caregiving transitions may be more complex than the current findings suggested. It is plausible that caregivers’ declining health precipitates IWDs’ placement, but it was not clear whether there was a health

706

Journal of Aging and Health 27(4)

recovery period after the actual transition or whether caregiver health just kept worsening following the occurrence of a caregiving transition. Longitudinal data with measurement occasions extending beyond IWDs’ placement are necessary to clarify these patterns. Therefore, a longer followup period with more measurement occasions is advantageous to precisely track individual caregivers’ health trajectories both prior to and long after their relatives’ placement. Last, the pseudo-R2 statistic based on proportionate reduction in error (PRE) for functional limitation, comparing the null model and the model with predictors at both Level-1 and Level-2 submodels (Model 4) was calculated (Raudenbush & Bryk, 2002). Thus, we were able to calculate the model effect size based on PRE, and quantify the magnitude of variance components reduction after adding predictors. Level-1 effect size was 0.46, suggesting a medium-to-large effect, and Level-2 effect size was 0.36, suggesting a medium effect (Cohen, 1992). Although the models showed significant associations between magnitude of ADS use, caregiving transitions, and caregiver health, it is not clear whether these associations were clinically significant. To what extent these significant effects are meaningful for caregivers in the long run can be triangulated in future studies. The findings on ADS use, caregiver health, and caregiving transition have implications for potential pathways to promote caregiver health and delay IWDs’ placement. On the one hand, community-based ADS programs have health benefits for caregivers; regular and sustained utilization is especially helpful in reducing care-related stressor exposures and creating opportunities for positive self-care among dementia family caregivers (Klein et al., 2014; Zarit et al., 2014; Zarit et al., 1998). On the other hand, it is possible to expand regular ADS programs to include a physical fitness module to promote moderate and age-appropriate exercise for both caregivers and patients. Such programs may have direct health benefits for caregivers and may also benefit caregivers indirectly by promoting patient health and functioning. For example, Woodhead, Zarit, Braungart, Rovine, and Femia (2005) examined the effects of different types of physical activity at ADS centers on behavior and psychological symptoms of dementia (BPSD) over 3 months. Three domains of BPSD were examined: restless behaviors, mood behaviors, and positive behaviors. The authors found that activities at ADS centers had positive benefits as IWDs’ positive behaviors increased for all participants. Basic ADS programs providing caregiving respite plus other components such as physical activity or behavior management may be most beneficial for both caregiving partners (Gitlin et al., 2006; Teri et al., 2003). How these innovative programs lay out over time will have implications for the well-being of the caregiving dyads, and the quality of life for the extended family.

Liu et al.

707

Declaration of Conflicting Interests The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by the National Institute on Aging (NIA; Grant No. RO1 AG031758, “Daily Stress and Health of Family Caregivers”).

References Aneshensel, C. S., Pearlin, L. I., Mullan, J. T., Zarit, S. H., & Whitlatch, C. J. (1995). Profiles in caregiving: The unexpected career. San Diego, CA: Academic Press. Bass, D. M., & Bowman, K. (1990). The transition from caregiving to bereavement: The relationship of care-related strain and adjustment to death. The Gerontologist, 30, 35-42. doi:10.1093/geront/30.1.35 Baumgarten, M., Battista, R. N., Infante-Rivard, C., Hanley, J. A., Becker, R., & Gauthier, S. (1992). The psychological and physical health of family members caring for an elderly person with dementia. Journal of Clinical Epidemiology, 45, 61-70. doi:10.1016/0895-4356(92)90189-T Baumgarten, M., Hanley, J. A., Infante-Rivard, C., Battista, R. N., Becker, R., & Gauthier, S. (1994). Health of family members caring for elderly persons with dementia: A longitudinal study. Annals of Internal Medicine, 120, 126-132. doi:10.7326/0003-4819-120-2-199401150-00005 Belle, S. H., Burgio, L., Burns, R., Coon, D., Czaja, S. J., Gallagher-Thompson, D., . . . Resources for Enhancing Alzheimer’s Caregiver Health (REACH) II Investigators. (2006). Enhancing the quality of life of dementia caregivers from different ethnic or racial groups: A randomized, controlled trial. Annals of Internal Medicine, 145, 727-738. doi:10.7326/0003-4819-145-10-200611210-00005 Buhr, G. T., Kuchibhatla, M., & Clipp, E. C. (2006). Caregivers’ reasons for nursing home placement: Clues for improving discussions with families prior to the transition. The Gerontologist, 46, 52-61. doi:10.1093/geront/46.1.52 Cho, S., Zarit, S. H., & Chiriboga, D. A. (2009). Wives and daughters: The differential role of day care use in the nursing home placement of cognitively impaired family members. The Gerontologist, 49, 57-67. doi:10.1093/geront/gnp010 Cohen, J. (1992). A power primer. Psychological Bulletin, 112, 155-159. doi:10.1037/0033-2909.112.1.155 Curran, P. J., Bauer, D. J., & Willoughby, M. T. (2006). Testing and probing interactions in hierarchical linear growth models. In C. S. Bergeman & S. M. Boker (Eds.), Methodological issues in aging research. Notre Dame series on quantitative methods (pp. 99-129). Mahwah, NJ: Lawrence Erlbaum. Elliott, A. F., Burgio, L. D., & DeCoster, J. (2010). Enhancing caregiver health: Findings from the resources for enhancing Alzheimer’s caregiver health

708

Journal of Aging and Health 27(4)

II intervention. Journal of the American Geriatrics Society, 58, 30-37. doi:10.1111/j.1532-5415.2009.02631.x Gaugler, J. E., Edwards, A. B., Femia, E. E., Zarit, S. H., Stephens, M.-A. P., Townsend, A., & Greene, R. (2000). Predictors of institutionalization of cognitively impaired elders: Family help and the timing of placement. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 55, P247-P255. doi:10.1093/geronb/55.4.P247 Gaugler, J. E., Jarrott, S. E., Zarit, S. H., Stephens, M.-A. P., Townsend, A., & Greene, R. (2003). Respite for dementia caregivers: The effects of adult day service use on caregiving hours and care demands. International Psychogeriatrics, 15, 3758. doi:10.1017/S1041610203008743 Gaugler, J. E., & Zarit, S. H. (2001). The effectiveness of adult day services for disabled older people. Journal of Aging and Social Policy, 12, 23-47. doi:10.1300/ J031v12n02_03 Gitlin, L. N., Reever, K., Dennis, M. P., Mathieu, E., & Hauck, W. W. (2006). Enhancing quality of life of families who use adult day services: Short- and longterm effects of the adult day services plus program. The Gerontologist, 46, 630639. doi:10.1093/geront/46.5.630 Goode, K. T., Haley, W. E., Roth, D. L., & Ford, G. R. (1998). Predicting longitudinal changes in caregiver physical and mental health: A stress process model. Health Psychology, 17, 190-198. doi:10.1037/0278-6133.17.2.190 Hays, R. D., & Stewart, A. L. (1992). Construct validity of MOS health measures. In A. L. Stewart & J. E. Ware (Eds.), Measuring functioning and well-being: The medical outcomes study approach (pp. 325-342). Durham, NC: Duke University Press. Katz, S., Ford, A. B., Moskowitz, R. W., Jackson, B. A., & Jaffe, M. W. (1963). Studies of illness in the aged, the index of ADL: A standardized measure of biological and psychosocial function. The Journal of the American Medical Association, 185, 914-919. doi:10.1001/jama.1963.03060120024016 Kessler, R. C., Andrews, G., Colpe, L. J., Hiripi, E., Mroczek, D. K., Normand, S.L. T., . . .Zaslavsky, A. M. (2002). Short screening scales to monitor population prevalences and trends in non-specific psychological distress. Psychological Medicine, 32, 959-976. doi:10.1017/S0033291702006074 King, A. C., Baumann, K., O’Sullivan, P., Wilcox, S., & Castro, C. (2002). Effects of moderate-intensity exercise on physiological, behavioral, and emotional responses to family caregiving: A randomized controlled trial. The Journals of Gerontology Series A: Biological Sciences and Medical Sciences, 57, M26-M36. doi:10.1093/gerona/57.1.M26 Klein, L. C., Kim, K., Almeida, D. M., Femia, E. E., Rovine, M. J., & Zarit, S. H. (2014). Anticipating an easier day: Effects of adult day services on daily cortisol and stress. The Gerontologist. doi:10.1083/geront/gne060 Kuzuya, M., Masuda, Y., Hirakawa, Y., Iwata, M., Enoki, H., Hasegawa, J., & Iquich, A. (2006). Day care service use is associated with lower mortality in communitydwelling frail older people. Journal of the American Geriatrics Society, 54, 13641371. doi:10.1111/j.1532-5415.2006.00860.x

Liu et al.

709

Lawton, M. P., & Brody, E. M. (1969). Assessment of older people: Self-maintaining and instrumental activities of daily living. The Gerontologist, 9, 179-186. doi:10.1093/geront/9.3_Part_1.179 Lawton, M. P., Moss, M., Hoffman, C., & Perkinson, M. (2000). Two transitions in daughter’s caregiving careers. The Gerontologist, 40, 437-448. doi:10.1093/ geront/40.4.437 Loeser, J. D., & Melzack, R. (1999). Pain: An overview. The Lancet, 353, 1607-1609. doi:10.1016/S0140-6736(99)01311-2 Mausbach, B. T., Roepke, S. K., Ziegler, M. G., Milic, M., von Känel, R., Dimsdale, J. E., . . .Grant, I. (2010). Association between chronic caregiving stress and impaired endothelial function in the elderly. Journal of the American College of Cardiology, 55, 2599-2606. doi:10.1016/j.jacc.2009.11.093 McCann, J. J., Hebert, L. E., Bienias, J. L., Morris, M. C., & Evans, D. A. (2004). Predictors of beginning and ending caregiving during a 3-year period in a biracial community population of older adults. American Journal of Public Health, 94, 1800-1806. doi:10.2105/ajph.94.10.1800 McCarty, H. J., Roth, D. L., Goode, K. T., Owen, J. E., Harrell, L., Donovan, K., & Haley, W. E. (2000). Longitudinal course of behavioral problems during Alzheimer’s disease: Linear versus curvilinear patterns of decline. The Journals of Gerontology Series A: Biological Sciences and Medical Sciences, 55, M200-M206. doi:10.1093/gerona/55.4.M200 McHorney, C. A., Ware, J. E. Jr., & Raczek, A. E. (1993). The MOS 36-Item ShortForm Health Survey (SF-36): II. Psychometric and clinical tests of validity in measuring physical and mental health constructs. Medical Care, 31, 247-263. doi:10.1097/00005650-199303000-00006 Navaie-Waliser, M., Feldman, P. H., Gould, D. A., Levine, C., Kuerbis, A. N., & Donelan, K. (2002). When the caregiver needs care: The plight of vulnerable caregivers. American Journal of Public Health, 92, 409-413. doi:10.2105/ AJPH.92.3.409 Pearlin, L. I., Mullan, J. T., Semple, S. J., & Skaff, M. M. (1990). Caregiving and the stress process: An overview of concepts and their measures. The Gerontologist, 30, 583-594. doi:10.1093/geront/30.5.583 Perkins, M., Howard, V. J., Wadley, V. G., Crowe, M., Safford, M. M., Haley, W. E., . . .Roth, D. L. (2013). Caregiving strain and all-cause mortality: Evidence from the REGARDS study. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 68, 504-512. doi:10.1093/geronb/gbs084 Pinquart, M., & Sörensen, S. (2003). Differences between caregivers and noncaregivers in psychological health and physical health: A meta-analysis. Psychology and Aging, 18, 250-267. doi:10.1037/0882-7974.18.2.250 Pinquart, M., & Sörensen, S. (2007). Correlates of physical health of informal caregivers: A meta-analysis. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 62, P126-P137. doi:10.1093/geronb/62.2.P126 Preacher, K. J., Curran, P. J., & Bauer, D. J. (2006). Computational tools for probing interactions in multiple linear regression, multilevel modeling, and latent

710

Journal of Aging and Health 27(4)

curve analysis. Journal of Educational and Behavioral Statistics, 31, 437-448. doi:10.2307/4122453. Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical linear models: Applications and data analysis methods (2nd ed.). Thousand Oaks, CA: Sage. Singer, J. D., & Willett, J. B. (2003). Applied longitudinal data analysis: Modeling change and event occurrence. New York, NY: Oxford University Press. Son, J., Erno, A., Shea, D. G., Femia, E. E., Zarit, S. H., & Parris Stephens, M. A. (2007). The caregiver stress process and health outcomes. Journal of Aging and Health, 19, 871-887. doi:10.1177/0898264307308568 Stewart, A. L., Greenfield, S., Hays, R. D., Wells, K., Rogers, W. H., Berry, S. D., . . . Ware, J. E. Jr. (1989). Functional status and well-being of patients with chronic conditions: Results from the Medical Outcomes Study. The Journal of the American Medical Association, 262, 907-913. doi:10.1001/jama.1989.03430070055030 Stewart, A. L., Ron, D. H., & Ware, J. E. Jr. (1988). The MOS short-form general health survey: Reliability and validity in a patient population. Medical Care, 26, 724-735. doi:10.1097/00005650-198807000-00007 Teri, L., Gibbons, L. E., McCurry, S. M., Logsdon, R. G., Buchner, D. M., Barlow, W. E., . . .Larson, E. B. (2003). Exercise plus behavioral management in patients with Alzheimer disease: A randomized controlled trial. The Journal of the American Medical Association, 290, 2015-2022. doi:10.1001/jama.290.15.2015 Vitaliano, P. P., Zhang, J., & Scanlan, J. M. (2003). Is caregiving hazardous to one’s physical health? A meta-analysis. Psychological Bulletin, 129, 946-972. doi:10.1037/0033-2909.129.6.946 Ware, J. E. Jr., & Sherbourne, C. D. (1992). The MOS 36-Item Short-Form Health Survey (SF-36): I. Conceptual framework and item selection. Medical Care, 30, 473-483. doi:10.1097/00005650-199206000-00002 Woodhead, E. L., Zarit, S. H., Braungart, E. R., Rovine, M. R., & Femia, E. E. (2005). Behavioral and psychological symptoms of dementia: The effects of physical activity at adult day service centers. American Journal of Alzheimer’s Disease and Other Dementias, 20, 171-179. doi:10.1177/153331750502000314 Zarit, S. H., Kim, K., Femia, E. E., Almeida, D. M., & Klein, L. C. (2014). The effects of adult day services on family caregivers’ daily stress, affect, and health: Outcomes from the Daily Stress and Health (DaSH) study. The Gerontologist, 54, 570-579. doi:10.1093/geront/gnt045 Zarit, S. H., Kim, K., Femia, E. E., Almeida, D. M., Savla, J., & Molenaar, P. C. M. (2011). Effects of adult day care on daily stress of caregivers: A within-person approach. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 66B, 538-546. doi:10.1093/geronb/gbr030 Zarit, S. H., Stephens, M. A. P., Townsend, A., & Greene, R. (1998). Stress reduction for family caregivers: Effects of adult day care use. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 53B, S267-S277. doi:10.1093/geronb/53B.5.S267 Zarit, S. H., Todd, P. A., & Zarit, J. M. (1986). Subjective burden of husbands and wives as caregivers: A longitudinal study. The Gerontologist, 26, 260-266. doi:10.1093/geront/26.3.260

Copyright of Journal of Aging & Health is the property of Sage Publications Inc. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.