The influence of community-based services on the burden of spouses caring for their partners with dementia.
Abstract: Despite the vast literature on caregiver stress, few studies have explored how community services affect the stress process for spousal caregivers. The current study explores the differential effects of emotional and tangible support provided by family and friends and by formal services, and caregivers' perceptions of community services on spousal caregivers' burden. The study used a survey design with a sample of 85 spousal caregivers caring for their partners with dementia in the community. The study found that in-home services, as they are currently offered, do little to reduce the burden of spouses caring for their partners with dementia. Rather the most effective service is the provision of adult day programs, which provide not only respite for the spousal caregiver, but also opportunities for social interaction for their partners with dementia. This study further found that spousal caregivers experience a relatively high level of service-related stress. However, when examined alongside care recipient behavioral challenges and frequency of day program use, caregivers' perceptions of and experiences with the service system did not uniquely explain their burden. Implications of the findings for policy, research, and practice are discussed.

KEY WORDS: caregiver burden; community services; dementia care; spousal caregivers
Article Type: Report
Subject: Caregivers (Psychological aspects)
Dementia (Care and treatment)
Married people (Social aspects)
Married people (Health aspects)
Community service (Influence)
Authors: Sussman, Tamara
Regehr, Cheryl
Pub Date: 02/01/2009
Publication: Name: Health and Social Work Publisher: National Association of Social Workers Audience: Academic; Professional Format: Magazine/Journal Subject: Health; Sociology and social work Copyright: COPYRIGHT 2009 National Association of Social Workers ISSN: 0360-7283
Issue: Date: Feb, 2009 Source Volume: 34 Source Issue: 1
Topic: Event Code: 290 Public affairs
Product: Product Code: 8340000 Community Services NAICS Code: 6242 Community Food and Housing, and Emergency and Other Relief Services SIC Code: 8300 SOCIAL SERVICES
Geographic: Geographic Scope: Canada Geographic Code: 1CANA Canada
Accession Number: 195265940
Full Text: Caring for a spouse with dementia in the community is fraught with emotion. First and foremost, spousal caregivers are grieving the loss of their partners, as a result of a long and devastating illness. Although the person with dementia is physically alive, his or her psychological presence in the relationship dissipates as the disease progresses (Boss, 1999). In addition to facing the loss of their life-long partners, spousal caregivers of people with dementia are more likely to struggle with their own health problems, spend more hours per week caregiving, and care for relatives with more severe behavioral problems than do other familial caregivers (Schofield, Murphy, Herrman, Bloch, & Singh, 1998; Seltzer & Wailing Li, 2000). Perhaps as a result of these contextual factors, spouses experience more negative effects from caregiving (Seltzer & Wailing Li, 2000) and benefit from different forms of social support than do adult children (Cortell, 1996; Wailing Li, Seltzer, & Greenberg, 1997).

With social and political pressures to "age in place" spousal caregivers, who comprise 38 percent of the nonpaid caregiving population in North America, are facing increased pressure to provide care to their partners at home for longer periods of time (Canadian Study of Health and Aging Working Group, 1994; Stone, Cafferata, & Sangl, 1987). It is important to note that these expectations come at a time when most Western industrialized governments, fearing a fiscal crisis due to the aging of the population, have been limiting rather than expanding the parameters of government-sponsored community support. As a consequence, spouses of individuals with dementia must offer physical care, negotiate challenging behaviors, provide continuous emotional support, and grieve while managing a complex and resource-limited system of health and support services (Furlini, 2001; Hills & MacRae, 1998).

Despite the vast literature on caregiver stress, few studies have explored how formal services such as professional support, homemaking services, and adult day programs can affect the stress process. Contrary to the prevailing belief of those offering services, the limited data available suggest that formal services exert little impact on the stress process for family caregivers (Aneshensel, Pearlin, Mullan, Zarit, & Whitlach, 1995; O'Rourke, & Tuokko, 2000) and may produce rather than reduce stress for caregivers (Ducharme et al., 2007; Miller et al., 2001). However, measurement and conceptual limitations restrict the interpretation of these findings. Most of these studies have tended to group together various types of family caregivers as if their experiences can be equated. Given that spouses have been found to differ from other familial caregivers in their use of and experiences with community services, that spouses are at particular risk of distress, and that most spouses care for their partners at home for the duration of their caregiving careers, the lack of focus on spousal caregivers, formal community services, and the stress process is striking (Perry & O'Connor, 2002; Sorensen, Pinquart, & Duberstein, 2002).

Previous research has also neglected to explore how different types of community services affect caregivers. Instead, research in this area tends to represent formal services by aggregating the total number of services or total amount of services used by caregivers and explore the associations between these sums and caregiver mental health. This "total score" approach may mask the unique effect of particular services on caregiver mental health. Given that emotional support has been found to be particularly beneficial for spouses, for example, one may expect that formal services offering emotional support (such as one-on-one professional support) would exert a unique effect on spousal caregivers (Wailing Li et al., 1997). Furthermore, because spouses have articulated a preference for in-home tangible support (homemaking services) over day program services, this type of service may provide specific relief for spousal caregivers (Cortell, 1996).

Finally, research has also neglected to review caregivers' perceptions of services and service systems alongside their report of service use. Measuring service use alone overlooks empirical reports that have continued to document caregivers' frustrations with accessing services that are seen as inflexible and fragmented and with overseeing service providers who can be experienced as patronizing and authoritative (Guberman & Maheu, 2000; Smale & Dupuis, 2003). Equating "use" with "usefulness" is also contrary to findings in the literature, which has consistently found that caregivers' perceptions of social support from family and friends is a much stronger predictor of caregiver distress than are more objective measures of support, such as network size (Clyburn, Stones, Hadjistavropoulos, & Tuokko, 2000). Given that spousal caregivers are often critical of the quality, timeliness, and benefits of formal care services (Cortell, 1996; Perry & O'Connor, 2002), accounting for spouses' perceptions of services in addition to types and amounts of services used seems critical to understanding the potential effect of community services on spousal caregivers' mental health.

Addressing the measurement issues prevalent in previous studies on service use and caregiver stress, the current study explores the differential effect of emotional and tangible support provided by family and friends, emotional and tangible support provided by formal services, and caregivers' perceptions of community services on spousal caregivers' burden. The study further accounts for the impact of care recipient functioning on spouses' mental health as these factors have been consistently found to affect caregiver distress (Clyburn et al., 2000).

METHOD

Hypotheses

Although sparse, previous research comparing spouses to adult children and exploring spouses' unique experiences with community services provide some evidence from which to generate hypotheses for the current study. This study hypothesized the following:

* Emotional and tangible support provided by family and friends will have an impact on caregiver burden.

* Emotional support provided by formal community services (that is in-home professional support and one-on-one caregiver support) will have an impact on caregiver burden.

* Tangible support provided by formal community services (that is, homemaking services, day program services, and institutional respite services) will have an impact on caregiver burden.

* Caregiver's perceptions of support from formal support services will have a differential impact on burden when compared with actual usage of support services.

* The functional level of the person with dementia will have an impact on caregiver burden.

Data Collection

Spousal caregivers were recruited for this study on the basis of the following four criteria: (1) they were actively providing care to a partner in the community, (2) they or their spouse were receiving at least one publicly funded community service, (3) they were providing care to someone with cognitive losses that were progressive and irreversible, and (4) they were caring for a partner who was 65 years of age or older. For the purpose of this study, publicly funded community services included homemaking services, institutional respite services, professional services (provided by an allied health professional other than a physician), caregiver support services (one-on-one or group support), and adult day program services, paid for in full or in part by government sources. Participants were recruited through mailings, bulletin board postings, support groups, educational sessions, health professionals, and other participants. All caregivers (N = 46) who were directly recruited by the researcher, responded to advertisements, or were referred by other research participants completed the questionnaire package. Only three caregivers referred by health care professionals (N = 39) decided not to participate, resulting in a 93 percent response rate for those individuals. A response rate for all individuals who were approached by health care professionals cannot be determined as the researcher did not have access to this information. In all, 85 spousal caregivers recruited from 17 different organizations, participated in the study.

Data were collected in two-hour face-to-face interviews at the location of the participant's choice. Seven caregivers elected to complete the questionnaires independently and thus were sent the study measures and the consent forms in a stamped self-addressed, envelope. There were no demographic differences found between those caregivers who chose to complete the questionnaire independently and those who participated in the face-to-face process. During the face-to-face interviews, caregivers spontaneously offered comments to explicate their experience. These comments were recorded in written form by the researchers and serve to enrich the findings by offering possible interpretations for the quantitative findings. The research was conducted in accordance with the standards of the Tri-Council Policy Statement for Ethical Conduct for Research Involving Humans (Canadian Institutes of Health Research, 1998) as required by all researchers in Canadian universities. Procedures were approved by the Health Sciences Research Ethics Board at University of Toronto.

Measures

Use of Community-Based Services. Each spousal caregiver in this study was asked to report whether or not they were currently using any or all of the following three tangible support services: adult day program services, institutional respite services, homemaking services, and any or all of the following three emotional-informational support services: in-home professional support services (including nursing services, occupational therapy services, or physiotherapy services), caregiver support services (including, one-on-one caregiver support, education, or case management) or interdisciplinary team support from a geriatric clinic. Caregivers were also asked to indicate how often they were accessing these six services.

Caregivers' Perceptions of Community Services. The Alberta Continuity of Services Scale for Mental Health (Adair et al., 2003) consists of 43 items and was originally developed to capture service continuity issues for community-residing adults with severe mental illness. It taps into four elements of service delivery: (1) system coordination/fragmentation (for example, "There don't seem to be links from one service to the next"), (2) user involvement in service planning and decision making (for example, "I was asked what I wanted out of treatment"), (3) system flexibility/rigidity (for example, "My services will change if my needs change"), and (4) access (for example, "After discharge, I had to wait a long time before I was seen in a community program"). For each item, respondents are asked to indicate whether they strongly agree (0), agree (1), are unsure (2), disagree (3), or strongly disagree (4). Total scores range from 0 to 172, with higher scores indicating less service continuity or more service-related stress. The Cronbach's alpha for the sample in this study was .85.

Informal Support from Family and Friends. The Expressive Support Scale (Aneshensel et al., 1995) is an eight-item scale asking caregivers about the emotional support available to them from family and friends. Sample items include "There is no one who really understands what you are going through and "You have someone you feel you can trust." Accompanying each item is a four-point Likert scale ranging from strongly disagree = 0 to strongly agree = 3. Total scores can range from 0 to 24, with higher scores representing more emotional support. In this study, the Cronbach's alpha was .88.

Tangible support received was measured by asking caregivers four dichotomous yes (1) or no (0) questions. Caregivers were asked whether they had a friend or relative who helped them on a regular basis, whether they needed more help from informal sources on a regular basis, whether they had a friend or relative who helped them periodically, and whether they needed more periodic help from informal sources.

Functioning of the Person with Dementia. Two measures were used to capture the level of functioning of the person with dementia. The Dementia Behavior Disturbance Scale (Baumgarten, Becker, & Gauthier, 1990) asks caregivers how often their partners exhibited any of 28 behaviors in the previous week, ranging from 0 = never to 4 = all of the time. Total scores can range from 0 to 112, with higher scores indicating more behavioral disturbance. Examples of behaviors include asking the same question over and over, showing a lack of interest in daily activities, and making unwarranted accusations. The Cronbach's alpha in this study was .68.

The activities of daily living/instrumental activities of daily living (ADL/IADL) subscale of the Older Americans Resources and Services Multi-dimentional Functional Assessment Questionnaire (George & Fillenbaum, 1985) includes 14 items that inquire about the extent the person with dementia can perform IADLs such as use the telephone and shop and basic ADLs such as bathe and use the bathroom. For each of the items, respondents are asked to report on their ability to perform the task without help = 1, with some help = 2, or not at all = 3. Total scores range from 14 to 42, with higher scores indicating the need for more assistance with ADLs. The Cronbach's alpha in this study was .90.

Caregiver Distress. The Zarit Burden Interview (Zarit, Reever, & Bach-Peterson, 1980) is a 22-item scale designed to provide a global measure of caregiver distress. The items reflect common areas of concern for caregivers of individuals with dementia. Caregivers are asked to report on the frequency with which these areas have suffered as a result of caring for their relative. A five-point scale ranging from never or not applicable = 0 to all the time = 4 is used. Total scores range from 0 to 88, with higher scores indicating more caregiver distress. Most studies examining caregiver distress for caregivers of individuals with cognitive impairment use this scale to measure burden (Bedard, Pedlar, Martin, Malott, & Stones, 2000). A significant advantage of the popularity of this measure is that results obtained across studies can be easily compared and synthesized. Cronbach's alpha reliability in this sample was .84

RESULTS

Demographics

Participants, averaged 76 years of age, 82 percent were female, 78 percent spoke English as a first language, 94 percent were white, and 55 percent perceived their annual income to be adequate to meet their needs. Fifty-five percent were college or university graduates. The median household income for spousal caregivers in this study fell within the $40,000 to 49,999 category. Compared with the Canadian national average for this age cohort, these spousal caregivers can be said to be overrepresentative of spousal caregivers who are female, well educated, white, and English speaking (Canadian Study on Health and Aging Working Group, 1994). For example, female spouses usually outnumber male spouses by only a 2:1 ratio (Canadian Study on Health and Aging Working Group, 1994). The 6:1 ratio present in this study overrepresents female spousal caregivers. Furthermore, the majority of caregivers in this sample (55.3 percent) completed either college or university. Participants in this study can be said, on average, to represent a highly educated group for their age cohort. The income distribution appears reflective of household incomes for Canadian seniors (Statistics Canada, 2005). Most spousal caregivers in this study were caring for spouses with Alzheimer disease (58.8 percent) or vascular dementia (31.4 percent), for an average of 5.1 years (SD = 3.3 years).

Caregiver Distress

The mean burden score for this sample on the Zarit Burden Interview was 35 (SD = 14.1) and a range of 11 to 76. A recent meta-analysis, which included 58 studies that used the Zarit Burden Interview, reported a mean burden level of 29.9 (SD = 9.3) (Pinquart & Sorensen, 2005). The difference between the current sample and others is statistically significant [t(84) = 3.32, p = .001], suggesting that the caregivers in this study can be said to be more burdened than those in a variety of other caregiving studies.

Use of Community-Based Services

Seventy-four percent of the study participants received some form of homemaking or personal support services such as help with bathing and light housekeeping. Like caregivers in other Canadian studies, caregivers in this study used homemaking services more than any other form of respite (Canadian Study on Health and Aging Working Group, 1994; Smale & Dupuis, 2003).

Sixty percent of caregivers in this study used adult day program services as a form of respite for a median of two days per week. The median for this group of caregivers is influenced, in part, by the condition put forth by most day centers in the study, which required a two-day minimum for day program use. Institutional respite was the least common form of respite services used by caregivers.

Forty-eight percent of caregivers in this study used in-home professional support services. Support was largely provided by occupational therapists and ++physiotherapists who were focused on the physical needs of the person with dementia. This service was predominantly short-term, with few caregivers (n = 10; 11.8 percent) receiving five visits or more. Forty-seven percent of caregivers in this study accessed information and support services from social workers and social service workers in community agencies or disease-specific organizations (such as the Alzheimer Society). Different from in-home professional services, this service was more focused on caregiver support. However, like in-home professional services, the information and emotional support provided to caregivers was of a short duration with only eight caregivers (9.4 percent) receiving more than five consultations over the course of the year. Finally, 35 (41.2 percent) caregivers in this study received nonmedical services from a geriatric clinic such as social work services or in-home safety assessments. Caregivers were not asked how frequently they received these services.

Bivariate analyses were conducted between caregiver burden and both tangible and emotional formal support services (see Table 1). Given the multiple relationships explored in each domain (for example, three variables representing use of tangible formal services, three variables representing use of emotional-informational formal services, three variables representing frequency of tangible services use; three variables representing frequency of emotional-informational services use) a Bonferroni step down (Holm) correction coefficient was used to correct for the increased possibility of finding a significant association with caregiver burden by chance (Type 1 error). None of the variables representing caregivers' use of professional services (emotional-informational formal support), frequency of use of professional services (frequency of emotional-informational formal support), or use of tangible services were associated with the Zarit Burden Interview. However, the frequency of day program use was associated with caregiver burden. Specifically, the more days caregivers used day program services the less burden they reported. This relationship remained significant when correcting for the increased probability of a Type 1 error. There was no association between the frequency of receiving other formal tangible support services such as homemaking services or institutional respite and burden. From these findings, it would appear that regularly scheduled out of home programming may offer spousal caregivers more relief from burden than regularly scheduled in-home programming. Contrary to study expectations, receiving emotional support from formal services and receiving in-home tangible services did not appear to affect spousal caregivers' mental health.

Caregiver Perceptions of Community Services

In the current study, higher scores on the Alberta Continuity of Services Scale indicate less perceived service continuity or more service-related stress. To compare the study mean with the average reported by the instrument developers, the coding scheme was reversed (see Adair et al., 2003). It is striking that the sample of spouses in the current study averaged 37.6 points lower than the original sample of adults with severe mental illnesses, a difference that is statistically significant [t(72) = 15.64, p > .001]. That spousal caregivers' experiences with community services are worse than those of a population considered to be "vulnerable" and "hard to serve" is clinically important.

Despite the chronic stress spousal caregivers in this study were exposed to, many indicated that they did not have access to an individual who was monitoring their services or their well-being. For example, 60 percent of the participants strongly disagreed or disagreed that their care was checked regularly and 49.4 percent strongly disagreed or disagreed that someone in the system called to check in on them.

In the bivariate analyses, the Alberta Continuity of Services Scale was associated with caregiver burden (r = .26, p = .02), suggesting that for this sample of spousal caregivers, more stress associated with the service system was associated with more distress as a caregiver.

Informal Supports from Family and Friends

Most caregivers in this study did not have regular help from family and friends (70 percent) and wished for more regular support (52 percent). Periodic help was more commonly available with 40 percent of the sample receiving some periodic help and 27 percent reporting the need for more periodic help. The average caregiver in this sample scored 17.7 out of a possible 24 on the Expressive Support Scale, suggesting a high level of emotional support. None of the categorical variables measuring informal support were associated with caregiver burden. Furthermore, converse to study expectations, perceived emotional support from family and friends was not associated with burden (r = .10, p = .36).

Functioning of the Person with Dementia

The caregivers in this study reported a medium-to-high range of behavioral challenges posed by the persons for whom they were caring on the Dementia Behavior Disturbance Scale. Behaviors reported most frequently by caregivers were lack of interest; asking the same question repeatedly; sleeping excessively; and losing, misplacing, and hiding things.

The mean score for the ADL/IADL subscale of the Older Americans Resources and Services Multidimensional Functional Assessment Questionnaire was 27.9 (SD = 5.6) and a range from 16 to 42. Over 90 percent of the sample reported that their partners required help with ADLs such as meal preparation, managing money, and doing housework; over 75 percent stated that their partners required assistance with basic ADLs such as bathing and dressing.

In keeping with the literature on caregiver stress, the behavioral problems of the person with dementia were significantly related to caregiver burden for this sample of caregivers (r = .38, p < .001). However, physical functioning was not related to caregiver burden (r = .06, p = .60).

The Influence of Service Use, Perceptions of Services, and Functioning of the Person with Dementia on Caregiver Burden

Hierarchical regression analysis was performed with only those variables found to be significant at the bivariate level. This analysis requires the researcher to predetermine the order of entry of variables. The results of the analysis with caregiver burden as an outcome are presented in Table 2. The first block represents the use of tangible formal services and includes one variable measuring the frequency of day program use. The second block represents caregivers' perceptions of services as measured by the Alberta Continuity of Services Scale. The final block represents the person with dementia's behavioral challenges. The model explains 26 percent of the variance of caregiver burden (adjusted [R.sup.2] = .26, F = 6.867, p < .001). In this model, only number of days in which the person with dementia participated in a day program and degree of behavioral difficulties had a statistically significant influence on burden.

DISCUSSION

With increased efforts to reduce health care spending, particularly that associated with inpatient care, and with a rising societal ethos that care of aging individuals is best offered in the community, family members are increasingly finding themselves in the role of supporting their infirm relatives at home. The documented emotional costs to family members are vast and include strains such as depression, anxiety, grief, and overload (Clyburn et al., 2000). One of the populations most vulnerable to these burdens of care are spouses who often view caring as an extension of their marital commitment and who are more likely to continue caring despite the limited services available to support them and the emotional suffering they experience (Perry & O'Connor, 2002).

Many factors contribute to the experience of strain, not the least of which is the level of behavioral changes exhibited by the care recipient. The participants in this study reported a medium to high range of behavioral challenges posed by the people for whom they were caring. As in other studies, these behaviors remained significantly associated with caregiver burden when examined alongside other variables. Behaviors most frequently challenging spousal caregivers in this study were repetitive questioning, lack of interest, excessive sleeping, and the frequent loss or misplacement of items. Although the high level of demand faced by caregivers in this study may represent a bias in this sample, it may also represent a trend noted in the literature that suggests caregivers connect with services late in their caregiving careers (Gottlieb & Johnson, 2000). Although program administrators and clinicians consider this trend to be potentially detrimental to caregivers, service regulations often encourage the late use of services. In Ontario, the province in which this study took place, for example, individuals cannot access public homemaking services unless their spouses require some assistance with personal care such as bathing. In the case of Alzheimer disease and related dementias, these limitations often occur later in the disease process. An important question arises therefore, regarding how service eligibility requirements can best facilitate access to services at an earlier stage of caregiving.

Service Use and Caregiver Burden

This research study specifically explored the associations among adult day program services, homemaking services, institutional respite services (tangible services), geriatric clinic services, and ongoing professional support services (emotional support services) with the Zarit Burden Interview. The only significant association found between these service types and caregiver burden was the association between the frequency of day program use and the burden on caregivers. That is, the more frequently a caregiver's spouse attended a day program, the less burden his or her caregiving spouse experienced.

Although this cross-sectional study cannot offer the conclusion that the frequency of day program use on the part of the person with dementia predicts the level of caregivers' burden, comments provided by caregivers in this study provide possible interpretations of this unique association. Most caregivers who used this service expressed relief at their ability to have a day in the house by themselves. Alongside this relief, however, was the satisfaction and peace of mind associated with knowing that the day program was good for their spouse. It was not simply a six-hour break; this period was a time when caregivers believed that their partners were receiving the stimulation and socialization they needed to maintain their levels of functioning and possibly to prevent rapid deterioration. The notion that the relief of caregivers is experienced, in part, because of the beneficial effect that day programs have on people with dementia has been found in other studies (Gottlieb & Johnson, 2000; Powell & Roberts, 2002).

In this study, the benefits that appear to be associated with day programs were not replicated by homemaking services. Even caregiving spouses who purchased many hours of in-home support, did not appear to benefit in the form of a sense of reduced burden. This finding is significant because homemaking services are more commonly used by caregivers (Smale & Dupuis, 2003) and are often the preferred form of respite for spousal caregivers (Cortell, 1996).

One of the big differences between adult day program services and homemaking services is that adult day programs allow caregivers to enjoy relief from caregiving without having to grapple with the potential sense of "invasion" that may arise when workers come into homes. It is possible, however, that the social benefits offered to people with dementia participating in day programs may also account for some of the differences in the outcomes for caregivers. In the face-to-face interactions with caregivers in this study, for example, many comments were made regarding the lack of available in-home social support for their spouses. Homemaking services were offered either to physically assist their spouses or to provide a combination of supervision and household support, but were rarely framed as opportunities to provide social stimulation to the caregivers' partners.

Unfortunately, the lack of availability of social support for older people with dementia suggests that in-home services may not be meeting the needs of spousal caregivers who cannot exit their roles unless they know the "respite" being offered to them is also socially and psychologically benefiting their partners (Strang & Haughey, 1998). From the comments provided by caregivers in this study and others, it appears that emphasizing the social benefits of day programs could encourage spousal caregivers to access day program services and that redesigning in-home services to better meet the social needs of individuals with dementia may provide more efficacious respite to spousal caregivers (Gottlieb & Johnson, 2000). Further research exploring spousal caregivers' perceptions of the benefits and limitations of these programs is warranted given that the potential efficacious features of each service were not systematically explored in the current study.

Service Stress and Caregiver Burden

One of the hypotheses of this research study was that caregivers' experiences with the community care system would be associated with caregiver stress. Despite the high levels of service stress experienced by the caregivers in this study, service stress was not significantly associated with caregiver burden when examined alongside service use and care recipient behaviors. Several factors may have limited the association between spousal caregivers' experiences with the formal care system and their experience of burden. It is possible, for example, that the impact of their overall loss was so momentous for this group of spousal caregivers that any additional stress generated through an inadequate service system was insignificant. Caregiver studies using grief as a framework for analysis have found that the losses experienced by caregivers are pervasive and significantly contribute to their distress (Boss, 1999; Sanders & Adams, 2005).

The stage of caregiving in which most of these caregivers found themselves may have also affected the relationship between service stress and caregivers' mental health. Frameworks that recognize caregiving as a career indicate that there are particular transitional points in caregiving that expose caregivers to different stressors (Aneshensel et al., 1995). The caregivers in this study provided an average of five years of caregiving, and most caregivers were contemplating the possibility of long-term institutional care (80 of 85 caregivers). Quite possibly, caregivers well into their careers have lowered their expectations about support from formal sources and are thereby less affected by service inadequacies. If this is so, caregivers' expectations regarding the availability and comprehensiveness of formal services may mediate the relationship between service continuity and caregivers' mental health.

Study Limitations and Contributions

There are several limitations to this study. First, this is a cross-sectional study that has not captured spousal caregivers' baseline levels of caregiver burden. Therefore, although it is likely that the number of days caregivers accessed day program services affected their levels of burden, it is also possible that caregivers who used more days of day programming were less burdened. Although this study cannot conclude that the frequency of day programs use reduces spousal caregivers' burden, the comments provided by spousal caregivers in this study do highlight that spousal caregivers perceive these programs to be helpful to them. Second, this study did not explore caregivers' expectations of the service system as a mediating variable between service stress and caregivers' experience of burden. Exploring caregivers' expectations toward service provision might have provided more information that could be used to understand how service systems differentially affect spousal caregivers. Furthermore, exploring how caregivers service expectations differ at various stages of the caregiving career could highlight important changes in caregivers' attitudes toward service provision. Third, this study did not ask caregivers to identify the length of time they have been using services. Therefore the notion that caregivers in this study connected with services late in their caregiving career is speculative. Finally, this study relied on a convenience sample of spouses already in contact with community services who may not be representative of caregivers who are randomly drawn from the community.

Alongside these limitations are some important contributions. This study is probably the first attempt to explore how spouses' use of and experiences with community services affects their mental health. A review of the literature indicates that previous studies on caregiver stress have either excluded formal services completely or collapsed all types of services into one variable. Using the Alberta Continuity of Services Scale to represent "service stress," this study incorporates caregivers' perceptions of services into a model of stress. Furthermore, this study makes important distinctions among service types and their differential associations with caregivers' mental health. It is important to note that differential associations between forms of respite and caregiver burden were found, suggesting that studies accounting for formal services in their models of stress should consider these services separately.

CONCLUSION

The research and service literature is replete with accounts and evidence about the burden experienced by those who care for loved ones with dementia. This awareness has prompted an investment in community-based health and social services aimed at assisting both the patient and the caregiver to reduce distress and ultimately avoid the need for institutional care. There is often an assumption that the services provided by such programs are meeting the needs of clients and are achieving their aims. However, there has been a lack of attention to evaluating the impact of specific elements of these services on different types of familial caregivers. This study found that in-home services, as they are currently offered, do little to reduce the burden of spouses caring for their partners with dementia. Rather, the most effective service is the provision of adult day programs, which provide not only respite for the spousal caregiver, but also opportunities for social interaction for their partners with dementia. Perhaps as a consequence of unmet service needs, spousal caregivers in this study reported relatively high levels of service-related stress. Considering the tremendous investment in community-based services, it is imperative that research addresses in greater detail how specific elements of services and programs can best support spousal caregivers at home and that policymakers and program administrators consider the unique needs of spousal caregivers when planning services for this strained population of caregivers.

Original manuscript received April 30, 2007

Final revision received February 4, 2008

Accepted June 6, 2008

REFERENCES

Adair, C. E., Wild, T. C., Joyce, A., McDougall, G., Gordon, A., Costigan, N., Mitton, C., & Lu, M. (2003). Continuity of Mental Health Services Study of Alberta. A research program on continuity of mental health care. Unpublished report.

Aneshensel, C. S., Pearlin, L. I., Mullan, J. T., Zarit, S., & Whitlatch, C. J. (1995). Profiles in caregiving. The unexpected career. San Diego: Academic Press.

Baumgarten, M., Becker, R., & Gauthier, S. (1990). Validity and reliability of the Dementia Behavior Disturbance Scale. Journal of the American Geriatrics Society, 38, 221-226.

Bedard, M., Pedlar, D., Martin, N. J., Malott, O., & Stones, M. J. (2000). Burden in caregivers of cognitively impaired older adults living in the community: Methodological issues and determinants. International Psychogeriatrics, 12, 307-332.

Boss, P. (1999). Ambiguous loss: Learning to live with unresolved grief. Cambridge, MA: Harvard University Press.

Canadian Institutes of Health Research, Natural Sciences and Engineering Research Council of Canada, Social Sciences and Humanities Research Council of Canada. (1998). Tri-council policy statement: Ethical conduct for research involving humans (with 2000, 2002, and 2005 amendments). Ottawa: Author.

Canadian Study on Health and Aging Working Group. (1994). Patterns of caring for people with dementia m Canada: The Canadian Study of Health and Aging. Canadian Journal on Aging, 13, 470-487.

Clyburn, L. D., Stones, M. J., Hadjistavropoulos, T., & Tuokko, H. (2000). Predicting caregiver burden and depression in Alzheimer's disease. Journals of Gerontology: Series B: Psychological Sciences and Social Sciences, 55B(1), S2-S13.

Cortell, V. (1996). Respite use by dementia caregivers: Preferences and reasons for initial use. Journal of Gerontological Social Work, 26(3/4), 35-55.

Ducharme, F., Levesque, L., Lachance, L., Gangbe, M., Zarit, S. H., Vezina, J., & Caron, C. D. (2007). Older husbands as caregivers. Factors associated with health and the intention to end caregiving. Research on Aging, 29(1), 3-31.

Furlini, L. (2001). The parent they knew and the "new" parent: Daughters' perceptions of dementia of the Alzheimer's type. Home Health Care Services Quarterly, 20(1), 21-38.

George, L. K., & Fillenbaum, G. G. (1985). OARS methodology. A decade of experience in geriatric assessment. Journal of the American Geriatrics Society, 33, 607-615.

Gottlieb, B. H., & Johnson, J. (2000). Respite programs for caregivers of persons with dementia: A review with practice implications. Aging & Mental Health, 4(2), 119-129.

Guberman, N., & Maheu, P. (2000). La rapport entre l'adequation des services de mantien a domicile et la trajectoire de travail des personnes soignantes [The relationship between the adequacy of formal services and the work trajectory of caregivers to elderly relatives]. Canadian Journal on Aging, 19(3), 380-408.

Hills, B., & MacRae, H. (1998). Managing feelings: Caregiving as emotion work. Research on Aging, 20(1), 137-160.

Miller, B., Townsend, A., Carpenter, E., Montgomery, R. V.J., Stull, D., & Young, R. (2001). Social support and caregiver distress: A replication analysis. Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 56B(4), S249-256.

O'Rourke, N., & Tuokko, H. (2000). The psychological and physical costs of caregiving: The Canadian Study of Health and Aging. Journal of Applied Gerontology, 19, 389-404.

Perry, J., & O'Connor, D. (2002). Preserving personhood: (Re)membering the spouse with dementia. Family Relations, 51, 55-62.

Pinquart, M., & Sorensen, S. (2005). Ethnic differences in stressors, resources and psychological outcomes of family caregiving: A meta-analysis. Gerontologist, 45, 90-106.

Powell, J., & Roberts, H. (2002). Identifying 'value' in day care provision for older people. Journal of the Royal Society for the Promotion of Health, 122(3), 158-164.

Sanders, S., & Adams, K. B. (2005). Grief reactions and depression in caregivers of individuals with Alzheimer's disease: Results from a pilot study in an urban setting. Health & Social Work, 30, 287-295.

Schofield, H., Murphy, B., Herrman, H., Bloch, S., & Singh, B. (1998). Carers of people aged 50 and over with physical impairment, memory loss and dementia: A comparative study. Ageing and Society, 13, 355-369.

Seltzer, M. M., & Wailing Li, L. (2000). The dynamics of caregiving. Transitions during a Three-year prospective study. Gerontologist, 40, 165-178.

Smale, B., & Dupuis, S. L. (2003). In their own voices: A profile of dementia caregivers in Ontario. Stage 1: Survey results .Unpublished report, University of Waterloo, Murray Alzheimer Research and Education Program, Waterloo, Ontario

Sorensen, S., Pinquart, M. H., & Duberstein, P. (2002). How effective are interventions for caregivers? An updated meta-analysis. Gerontologist, 42, 356-367.

Statistics Canada. (2005). Income after tax by economic family type (CANSIM Tables 202-0603, Catalogue no. 75-202-XIE). Ottawa: Author.

Stone, R., Cafferata, G. L., & Sangl, J. (1987). Caregivers of the frail elderly: A national profile. Gerontologist, 27, 616-626.

Strang, V. R., & Haughey, M. (1998). Factors influencing caregivers' ability to experience respite. Journal of Family Nursing, 4(3), 231-254.

Wailing Li, L., Seltzer, M. M., & Greenberg, J. S. (1997). Social support and depressive symptoms: Differential patterns in wife and daughter caregivers. Journals of Gerontology, Series B: Psychological Sciences and Social Sciences, 52B, S200-S211.

Zarit, S. H., Reever, K. E., & Bach-Peterson, J. (1980). Relatives of the impaired elderly. Correlates of feelings of burden. Gerontologist, 20, 649-655.

Tamara Sussman, PhD, is assistant professor, School of Social work, McGill University, 3506 University Street, Montreal, Canada H3A 2A7; e-mail: tamara.sussman@mcgill.ca. Cheryl Regehr, PhD, is dean, Factor-Inwentash Faculty of Social Work, Toronto. The authors acknowledge the funding provided by the Nathan Markus Fellowship in Gerontological Social Work. The authors also thank Drs. Sheila Neysmith and David Streiner for their contributions to the doctoral dissertation that informed this article. An earlier version of this article was presented at the Fourth Annual National Gerontological Social Work Conference, February 16-19, 2006, Chicago.
Table 1: The Relationship between Service Use and Caregiver
Distress (N = 85)

                                       M (Zarit
                                        Burden
Variable                              Interview)    SD

In-home professional
  Yes                                   34.42      12.11
  No                                    35.48      15.82

Frequency of in-home professional
Geriatric clinic
  Yes                                   34.10      13.19
  No                                    35.69      12.62

Info/support
  Yes                                   35.54      14.15
  No                                    34.43      14.15

Frequency of info/support
Day program
  Yes                                   3G.7G      14.05
  No                                    32.2G      13.88

Frequency of day program
Homemaking
  Yes                                   35.44      13.40
  No                                     33.6      16.14

Homemaking Hours
Uses institutional respite
  Yes                                   35.47      11.42
  No                                    34.84      14.74

Frequency of institutional respite

                                                     Sig.
Variable                                T      df    of T

In-home professional
  Yes                                  .83    83     .73
  No

Frequency of in-home professional
Geriatric clinic
  Yes                                  .44    53     .66
  No

Info/support
  Yes                                  .36    83     .72
  No

Frequency of info/support
Day program
  Yes                                 1.45   83      .15
  No

Frequency of day program
Homemaking
  Yes                                 -.53   83      .60
  No

Homemaking Hours
Uses institutional respite
  Yes                                  .17     .83   .87
  No

Frequency of institutional respite

                                      Pearson   Spearman    Sig.
Variable                                 r       [rho]      of r

In-home professional
  Yes
  No

Frequency of in-home professional                 -.06      .72
Geriatric clinic
  Yes
  No

Info/support
  Yes
  No

Frequency of info/support                          .28      .09
Day program
  Yes
  No

Frequency of day program               -.34                 .015 *
Homemaking
  Yes
  No

Homemaking Hours                       -.21                 .10
Uses institutional respite
  Yes
  No

Frequency of institutional respite      .33                 .10

Notes: Holm correction was significant only for frequency of day
program. Variable remained significant when correcting for
multiplicity (critical p value = .017).

* Pearson correlation coefficient was significant at p < .05.

Table 2: The Influence of Service Use, Perceptions of Services, and
Functioning of the Person with Dementia on Caregiver Burden (N = 85)

Variable                            B      SE B   [beta]     p

Block One
  Days in day program             -.3.99   1.5     -.33    .01 **
Block Two
  Perception of service system     .16      .09     .22    .08
Block Three
  Behavioral difficulties          .49      .16     .36    .01 **

                                                             p
                                              Adjusted    Adjusted
Variable                          [R.sup.2]   [R.sup.2]   [R.sup.2]

Block One
  Days in day program               .12         .10        .015
Block Two
  Perception of service system      .17         .14        .071
Block Three
  Behavioral difficulties           .31         .26        .005

** Coefficients more than twice their standard error at p < .01.
Gale Copyright: Copyright 2009 Gale, Cengage Learning. All rights reserved.