## Abstract

Purpose:Family caregivers comprise the backbone of long-term-care provision in the United States, yet little is known about how the composition and experience of family caregiving has changed over time. Design and Methods:Data are drawn from the 1989 and 1999 National Long-Term Care Survey and Informal Caregiver Survey to develop nationally representative profiles of disabled older adults and their primary informal caregivers at two points in time. Results:The proportion of chronically disabled community-dwelling older adults who were receiving informal assistance from family or friends declined over the period of interest, whereas the proportion receiving no human help increased. On average, recipients of informal care were older and more disabled in 1999 than in 1989. Primary caregivers were children (41.3%), spouses (38.4%), and other family or friends (20.4%); children were more likely and others less likely to serve as primary caregivers in 1999 relative to 1989. Primary caregivers provided frequent and high levels of help at both points in time. A striking increase was found (from 34.9% to 52.8%) in the proportion of primary caregivers working alone, without secondary caregiver involvement. Implications:In the context of projected demographic trends and budgetary constraints to public health insurance programs, these data underscore the importance of identifying viable strategies to monitor and support family caregivers in the coming years.

It is well established that, in the United States, family caregivers comprise the backbone of the long-term-care workforce. Unfortunately, there is little information to suggest how, in terms of composition and experience, family caregiving has changed over time. The National Long-Term Care Survey (NLTCS) and its accompanying Informal Caregiver Survey represent a rich, nationally representative source of information regarding both caregivers and the disabled older adults in their care. A national profile of caregivers to frail elders was previously published from this data source, although this profile was based on data that are now more than 20 years old (Stone, Cafferata, & Sangl, 1987). Since then, two additional waves of the NLTCS and its Informal Caregiver Survey have been fielded. In this study we draw from the 1989 and 1999 NLTCS and Informal Caregiver Survey to update the profile of primary informal caregivers to disabled older adults, and to examine changes in the composition and experience of this group across a 10-year period.

At this juncture there is a particularly great need for comprehensive information about caregivers to older adults. First, increases in longevity (Federal Interagency Forum on Aging-Related Statistics, 2004), declining household size (U.S. Census Bureau, 2005), and a general shortage of direct health care workers (Buerhaus, Staiger, & Auerbach, 2000) collectively point to fewer caregivers for growing numbers of older adults in the coming years. Recent reports of declines in unpaid assistance and greater reliance on paid help (Liu, Manton, & Aragon, 2000; Spillman & Pezzin, 2000) underscore the practical implications of this issue. Moreover, estimates of the economic value provided by family caregivers range from $45 billion to more than$200 billion annually (U.S. Department of Health and Human Services [DHHS], 1998; Arno, Levine, & Memmott, 1999) and highlight potential financial consequences to society should the efforts of family caregivers be increasingly replaced with paid assistance.

A broad understanding of caregiver circumstances also is essential to monitoring the effects of relevant policy and legislation. For example, changes to Medicare's acute and postacute reimbursement have created incentives to discharge medically complex patients back to the community (and, therefore, to their caregivers) “quicker and sicker.” The 1990 Americans with Disabilities Act and the later Supreme Court Olmstead v. LC (1999) ruling represent a broad societal shift toward “rebalancing” the long-term-care system to reduce institutional care and shift resources to community settings. Related federal initiatives, such as New Freedom and the National Family Caregiver Support Program, have implications for family caregivers (Feinberg & Newman, 2004; Rosenbaum, 2001), yet without empirical data it is impossible to monitor the collective impact of these programs.

Recent studies have drawn on the NLTCS to examine issues related to disability and personal care; however, these studies have largely examined the changing dynamics of disability, or the totality of assistance received by recipients (Liu et al., 2000; Manton & Gu, 2001; Spillman & Pezzin, 2000; Stone & Kemper, 1989). This study is unique in that its focus is on caregivers to disabled older adults. Our goal is to describe, for the two time periods of 1989 and 1999, the composition of disabled older adults' primary informal caregivers, including their sociodemographic characteristics; the intensity, frequency, duration, and scope of assistance they provide; the provision of help from others; the competing demands for their time; and the recipients to whom they provide assistance.

## Methods

### Data

Data for this study were drawn from two linked national surveys that together provide information regarding both disabled older adults (the NLTCS) and their primary caregivers (the Informal Caregiver Survey). The NLTCS is a nationally representative survey of Americans aged 65 years and older that was specifically designed to study the prevalence of chronic disability (Manton, Corder, & Stallard, 1997; Manton & Gu, 2001). Medicare enrollment files constituted the sampling frame, and screening interviews were conducted to determine eligibility for a detailed interview. Among individuals who were screened, all who were identified as chronically disabled (indicated by institutional residence or having any problem with activities of daily living, i.e., ADLs, or instrumental activities of daily living, i.e., IADLs, that had lasted or was expected to last at least 3 months) as well as a subsample of unimpaired individuals were selected for a detailed interview. The NLTCS was first conducted in 1982 and has been repeated every 5 years since 1984.

The Informal Caregiver Survey was conducted among informal primary caregivers to community-dwelling NLTCS participants in 1982, 1989, and 1999. Eligible caregivers included relatives or unpaid nonrelatives who had provided one or more hours of help to NLTCS participants with ADL or IADL activities in the week prior to the community interview. When more than one helper met eligibility criteria, the primary caregiver was designated as the helper who provided the greatest number of hours of ADL assistance to the participant in the previous week, or if no ADL help was provided, the helper who provided the greatest hours of help with IADL tasks. Primary caregivers were questioned on a variety of topics, including the nature and intensity of help provided to recipients, help provided by others, perceptions of their role in providing help, living arrangements, employment, and income.

Ideally, in our analyses we would have drawn on all three waves of the Informal Caregiver Survey, but several modifications to its design impede comparisons between the 1982 survey and later waves. Most notably, the 1982 Informal Caregiver Survey was restricted to ADL caregivers (i.e., IADL caregivers were excluded), but it included both primary and secondary ADL caregivers. The 1989 and 1999 waves included both ADL and IADL caregivers, but these waves were restricted to primary caregivers (i.e., secondary caregivers were excluded). For these reasons, our analyses focus on a comparison of the 1989 and 1999 surveys.

### Study Populations: Defining Disabled Elderly Care Recipients and Their Primary Caregivers

Comparisons across the 1989 and 1999 waves necessitate careful attention to the construction of comparable study samples. One consideration for cross-wave comparisons is that the 1989 Informal Caregiver Survey targeted eligible caregivers only for a subsample of NLTCS participants. Specifically, in 1989, a subsample of all participants who were screened (6,701 of 17,565 individuals) was designated as the sample from which Informal Caregiver Survey interview participants would be drawn (pending eligibility). Given our interest in making comparisons between 1989 and 1999, it was important for us to comprehensively understand sampling strategies employed by the NLTCS and Informal Caregiver Survey across the two time periods. To evaluate and address potential threats to study-sample comparability, we undertook a careful review of pertinent issues.

In this study, we identified chronically disabled care recipients as individuals who both (a) screened into the NLTCS as chronically disabled and (b) indicated that they received active or standby help with at least one ADL activity (eating, bathing, dressing, toileting, transferring, and moving about indoors), or received help with at least one IADL task (grocery shopping, transportation, laundry, light housework, money management, meal preparation, telephone use, medication management, and outdoor mobility) that they could not perform as a result of a health problem or disability. We specified these criteria because a proportion of Informal Caregiver Survey respondents indicated providing assistance to individuals who were not identified as chronically disabled in the NLTCS screener interview. Additionally, the Informal Caregiver Survey eligibility criteria did not distinguish the reason caregivers provided assistance, and some Informal Caregiver Survey respondents were helping NLTCS participants for reasons other than a health problem or disability. Failing to clearly specify chronic disability and the underlying reason for help in defining the recipients of care could impede consistency in study-sample selection across years.

For the purposes of this study, primary caregivers were identified as relatives or unpaid nonrelatives providing ADL or IADL help to a chronically disabled older adult as defined herein. As discussed, caregivers were excluded in this study if they were found to be providing help to individuals who were not chronically disabled, or if they provided help for reasons other than a health problem or disability. However, in some instances caregivers were excluded from the 1989 and 1999 Informal Caregiver Survey study samples but were considered primary informal caregivers for the purposes of this study. This discrepancy arose from two distinct data issues.

First, eligibility for the Informal Caregiver Survey was contingent on the reported provision of one or more hours of help in the week prior to the detailed community interview. When information on hours was missing, for example, if the number of hours was unknown, otherwise eligible primary caregivers were excluded from the Informal Caregiver Survey study sample. In this study we have included otherwise eligible primary caregivers with missing hours; they are included with other nonrespondents (rather than being considered ineligible) and weighted accordingly in analyses of caregivers. This issue is relevant to the comparability of estimates across years, because missing helper hours was more pervasive and therefore disproportionately influential in 1999 relative to 1989 (7.6% in 1989 vs 11.7% in 1999).

A second and more difficult challenge arose concerning chronically disabled individuals who reported receiving assistance but for whom helper payment status and relationship were not recorded. As a result, whether the primary caregiver was informal could not be determined. The proportion of chronically disabled community-dwelling elderly with helpers of indeterminate eligibility status was 1.2% and 11.5% in 1989 and 1999, respectively. These individuals were judged to be not eligible for the Informal Caregiver Survey. In this study, we addressed this issue by presenting two sets of estimates regarding characteristics of chronically disabled older people who received help: one that excluded individuals with helpers of indeterminate eligibility, and another set that included individuals with helpers of indeterminate eligibility. Information regarding caregiver characteristics necessarily excludes helpers of indeterminate eligibility.

Table 1 provides weighted and unweighted estimates of chronically disabled community-dwelling older adults and their helper status. These data indicate that the overall population of chronically disabled community-dwelling older adults increased from 5.3 million in 1989 to 5.6 million 1999. The proportion of chronically disabled individuals who received no human help increased from 22.3% to 29.5%, and the proportion who relied on help from a paid primary caregiver increased from 6.3% in 1989 to 7.6% in 1999. Applying this study's eligibility criteria yielded a total of 2,552 chronically disabled older adults who received assistance from a primary informal caregiver in 1989 (of which 977 were in the designated subsample for Informal Caregiver Survey interview) and 1,622 such individuals in 1999. These unweighted samples were representative of 3,733,000 and 2,880,000 community-dwelling disabled older adults with primary informal caregivers in 1989 and 1999, respectively. Interviews were completed with 814 of 977 eligible caregivers in 1989 (83.3% response rate), and 1,149 of 1,622 eligible caregivers in 1999 (70.8% response rate).

### Construction of Caregiver Weights for National Estimates

Observations from the NLTCS and Informal Caregiver Survey must be weighted to produce nationally representative estimates of the number and characteristics of primary caregivers, and to account for the complex survey design of the NLTCS. Caregiver weights were released with both the 1989 and 1999 Informal Caregiver Survey. Although the NLTCS-detailed community weights served as the starting point for the caregiver weights in both years, the methods by which the caregiver weights were constructed differed. Specifically, in 1999 a nonresponse adjustment was made that accounted for differences in response status by recipients' age, gender, and disability. The 1989 caregiver weights did not include a nonresponse adjustment but were inflated by a constant factor to account for the selection of a subsample of the NLTCS-screened population from which those eligible for the Informal Caregiver Survey were identified. In both years, the caregiver weights released with the data included weights for caregivers to individuals who were unimpaired at the time of the screening interview as well as individuals who were receiving help for reasons other than disability.

To ensure comparability across time periods, new caregiver weights were constructed for this study. In both years, detailed community weights for NLTCS respondents who met study disability criteria were the starting point. A nonresponse adjustment was applied to the weights in both years, using the 1999 approach (response status of recipients by age, i.e., 65–74, 75–84, and 85 years or older; gender; and disability level, i.e., IADLs only, one or two ADLs, three or four ADLs, and five or six ADLs). In addition, we defined nonrespondents to include not only those caregivers who were targeted and did not complete the Informal Caregiver Survey interview, but also caregivers who were not designated as eligible for interview because hours of help were not recorded (discussed in earlier paragraphs and in Table 1). Information on the original nonresponse adjustment strategy for 1989 was obtained from the Center for Demographic Studies at Duke University (R. Schwartz, personal communication, March 16, 2005).

### Data Analysis and Estimation

The analysis first compared demographic characteristics and the functional health of disabled older adults who were receiving assistance from family or friends in 1989 and 1999. We then examined the composition, characteristics, and involvement of primary caregivers for the two time periods. Given the recognized importance of relationship to caregiving dynamics (Shanas, 1979; Stone et al., 1987), we stratified analyses of caregiver characteristics, the assistance provided and received from others, and competing demands by primary caregiver relationship to recipient (spouse, child, other). We observed striking changes in secondary caregiver involvement during the decade of interest; therefore, we performed additional analyses to further elucidate circumstances surrounding these changes.

Standard errors of estimates were generated using the generalized variance function method described in the “Source and Accuracy Statement” released by the Center for Demographic Studies to account for the NLTCS sampling strategy. Estimates that cannot be considered reliable (relative standard errors exceeding 30%) are noted in our tables. Statistical tests of difference were calculated for differences in estimates across the two time periods. Unless specifically mentioned in the text, we limited notable differences to statistically significant findings at the.05 level. Missing data were minimal for recipient characteristics (< 1.0% for all measures) and were generally low for spouse and child caregivers (0.0% to 5.5%), but were somewhat higher for other caregivers.

## Results

### Who Are the Care Recipients?

Table 2 shows the characteristics of chronically disabled community-dwelling older adults overall, as well as those receiving assistance from an informal primary caregiver in 1989 and 1999. Two sets of estimates are provided (excluding and including primary caregivers of unclear eligibility), and although both indicate declines in the number of community-dwelling disabled older adults who received assistance from family or unpaid friends, the magnitude of the decline is smaller when all helpers of unknown eligibility are assumed to be primary informal caregivers (3.8 to 3.5 million vs 3.7 to 2.9 million). In both sets of estimates, the magnitude of decline in the proportion of disabled older adults with informal assistance was smaller among those disabled in five or six ADLs (from 93.4% to 85.0% among those with an eligible primary caregiver) compared with those with IADL disability only (51.9% to 26.3%; data not shown).

Sociodemographic characteristics of chronically disabled older adults receiving assistance did not vary for the most part between 1989 and 1999. The exceptions were increases in age and disability level. There was a significant decline in the percentage of care recipients with IADL disability and an increase in the percentage with five or six ADLs between 1989 and 1999. The magnitude of this change was smaller but remained significant when we included individuals with helpers of unknown eligibility, because these individuals tended to be less disabled.

Focusing on older adults with primary caregivers eligible for this study (the sample of interest for all remaining analyses) in 1999, we found that recipients were, on average, 80 years of age and were predominantly female (66.1%), married (45.4%), or widowed (45.1%), and most often lived with a spouse (35.7%), with a spouse or children (29.5%), or alone (25.4%). Recipients reported a substantial burden of disability; 72.4% received task assistance with one or more ADLs and 25.6% received assistance with five or six ADLs.

### Who Are the Primary Caregivers?

The proportion of spouse primary caregivers was relatively stable across the decade of interest, but the proportions of child and other primary caregivers shifted, with children being more likely to take on this role (41.3% in 1999, up from 35.9% in 1989) and others being less likely (20.4% in 1999, down from 24.3% in 1989; Table 3). By 1999, nearly 45% of all primary caregivers were older than 65 years of age, with 47.4% of spousal primary caregivers being 75 years or older. In 1999, approximately two thirds of primary caregivers were female, although male caregivers were increasingly represented across all relationship types over the decade of interest (p <.10). The distribution of primary caregivers' marital status, self-rated health, and distance to recipients' residence was generally stable within relationship types across the two time periods. More than 95% of all primary caregivers lived within 30 min of recipients in both years; this close proximity likely reflects the high disability levels and related needs of disabled older care recipients.

### Types and Intensity of Caregiver Assistance

Although there was long-standing primary caregiver involvement that required substantial time commitments at both points in time, there were some notable shifts (Table 4). In particular, there were significant increases in the proportion of caregivers providing help on a daily basis, and decreases in the proportion providing help fewer than three days per week, with both trends being most notable among children. While fewer than one in five caregivers had been providing help for less than 1 year at either time point, a significantly greater proportion of arrangements were of shorter duration in 1999, a trend that was evident across all relationship types but that was most pronounced among children (p <.10). Lastly, although spousal caregivers continued to provide help with the same frequency in both time periods, they were generally less likely to provide more than 40 hr of help and more likely to provide fewer than 10 hr of help per week in 1999 relative to 1989.

Primary caregivers most commonly reported providing help with shopping or transportation (85.3%) and household tasks (77.7%), but high levels of assistance were evident across all tasks. Primary caregivers were significantly more likely to provide assistance with indoor mobility and medication administration but were less likely to report providing help with household tasks and shopping or transportation in 1999 as compared with 1989; these trends were observed across all relationship types.

### Competing Demands and Secondary Assistance

There was considerable variation in the proportion of primary caregivers with competing demands across relationship types (Table 5). For example, 31.6% of primary caregivers were employed in 1999, but this ranged from a low of 8.2% among spouses to a high of 50.4% among children. Approximately half of the employed primary caregivers experienced some level of work conflict in terms of rearranging their work schedules, working fewer hours, or taking time off without pay. These trends were fairly constant during the two time periods, with the exception of employment and work conflict among spouses, both of which declined over the decade of interest.

Perhaps the most notable secular trend was the striking increase across all relationships in the proportion of primary caregivers “going it alone” without help from family, friends, or paid caregivers. The proportion of primary caregivers without any help from others increased significantly from 34.9% in 1989 to 52.8% by 1999. The proportion of caregivers who received help from family or friends declined from 38.5% in 1989 to 28.0% by 1999, and the proportion who relied on paid assistance either alone or in conjunction with unpaid help declined from 26.6% in 1989 to 19.2% in 1999. Although declines in the proportion of primary caregivers with paid help was evident across all relationship types, the drop in assistance from family or friends was more evident among spouses and children relative to other caregivers. As in 1989, spouses in 1999 remained most likely to serve as the sole caregiver (67.6%), with other caregivers being least likely (33.9%).

### Secondary Support by Type and Intensity of Caregiver Assistance

To further elucidate circumstances surrounding the shifts in secondary caregiver involvement between 1989 and 1999, we examined primary caregiver involvement by presence of secondary caregivers (Table 6). Not surprisingly, primary caregivers going it alone were most likely to be caring for the least impaired recipients (33.9%) and least likely to be caring for the most disabled ones (18.8%). In contrast, primary caregivers with paid help alone or in conjunction with unpaid help were least likely to care for the least disabled recipients (14.5%) and most likely to provide help to individuals with disability in five or six ADLs (40.1%). With that said, the shift from lower to higher levels of recipient disability was significant across all groups of primary caregivers, regardless of secondary caregiver involvement.

Despite increases in recipient disability, primary caregiver involvement was remarkably stable across the two time periods within categories of secondary help. Primary caregivers with unpaid help provided an average of 29 hr of help per week, which is 1 hr less than the average of 30 hr among those who served as sole caregivers or who relied on paid help. An examination of primary caregiver hours in conjunction with overall hours from all helpers highlights the impact of secondary caregiver support. Total hours of help in 1999 ranged from 30 hr within arrangements where primary caregivers received no help, to 43 hr within arrangements where primary caregivers' efforts were supplemented with unpaid help from family and friends, and 62 hr within arrangements where primary caregivers' efforts were supplemented with paid help. Whereas assistance from all helpers increased from 47 to 62 hr per week among arrangements with paid help, there was a decline in total hours from 46 to 43 hr among arrangements that relied exclusively on unpaid secondary help.

Lastly, given the striking decline in secondary caregiver involvement between 1989 and 1999, we examined the extent to which this finding was robust to varied assumptions regarding primary caregiver eligibility. Relying on unweighted data (the 1989 sampling strategy precludes the use of sampling weights) and making the most extreme assumption that all helpers of indeterminate eligibility were in fact eligible primary informal caregivers with secondary caregiver involvement, we observed an increase in the proportion of primary caregivers going it alone from 34.9% to 42.5% (as compared with unweighted estimates of 35.3% to 51.9% when helpers of unclear eligibility were excluded; data available upon request). Thus, the finding regarding declines in secondary caregiver involvement does appear to be real, despite some uncertainty regarding its magnitude.

## Discussion

This study updates a profile of primary caregivers to community-dwelling disabled older adults and the individuals in their care and examines changes across two time periods—1989 and 1999. Given that family caregivers comprise the backbone of long-term-care provision in this country, there is remarkably little nationally representative information about this group of individuals, and even less information about changes in their composition and experience over time. Our objective in this study was to fill this void, drawing from a well-defined nationally representative sample of disabled older adults at two points in time, a decade apart. Although findings related to the composition of caregivers and the assistance provided are in some respects predictable and stable, several important trends merit attention.

### Declines in Secondary Caregiver Involvement

There were notable increases in age and disability in this sample of disabled older adults living in the community and receiving assistance from family or friends. By 1999, nearly one third of recipients were at least 85 years of age and more than one in four recipients were disabled in five or six ADL activities. In light of increases in the recipients' levels of disability, the observed increases in the proportion of primary caregivers who reported going it alone were particularly striking. The percentage of primary caregivers serving as sole caregivers increased from 34.9% in 1989 to 52.8% in 1999, with declines in assistance from secondary caregivers evident for family and friends, as well as paid helpers. These findings must be interpreted with caution in light of the high proportion of disabled older adults in 1999 who reported receiving help but whose helpers were excluded from the study as a result of missing information on payment status and relationship. However, we found that this trend persisted even under the conservative assumption that all unidentified helpers were informal primary caregivers with secondary caregiver involvement. Declining secondary caregiver involvement has been reported by others (Spillman & Pezzin, 2000), and it has many possible causes, including greater geographic mobility, greater female workforce participation, increasing reliance on child primary caregivers who may have previously provided secondary support, and environmental and technologic changes (e.g., assistive technology) that may generally diminish a person's need for human help.

Declines in the proportion of primary caregivers with secondary helpers who were paid also were sizeable between 1989 and 1999, and seemingly are inconsistent with earlier evidence suggesting greater reliance on paid care among family caregivers (Liu et al., 2000; Spillman & Pezzin, 2000). Although the percentage of chronically disabled older adults relying on a paid primary caregiver increased slightly over this time period, after taking into account trends in secondary paid help, we found that paid help declined overall. The time period of interest in each analysis is an important consideration. Studies showing increased reliance on paid care (Liu et al.; Spillman & Pezzin) examined time periods prior to 1997, when financing of Medicare's home-health benefit shifted from cost-based to prospective reimbursement. There is strong evidence that changes to the financing of Medicare's home-health benefit decreased services overall and shifted the mix of home-health services away from personal care services and toward skilled care (McCall, Korb, Petersons, & Moore, 2003; Murtaugh, McCall, Moore, & Meadow, 2003). Although changes in Medicare reimbursement may have contributed to observed declines in paid help, establishing such a relationship is beyond the scope of these data.

### Changing Caregiver Experiences

The information presented herein confirms broad differences in the caregiving experience by virtue of relationship. In particular, spousal primary caregivers were most likely to serve as sole caregivers and to provide assistance with the greatest frequency and intensity. Child primary caregivers were most likely to face the challenge of juggling caregiving responsibilities along with either work or childrearing. It has been posited that the role of primary caregiver is typically assumed in a hierarchy: first spouses, then children, and finally other relatives or friends (Shanas, 1979). Our findings are generally consistent with this hypothesized hierarchy, although there was a shift over the decade of interest toward representation of men and greater reliance on children. Interestingly, these data also indicate that other primary caregivers were more likely to provide help of greater intensity, both in terms of hours of care and types of tasks (e.g., personal hygiene) in 1999 relative to 1989. Unfortunately, little detail was available regarding the specific relationship of these other primary caregivers to recipients.

It is counterintuitive that, in the face of significant increases in disability among recipients and decreases in the proportion of primary caregivers with secondary helpers, the average hours of help provided by primary informal caregivers decreased from 31 hr to 30 hr between 1989 and 1999. In fact, these data indicate that given the greater task needs of recipients, primary caregivers have shifted the nature and frequency of assistance provided. Relative to 1989, primary caregivers were more likely to provide assistance with greater frequency and to help with administration of medicine and indoor mobility, but were less likely to provide help with household tasks, shopping, or transportation in 1999. Given the generally high levels of care being provided, caregivers may necessarily make trade-offs about what types of assistance to supply within a finite amount of time, shifting help away from what are perhaps more time-consuming and discretionary tasks to those deemed most essential.

We found a greater proportion of chronically disabled older adults who received no help, along with a decrease in the number who relied on primary informal caregivers between 1989 and 1999. Decreases in reliance on human help (Freedman et al., 2004; Spillman, 2004) and informal help (Wolf, Hunt, & Knickman, 2005) have been reported by others and may be attributed to several factors. First, there is evidence to suggest that there have been declines in IADL disability among community-dwelling older adults in recent years (Freedman et al., 2004; Freedman, Martin, & Schoeni, 2002; Manton & Gu, 2001; Spillman), which could reduce the need for human help. Growth in assisted living as a residential arrangement also could partially account for fewer caregivers in this particular study, in that individuals living with three or more unrelated individuals in settings with medical supervision were targeted for the NLTCS institutional interview and were therefore excluded from the Informal Caregiver Survey. Other plausible causes include increased use of technology or assistive devices, or changes to the built environment that might simplify some tasks and diminish a person's need for human help. Whether these observed declines reflect increased unmet need also is an important question and cannot be ruled out.

### Limitations

It is important to emphasize that this study was restricted to primary informal caregivers rather than all informal caregivers who provide assistance to disabled older adults. That said, it is recognized that primary caregivers typically provide higher levels of assistance than other caregivers (Donelan et al., 2002; Spillman & Pezzin, 2000; Stone et al., 1987). To the extent that primary caregivers provide help with the greatest intensity, are more likely to coreside with recipients (Stone et al.), and may serve as the critical link in maintaining community residence, they are an important target for government, health plan, or employer policies that provide financial support or other assistance to caregivers.

The findings of this study vary from other reports of family care in several important ways. First, the study is restricted to assistance provided to disabled older adults living in the community rather than the entirety of individuals of all ages and in all settings who require assistance with daily tasks. Second, this study draws from a well-defined target population of individuals receiving assistance and primary caregivers actively providing help. Because of these issues, the number of caregivers identified here as providing assistance is substantially lower than has been reported elsewhere (DHHS, 1998; National Alliance for Caregiving [NAC] & AARP, 2004).

Despite differences in definitions of family caregiving, our findings are consistent with other published reports in several respects. For example, the disproportionate representation of women, spouses, and children among primary caregivers as well as the reported duration and intensity of assistance are comparable with the findings of other studies (Donelan et al., 2002; NAC & AARP, 2004; Stone et al., 1987). Findings related to the extent to which primary caregivers were employed (NAC & AARP) or living at a distance are lower than has been reported elsewhere (Donelan et al.; NAC & AARP), but they are not unreasonable given the high levels of disability in the sample population. The finding that about half of all employed caregivers experienced some level of work conflict is consistent with other studies describing the considerable economic and personal consequences that often accrue to employed family caregivers (Covinsky et al., 2001; Doty, Jackson, & Crown, 1998).

In order to compare data from 1989 and 1999, we took care to construct comparable study samples and to provide transparency in the treatment of potential threats to comparability posed by sampling and data-collection processes of the NLTCS and Informal Caregiver Survey. Nonetheless, different assumptions regarding the construction of the study samples and the eligibility of helpers of indeterminate eligibility could change the magnitude of study findings. For example, weighted estimates indicate that the proportion of primary caregivers without secondary caregiver involvement increased from 34.9% to 52.8% between 1989 and 1999. Sensitivity analyses based on the most extreme assumption that all helpers of indeterminate eligibility were in fact eligible informal primary caregivers and that all had secondary caregiver involvement reduced the magnitude of this difference from 18% to 8%. Similarly, sensitivity analyses related to key findings regarding declining numbers of primary caregivers and increasing disability levels of chronically disabled older care recipients persisted under the most conservative assumptions, albeit reduced in magnitude (Table 1). Thus, despite a number of important data issues, we believe the general trends reported here are robust to the most obvious concerns regarding cross-wave comparisons.

### Policy Implications

This comprehensive profile of primary caregivers and the disabled older adults to whom they provide assistance has considerable relevance to policy makers charged with crafting sustainable long-term-care systems. The cost of long-term care in 2000 was estimated at $123 billion, and this cost is expected to more than double to$295 billion by 2030 (Congressional Budget Office, 1999). These figures are particularly impressive in that they exclude all accounting of family caregivers' efforts. The implications of reductions in family caregivers' help could be enormous. Other countries have moved to explicitly recognize and reward family care as part of the long-term-care continuum (Geraedts, Heller, & Harrington, 2000). There is some momentum in this country toward adopting consumer-directed models that afford greater flexibility in directing payments to nonagency caregivers, including family, though efforts thus far have largely been concentrated within the Medicaid program (Benjamin, 2001; Foster, Brown, Phillips, & Carlson, 2005; Foster, Brown, Phillips, Schore, & Carlson, 2003).

The National Family Caregiver Support Program is the first federal initiative to explicitly recognize and support family caregivers. Stated goals of the program include dissemination of information to caregivers so that they may better access supportive services, organization of support groups, as well as the provision of individual counseling, training, respite care, and other supplemental services on a limited basis. National program funding totals \$155 million, which is small relative to the estimated contributions of family caregivers (e.g., Arno, Levine, & Memmott, 1999; DHHS, 1998). Preliminary reports indicate considerable variability in the program administration, design, and service delivery of the National Family Caregiver Support Program, but that it has increased awareness of family caregivers (rather than only recipients) as legitimate consumers of community services (Feinberg & Newman, 2004).

In fact, looking more broadly than the National Family Caregiver Support Program, there is little consensus regarding how to best support caregivers. Numerous caregiver interventions have cumulatively failed to generate a strong evidence base about how to best train and support caregivers in their diverse roles (Knight, Lutzky, & Macofsky-Urban, 1993; Sorensen, Pinquart, & Duberstein, 2002; Zarit, Gaugler, & Jarrott, 1999). In addition, most intervention studies have targeted narrow populations of caregivers who provide assistance to individuals with a specific index disease (Brouwer et al., 2004; Han & Haley, 1999; Harding & Higginson, 2003; Schulz et al., 2002), limiting the ability to translate study findings more broadly. Fragmented financing, multiple points of entry, variable service offerings, and lack of a standardized approach to caregiver assessment cumulatively impede family caregivers' access and use of existing community services.

This study confirms the commitment and contributions of primary informal caregivers to caring for disabled older adults living in the community. Moreover, findings indicate that despite substantial increases in age and disability among care recipients, there were striking increases in the proportion of primary caregivers providing assistance alone. These data, in conjunction with projected increases in the numbers of older people living with chronic illness and disability, underscore the importance of improving our understanding of what approaches most effectively sustain family caregivers. The findings indicate a very real need to systematically and comprehensively monitor family caregiving and the impact of caregiver-support initiatives in the coming years.

This project was funded by the National Institutes of Health/National Institute of Aging under Grant 1 R03 AG025153-01 and an unrestricted grant from Pfizer. We gratefully acknowledge the advice, comments, and insight of Brenda Spillman, Richard Schwartz, and Barbara Priboth.

1

Department of Health Policy and Management and the Roger C. Lipitz Center for Integrated Health Care, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD.

2

Division of Geriatrics and Gerontology, Johns Hopkins University School of Medicine, Baltimore, MD.

Decision Editor: Linda S. Noelker, PhD

Table 1.

Study Status, Unweighted Sample, and Weighted National Estimates.

1989

1999

Variable Unweighted (NUnweighted (%) Weighted (N)a Weighted (%) Unweighted (NUnweighted (%) Weighted (N)a Weighted (%)
Helper status of chronically disabled elderly people
Total chronically disabled community-dwelling individuals aged 65+ 3,547 100.0 5,321,000 100.0 3,115 100.0 5,621,000 100.0
Chronically disabled, does not receive help 713 20.1 1,188,000 22.3 872 28.0 1,656,000 29.5
Chronically disabled, receives help 2,834 79.9 4,133,000 77.7 2,243 72.0 3,965,000 70.5
Primary caregiver is paidb 240 6.8 335,000 6.3 264 8.5 425,000 7.6
Primary caregiver is family or unpaid (eligible for this study)c 2,552 71.9 3,733,000 70.2 1,622 52.1 2,880,000 51.2
Eligibility of primary caregiver is uncleard 42 1.2 64,400 1.2 357 11.5 660,000 11.8
National Informal Caregiver Survey interview status
Primary caregiver is family or unpaid (eligible) 977e 100.0 3,733,000 100.0 1,622 100.0 2,880,000 100
Completed Informal Caregiver Survey interview 814 83.3 — — 1,149 70.8 — —
No Informal Caregiver Survey interview
Caregiver nonresponse 89 9.1 — — 283 17.4 — —
No interview attempted (no hours reported for helper) 74 7.6 — — 191 11.8 — —
1989

1999

Variable Unweighted (NUnweighted (%) Weighted (N)a Weighted (%) Unweighted (NUnweighted (%) Weighted (N)a Weighted (%)
Helper status of chronically disabled elderly people
Total chronically disabled community-dwelling individuals aged 65+ 3,547 100.0 5,321,000 100.0 3,115 100.0 5,621,000 100.0
Chronically disabled, does not receive help 713 20.1 1,188,000 22.3 872 28.0 1,656,000 29.5
Chronically disabled, receives help 2,834 79.9 4,133,000 77.7 2,243 72.0 3,965,000 70.5
Primary caregiver is paidb 240 6.8 335,000 6.3 264 8.5 425,000 7.6
Primary caregiver is family or unpaid (eligible for this study)c 2,552 71.9 3,733,000 70.2 1,622 52.1 2,880,000 51.2
Eligibility of primary caregiver is uncleard 42 1.2 64,400 1.2 357 11.5 660,000 11.8
National Informal Caregiver Survey interview status
Primary caregiver is family or unpaid (eligible) 977e 100.0 3,733,000 100.0 1,622 100.0 2,880,000 100
Completed Informal Caregiver Survey interview 814 83.3 — — 1,149 70.8 — —
No Informal Caregiver Survey interview
Caregiver nonresponse 89 9.1 — — 283 17.4 — —
No interview attempted (no hours reported for helper) 74 7.6 — — 191 11.8 — —

aWeighted with the NLTCS sample weights for participants who completed the detailed community interview.

bIncludes (a) participants with paid helpers only and (b) participants with hours for a paid helper only. Not eligible for the Informal Caregiver Survey. Note that a small number of individuals in this category reported both paid and unpaid hours; when the helper with the greatest number of hours was paid, these individuals were excluded from the sample (n = 25 in 1989 and n = 6 in 1999). Although the percentage of chronically disabled elderly people with paid primary caregivers increased slightly between 1989 and 1999, paid help declined overall when involvement of secondary paid caregivers was considered (see Table 5).

cDesignated as eligible for the Informal Caregiver Survey.

dNot designated as eligible for the Informal Caregiver Survey. Information regarding helper's payment status and relationship to NLTCS participant is unknown.

e1989 Informal Caregiver Survey-eligible caregivers were identified for a subsample of participants who completed the NLTCS screening interview whereas 1999 Informal Caregiver Survey eligibility was ascertained for all participants who completed the NLTCS screening interview. A nonresponse adjustment was made to sample weights that reflects age, gender, and disability level of care recipients, following the protocol used by R. Schwartz (personal communication, March 16, 2005).

Table 2.

Characteristics of Chronically Disabled Community-Dwelling Older Adults Overall and With Informal Primary Caregivers Using Alternative Sample Definitions.

Community-Dwelling Chronically Disabled, Receives Help, and:

Total Chronically Disabled Adults Aged 65+ Living in the Community

Primary Caregiver Eligible for Informal Caregiver Survey

Primary Caregiver Eligible or Eligibility Uncleara

Care Recipient Characteristics 1989 1999 1989 1999 1989 1999
Total number, n (%) 5,320,951 5,621,041 3,733,000 2,880,000 3,798,000 3,540,000
100.0 100.0 100.0 100.0 100.0 100.0
Age in years (%)
65–74 36.6 29.7** 33.5 28.6** 33.9 28.9**
75–84 42.2 42.9 43.0 41.3 42.7 41.1
85+ 21.2 27.4** 23.5 30.0** 23.3 30.0**
Mean age 77.9 79.4 78.5 79.7 78.4 79.7
Gender (%)
Female 67.8 68.2 67.4 66.1 67.2 67.1
Male 32.2 31.8 32.6 33.9 32.8 32.9
Marital status (%)b
Married 42.0 38.7* 44.0 45.4 44.1 42.9
Widowed 45.6 49.7** 45.8 45.1 45.5 47.0
Divorced or separated 7.3 7.2 5.8 5.6 5.8 5.8
Never married 4.7 4.1 3.9 3.6 4.1 3.9
Living arrangements (%)
Lives alone 36.4 37.8 29.1 25.4* 29.1 29.2
Lives with spouse only 35.1 31.1** 35.9 35.7 35.9 34.2
Lives with spouse and/or children 21.0 22.2 26.9 29.5 26.7 27.8
Other arrangements 7.5 8.9 8.1 9.4 8.2 8.8
Perceived health status (%)b
Excellent 11.9 10.2 10.4 7.7** 10.5 8.4*
Good 31.4 32.7 28.9 28.4 29.1 29.3
Fair 34.3 34.5 33.8 34.8 33.7 34.5
Poor 21.7 20.7 26.2 27.6 26.1 25.8
IADLs only 53.3 53.9 39.5 27.6** 39.4 35.0
1–2 ADLs 23.9 20.6** 30.3 30.5 30.6 29.0
3–4 ADLs 9.7 10.1 12.8 16.3 12.7 14.6
5–6 ADLs 13.1 15.4* 17.5 25.6** 17.3 21.4**
Community-Dwelling Chronically Disabled, Receives Help, and:

Total Chronically Disabled Adults Aged 65+ Living in the Community

Primary Caregiver Eligible for Informal Caregiver Survey

Primary Caregiver Eligible or Eligibility Uncleara

Care Recipient Characteristics 1989 1999 1989 1999 1989 1999
Total number, n (%) 5,320,951 5,621,041 3,733,000 2,880,000 3,798,000 3,540,000
100.0 100.0 100.0 100.0 100.0 100.0
Age in years (%)
65–74 36.6 29.7** 33.5 28.6** 33.9 28.9**
75–84 42.2 42.9 43.0 41.3 42.7 41.1
85+ 21.2 27.4** 23.5 30.0** 23.3 30.0**
Mean age 77.9 79.4 78.5 79.7 78.4 79.7
Gender (%)
Female 67.8 68.2 67.4 66.1 67.2 67.1
Male 32.2 31.8 32.6 33.9 32.8 32.9
Marital status (%)b
Married 42.0 38.7* 44.0 45.4 44.1 42.9
Widowed 45.6 49.7** 45.8 45.1 45.5 47.0
Divorced or separated 7.3 7.2 5.8 5.6 5.8 5.8
Never married 4.7 4.1 3.9 3.6 4.1 3.9
Living arrangements (%)
Lives alone 36.4 37.8 29.1 25.4* 29.1 29.2
Lives with spouse only 35.1 31.1** 35.9 35.7 35.9 34.2
Lives with spouse and/or children 21.0 22.2 26.9 29.5 26.7 27.8
Other arrangements 7.5 8.9 8.1 9.4 8.2 8.8
Perceived health status (%)b
Excellent 11.9 10.2 10.4 7.7** 10.5 8.4*
Good 31.4 32.7 28.9 28.4 29.1 29.3
Fair 34.3 34.5 33.8 34.8 33.7 34.5
Poor 21.7 20.7 26.2 27.6 26.1 25.8
IADLs only 53.3 53.9 39.5 27.6** 39.4 35.0
1–2 ADLs 23.9 20.6** 30.3 30.5 30.6 29.0
3–4 ADLs 9.7 10.1 12.8 16.3 12.7 14.6
5–6 ADLs 13.1 15.4* 17.5 25.6** 17.3 21.4**

Notes: ADLs = activities of daily living; IADLs = instrumental ADLs. ADLs include eating, bathing, toileting, transferring, and indoor mobility. IADLs include grocery shopping, transportation, laundry, lighting housework, money management, meal preparation, telephone use, medication management, and outdoor mobility. Source: 1989 and 1999 National Long-Term Care Survey and Informal Caregiver Survey.

aIndividuals who reported receiving assistance but neither relationship nor payment status of helper was recorded; eligibility as an informal caregiver could not be determined.

bInformation was missing for a small proportion of NLTCS participants (< 1.0% in both years).

*Statistically significant from the 1989 estimate at the 10% level of significance.

**Statistically significant from the 1989 estimate at the 5% level of significance.

Table 3.

Characteristics of Primary Informal Caregivers to Chronically Disabled Elderly Care Recipients.

All

Spouse

Child

Other

Caregiver Characteristics 1989 1999 1989 1999 1989 1999 1989 1999
Population 3,733,000 2,880,000 1,487,000 1,105,000 1,341,000 1,190,000 906,000 588,000
Percent 100.0 100.0 39.8 38.4 35.9 41.3** 24.3 20.4**
Age, in years (%)
14–44 13.8 11.2 0.2 1.1 24.3 17.3** 20.7 17.6
45–64 30.9 40.3 14.2 10.0* 57.9 68.0** 18.2 41.1
65–74 27.2 21.8 47.0 37.6** 15.6 11.7 11.7 12.3
75 18.0 22.1 38.2 47.4** 1.1 2.1 10.1 14.9
Mean age 61.6 62.5 72.1 73.8 53.0 54.5 53.7 57.1
Gender (%)
Male 28.6 31.4* 40.6% 42.7 23.5 26.7 16.3 19.4
Female 71.2 67.4* 59.4% 57.0 76.5 73.3 82.7 74.8
Marital status (%)
Married 70.8 71.2 100.0% 99.1 54.9 56.6 46.6 48.1
Widowed 7.6 6.1 0.0% 0.0 9.6 7.7 16.9 14.1
Divorced or separated 9.5 11.4 0.0% 0.4 17.0 20.6 13.8 13.3
Never married 10.2 9.3 0.0% 0.3 17.1 14.8 16.7 15.2
Perceived health status (%)
Excellent 25.6 22.9 19.9% 14.0* 27.4 28.3 32.2 28.5
Good 41.5 41.9 39.9% 46.9** 42.5 40.3 42.8 35.6
Fair 24.1 24.1 30.6% 27.1 21.7 21.7 16.9 23.1
Poor 7.6 5.1 8.7% 7.1 7.0 4.2* 6.9 3.4
Distance to recipient (%)
Coreside 68.9 68.3 100.0% 97.3 54.7 51.0 38.7 48.2
≤ 10 min 20.0 18.4 0.0% 0.6 28.7 31.9 40.0 24.5
11–30 min 6.6 8.6* 0.0% 0.0 10.2 12.4 12.2 17.0
31–60 min 1.8 1.2 0.0% 0.0 2.0 2.1 4.5 1.3
> 1 hr 2.1 1.4 0.0% 0.8 4.3 2.3* 2.3 0.6
All

Spouse

Child

Other

Caregiver Characteristics 1989 1999 1989 1999 1989 1999 1989 1999
Population 3,733,000 2,880,000 1,487,000 1,105,000 1,341,000 1,190,000 906,000 588,000
Percent 100.0 100.0 39.8 38.4 35.9 41.3** 24.3 20.4**
Age, in years (%)
14–44 13.8 11.2 0.2 1.1 24.3 17.3** 20.7 17.6
45–64 30.9 40.3 14.2 10.0* 57.9 68.0** 18.2 41.1
65–74 27.2 21.8 47.0 37.6** 15.6 11.7 11.7 12.3
75 18.0 22.1 38.2 47.4** 1.1 2.1 10.1 14.9
Mean age 61.6 62.5 72.1 73.8 53.0 54.5 53.7 57.1
Gender (%)
Male 28.6 31.4* 40.6% 42.7 23.5 26.7 16.3 19.4
Female 71.2 67.4* 59.4% 57.0 76.5 73.3 82.7 74.8
Marital status (%)
Married 70.8 71.2 100.0% 99.1 54.9 56.6 46.6 48.1
Widowed 7.6 6.1 0.0% 0.0 9.6 7.7 16.9 14.1
Divorced or separated 9.5 11.4 0.0% 0.4 17.0 20.6 13.8 13.3
Never married 10.2 9.3 0.0% 0.3 17.1 14.8 16.7 15.2
Perceived health status (%)
Excellent 25.6 22.9 19.9% 14.0* 27.4 28.3 32.2 28.5
Good 41.5 41.9 39.9% 46.9** 42.5 40.3 42.8 35.6
Fair 24.1 24.1 30.6% 27.1 21.7 21.7 16.9 23.1
Poor 7.6 5.1 8.7% 7.1 7.0 4.2* 6.9 3.4
Distance to recipient (%)
Coreside 68.9 68.3 100.0% 97.3 54.7 51.0 38.7 48.2
≤ 10 min 20.0 18.4 0.0% 0.6 28.7 31.9 40.0 24.5
11–30 min 6.6 8.6* 0.0% 0.0 10.2 12.4 12.2 17.0
31–60 min 1.8 1.2 0.0% 0.0 2.0 2.1 4.5 1.3
> 1 hr 2.1 1.4 0.0% 0.8 4.3 2.3* 2.3 0.6

Notes: “Child” refers to sons or daughters; sons-in-law and daughters-in-law were categorized as “Other.” Column percentages may not sum to 100% due to missing data. Results of tests of statistical significance are not shown when missing data exceeded 5% (age and perceived health status for “All Primary Caregivers (PCGs),” perceived health status for “Child PCG,” and for all measures among “Other PCG.” Source: 1989 and 1999 National Long-Term Care Survey and Informal Caregiver Survey.

*Statistically significant from the 1989 estimate at the 10% level of significance.

**Statistically significant from the 1989 estimate at the 5% level of significance.

Relative standard error ≥ 30%.

Table 4.

Types and Intensity of Assistance Provided by Informal Primary Caregivers by Relationship to Recipient.

All

Spouse

Child

Other

Type and Intensity of Assistance 1989 1999 1989 1999 1989 1999 1989 1999
Length of Caregiving, in years (%)
< 1 14.3 18.2** 14.2 18.0 11.9 16.7* 18.3 21.6
1–4 33.9 30.8 31.7 30.7 37.1 31.2 32.9 30.1
> 4 50.5 45.9** 52.9 47.0* 49.7 47.2 47.6 41.0
Average days per week (%)
< 3 20.4 15.7** 7.1 6.9 23.7 17.4** 37.2 28.2**
3–6 10.0 9.1 3.6 3.3 14.1 12.0 14.5 13.6
7 69.6 75.0** 89.3 89.5 62.2 69.5** 48.4 55.4
Average hours per week (%)
≤ 10 30.4 33.8 17.1 24.4** 32.6 37.8 49.0 41.7
> 10 ≤ 20 18.2 17.2 15.6 13.6 21.4 20.5 17.9 16.3
> 20 ≤ 40 22.7 22.7 25.0 27.4 22.5 19.5 19.0 19.4
> 40 28.0 26.3 41.2 34.3** 22.9 21.2 13.8 20.4*
Mean PCG hours 31 30 41 38 29 25 19 24
Mean overall hours 41 38 48 41 39 37 31 33
Type of assistance provideda
Shopping and/or transportation 91.4 85.3** 89.5 86.1 97.3 88.8** 85.7 76.0**
Household tasks 81.5 77.7** 88.4 83.2** 80.0 75.1 72.4 72.1
Finances 49.4 49.4 53.2 52.8 54.0 55.8 36.4 29.7
Personal care and nursing 44.9 48.5* 55.8 55.7 42.1 43.6 31.2 44.6**
Administration of medicine 36.6 41.8** 43.5 46.7 36.4 39.6 25.7 36.9**
Indoor mobility 29.1 35.1** 31.9 39.7** 28.2 33.3* 25.7 29.8
All

Spouse

Child

Other

Type and Intensity of Assistance 1989 1999 1989 1999 1989 1999 1989 1999
Length of Caregiving, in years (%)
< 1 14.3 18.2** 14.2 18.0 11.9 16.7* 18.3 21.6
1–4 33.9 30.8 31.7 30.7 37.1 31.2 32.9 30.1
> 4 50.5 45.9** 52.9 47.0* 49.7 47.2 47.6 41.0
Average days per week (%)
< 3 20.4 15.7** 7.1 6.9 23.7 17.4** 37.2 28.2**
3–6 10.0 9.1 3.6 3.3 14.1 12.0 14.5 13.6
7 69.6 75.0** 89.3 89.5 62.2 69.5** 48.4 55.4
Average hours per week (%)
≤ 10 30.4 33.8 17.1 24.4** 32.6 37.8 49.0 41.7
> 10 ≤ 20 18.2 17.2 15.6 13.6 21.4 20.5 17.9 16.3
> 20 ≤ 40 22.7 22.7 25.0 27.4 22.5 19.5 19.0 19.4
> 40 28.0 26.3 41.2 34.3** 22.9 21.2 13.8 20.4*
Mean PCG hours 31 30 41 38 29 25 19 24
Mean overall hours 41 38 48 41 39 37 31 33
Type of assistance provideda
Shopping and/or transportation 91.4 85.3** 89.5 86.1 97.3 88.8** 85.7 76.0**
Household tasks 81.5 77.7** 88.4 83.2** 80.0 75.1 72.4 72.1
Finances 49.4 49.4 53.2 52.8 54.0 55.8 36.4 29.7
Personal care and nursing 44.9 48.5* 55.8 55.7 42.1 43.6 31.2 44.6**
Administration of medicine 36.6 41.8** 43.5 46.7 36.4 39.6 25.7 36.9**
Indoor mobility 29.1 35.1** 31.9 39.7** 28.2 33.3* 25.7 29.8

Notes: “Child” refers to sons or daughters; sons-in-law and daughters-in-law were categorized as “Other.” Note: Column percentages may not sum to 100% due to missing data. Missing data were less than 5% of the total for all measures except length of caregiving among “Other” primary caregivers. Source: 1989 and 1999 National Long-Term Care Survey and Informal Caregiver Survey.

aShopping and/or transportation refers to shopping for groceries, running errands, outside mobility, and transportation; household tasks refers to meal preparation, light housework, and laundry; finances refers to help with managing money; personal care and nursing includes eating, dressing, bathing, and toileting; administration of medicine includes giving shots, injections, medicine, pills, or changing bandages; indoor mobility refers to transferring and getting around.

*Statistically significant from the 1989 estimate at the 10% level of significance.

**Statistically significant from the 1989 estimate at the 5% level of significance.

Table 5.

Competing Demands and Secondary Assistance by Relationship of Informal Primary Caregiver to Recipient.

All

Spouse

Child

Other

Caregiver Characteristics 1989 1999 1989 1999 1989 1999 1989 1999
Coresiding children < 15 years (%) 11.8 11.3 2.8 3.8 18.3 18.0 17.1 12.2
Employment (%)
Employed for pay 33.5 31.6 12.2 8.2* 52.0 50.4 41.1 37.6
Employed PCGs who:
Rearranged schedules 38.9 38.6 39.9 29.9** 35.1 41.5* 45.3 34.2**
Worked fewer hours 22.8 23.3 32.3 19.9** 25.4 25.8 13.3 18.0
Took time off without pay 19.6 20.2 19.2 13.4** 21.3 22.5 16.5 16.7
Work conflict (any of the above) 49.2 49.0 48.6 39.6** 50.2 52.2 47.6 44.3
Secondary caregiver involvement (%)
None 34.9 52.8** 46.0 67.6** 28.8 48.3** 25.6 33.9*
Other family or friends 38.5 28.0** 31.7 18.1** 43.0 30.0** 43.0 42.6
Family, friends, and paid help 26.6 19.2** 22.3 14.0** 28.2 21.8** 31.4 23.5*
All

Spouse

Child

Other

Caregiver Characteristics 1989 1999 1989 1999 1989 1999 1989 1999
Coresiding children < 15 years (%) 11.8 11.3 2.8 3.8 18.3 18.0 17.1 12.2
Employment (%)
Employed for pay 33.5 31.6 12.2 8.2* 52.0 50.4 41.1 37.6
Employed PCGs who:
Rearranged schedules 38.9 38.6 39.9 29.9** 35.1 41.5* 45.3 34.2**
Worked fewer hours 22.8 23.3 32.3 19.9** 25.4 25.8 13.3 18.0
Took time off without pay 19.6 20.2 19.2 13.4** 21.3 22.5 16.5 16.7
Work conflict (any of the above) 49.2 49.0 48.6 39.6** 50.2 52.2 47.6 44.3
Secondary caregiver involvement (%)
None 34.9 52.8** 46.0 67.6** 28.8 48.3** 25.6 33.9*
Other family or friends 38.5 28.0** 31.7 18.1** 43.0 30.0** 43.0 42.6
Family, friends, and paid help 26.6 19.2** 22.3 14.0** 28.2 21.8** 31.4 23.5*

Notes: PCG = primary caregiver. “Child” refers to sons or daughters; sons-in-law and daughters-in-law were categorized as “Other.” Source: 1989 and 1999 National Long-Term Care Survey and Informal Caregiver Survey.

*Statistically significant from the 1989 estimate at the 10% level of significance.

**Statistically significant from the 1989 estimate at the 5% level of significance.

Table 6.

Aspects of Informal Caregiving by Presence of Secondary Caregivers.

Secondary Caregivers

None

Paida

Aspects of Informal Caregiving 1989 1999 1989 1999 1989 1999
Population, in thousands 1,301,000 1,520,000 1,438,000 807,000 994,000 552,000
Percent 100.0 100.0 100.0 100.0 100.0 100.0
Care recipient disabled in: (%)
IADLs only 45.8 33.9** 41.4 24.8** 28.3 14.5**
1–2 ADLs 33.6 32.8 26.3 29.3 31.7 26.0
3–4 ADLs 10.4 14.5* 13.8 17.5 14.5 19.5
5–6 ADLs 10.2 18.8** 18.6 28.4** 25.5 40.1**
Average days per week (%)
< 3 17.6 13.7* 21.3 15.6* 22.7 20.8
3–6 8.2 7.8 9.4 11.7 13.1 8.4*
7 74.1 76.9 69.3 72.4 64.2 70.3
Average hours per week (%)
≤ 10 27.3 31.8 29.9 36.2* 35.2 34.1
> 10 ≤ 20 18.9 16.6 19.3 17.1 15.8 18.3
> 20 ≤ 40 25.9 23.3 19.6 21.4 22.9 22.0
> 40 27.3 27.0 31.0 24.7 24.5 25.6
Hours from all helpers 30 30 46 43 47 62
Primary caregiver hours 30 30 34 29 29 30
Type of assistance provided (%)b
Shopping and/or transportation 91.3 87.9* 92.2 83.4** 90.2 80.7**
Household tasks 84.0 79.3* 81.9 79.1 77.6 71.1*
Finances 48.8 51.8 48.3 40.0** 51.8 56.3
Personal care and nursing 43.4 45.9 44.0 51.7** 48.1 50.9
Administration of medicine 30.5 39.0** 38.4 42.2 42.1 48.9
Indoor mobility 21.6 30.1** 31.4 38.9** 35.5 43.2*
Secondary Caregivers

None

Paida

Aspects of Informal Caregiving 1989 1999 1989 1999 1989 1999
Population, in thousands 1,301,000 1,520,000 1,438,000 807,000 994,000 552,000
Percent 100.0 100.0 100.0 100.0 100.0 100.0
Care recipient disabled in: (%)
IADLs only 45.8 33.9** 41.4 24.8** 28.3 14.5**
1–2 ADLs 33.6 32.8 26.3 29.3 31.7 26.0
3–4 ADLs 10.4 14.5* 13.8 17.5 14.5 19.5
5–6 ADLs 10.2 18.8** 18.6 28.4** 25.5 40.1**
Average days per week (%)
< 3 17.6 13.7* 21.3 15.6* 22.7 20.8
3–6 8.2 7.8 9.4 11.7 13.1 8.4*
7 74.1 76.9 69.3 72.4 64.2 70.3
Average hours per week (%)
≤ 10 27.3 31.8 29.9 36.2* 35.2 34.1
> 10 ≤ 20 18.9 16.6 19.3 17.1 15.8 18.3
> 20 ≤ 40 25.9 23.3 19.6 21.4 22.9 22.0
> 40 27.3 27.0 31.0 24.7 24.5 25.6
Hours from all helpers 30 30 46 43 47 62
Primary caregiver hours 30 30 34 29 29 30
Type of assistance provided (%)b
Shopping and/or transportation 91.3 87.9* 92.2 83.4** 90.2 80.7**
Household tasks 84.0 79.3* 81.9 79.1 77.6 71.1*
Finances 48.8 51.8 48.3 40.0** 51.8 56.3
Personal care and nursing 43.4 45.9 44.0 51.7** 48.1 50.9
Administration of medicine 30.5 39.0** 38.4 42.2 42.1 48.9
Indoor mobility 21.6 30.1** 31.4 38.9** 35.5 43.2*

Notes: ADLs = activities of daily living; IADLs = instrumental ADLs. ADLs include eating, bathing, toileting, transferring, and indoor mobility. IADLs include grocery shopping, transportation, laundry, light housework, money management, meal preparation, telephone use, medication management, and outdoor mobility. Column percentages may not sum to 100% due to missing data; missing data were less than 5% for all measures. Source: 1989 and 1999 National Long-Term Care Survey and Informal Caregiver Survey.

aIndicates the use of paid help either alone or in conjunction with unpaid help from family and/or friends.

bShopping and/or transportation refers to shopping for groceries, running errands, outside mobility, and transportation; household tasks refers to meal preparation, light housework, and laundry; finances refers to help with managing money; personal care and nursing includes eating, dressing, bathing, and toileting; administration of medicine includes giving shots, injections, medicine, pills, or changing bandages; indoor mobility refers to transferring and getting around.

*Statistically significant from the 1989 estimate at the 10% level of significance.

**Statistically significant from the 1989 estimate at the 5% level of significance.

## References

Americans With Disabilities Act of 1990, 42 U.S.C.A. § 12101 et seq. (West 1993).
Arno, P., Levine, C., & Memmott, M., (
1999
). The economic value of informal caregiving.
Health Affairs (Millwood),

18,
182
-188.
Benjamin, A., (
2001
). Consumer-directed services at home: A new model for persons with disabilities.
Health Affairs,

20,
80
-95.
Brouwer, W., van Exel, N., van de Berg, B., Dinant, H., Koopmanschap, M., & van den Bos, G., (
2004
). Burden of caregiving: Evidence of objective burden, subjective burden, and quality of life impacts on informal caregivers of patients with rheumatoid arthritis.
Arthritis and Rheumatism,

51,
570
-577.
Buerhaus, P., Staiger, D., & Auerbach, D., (
2000
). Implications of an aging registered nurse workforce.
Journal of the American Medical Association,

283,
2948
-2954.
Congressional Budget Office. (
1999
). CBO memorandum: Projections of expenditures for long-term care services for the elderly. Washington DC: Author.
Covinsky, K., Eng, C., Lui, L., Sands, L., Sehgal, A., & Walter, L., et al (
2001
). Reduced employment in caregivers of frail elders: Impact of ethnicity, patient clinical characteristics, and caregiver characteristics.
Journal of Gerontology: Medical Sciences,

56A,
M707
-M713.
Donelan, K., Hill, C., Hoffman, C., Scoles, K., Feldman, P., & Levine, C., et al (
2002
). Challenged to care: Informal caregivers in a changing health system.
Health Affairs (Millwood),

21,
222
-231.
Doty, P., Jackson, M., & Crown, W., (
1998
). The impact of female caregivers' employment status on patterns of formal and informal eldercare.
The Gerontologist,

38,
331
-341.
Federal Interagency Forum on Aging-Related Statistics. (
2004
). Older Americans 2004: Key indicators of well-being. Washington, DC: U.S. Government Printing Office.
Feinberg, L., & Newman, S., (
2004
). A study of 10 states since passage of the National Family Caregiver Support Program: Policies, perceptions, and program development.
The Gerontologist,

44,
760
-769.
Foster, L., Brown, R., Phillips, B., & Carlson, B., (
2005
). Easing the burden of caregiving: The impact of consumer direction on primary informal caregivers in Arkansas.
The Gerontologist,

45,
474
-485.
Foster, L., Brown, R., Phillips, B., Schore, J., & Carlson, B., (
2003
). Improving the quality of Medicaid personal assistance through consumer direction.
Health Affairs (Millwood),

22,
(Suppl.),
W3-162
-W3-175.
Freedman, V. A., Crimmins, E., Schoeni, R. F., Spillman, B. C., Aykan, H., & Kramarow, E., et al (
2004
). Resolving inconsistencies in trends in old-age disability: Report from a technical working group.
Demography,

41,
417
-441.
Freedman, V., Martin, L., & Schoeni, R., (
2002
). Recent trends in disability and functioning among older adults in the United States.
Journal of the American Medical Association,

288,
3137
-3146.
Geraedts, M., Heller, G., & Harrington, C., (
2000
). Germany's long-term-care insurance: Putting a social insurance model into practice.
Milbank Quarterly,

78,
375
-401,
340
.
Han, B., & Haley, W., (
1999
). Family caregiving for patients with stroke. Review and analysis.
Stroke,

30,
1478
-1485.
Harding, R., & Higginson, I.J., (
2003
). What is the best way to help caregivers in cancer and palliative care? A systematic literature review of interventions and their effectiveness.
Palliative Medicine,

17,
63
-74.
Knight, B. G., Lutzky, S. M., & Macofsky-Urban, F., (
1993
). A meta-analytic review of interventions for caregiver distress: Recommendations for future research.
The Gerontologist,

33,
240
-248.
Liu, K., Manton, K., & Aragon, C., (
2000
). Changes in home care use by disabled elderly persons: 1982–1994.
Journal of Gerontology: Social Sciences,

55B,
S245
-S253.
Manton, K., Corder, L., & Stallard, E., (
1997
). Chronic disability trends in elderly United States populations: 1982–1994.
Proceedings of the National Academy of Sciences,

94,
2593
-2598.
Manton, K., & Gu, X., (
2001
). Changes in the prevalence of chronic disability in the United States Black and non-Black population above age 65 from 1982 to 1999.
Proceedings of the National Academy of Sciences,

98,
6354
-6359.
McCall, N., Korb, J., Petersons, A., & Moore, S., (
2003
). Reforming Medicare payment: Early effects of the 1997 Balanced Budget Act on postacute care.
Milbank Quarterly,

81,
(2),
277
-303,
172
-273.
Murtaugh, C., McCall, N., Moore, S., & Meadow, A., (
2003
). Trends in Medicare home health care use: 1997–2001.
Health Affairs (Millwood),

22,
146
-156.
National Alliance for Caregiving & AARP. (
2004
). Caregiving in the U.S. Bethesda, MD: Author.
Olmstead v. L.C., 527 U.S 581 (1999).
Rosenbaum, S., (
2001
). Olmstead v L.C.: Implications for family caregivers (Policy Brief 8). San Francisco: Family Caregiver Alliance, National Center on Caregiving.
Schulz, R., O'Brien, A., Czaja, S., Ory, M., Norris, R., & Martire, L., et al (
2002
). Dementia caregiver intervention research: In search of clinical significance.
The Gerontologist,

42,
589
-602.
Shanas, E., (
1979
). Social myth as hypothesis: The case of the family relations of old people.
The Gerontologist,

19,
3
-9.
Sorensen, S., Pinquart, M., & Duberstein, P., (
2002
). How effective are interventions with caregivers? An updated meta-analysis.
The Gerontologist,

42,
356
-372.
Spillman, B., (
2004
). Changes in elderly disability rates and the implications for health care utilization and cost.
Milbank Quarterly,

82,
(1),
157
-194.
Spillman, B., & Pezzin, L., (
2000
). Potential and active family caregivers: Changing networks and the “sandwich generation.”.
Milbank Quarterly,

78,
(3),
347
-374.
Stone, R., Cafferata, G., & Sangl, J., (
1987
). Caregivers of the frail elderly: A national profile.
The Gerontologist,

27,
616
-626.
Stone, R., & Kemper, P., (
1989
). Spouses and children of disabled elders: How large a constituency for long-term care reform?
Milbank Quarterly,

67,
(3–4),
485
-509.
U.S. Census Bureau. (
2005
). Households by size: 1960 to present. Washington, DC: Author.
U.S. Department of Health and Human Services. (
1998
). Informal caregiving: Compassion in action. Washington DC: Author.
Wolf, D. A., Hunt, K., & Knickman, J., (
2005
). Perspectives on the recent decline in disability at older ages.
Milbank Quarterly,

83,
(3),
365
-395.
Zarit, S., Gaugler, J., & Jarrott, S., (
1999
). Useful services for families: Research findings and directions.
International Journal of Geriatric Psychiatry,

14,
165
-178.