Introduction

Anyone who has been in Gerontology in recent years knows that there has been tremendous interest in family caregiving. The enduring interest in this topic has to do with the central role the family plays in our lives. There is a timeless feel to the issues of intergenerational relationships and support, and to the struggles of families to do the best they can for their older relatives, while being tugged in many different directions.

With all that has been written and said about caregiving, it is easy to conclude that almost every aspect has been probed, that there is nothing left to say, or that there are no new research directions. What I hope to accomplish is to take a fresh look at some issues and to suggest there are still important questions that need to be explored. My comments are organized around 3 issues:

• The social context of caregiving
• Main findings on the stress process in caregiving
• Promising strategies for intervention research.

While much of my emphasis is on the last point, an appreciation of the social context and stress process of caregiving is a necessary foundation for the next generation of research.

The social context of caregiving

It was not so long ago, the 1950s and 1960s, that family scholars were predicting that the extended family was a thing of the past, and it was expected that elderly people would largely be left on their own. Rather than withering away, however, the extended family has persisted—in changed, but nonetheless vital forms. Families are involved not just in care of children, but in continuing, complex, sometimes difficult, sometimes rewarding relationships with older generations. And they are involved in their care.

A dramatic demonstration of the continued involvement of families comes from a study my colleagues and I conducted comparing family help for older people in the United States and Sweden (Shea, et al., 2003). We expected the extensive system of formal services in Sweden would reduce family contributions to care. The data shown in Figure 1 show the sources of help for people with one or more dependency in activities of daily living. As expected, older Swedes received more help from the state, but family involvement in providing care was high in both countries. Perhaps the most striking difference, one that we did not anticipate, was that more people in the US received no help from anyone, despite having ADL dependencies. This vulnerable group was comprised mainly of people who had no close relatives.

Sources of Assistance Provided in US and Sweden

[Figures given in percent. Source: Shea et al., 2003]

These data show the continuing involvement of the family—over half of elders get all their care from their family, even in a welfare state. AND where public services are not as widely available, it is the people without families who struggle without sufficient help. The family remains, to use Christopher Lasch’s (1977) words, a haven in a heartless world.

The family today is different from what people meant in the past. Families come in all shapes and sizes, with great variability in size, structure and in the relationships among family members. From small families with a few people in each generation—called “beanpole” families by Burton and Bengtson (1985), to large extended families. The bigger the family, the more potential resources to help with caregiving. Factors such as women’s participation in the workforce, divorce, remarriage, step relationships and other family forms, and the ways these arrangements are worked out also affect how and how much help is given. The past patterns of exchange within a family, including the help that parents continue to give their adult children, may also affect how care in later life is carried out.

One contextual factor that has sometimes been overlooked in research is the relationship between caregiver and care receiver. As the excellent work of Baila Miller (1990) and others have shown, husbands and wives differ considerably in how they approach caregiving. In turn, spouses probably differ from adult children in their commitment and involvement, including such policy relevant dimensions as nursing home placement.

I now turn to findings on the stress process in caregiving. This information should be very familiar and so I will provide a brief, conceptual overview.

Most people approach the issue of carer stress with what might be called a naïve or simple model. This model assumes that care-related stressors, such as behavior problems, have a direct and immediate impact on outcomes such as depression and poor health. Although there is a certain common sense appeal to this idea, it has not been supported by most research. Probably the most central and enduring finding in caregiving research is that stressors have a surprising modest association with outcomes (e.g., Anshensel, Pearlin, Mullan, Zarit, & Whitlatch, 1995; Pearlin, Mullen, Semple & Skaff, 1990; Zarit, Reever, & Bach-Peterson, 1980). Instead, findings suggest that carer stress is a multidimensional process, with outcomes determined by the interplay of many different factors. As illustrated in Figure 2, the disease and disabilities create a context in which carers may experience distress, but the extent to which they do depends on other risk factors and resources in the situation. These factors include the meanings that people give to stressors (Lazarus & Folkman, 1984; Pearlin et al., 1990), the spillover of care-related stressors to other areas of life (Pearlin et al.), and the resources which can contain the impact of stressors on the caregiver’s life and well-being.

Finally, it is the sum of this process, not just the occurrence of stressors that leads to critical outcomes: the impact on the health and well-being of carers or nursing home placement of the care receiver. I also want to note that different combinations of risks and resources can lead to the same outcome, a point that I will return to later.

This brings us to the third topic I want to address, promising strategies for intervention. While some studies have had positive outcomes (e.g., (Marriott, Donaldson, Tarrier, & Burns, 2000; McCurry et al., 2005; Mittelman et al., 1995; 1996; 2004; Teri et al., 1997; Whitlatch, Zarit & von Eye, 1991; Zarit, Stephens, Townsend & Greene, 1998), many others have reported disappointing results of interventions. One possibility, of course, is that there may be little leverage for change. Carers may be too embedded in the relationship to make changes, or the stressors may just be too demanding for all but a small handful of carers to manage?

An alternative interpretation is that caregiver interventions are not provided in an optimal way. A review of the literature suggests several problems in design and measurement may have contributed to modest or nil findings.

Turning first to design, randomized trials are considered the “gold standard” for evaluating the efficacy of a treatment, but they have been applied to caregiver interventions without considering if modifications might be needed to address unique characteristics of the population or the questions being posed. The design shown in Figure 3a is typical for a randomized trial for treatment of people suffering from depression. People who meet certain eligibility criteria, such as a diagnosis of major depression, are randomly assigned to one or more treatment and control conditions. Improvement is assessed in terms of reductions in depressive symptoms.

But now consider the design shown in Figure 3b. This is the design that has most often been used in caregiver treatment studies. People are enrolled into the study because they are carers. They are then treated for a problem (e.g., burden, depression), but without consideration of whether they have that problem in the first place. Let’s focus on depression, which is the most widely used outcome in caregiver interventions. Most studies of the rate of depression among carers show that between 20 percent and 50 percent have clinically significant symptoms. That is a high prevalence of depressive symptoms, but it also means that between 50 and 80 percent do not have the problem that the treatment addresses. By recruiting carers regardless of their status on a depression measure, it assures that some people in the treatment will not be depressed and thus cannot show improvement. The result is a loss of statistical power to detect change.

The same reasoning applied to the risks and protective factors or resources targeted by the intervention. Caregiver treatment studies have tended to use interventions that target one or two risk factors, for example, how carers appraise problem behaviors or their lack of social support. Again, the assumption is made that carers will have these problems, without assessing to determine that they actually do.

This problem could be viewed in another way. Interventions for carers have generally failed to take into account the extent of individual differences that are found at every step in the stress process. Carers vary in their exposure to stressors, in their subjective responses to those stressors, in the resources they have for managing or containing stressors, and in the outcomes they experience. The “one size fits all” approach does not generally work for clothing, and it does not work for caregiver interventions. It should not be surprising, then, that the outcomes reported by the more flexible interventions (interventions like those done by Mittelman’s group) have generally been positive. In contrast to “one size fits all,” these approaches tailor the components of treatment and the dosage of those components to the needs of the individual caregiver.

A new type of treatment design called “adaptive intervention,” has been developed to deal with situations that are similar to caregiving where there are multiple risks that can lead to the same outcome (Collins, Murphy, & Bierman, (2004). In an adaptive intervention, participants are assessed for the specific risk and protective factors that contribute to the targeted outcome (e.g., relief of depressive symptoms). They then receive treatment modules that address their specific risks. Adaptive treatment, in effect, captures the art of good clinical practice, where interventions are modified to fit the person and his/her needs. Despite the fact that each person would receive a somewhat different mix of treatment components and dosage, it is possible to test the efficacy of an adaptive trial with the same scientific rigor as any other intervention. This is a promising strategy –one that my colleagues and I have begun testing with carers.

Two other issues need to be considered concerning the relation of risks and outcomes. First, it needs to be determined that the treatment actually affects the targeted risks. For example, one goal of support groups is that they increase the support available to carers. Does that actually happen? If they are not exposed to the therapeutic processes that are supposed to take place in a support group or any other intervention, then it is unrealistic to expect that they will experience positive changes on the outcome variables. Second, it needs to be determined that the proposed mechanism of treatment (e.g., increasing support) actually affects the targeted outcomes. Treatments may fail to meet their objectives for either reason, and we can learn much by disentangling these processes.

It may also be the case that a proximal outcome of a treatment, for example, that support groups provide support, may be a good thing for carers, whether or not support has an effect on public health indicators, such as depression or health.

One final point is the magnitude or dosage of the intervention. While dosage has long been a central issue in drug trials, the amount of exposure to behavioral and social interventions has rarely received systematic consideration. Among carer interventions, the most effective interventions have generally been those that were more powerful (e.g., Mittelman et al., 1995; 1996; Zarit et al., 1998).

There is much work that still needs to be done on family caregiving. Important questions need to be addressed about family structure and relationships, and about how the stress process unfolds in different contexts and setting. We especially need more and better tests of novel treatments that are designed to make the everyday experience of caregiving families a bit more manageable. I hope that as a field, we approach these tasks with respect for the enormous effort that families make, the sacrifices they are willing to make, the courage they have, and their ability to overcome bad advice from friends, family, physicians and other service providers. These families show us everyday how the ties of affection and obligation transcend memory loss and disability. Working with families, we can find to best ways to support their primary goal--helping the disabled elder to preserve dignity and meaning for as long as possible.

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

Burton, L. M., & Bengtson, V. L. (1985). Black grandmothers: Issues of timing and continuity of roles. In V.L. Bengtson & J.F. Robertson (Eds.) Grandparenthood (pp. 61-77). Beverley Hills, CA: Sage.

Collins, L. M., Murphy, S. A. & Bierman, K. L. (2004). A conceptual framework for adaptive preventive interventions. Prevention Science, 5, 185-196 .

Lasch, C. (1977) Haven in a heartless world: The family besieged. New York: Basic Books.

Lazarus, R. S., & Folkman, S. (1984). Stress, appraisal, and coping. New York: Springer.

Marriott, A., Donaldson, C., Tarrier, N., & Burns, A. (2000). Effectiveness of cognitive-behavioral family intervention in reducing the burden of care in carers of patients with Alzheimer’s disease. British Journal of Psychiatry, 176, 557-562.

Miller, B. (1990). Gender Differences in Spouse Caregiver Strain: Socialization and Role Expectation. Journal of Marriage and the Family, 52, 311-321.

Mittelman, M. S., Ferris, S. H., Shulman, E., Steinberg, G., Ambinder, A., Mackell, J., & Cohen, J. (1995). A comprehensive support program: Effect on depression in spouse-caregivers of AD patients. Gerontologist, 35, 792-802.

Mittelman, M. S., Ferris, S. H., Shulman, E., Steinberg, G., & Levin, B. (1996). A family intervention to delay nursing home placement of patients with Alzheimer disease: A randomized controlled trial. Journal of the American Medical Association, 276(21) 1725-1731.

Mittelman, M. S., Roth, D. L., Coon, D. W., & Haley, W. E. (2004). Sustained benefit of supportive intervention for depressive symptoms in caregivers of patients with Alzheimer’s disease. American Journal of Psychiatry, 161(5), 850-856.

Pearlin, L. I., Mullan, J. T., Semple, S. J., & Skaff, M. M. (1990). Caregiving and the stress process: An overview of concepts and their measures. Gerontologist, 30(5), 583-594.

Shea, D. G., Davey, A., Femia, E. E., Zarit, S. H., Sundstrom, G., Berg, S. & Smyer, M. A. (2003). Exploring assistance in Sweden and the United States. Gerontologist, 43, 712-721.

Whitlatch, C. J., Zarit, S. H., & von Eye, A. (1991). Efficacy of interventions with caregivers: A reanalysis. Gerontologist, 31, 9-14.

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

Zarit, S. H., Stephens, M.A.P., Townsend, A., & Greene, R. (1998). Stress reduction for family caregivers: Effects of day care use. Journal of Gerontology: Social Sciences, 53B, S267-S277.

Zarit, S. H., & Zarit, J. M. (2006). Mental disorders in older adults: Fundamentals of assessment and treatment. 2nd ed. New York: Guilford.