Family-based therapy for dementia caregivers: clinical observations

Abstract

Family caregiving for dementia patients is a major social and clinical problem. Family caregivers face major stressful emotional, social and economic burdens, and the negative consequences associated with caregiving are well documented. Given the projected increase in the number of people with dementia, there is a need to identify approaches that will help families manage the challenges of caregiving. Social support from friends and family members has consistently been found to mediate caregiver outcomes, yet many caregivers face problems with isolation and estrangement from family members. In this regard, family-based therapy is a promising intervention for increasing social support for caregivers, and enhancing their quality of life and ability to provide care.This paper will discuss how family-based therapy can be applied as an intervention for family caregivers of dementia patients.The clinical implications of specific interactional patterns will be presented via case examples from an ongoing clinical trial with white American and Cuban American caregivers of dementia patients.The intent is to demonstrate how identification of interactional patterns is a valuable tool for implementing family-based interventions.

Introduction

Currently about 4 million Americans suffer from Alzheimer's disease (AD), and by the year 2040 this number is expected to increase to 9 million or about one in 30 Americans (Brody & Cohen, 1989). The majority of people with AD live at home and are cared for by family members such as spouses, daughters or daughters-in-law (Zarit et al., 1998). Caring for a family member with dementia creates significant challenges, and many caregivers experience considerable burden and stress. The negative consequences associated with caregiving are well documented and include depression and anxiety, increased use of psychotropic medications, immunological dysfunction and increased use of medical services.

Because individual caregivers respond differently to the stressors of caregiving situations (Zarit et al., 1998), researchers have sought to identify factors that impact the relationship between the demands of caregiving and caregiver distress.These factors include care recipient impairment, caregiver characteristics and contextual factors such as social support. Among care recipient impairment variables, the level of behavioral symptoms, such as wandering and repetitive questioning, has emerged as the strongest predictor of caregiver distress (Haley et al., 1987; Teri et al., 1992). Caregiver outcomes also vary as a function of gender (females experience more burden than males) (Schulz & Williamson, 1994), age (older caregivers experience more burden) and ethnicity (African-American caregivers typically report less burden than Caucasian caregivers) (Fredman et al., 1995). In addition, caregivers who employ a problem-focused coping style have been found to have better outcomes than those who employ emotion-focused coping (Kramer, 1997; Rose et al., 1997).

Clearly, the relationship between caregiver burden and outcomes is complex and there appear to be considerable individual differences in response to the demands of caregiving (Schulz et al., 1995). Understanding factors that impact on caregiver burden and distress is critical to the development of efficacious intervention strategies. As noted, social support has consistently been found to have an effect on caregiver outcomes. Generally emotional support and receiving assistance with caregiving tasks is associated with lowered feelings of burden and depression and higher wellbeing (Franks & Stephens, 1992; Teri, 1994). Also, caregivers who engage in social and recreational activities and interact with family and friends are less depressed than caregivers who are more socially isolated (as reported in Haley, 1997).

One critical source of social support for caregivers is the family. Family members can offer instrumental and emotional support to caregivers and ease the burdens associated with the demands of caregiving. However, family interactions and pre-existing family conflicts can also exacerbate caregiver distress. Family variables that are related to caregiver distress include: marital conflict among caregiving daughters (Creasey et al., 1990), family conflict (Sample, 1992), boundary ambiguity (Boss, 1993), problematic communication patterns (Speice et al., 1998) and incongruence of family members' appraisal of the care recipient (Pruchno et al., 1997). As discussed by Cohen et al. (1998), the beliefs that family members have about what influences events and circumstances during the course of dementia are powerful forces in determining how well caregivers cope with the demands of the illness.

Despite the importance of understanding the impact of family dynamics on caregiving and the obvious disruption to family life associated with Alzheimer's disease, the literature on the relationship between family factors and caregiver outcomes is sparse (Teri & Logsdon, 1992). Furthermore, there is little synergy between the fields of family systems and gerontology. A review of articles published in family therapy journals between the years 1986–1994 found that only 3% had an aging focus (van Amburg et al., 1996), and reviews of the dementia caregiving literature have yielded similar results. Although caregivers express a desire for family counseling (Wykle, 1996), the approach to caregiver interventions has traditionally been via group support or individual psychoeducational models (Teri & Logsdon, 1992). In fact, some theorists cite the ‘paradox’ of the lack of family interventions for caregiving, which is a prototypical family systems issue (Knight & McCallum, 1998).

The scarce literature notwithstanding, there is support in the field for the family therapy approach for caregivers of dementia patients (Cohen et al., 1998; Czaja et al., 2000; Knight & McCallum, 1998; Teri & Logsdon, 1992) and reason to believe in its promise for helping families adjust to the challenges of caregiving. Generally, family therapy can increase support provided to caregivers by family members and decrease family conflict associated with caregiving. Family-based interventions have been found to be effective in the prevention, treatment and management of various mental disorders such as substance abuse and schizophrenia (Bischoff et al., 1996). There is also growing evidence that family interventions are effective for the treatment and management of physical illnesses such as pediatric asthma and diabetes, cardiovascular and neurological disorders, obesity and anorexia nervosa (Campbell & Patterson, 1995).

The goal of this paper is to demonstrate how family-based therapy can be applied as an intervention for family caregivers of dementia patients. Specifically, the clinical implications of family interactional patterns will be presented via case examples from an ongoing clinical trial (REACH) with white American and Cuban American caregivers. REACH (Resources for Enhancing Alzheimer's Caregiver Health) is a multi-site project, funded by the National Institute on Aging and the National Institute of Nursing Research, which is evaluating the efficacy of a variety of psychosocial and behavioral interventions on reducing the burden and improving the wellbeing of family caregivers of persons with Alzheimer's disease. These interventions encompass a variety of techniques including psycho/social/educational services, behavioral interventions, innovations in community services, environmental modifications and technology-based interventions.

A family-based therapy, Structural Ecosystems Therapy (SET), developed by Szapocznik and colleagues (Mitrani et al., 2000; Szapocznik et al., 1994) is one of the interventions being evaluated at the Miami REACH site.This intervention differs from the other interventions being evaluated within REACH in that it attempts to address the needs of the entire family within a conjoint context. It also focuses on transforming the process of family relationships as a mediator to improving caregiver support. The aim of the SET intervention is to improve the caregiver's interactions within her or his entire social ecosystem (family, community, health providers, etc.) to increase the extent to which the caregiver's emotional, social and instrumental needs are met and, in turn, improve psychological adjustment.

The specific intent of this paper is to demonstrate how the identification of family interactional patterns is a valuable clinical tool for implementing family-based interventions.Attention will be also be given to potential differences in interactional patterns between white American and Cuban American families.This paper will focus exclusively on patterns of interaction that occur within the family, and will not discuss the extra-familial, or ecosystemic, aspects of the intervention.

The application of SET within the Miami REACH project

As noted, SET is one of the interventions being evaluated at the Miami site of the REACH project. The participants (224 family caregivers of dementia patients, including 112 white American caregivers and 112 Cuban American caregivers) are randomly assigned to one of the three treatment conditions. The other intervention conditions include SET enhanced with a computer integrated telephone system (SET+CTIS), and a minimal contact telephone support control. The CTIS system serves to augment the family therapy intervention by facilitating the caregivers' ability to communicate with family members, friends, other caregivers and their therapist. The participants are recruited from two memory disorder clinics affiliated with the Department of Psychiatry and Behavioral Sciences, and from the community via newspaper and radio advertisement. In almost all cases, it is the primary caregiver who makes the initial call to the project.

All participants are required to meet the REACH inclusion/exclusion criteria, as well as the sitespecific criteria.The care recipients are 60+ years of age and have a DSM-IV diagnosis of dementia or a score of 23 or less on the Mini-Mental State Examination (MMSE; Folstein et al., 1975), and the presence of two IADL or one ADL impairments.Among other criteria, the primary caregivers are required to have been in the caregiver role for at least six months, live with the care recipient, and provide at least four hours of assistance per day. For the Miami site, at least one other family member must participate in the project. Other family members include actual or fictive family who provide emotional or instrumental support.

Parameters of the SET intervention and pragmatic considerations

The SET intervention in REACH takes place over a course of 12 months, with weekly sessions for the first four months, bi-weekly sessions for the subsequent two months and monthly sessions for the final six months of treatment. This schedule was selected to provide a standard dosage and to allow sufficient time to meet treatment goals. The tapering of treatment allows for an extended treatment period in which new caregiving problems are likely to emerge and facilitates generalization of treatment gains in the face of these emerging problems.

The majority of sessions are conducted in the caregiver's home or other settings (e.g. a day care center, hospital or adult child's home). Because this is a funded research project, services are provided at no charge to the families, and cost studies are needed to determine the feasibility of offering these services on a fee for service basis. In the SET+CTIS intervention, some family sessions are conducted via the computer telephone system, and show promise for more economical treatment delivery.

The standard protocol is for therapy sessions to be videotaped, although agreeing to be videotaped is not a condition for inclusion in the project. The most common reason given by families for videotape refusal is concern that it will disturb the care recipient. Some families have also expressed privacy concerns. The majority of families, however, agree to be videotaped and, given the high level of disclosure of intimate material, do not appear to be constricted by the presence of the camera. The purpose for videotaping sessions involves both research and clinical considerations. Research uses include treatment adherence evaluations and process study. Clinically, session recordings are useful for supervision/team consultation, which is highly recommended for helping the therapist remain true to the SET intervention.

Overview of Structural Ecosystems Therapy (SET)

SET is a family-based intervention initially developed for the families of adolescent substance abusers, and adapted to enhance support for caregivers. SET is derived from a combination of two theoretical approaches: (1) the structural/systemic approach (Minuchin, 1974; Minuchin, et al., 1967; Szapocznik & COSSMHO, 1994; Szapocznik & Kurtines, 1989; Kurtines & Szapocznik, 1996) and (2) the ecosystemic approach (Bronfenbrener, 1979).

As a systemic approach, the SET model views the behavior of family members as interdependent and repetitive. In some instances the repetitive patterns of family interactions may be maladaptive or obsolete, leading to symptoms such as caregiver distress. Other family patterns of interaction may be particularly adaptive and relieve caregiver burden. SET is particularly appropriate for minority families because it recognizes the importance of culture as a contextual variable which can have a pronounced influence on family interactions.

The primary caregiver is the central figure in the SET intervention.This is the individual who is bearing the most practical day-to-day caregiving burden, and therefore is the most motivated to affect change in the family interactions that are related to caregiving. The primary caregiver is also typically the gatekeeper who controls access to the AD patient and her or his caregiving needs. Therefore, while other family members are typically distressed by the current family functioning, and certainly the aim of SET is to enhance the adaptation of the entire family to the caregiving situation, the primary caregiver is the initial entry point of the therapist into the family and usually remains the primary contact person throughout the intervention.

Generally, family interactional changes that are beneficial to the caregiver (e.g. by increasing available support) also benefit other family members (e.g. by improving their relationship with the caregiver, or by meeting their goals for improved caregiving). In fact, it is only through this mutual satisfaction that interactional changes can be maintained, since the absence of reinforcement on either side would cause the new interaction to be extinguished. There are cases (particularly among caregivers with borderline or narcissistic features) in which primary caregivers have unreasonable expectations of family support. In such cases, the therapist seeks to balance the caregiver's need for support while helping other family members to set limits on overwhelming demands from the caregiver.

In the initial sessions, the therapist usually meets with the caregiver alone or the caregiver and other family members, and gradually gathers the history of the family, AD and the caregiving experience. The goal of these initial sessions is to build rapport, establish therapeutic goals, and begin to assess the caregiver's family interactions.The therapist encourages family members to interact with each other in session to assess and transform interactions as they occur.

A central tenet of the SET approach is its emphasis on the processes of family interaction rather than on content. Process is comprised of how people interact, while content is the topic that they are interacting about. For instance, suppose a therapist learns from the caregiver that she and her daughter are in conflict regarding the AD patient's medication such that they have stopped speaking to each other. In this case, the content is the patient's medication, while the process is mother and daughter's inability to maintain communication in the face of disagreement. SET aims to transform the interactional processes (e.g. the inability to negotiate disagreement) that block the resolution of content concerns (medication management).

The therapist assesses interactions, determines which of these might be targeted in treatment and establishes a plan to transform interactions. This approach involves: (1) developing a clear understanding of the nature of supportive and problematic interactions, (2) understandinghow these interactions are related to the family's current level of functioning, and (3) intervening in a very deliberate fashion to enhance supportive and reduce maladaptive interactions.

In the example given above, once the therapist has determined that there is a caring bond between mother and daughter, or at least between each of them and the AD patient, a planned set of interventions might include: (1) speaking individually with the women to engage and prepare them for a conjoint session, (2) strengthening their alliance by having them talk together about a common concern (e.g. the patient's agitation which is resulting in complaints from the day care center), (3) having the women work together on a practical problem-solving plan while helping them maintain their communication despite disagreement (e.g. by blocking escalating conflict or personal attacks), (4) following up on the plan in a subsequent session and congratulating mother and daughter on their teamwork, or exploring barriers that impeded progress (these might be practical, relational or intrapsychic barriers), and so on.

Family interactional patterns and case examples

The assessment of interactional patterns is conducted along four dimensions of family functioning that are particularly related to caregiver functioning: structure (family roles), developmental stage, resonance (interpersonal boundaries), and conflict resolution. Table 1 gives examples of maladaptive family interactions that fall within these four dimensions.

Table 1.

Caregiver family interactional patterns

Pattern	Interaction
Structure
Hierarchy	Family functions not accomplished due to a leadership void
	Family members usurp the caregiver's leadership
	Power struggles
Alliance	Lack of collaboration among family members
	Family members frequently pulled into each other's conflicts
Communication	Lack of direct verbal communication between any two family members
	Ineffective communication (e.g. vague, lecturing, over-emotional)
Developmental stage	Adjustment difficulties related to dementia
	Adjustment difficulties related to other transitions (e.g. retirement, relocation, divorce)
	A family member is expected to function beyond his or her capacity
	A family member is expected to function below his or her capacity
Resonance
Enmeshment	The caregiver is emotionally over-involved with the care recipient (e.g. is personally offended by dementia symptoms, unable to separate)
	Emotional over-involvement between other family members
	Family members invade each other's boundaries (e.g. eavesdropping)
	Family members speak simultaneously or interrupt each other
Disengagement	Pre-existing detachment between the caregiver and care recipient
	Detachment between the caregiver and care recipient due to dementia
	Detachment between other family members
Conflict resolution	Family denies or avoids conflicts/problems
	Conflicts/problems remain unresolved because focus shifts across issues
	Family interactions are frequently negative/conflictive

This section will describe the interactional patterns listed inTable 1 and provide case examples to illustrate their impact on the caregiver and how they can be approached in therapy. Identifying details in the case examples have been changed to protect client confidentiality. For purposes of highlighting a particular dimension of interactional functioning, each case presents only one problematic interactional pattern. Although some families have a single or dominant problematic pattern, most families have multiple problematic patterns. Moreover, a problem in one dimension of family interaction typically has an impact on other dimensions. For example, the lack of direct verbal communication (structure) may indicate disengagement (resonance) and has an impact on the family's ability to confront problems (conflict resolution). Fortunately, a single intervention can often tackle interrelated problems across dimensions.

Structure

The caregiving system, which consists of all individuals who play caregiving roles, must be organized to distribute caregiving responsibilities and for mutual support. Caregiver wellbeing is related to the caregiver's ability to serve, and be validated by the family, as a leader of the caregiving system. Caregiving structure involves family hierarchy, alliances and communication flow. In a family with effective structure, the caregiver is able to successfully distribute caregiving tasks, lead decision-making and collaborate with other family executives, and maintain alliances with other family members.

Culture can have an impact on caregiving structure. Cuban families often hold a hierarchical relational orientation (Szapocznik et al., 1978) and adhere to traditional gender roles. In Cuban families, older female caregivers may have particular difficulty in adopting some of the leadership roles of caregiving, especially vis-à-vis their care recipient husbands. Longstanding family roles can also have an impact on the ability of other family members, even those who are relatively acculturated, to accept the leadership of an older female caregiver.

An example is the case of Mrs L, the primary caregiver for her husband. The L's were a Cuban American couple, who, like many of their generation living in Miami, were not highly acculturated. Mr L had formerly been the capable and uncontested leader of the family, commanding submission from his wife because of his intimidating manner. AD exacerbated Mr L's aggressive and demanding style, so that Mrs L feared for her physical safety if she were to attempt to restrict Mr L's access to the car or to the household finances. Despite the fact that Mr L was clearly incapable of driving or managing financial matters (his mismanagement had already cost the family devastating losses), Mrs L felt helpless in controlling her husband's behavior.The L's son had gradually taken the reins of the family business and could not understand why his mother could not do the same at home. Feeling helpless and invalidated, Mrs L was depressed, anxious and suffered debilitating gastro-intestinal problems. She ceased all social interactions due to her fear of leaving Mr L alone, since he refused any substitute companion. Mrs L's relationship with her son was strained due to her frequent complaints about Mr L (often in the form of desperate telephone calls), and the son's minimization of the problem.

In the early sessions, the therapist determined that Mrs L's low status in her family's hierarchy was a central problematic interaction and contracted with Mrs L to work on enhancing her effectiveness at managing Mr L's behavior, and gaining respect and support from her son. The therapist recognized that the L's son would be a valuable member of the family leadership structure and aimed to facilitate an executive alliance between Mrs L and her son. Mrs L, however, was hesitant to include her son in family therapy. She said that he was too busy and politely refused the therapist's offer to call and invite him to a session.The therapist recognized Mrs L's reluctance as a symptom of the status quo in the hierarchy, and deferred challenging this behavior until the therapeutic alliance was more solidly established.

The therapist recognized an opportunity for change when Mrs L reported that her son was going to take Mr L on a business trip.The trip had an eye-opening impact on the son's perspective regarding Mr L's level of functioning. With five days of constant contact, the son was able to experience Mr L's difficulties in managing his own care, his disorientation and his insistence that he could be self-reliant. Mrs L gained tremendous satisfaction from her son's acknowledgement of her daily struggle, and agreed to allow the therapist to call him.

The L's son was appreciative of the therapist's expression of sympathy regarding his experiences on the business trip, and accepted her invitation to attend a session.The therapist's goal in this session was to engage Mrs L and her son in a collaborative relationship on managing Mr L's behavior and other caregiving needs, and to strengthen their supportive bond and alliance. Potential barriers included Mrs L's helpless manner, as well as the son's impatience and invalidating stance.

To overcome these barriers, the therapist looked for opportunities to highlight the son's dedication and Mrs L's successes in bypassing Mr L's unreasonable demands. The therapist helped Mrs L and her son discuss her complaints, helping the son describe his frustration at not being able to solve the problems that Mrs L related, and helping Mrs L describe her feelings of abandonment when her son hurried her off the telephone.To help build the caregiving alliance and lessen the negativity, the therapist focused on Mrs L's and her son's common goals (e.g. the L's safety and financial security, Mrs L's need for a sense of control and the son's need to be relieved of his mother's frantic phone calls).

With the therapist's guidance, mother and son worked together to generate a solution to the problem of Mr L's driving (the son would take the car keys and they would tell Mr L that the car was being repaired). The issue of household finances was more challenging as the son's solution was to take over their management, while Mrs L wanted to be more involved.The therapist was encouraged by the assertive manner in which Mrs L expressed her desire to participate in this leadership function. She guided Mrs L and her son in creating a list of finance tasks, dividing up some short-term tasks (e.g. consulting an attorney, searching for lost documents) and offered to follow-up in a subsequent session.

The son continued to participate in the therapy process intermittently until termination. Later sessions were aimed at maintaining and continuing to build the collaborative caregiving alliance between Mrs L and her son. By termination, the L's son accepted the role of ‘financial advisor’ to his mother while Mrs L established a routine for paying bills. Mrs L hired a housekeeper who she could trust to watch over Mr L (and whom Mr L accepted, since she was obstensibly present for housekeeping rather than as his companion), freeing her for social contacts and to maintain regular attendance in a support group. Mrs L's contacts with her son were also more positive, and he visited more regularly.

In this example, the problematic family interactional pattern was primarily one of ineffective structure. The caregiver lacked credibility as a family leader, did not have her son's collaboration in leadership functions and therefore was overwhelmed by the patient's behavior and needs. By gaining access to the son and highlighting mother and son's common goals and complementary strengths, the therapist helped to form a new ‘executive sub-system’ that was better equipped to manage caregiving.

Developmental stage

Families, like individuals, go through a series of developmental stages, each with its own tasks that must be completed. There are expected life cycle events such as marriage, the birth of children, retirement and death in old age. Other significant life cycle events are not so expected such as divorce, untimely death, migration and illness. The onset of dementia marks the beginning of a new developmental stage for the family, which involves both expected and unexpected life cycle events.

Stress is often greatest at transition points from one stage to another of the family developmental process (Carter & McGoldrick, 1989). This stress is due in large measure to the need to renegotiate and otherwise adjust to some fundamental changes in roles within the family. The ease or difficulty of the family in incorporating these new roles determines their adjustment to the reality of living with a dementia patient.

There are several types of interactional adjustments that a caregiver family must make. The first type of adjustment is the filling of family roles that the patient played prior to her or his illness. While families may be attentive to instrumental roles (e.g. cooking or paying bills) that need to be filled, they are often less aware of relational roles (e.g. leadership, arbitration, communication, nurturing or family gathering).This type of adjustment consists of several processes, including recognizing the need to fill the roles once played by the care recipient, coming to terms with the need to ‘displace’ the care recipient from prior roles, managing the care recipient's resistance to giving up old roles and functions,and moving other persons into the roles once filled by the care recipient, without unduly burdening the caregiver. A common symptom of developmental adjustment difficulties is unrealistic expectations regarding the functioning of the AD patient, or inappropriately low expectations from other family members.

The case of Mr F illustrates this latter pattern. Mr F was a frail 87-year-old caregiver for his wife, who was in an advanced stage of AD. Although Mr F was a competent manager of Mrs F's care, hiring and supervising round-the-clock aides and making sound choices in her health care, his own health was faltering.The F's had two grown children living nearby. Although the children wanted to assist in caring for their parents, Mr F did not trust their judgement due to errors they had made in managing their own lives.

The therapist's assessment was that this family pattern of under-functioning adult children presented a severe threat in the event that Mr F would become incapacitated, and prevented the children from playing the caring role in their family of origin that they very much desired.To motivate Mr F in transforming this interaction, the therapist allied himself with Mr F's practical side, which dictated that the family be secure in the event of an emergency. Mr F agreed with this preventive goal, and contracted to work with the therapist to help elevate the children to productive adult status in the family.

Mr F's only son, who lived with his parents following a divorce and business failure, was particularly available for assistance and regretted his father's lack of confidence. Two months after the start of treatment, the therapist led a session with Mr F and his two children in which the son expressed his wish to be closer to Mr F and to receive his guidance on personal matters. For the first time ever, Mr F and his 50-year-old son discussed sex, the son's alcoholism, and his business failure. With much emotion, Mr F also explained his philosophy about hard work and frugality. As the discussion ensued, Mr F gained a new understanding of the choices that had led to his children's troubles, and the children expressed their respect for Mr F and their desire to help. At the following session, Mr F again met with his two children and spontaneously began to give detailed burial instructions for himself and Mrs F. This discussion marked the promotion of these two ‘incompetant’ children to adult roles in the family.

The developmental stage dimension is concerned with the family's adjustment to role transitions made necessary by the dementia of one of its members. In the preceding example, the adult children were functioning at a level that was not commensurate with their abilities, their desire to be of help, nor the needs of the caregiver and care recipient.This pattern of under-functioningwas transformed via an approach which redefined the family as ‘at risk’ in the event of caregiver disability, and guided intimate communication, which increased the father's confidence in his children.

Resonance

An important factor in the caregiver's ability to successfully balance her or his own personal needs with the demands of caregiving is the nature of the boundaries between the caregiver and care recipient. A high degree of emotional and psychological closeness, or enmeshment, between caregiver and care recipient makes it difficult for the caregiver to be an effective manager of the caregiving system because of a loss in objectivity and unwillingness to delegate caregiving tasks.

The resonance dimension is strongly influenced by culture.WhereasAmerican culture places a high value on individuality and independence, Cuban culture values collectivism and gives precedence to the needs of the family over the needs of the individual (Santisteban et al., in press). Therefore, the degree of interdependencyfound in Cuban families is typically much greater than that found in white American families. Fittingly, the pattern of caregiver–care recipient ‘enmeshment’ is observed more frequently among the Cuban families in the sample.

An example is the case of Mr S, the primary caregiver of his wife, who suffered later stage AD. Both Mr and Mrs S were Cuban American.Thirty years earlier, Mr S. had left his first wife and children to marry Mrs S, who he described as ‘the love of my life’. Mr S was fiercely protective of Mrs S. In the first session, the therapist was moved by Mr S's tender and constant attention to Mrs S. Although she could not respond and seemed completely oblivious to the content of the conversation, Mr S. attempted to include her in the discourse and whispered protectively when speaking with the therapist about his concerns for Mrs S's health.

Mr S's devotion to his wife came at the expense of considerable personal sacrifice. He was with her constantly and had completely cut off a normally active social life due to Mrs S's increased infirmity. Mr S. would not entrust his wife to the care of an aide, and only left the house for brief errands while a trusted neighbor sat with Mrs S. Mr S had also failed to attend to his own health needs, having gone years without a check-up, and showing visible signs of poor health.

Mr S was isolated from his family. He had a distant relationship with the children from his first marriage and with his brother, who lived reasonably nearby. Mr S felt close to Mrs S's daughter and consulted with her regarding Mrs S's health needs, but she lived in another state. It was this daughter who had sought our services.

Mr S was initially very wary of the therapist's involvement.The therapist was careful not to challenge Mr S's extreme dedication to his wife, yet recognized that this pattern was undermining Mr S's mental and physical health as well as his ability to sustain long-term caregiving. It was this latter point that the therapist employed to interest Mr S in seeking additional support.

After a month of cultivating a cordial relationship with Mr S, the therapist had an opportunity to intervene when Mrs S was hospitalized for a pulmonary problem. The therapist noted Mr S's agitation when he called to cancel their appointment, and offered to meet Mr S in the hospital where he found Mr S complaining loudly and bitterly to a nurse for her allegedly callous treatment of Mrs S.When they were alone together, the therapist persuaded Mr S to adopt a ‘strategic’ stance (to make requests in a manner which would not alienate others) toward the hospital so that his wife would receive the best possible care.

This hospital visit helped the therapist gain Mr S's confidence, who invited the therapist to call Mrs S's daughter. The therapist expressed his concerns to the daughter regarding Mr S's failing health and mobilized her to reinforce Mr S's positive communications with health care workers and to act as intermediary whenever possible.

The consequences of Mr S's negligence of his own health care became evident in the fourth month of therapy when Mr S was diagnosed with advanced prostate cancer. Mr S accepted the therapist's suggestion to contact his brother and son to inform them of his condition (he could not bring himself to contact his daughter). While the son did not respond to Mr S's overture, the brother did begin to visit Mr S more frequently and even accompanied him to radiation treatments.

By the seventh month of therapy, Mrs S's condition deteriorated and she was placed in hospice care. Despite the gravity of Mrs S's condition, Mr S was able to maintain assertive and appropriate interactions with the hospice workers. Mrs S died shortly thereafter. Mr S's brother and stepdaughter maintained their support, but he died six months later.

Family patterns regarding boundaries are typically deeply entrenched and tied to culturally-based values. Transforming resonance,therefore, requires a delicate approach. A therapist who threatens the family's dearly held beliefs regarding boundaries, whether they be cohesiveness or independence,will quickly be ejected from the therapeutic system.The reader is directed to the case of the R's below for an example of a relatively disengaged caregiver–care recipient dyad.

Conflict resolution

Having a family member with dementia presents a multitude of logistical problems, and precipitates crises that may cause friction and disagreement between and among family members. In many cases, family members naturally search for a person or thing upon which to direct their anger, pain and frustration, leading to negativity that can be diffuse or directed at one particular person. Sometimes family members blame others with little recognition of their own contribution to the problem. Conflict resolution refers to the family's capacity for effectively resolving differences of opinion with low levels of negativity and without the use of intermediaries or the forming of destructive coalitions. This includes the extent to which family members: (1) allow differences of opinion and painful issues to emerge, (2) can manage these differences without personal attacks or losing sight of the issue to be resolved, and (3) can find adequate solutions that have the support of the family members involved. An important aspect of conflict resolution is the capacity of the family to allow the caregiver to assert disagreements and/or criticisms clearly and directly, without leading to the dissolution of relationships.

An example of a problematic pattern of conflict resolution is found in the case of Mrs R, who was caring for her husband in the early stages of AD. Mr R had been forced to leave his law practice because of his dementia. Over the years, Mrs R had developed many independent interests and a full social life of her own. Mrs R was deeply distressed by Mr R's new need for her companionship and was unwilling to accept the reality of his disability. Mrs R became impatient whenever Mr R could not keep up with their conversation and would press him for a response, causing Mr R to become agitated to the point of violence.

The R's had two daughters living nearby. Mrs R had a tense relationship with one daughter and had completely cut off ties with the other due to a perceived slight several months earlier. Mrs R criticized her daughters for their selfishness and insensitivity. She felt that they did not give her and Mr R their due respect and attention.

The therapist had great difficulty in establishing a working relationship with Mrs R, whom she found demanding, hostile and disdainful. The therapist was also dismayed by Mrs R's harsh and disparaging treatment of Mr R. In supervision, the therapist was helped to see Mrs R's behavior as a symptom of her despair at the upheaval in her life, and was able to return to the next session with a more supportive stance. Mrs R responded well to the therapist's support and the opportunity to freely express her distress without being reminded of her obligation to Mr R. She admitted that she sometimes acted impulsively and inappropriately towards Mr R and thanked the therapist for listening to her.

It was clear that the therapist's initial interaction with Mrs R mirrored the interactions that Mrs R had with her family. Mrs R felt her own pain so deeply that she failed to see the impact of her behavior on others and then blamed them for their insensitivity. Interactions between Mrs R. and her family were such that even seemingly minor disagreements or disapointments could easily lead to escalating conflict or a complete shut-down in communication.The therapist developed a strategy for intervening in Mrs R's interactions with her children—if she could construct situations in which Mrs R was free to express her own needs, then the ground would be set for Mrs R to be more receptive to the input of others.

Managing the conflict between Mrs R and her daughters required a delicate and patient approach. There was a history of negativity and intermittent cutoffs between Mrs R and the daughters which was exacerbated by their distress regarding Mr R's illness. The therapist planned to use the first conjoint session to highlight the women's common loss as a way to bring them closer. Mrs R had a different agenda in mind, however. Her goal was to use the session to list all of her complaints about her daughters, especially the one who had slighted her by canceling a dinner invitation. The therapist feared that the negativity would become so intense that the daughters would no longer be willing to participate in therapy and that the rift between them and their mother would widen.

Fortunately, the therapist was able to offer support to the daughters without antagonizing Mrs R, and once Mrs R had finished with her ‘laundry list’ of complaints, the therapist guided her in expressing her pain regarding Mr R's condition and her lost independence.The daughters joined in this expression of grief and each one offered a specific act of support (one daughter offered to take Mr R to an event at her son's school the following day; the other made plans for a mother–daughter shopping trip). More conjoint sessions followed in which Mrs R was guided in presenting her needs in a less aggressive manner and the daughters were guided in supportively presenting their dissenting opinions to Mrs R (e.g. about Mrs R's treatment of Mr R). By termination, one daughter was regularly attending a caregiver support group with Mrs R, and the other had linked Mrs R to her synagogue.

Although we have found conflictive family interactions in both ethnic groups, these appear to be more common among the white American families, particularly between the caregivers and their adult children. Problems of conflict resolution among Cuban American families are often manifested in a pattern of denial of AD symptoms or reluctance of family members to discuss this topic. This process was illustrated in the case of Mrs L, above.

Conclusion

Caring for a family member with an illness such as dementia is extremely challenging, and many caregivers and family members experience negative outcomes such as depression and stress. Given the projected increase in the number of people with dementia, and the fact that most dementia patients are cared for at home by family members (Schulz & O'Brien, 1994), there is a need for strategies that help caregivers meet the demands of caregiving. Social and instrumental support from family members has been found to mediate caregiver outcomes such as feelings of depression or burden (e.g. Aneshensel et al., 1995), yet many caregivers are isolated and estranged from family members and friends. Despite the importance of family support for caregivers, we have a limited understanding of strategies which are efficacious in terms of maximizing family resources.

Although there is an extensive literature documenting the effectiveness of family therapy for various mental and health-related problems, the effectiveness of family therapy for caregivers is largely understudied. The intent of this paper was to demonstrate how a family-based therapy may be applied to family caregivers of dementia patients. Specifically, we sought to delineate several family interactional patterns that are particularly relevant to caregiving situations, and illustrate how these patterns could be addressed via a theoretically-based family therapy approach.

Assessment of family interactional patterns provides a road map for identifying maladaptive processes to be targeted in treatment.These maladaptive patterns block the ability of the caregiver to receive support from family members, and the ability of family members to offer relief to the caregiver. Assessment of interactional patterns also necessitates understanding contextual factors such as ethnicity which shape these interactions.Therefore, one focus in the reported case studies was on differences between white American and Cuban American caregivers. Attention to ethnic differences in attitudes towards caregiving and responses to caregiving responsibilities is critical given the increased ethnic diversity of the population. As illustrated, Cuban American caregivers have different perceptions of family and family obligations, as well as different family structures than white Americans. In addition, it is important to tailor interventions so that they are culturally congruent.

The REACH study illustrates an attempt to apply a family therapy model to caregiver populations. As such, the results from this project will provide valuable insight into the role of family support in caregiver outcomes. Furthermore, we will gather information on the protective/risk impact of family interactional patterns for caregiver distress. This type of information will provide guidance for refining the family therapy approach to better meet the specific needs of caregiver populations.