REVIEW OF THE LITERATURE ON MENTAL HEALTH AND HEALTH CARE UTILIZATION IN HIGH-RISK WOMEN

Women with an increased risk of breast cancer experience cancer-specific and generalized psychological distress (Lerman & Schwartz, Reference Lerman and Schwartz1993). Women who are positive for a family history of breast cancer and whose mothers died of the disease have significantly higher breast-cancer-related distress than women who are family-history positive but whose mothers have not died of breast cancer or than women who are family-history negative (Erblich et al., Reference Erblich, Bovbjerg and Valdimorsdottir2000). However, when screened using modules of the Structured Clinical Interview of the DSM-IV, less than 10% of the women enrolled in a high-risk registry met criteria for a clinical disorder (Coyne et al. Reference Coyne, Benazon and Gaba2000). In fact, when compared to two age- and education-matched samples, women in a high risk registry showed clinically significantly lower (20% vs. 28%) anxiety/depression than did women who attended a general medical practice clinic (Butow et al., Reference Butow, Meiser and Price2005). Different methods of assessment may yield different findings about high-risk women. For instance, in a sample in a Family History Clinic (High Risk), 13% of the women met diagnostic criteria for an affective disorder by structural interviewing using standard assessment and diagnostic criteria, whereas 26% did so by a standardized psychological questionnaire (Goldberg et al., Reference Goldberg and Williams1988).

Turning to cancer-specific worries, approximately 66% of a high-risk breast cancer clinic sample reported that worries about breast cancer interfered with their functioning across several life domains, including mental health and role functioning (Trask et al., Reference Trask, Patterson and Wang2001). A systematic study of cancer-related post-traumatic stress disorder (PTSD) in high-risk women revealed rates of current threshold and subthreshold PTSD to be 16% whereas 24% reported past (threshold and subthreshold) PTSD using DSM-IV criteria (Hamann et al., Reference Hamann, Somers and Smith2005). This differed only slightly from women in the clinic sample who had a diagnosis of breast cancer (current, 17%, past, 28%; Hamann et al., Reference Hamann, Somers and Smith2005).

Several studies examining health care utilization by high-risk women including screening for BRCA genes, mammographic screening, and breast self-examination reported relationships between health care and mental health.

A meta-analysis of controlled trials in BRCA genetic screening found that genetic counseling improved the knowledge of cancer genetics (pooled short-term difference = 0.70 U, 95% confidence level [CI] = 0.15 to 1.26 U), but did not alter the level of perceived risk (pooled short-term difference = −0.10 U, 95% CI = −0.23 to 0.04 U). No effect was noted in controlled trials on general anxiety (pooled long-term effect = 0.05 U, 95% CI = −0.21 to 0.31 U) or on cancer-specific worry (pooled long-term effect = −0.14 U, 95% CI = −0.35 to 0.06 U) (Braithwaite et al., Reference Braithwaite, Emery and Walter2004).

However, in another meta-analysis of the impact of genetic counseling, generalized anxiety was significantly reduced (average effect size r = −0.17, p < .01), as was accuracy of perceived risk (average effect size r + 0.56, p < .01; Meiser & Holliday, Reference Meiser and Holliday2002).

As regards mammography, there is a generally positive correlation between a patient's perceived risk and her compliance with recommended mammography screening (N = 52,766, g = 0.19; Katapodi et al., Reference Katapodi, Lee and Facione2004). In a systematic review, women at elevated risk for breast cancer do not appear to experience high levels of anxiety associated with mammography compared to women at normal levels of risk.

Furthermore, women recalled for additional mammographic studies experienced elevated levels of anxiety whether or not they belonged to the elevated risk group, although there were significant differences between the two groups (Watson et al., Reference Watson, Henderson and Brett2005).

In studies of the effects of anxiety on breast self-examination (BSE), women who are at higher risk for breast cancer and who have higher anxiety are less willing to perform monthly self-breast exams and do not come to the doctor's office for clinical breast examinations (p < .02; Kash et al., Reference Kash, Holland and Holger1992). In a study evaluating BSE in a cohort of high risk women, 57% of them report performing monthly BSEs, 20% report doing the exams less than once a month, 10% of the women never did them, and 13% of the women examined themselves at least once a week. Women below age 40 who examined their breasts more frequently than recommended (i.e., at lest once per week) were significantly (p = .03) more distressed than other women in the sample (van Dooren et al., Reference Van Dooren, Rijnsburger and Seynaeve2003).

HIGH RISK CLINIC DESCRIPTION

The Revlon–UCLA High Risk Clinic was founded in 1993. Its purpose is to care for the physical and emotional needs of women who are at heightened risk of breast cancer (e.g., because of family history), but who do not themselves specifically have breast cancer or a history of breast cancer. These women may have one or more first-degree female relatives with breast cancer or have physical conditions that elevate risk (ADH, LCIS) or both. They are followed by a team including a nurse practitioner, a medical oncologist, a clinical geneticist, a psychologist, a physician specializing in lifestyle management/nutrition, and mammography staff especially focused on high-risk women. Patients typically come at 6-month intervals, but may come more often if they present substantial clinical anxiety about their condition or situation.

The initial intake visit includes the following: (1) collection of detailed family histories (both maternal and paternal) and development of a family genetic pedigree; this generates a cancer risk analysis calculation (Gail et al., Reference Gail, Brinton and Byar1989), which is then presented to and discussed with the patient; (2) medical counseling about issues such as use of estrogen replacement and possible medical interventions (i.e., Tamoxifen) for risk reduction; (3) learning breast self-examination, first from an actual breast model, and then on oneself if possible; (4) exercise and nutritional planning to lower risk by lowering the basic blood level of circulating estrogen and planning for weight reduction/weight maintenance; and (5) exploring psycho-social life history of traumas related to familial breast cancer and exploring barriers to implementation of lifestyle changes.

A demographic profile of a subset (n = 160) of the High Risk Clinic population is presented in Table 1. As this table shows, most of the women in this sample rated their grief related to familial breast cancer as “moderate to high.” Table 1 also shows that half of the women have experienced a death due to breast cancer, with the most having had their mother as the breast cancer patient.

Table 1. High risk clinic patient demographics (N = 160)

Changes in psychological test results between the initial clinic visit and the first follow-up visit at 6 months are depicted in Table 2 for a subset of the High Risk Clinic population (n = 77). The data reflect a significant reduction in Personal Risk Rating (p = .001) over the 6 months. There was a trend toward reduction in state (reactive) anxiety, but with little change in trait (characterlogical) anxiety. This is to be expected, because desensitization to the clinic usually occurs between these two visits. The anxiety score of 41st percentile is clinically significant (40th percentile is the cutoff score; Spielberger, Reference Spielberger1983). Depressive symptoms (CES-D) shifted slightly between the two visits. However, the depressive symptoms are substantially below the cutoff score for the CES-D of 16, and in the normal range for depressive symptoms (Spielberger, 1983). Thus, these data show that it is easier to shift the patients' perception of risk than to shift their emotional states (anxiety and depressive symptoms) characteristically presented in our High Risk Clinic.

Table 2. Changes on psychological testing from baseline to first follow-up visit (N = 77 patients)

^a40th percentile indicates significant symptoms of clinical anxiety state.

^b40th percentile indicates significant core (chronic) anxiety.

^c16 is cutoff score for significant likelihood of diagnosis of depression.

COMPARISON WITH LICHTENTHAL'S FINDINGS

Lichtenthal modeled her concept of complicated grief on the work of Horowitz et al. (Reference Horowitz, Siegal and Holen1997) and Prigerson and Jacobs (Reference Prigerson, Jacobs, Strobe and Harrison2001). These two research teams designated the term “complicated grief” to distinguish it from the “normal” and painful emotional, behavioral, and cognitive responses to loss that usually subside after about 6 months. In complicated grief, unresolved responses continue to disrupt the individual's ability to function normally beyond the 1-year anniversary of the death, and even 14 months after the loss as measured by Horowitz et al.'s (1997) assessment analyses at 6 and 14 months. Those time points, rather than 12 months, were chosen in order to avoid confounding the data with anniversary reactions (Horowitz et al., 1997). Bereaved individuals variously describe strong yearnings for the deceased, feelings of aloneness/emptiness, difficulty in sleeping, and strong avoidance of reminders of the deceased.

Our results with our study population at the Revlon–UCLA High Risk Clinic are in broad agreement with Lichtenthal et al.'s (2004) description of the constellation of symptoms, as well as with the criteria for duration, which can last a lifetime. A large subset of women in our population does not experience the relief from normal grieving that usually occurs by 12 or 14 months after the death, as reflected by the large percentage (56%) of self-rated grief scores in the moderate to high range (see Table 1), whereas only 18% of patients rated their depression as moderate to high. Therefore, they are clearly differentiating between the experience of grief and that of depression (Clayton et al., Reference Clayton, Desmarais and Winokur1968, Reference Clayton, Hirjanic and Murphy1974) In fact, they go on to experience the onset of complicated grief, which waxes and wanes for the duration of the high-risk woman's lifetime. This is partially because of her biologic/genetic legacy and partially because of the particular nature of the (deceased) mother–daughter relationship. The experience of being at high risk never ceases and may thus maintain indefinitely the complicated grief phenomena.

PSYCHOLOGICAL RISK FACTORS FOR COMPLICATED GRIEF

Several investigators have demonstrated certain risk factors for complicated grief that extend back into childhood (Prigerson et al., Reference Prigerson, Bierhals and Kasl1997; Kelly et al., Reference Kelly, Edwards and Synott1999; van Dooren et al., 1998). One major factor is an insecure attachment style (Bowlby, Reference Bowlby1973), characterized by excessive dependency, compulsive caregiving, and defensive separation (protective distancing). As noted in developmental psychology work by Bowlby (Reference Bowlby1973, Reference Bowlby1980), each child develops an “internal working model” on which he or she relies to created a basic theory of self, the other, and the world. These models derive from interactions between the child and his or her caretakers, the primary attachment figures. Out of these interactions, a child forms mental schemas that help him or her interpret interpersonal experiences. The secure attachment model provides a child with an internalized self structure, the belief that a protective, supportive figure is available to the child, and the belief that, with time, the child will become competent to deal with the world (Neimeyer et al., Reference Neimeyer, Prigerson and Davies2002, p. 243). An insecure attachment model develops when parents are unavailable and/or rejecting, leading the child to feel incompetent and devalued. In these circumstances, the world seems unpredictable, and access to supportive, competent individuals is uncertain. Consequently, the child feels insecure about his or her ability to meet the world's demands successfully. In adulthood, such individuals may move between attachment styles variously characterized by excessive dependency in marriage, compulsive caregiving, or defensive separation in which the person is emotionally alienated and overly self-reliant.

STYLES OF INSECURE ATTACHMENT

Among the population of women with complicated grief at our high risk clinic, we have noted two different styles of insecure attachment. At one end of the continuum are women who seemingly cannot tolerate any separation from their children. They home-school their children, do not work outside of the home themselves, and do not engage in expected normal separations from their children in order to exercise, take classes, go on outings with adult friends, or spend time with their own partners/spouses. One woman told us, ”My mother left me from the time I was seven (the date of the initial cancer diagnosis), and I never want my children to feel that.” She was the designated caregiver for her mother, as well as for her sister and her father until the mother died when our patient was 13 years old.

At the other end of the continuum are women unable to form secure attachments with anyone, be it a friend, spouse, or child. Their sense of foreboding about their own longevity, coupled with the grief they experienced over the loss of their mothers, leads them to avoid emotional intimacy and enduring relationships. Instead, they develop a defensive emotional self-reliance and an avoidance of any affectional ties or find themselves caught in an approach/avoidance dilemma about attachments in which these women could be characterized as unstable or temperamental.

It should be noted that, in the van Doorn et al. (Reference Van Doorn, Kasl and Beery1998) study of risk factors for complicated grief, their subjects were individuals caring for their terminally ill spouses. The mean age of the spousal caretakers was 62 years, whereas the mean age of our high risk clinic population is 42 years old. Hence the developmental stages of their and our populations are different, with different ramifications for overattachment (i.e., to the spouse vs. to dependent children).

RECONSTRUCTION OF ONE'S PERSONAL NARRATIVE

A fundamental issue for our high risk population concerns their success or failure in reconstructing their sense of self after their mother's death. We view our population as susceptible to complicated grief because of the dual challenges posed by reconstructing their personal narrative (their own construction of their sense of self, i.e., their story) after their mother's death and by their overidentification with the mother's biological and physical being (Neimeyer, Reference Neimeyer2001).

The work of reconstruction requires that such a woman acknowledge uncertainty about her own survival and, hence, about achieving life goals such as marriage, creating a family, and having a career. The woman must shift her belief system, such that the seeming “inevitability” of a shorter life span becomes replaced by the “possibility” of a longer life. With that shift, the woman can “take the risk” of reinvesting in the world of relationships, marriage, work, and children despite the continued risk of loss. Key to this reconstruction is successfully working through the loss of the mother who provided a sense of continuity and affirmative support. Such success includes developing the ability to remember her without being mired in pain, anxiety, and reveries about the relationship severed by death. Another sign of success is the surviving daughter gently distancing herself from total identification with her (deceased) mother's being, body, biology, and fate.

This psychological work is demanding in that we do not imply that the daughter should “forget” or deny her mother's experience. Specifically, the daughter can realize that she is not bound to repeat her mother's experience. We want these daughters to acknowledge their own experience of surviving their mothers' illness and to recognize that “biology is not destiny” in every situation. If they are released from that conviction, then they can remember the memory of their mother's illness and death without being dominated by it. Furthermore, they do not have to internalize their mother's entire experience of illness and death in order to maintain a successful attachment to her. The result of successful grieving would allow the daughter to remember her mother and not be retraumatized. Unsuccessful grieving, which can devolve into complicated grief, leaves one mired in reveries about the past and its lost relationships. These kinds of reveries promote overattachment to the lost person, as well as an inability to create a world with new opportunities for attachment. Such unsuccessful grieving limits participation in the present.

The reconstruction of one's self-narrative can either be progressive or regressive. If it is regressive, the daughter becomes immersed in reveries or reconstructions of past events. We have seen vivid examples of this during high risk cancer clinic visits. In one extreme case, a daughter refused to alter any aspect of her family's home, even with the permission of her father, the bereaved spouse. She wore her deceased mother's clothing, despite the fact that her 53-year-old mother's style was not flattering to this 22-year-old woman. A more positive “relearning of the world” after her mother's death involving rational selective retention could have allowed this young woman to choose pieces of her mother's jewelry that held special meaning for her and to change the home in ways that reflected the tastes of the present occupants, incorporating treasured items of furniture that her mother liked. In this way, the daughter could have affirmed continuity with her mother while acknowledging her own separate identity after her mother's death.

Neimeyer (Reference Neimeyer2001) has examined the construction of an individual's self-concept based on her relationship with parental figures and has studied the impact of bereavement on the self-concept. In the case of the young woman just described, she put aside her own developing self and took on her mother's garments and house as a way of keeping her mother alive and not disrupting her own sense of herself by having to define herself in the absence of her mother. According to Neimeyer (Reference Neimeyer2001), in order to find meaning for herself in this new psychological and actual world, the young woman has to reconstruct her identity and to to “interbraid” the loss of another with the loss of oneself (the self mirrored by that other person). Involved in this reconstruction is a choice of significant aspects of one's mother and of memories of her that can then be amalgamated into a new sense of self. Critical to this focus for our population of high-risk women is the daughter's need to thoughtfully and not compulsively incorporate aspects of her mother into her “internal self” so as to live her own life and not erase her own identity in a frightened defense against the anxiety of totally “losing her mother” to death. The main goal is thus to support choice over compulsion.

Another example of regressive reconstruction emerged in the story of a daughter who so identified with her deceased mother that the daughter missed several annual clinic visits, missed her annual mammography appointments, and could not bring herself to do breast self exams. She reported that her mother had so delayed her own care of her breast tumor that her disease was widely metastatic by the time that she sought health care and informed her family.

THE RELATIONSHIP OF COMPLICATED GRIEF AND POST-TRAUMATIC STRESS DISORDER IN OUR POPULATION

Lichtenthal et al. (Reference Lichtenthal, Cruess and Prigerson2004) postulated that the key element of complicated grief involves the protracted nature of separation anxiety and distress. Coping efforts typically involve trying to reorient oneself in a world without the deceased. Lichtenthal et al. (Reference Lichtenthal, Cruess and Prigerson2004) distinguish complicated grief from PTSD by stating that “traumatized individuals are typically anxious about the threat related to the traumatic event, whereas bereaved individuals experience separation anxiety” (p. 648).

Our thesis, based on our population, differs from Lichtenthal et al. (Reference Lichtenthal, Cruess and Prigerson2004) on this issue of separation of PTSD from complicated grief. For our population, we observe that both of these emotional states are present and must therefore be jointly considered. The typical bereaved daughter in our clinical population experiences not only the loss of her mother but also anxiety over the risk of possibly being diagnosed with breast cancer herself.

When we review the DSM-IV diagnostic criteria for post-traumatic stress disorder, we note first that our population has been exposed to a traumatic event involving not only a threat to their mother's physical integrity (the diagnosis of breast cancer), but also their mother's actual death from breast cancer. These daughters report responses involving fear and helplessness. A second component of the PTSD diagnosis involves persistent reexperiencing of the traumatic event. Our population of women does describe repeated and intrusive distressing recollections of their mothers' illnesses and deaths, although they usually report repeated distressing dreams of their parents' deaths.

Another dimension of reexperiencing the event occurs when these women who are patients at our high risk breast cancer clinic visit the clinic itself or attend appointments in the adjacent hospital. Although they do not experience flashbacks, they do feel as though they are reliving segments of their mothers' medical visits. When our women subjects attend the clinic, they often report psychological distress at the sight of the hospital, the clinic and adjoining examination rooms, the physicians' white coats, the mammography machines, and procedures associated with breast examinations. We conclude that our population physiologically reacts to internal or external cues symbolizing the event of their mother's illness reflected by their frequent tearing up or crying during clinic visits.

Persistent avoidance of stimuli associated with the trauma and numbing of general responsiveness do exist in our population. Our patients frequently tell us that they have never discussed their thoughts or feelings about their mother's illness and death with anyone prior to their visit to our clinic. Many women in our population admit being reluctant to come to the high risk clinic for their own care because it carries too many painful associations for them. We note that they postpone receiving appropriate health care for themselves. Our women subjects are rarely unable to recall an important aspect of their mother's illness and/or death, unless they were very young at the time of her mother's acute illness. Most women in our population do not report feeling detached or estranged from others, nor do they usually experience markedly diminished interest in significant activities. These daughters are not emotionally constricted. However, a significant hallmark of their experience is a sense of a foreshortened future. In one section of our intake questionnaire, nearly every woman tell us that she not only fears getting breast cancer, she is certain that she will get it and will not have children or will not see them grow up or will not have a long marriage. This sense of certainty about developing breast cancer seems to heighten our women's vigilance about changes in the appearance of their breasts. Our subjects experience these symptoms for the duration of their lives, and they acknowledge that these symptoms may interfere with their procuring timely health care. Their dread of developing breast cancer can also interfere with their social and work lives because of their ambivalence about investing in work and in significant relationships, given the risk that work and relationships could be interrupted or severed by the development of breast cancer.

Thus, the actual death of a mother from breast cancer creates the separation anxiety typically related to complicated grief, whereas the risk of breast cancer to the daughter creates avoidance and experiences of distress typically related to PTSD. For our clinical population, images of the deceased mother paradoxically combine a sense of comfort with anxiety. Anxiety casts a shadow on images of the deceased, linked to our daughters' intrusive memories of their mothers' disfigurement, pain, debility and suffering. Our women feel that their own future is written in their mothers' experiences with breast cancer. For some of our women it feels like more than just a statistical probability: It feels as if their fate is actually sealed.

TREATMENT INTERVENTIONS FOR HIGH-RISK WOMEN WITH COMPLICATED GRIEF

We conclude that this group of women face dual challenges. On the one hand, they must resolve their attachment issues with their mother and create a new and meaningful world without her. On the other hand, as they create this new world, they must learn how to adapt to the threat of a diagnosis of cancer in themselves. The clinical challenge for us as therapists is to alter their belief system about biology being destiny—that is, the daughters' sense that they too will develop and die of breast cancer as did their mothers. Altering this belief system would permit the daughters to feel less trapped in the past and to devote more energy to their present and future.