Psychosomatic Medicine
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Published online before print December 24, 2007, 10.1097/PSY.0b013e31815c25cf
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Weihs, K. L.
Right arrow Articles by Simmens, S. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Weihs, K. L.
Right arrow Articles by Simmens, S. J.
Related Collections
Right arrow Social Support
Right arrow Cancer
Psychosomatic Medicine 70:117-124 (2008)
© 2008 American Psychosomatic Society

ORIGINAL ARTICLES

Close Relationships and Emotional Processing Predict Decreased Mortality in Women with Breast Cancer: Preliminary Evidence

Karen L. Weihs, MD, Timothy M. Enright, PhD and Samuel J. Simmens, PhD

From the Department of Psychiatry (K.L.W.) and the Arizona Cancer Center, University of Arizona, Tucson, Arizona; Department of Psychiatry and Behavioral Sciences (T.M.E.), The George Washington University, Washington, DC; and the Departments of Epidemiology and Biostatistics (S.J.S.), The George Washington University, Washington, DC.

Address correspondence and reprint requests to Karen L. Weihs, Department of Psychiatry, University of Arizona, 1501 N. Campbell Ave, Tucson, AZ 85724-5002, Phone: (520) 626-8940. E-mail: weihs{at}email.arizona.edu


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 NOTES
 REFERENCES
 
Objective: To examine close relationships and emotional processing as predictors of breast cancer mortality.

Methods: Ninety women were enrolled at 14 ± 5 months after diagnosis of Stage II/III breast cancer. The Nottingham Prognostic Index (NPI) quantified disease severity. Cox proportional hazards analyses were used to predict mortality using standardized variables.

Results: Twenty-one subjects developed recurrent disease and 16 died during an 8-year follow-up. NPI predicted increased mortality: risk ratio (RR) = 1.60 (CI = 1.05–2.41). Decreased mortality was predicted by confiding marriage (CONF): RR = 0.31 (CI = 0.10–0.99), and number of dependable, nonhousehold supports (SUPP): RR = 0.41 (CI = 0.21–0.80). A composite measure of close relationships (standardized CONF + SUPP = SUPPCONF) had a strong protective effect: RR = 0.30 (CI = 0.13–0.69). Two emotion processing variables, acceptance of emotion and emotional distress (POMS-TOT) were found to be negatively correlated (r = –.49). Acceptance of emotion predicted decreased mortality (RR = 0.46 (CI = 0.24–0.86)) when analyzed together with emotional distress, but not separately. There was a trend for a protective effect of emotional distress: RR = 0.37 (CI = 0.12–1.09) in the same analysis. RRs for mortality in a multivariable analysis were: SUPPCONF: RR = 0.55 (CI = 0.30–1.00); acceptance of emotion: RR = 0.48 (CI = 0.25–0.91); and emotional distress: RR = 0.40 (CI = 0.14–1.19).

Conclusions: Two aspects of close relationships—marital confiding and dependable, nonhousehold supports—were protective against breast cancer progression. Acceptance of emotion, after controlling for emotional distress, also predicted decreased mortality. Analysis of close relationships together with emotion processing variables suggested unique protective effects against mortality, but a larger study is necessary to determine whether this is the case.

Key Words: social support • marital quality • emotion processing • acceptance • breast cancer • survival

Abbreviations: CI = confidence interval; CONF = marital confiding; NPI = Nottingham Prognostic Index; RR = Risk Ratio; POMS-TOT = Total Mood Disturbance Score, Profile of Mood States; SUPP = number of dependable, nonhousehold supports; SUPPCONF = close relationships total.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
METHODS
 RESULTS
 DISCUSSION
 NOTES
 REFERENCES
 
Social and emotional processes may alter the risk of disease progression in breast cancer patients (1). Further understanding of these processes could improve prognostic accuracy and may provide new targets for intervention to improve disease outcome.

This prospective, longitudinal study examines the unique and combined effects of close relationships and emotional processing variables as protective factors against breast cancer progression. We previously reported that more dependable, nonhousehold relationships predicted decreased mortality in this sample of breast cancer patients—all of whom were mothers (2). A subsequent report from the Nurses Health Study described similar results in women with and without offspring (3). We examine the impact of close relationships on disease progression using a broader index, which includes marital confiding, along with the number of dependable nonhousehold relationships. We also examine two indicators of emotional processing—self-reported distress and acceptance of emotions—for protective effects against early mortality. Multivariate survival analyses including both close relationship and emotional processing variables provide information about their combined influence on breast cancer progression.

Marital Confiding and Dependable, Nonhousehold Support
Married women often identify their spouses as their most important source of emotional as well as instrumental support (4). Although adequacy of emotional support from all sources has been linked to improved breast cancer outcome (5), marital status alone has not been found to predict increased survival in the majority of prospective studies of breast cancer patients, after taking biologic risk factors into account (5–8). Two studies have even identified marriage as a risk factor for early mortality from breast cancer (9,10). Investigation of differences in marital quality may be needed to detect the influence of marriage on breast cancer outcomes, because marital quality rather than marital status has been found to be the "active ingredient" in the salutary effect of marriage on health in the general population (11). Evidence from studies of congestive heart failure (12), coronary artery disease (13) and end-stage renal disease (14) suggests that marital quality is protective against mortality. These studies found the protective effect of marital quality to be stronger for women than for men, suggesting that marital quality may be a particularly important protective factor in breast cancer patients.

Emotional Processing
Emotional processing involves acknowledging and accepting the emotions generated by one’s life experience, tolerating the associated arousal, and then exploring, reflecting on and making sense of the experience that generated the emotions (15). Emotional processing facilitates resolution of distress and decreases psychopathology over time, as demonstrated in both laboratory and clinical studies (16). Conversely, experiential avoidance inhibits emotional processing and is likely to prolong emotional distress, even though it may reduce awareness of distress in the acute setting (17). Coping with emotions through acceptance is increasingly being understood to have a salutary effect on the resolution of distress (15).

The life-threatening nature of breast cancer understandably contributes to emotional distress, which dissipates after the first year for most, but not for all breast cancer patients (18). From an emotion processing perspective, a breast cancer patient’s self-reported distress is more than an indicator of how much she is disturbed by the illness. It also reflects the extent to which she is processing her feelings in this difficult situation. Women who are processing rather than avoiding or suppressing their emotions would be expected to have moderately elevated scores on distress measures during the first year after diagnosis when they are coping with the stress of noxious treatments and life changes.

Consideration of both emotional responses and emotional processing is important when attempting to interpret studies of self-reported distress and its influence on breast cancer progression. Numerous prospective longitudinal studies have failed to detect an association between increased distress and breast cancer progression (10,19–22). Low levels of reported distress during cancer treatment have been linked to adverse disease outcome in a few studies (22–24), and one breast cancer study found modestly increased, rather than very high or very low, emotional distress to be protective against early mortality (25). Low levels of self-reported distress are found in conjunction with emotional suppression or repression. The finding that repressive coping predicts more rapid disease progression in breast cancer patients (26) is therefore consistent with the hypothesis that lower emotional processing, as indicated by low self-reported distress and emotional suppression/repression, is associated with increased mortality risk. Similarly, a large study of 847 breast cancer patients using self-report measures (27) found that suppression of emotions predicted increased mortality during a 9-year follow-up, whereas expression of emotion predicted decreased mortality. We are not aware of previous studies of acceptance of emotions as a protective factor in breast cancer. To make assumptions about the effects of mood on disease progression based only on "how much" distress exists, rather than also exploring "what is done with the distress," is potentially misleading. An emotion processing framework helps to integrate previous research for understanding how distress and methods of coping with distress affect disease outcome.

This study investigates distress and acceptance of emotions to test the hypothesis that these aspects of emotional processing are protective against disease progression in breast cancer patients.

Mutual Effects of Close Relationships and Emotional Processing on Disease Outcome
Spousal support has long been identified by social support and family researchers as playing a central role in patient emotional adjustment (28). Supportive communication in close relationships is known to improve psychological well-being (29). Reynolds et al. found an interaction of emotional expression and social support, such that those low in both of these factors were at more than twice the risk of mortality compared with those with high expression and greater social support (27).

Hypotheses for the Current Study

1) Emotional processing, measured as acceptance of emotion and elevated distress near the completion of treatment for breast cancer, will predict decreased mortality.
2) Close relationships, including marital confiding and more dependable, nonhousehold supports, will predict decreased mortality. Marital status will not predict disease outcome.


   METHODS
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
RESULTS
 DISCUSSION
 NOTES
 REFERENCES
 
Brief descriptions of study methods are provided here. Complete information on study design, subject recruitment, and sample characteristics is contained in our previous publication from this study (2).

Study Design
Psychosocial variables were assessed within 6 months after completion of adjuvant chemotherapy in women with Stage II or III breast cancer. Subjects were monitored for disease outcome for 8 years. Survival analyses were used to predict disease progression from psychosocial variables, after controlling for disease severity at diagnosis.

Eligibility and Recruitment
Subjects were recruited in 1991 to 1993 from four medical centers in the Washington, DC area. Disease status was measured annually until the end of 2000. Enrollment was limited to subjects diagnosed with Stage II and III (30) breast cancer within 6 months after completion of adjuvant chemotherapy. Because these subjects were recruited for a family coping study, eligible subjects had at least two other family members who would participate, including at least one son or daughter aged ≥10 years. Presence of a spouse was not a criterion for eligibility. Only data collected from patients are used in this report. The study was approved by the Human Subjects Protection Boards of all institutions where the subjects were recruited. Informed consent was obtained from all subjects before data collection.

Consecutive screening at four breast cancer treatment facilities identified 383 eligible breast cancer patients based on their time after diagnosis and treatment. A total of 270 of these 383 patients met the criteria for Stage II or III disease. Of these 270 patients, 183 patients were found to be eligible for the study based on family composition as well as disease stage and time after diagnosis. Ninety of these 183 eligible patients were enrolled in the study. Ninety-three eligible patients did not enroll, either because the patient or her family members were not interested or because of logistical circumstances such as insufficient contact information or poor proximity to the research center.

A previous publication (2) provided complete information on the demographic and clinical characteristics of all eligible patients in this study. There were no differences in clinical disease or treatment characteristics between those who did or did not enroll in the study. The only demographic difference between the groups was that the enrollees were more likely to have completed some college education.

Measures

Acceptance of Emotion
The Acceptance of Emotions Scale (AE) assesses the extent to which subjects are accepting, friendly, and nurturing toward their feelings (31). Thirteen items include statements such as "I naturally and easily attend to my feelings." Responses are based on the percentage of time each statement is true, in increments of 10 ranging from 0 for never/not at all to 100 for always/perfectly. The total score is the mean of ratings on each of 13 items. Internal consistency as measured by Cronbach’s {alpha} was 0.92 in the current sample and test retest reliability measured 15 months after the assessment used here was 0.58 (Pearson correlations).

Emotional Distress
The Profile of Mood States (POMS) (32) has 65 items and measures six affective states on a 5-point scale ranging from "not at all" to "extremely." Respondents are instructed to answer based on how they felt during the previous week. The POMS-Total Mood Disturbance Score (POMS-TOT) is a valid measure of mood states, which has been used to measure changes in mood related to psychotherapeutic interventions for cancer patients (33). It has six subscales and, because of reverse coding of one subscale, the total scale score (POMS-TOT) can range from –55 to 270. Cronbach’s {alpha} for the POMS-TOT is 0.90 (34).

Nonhousehold Confidants and Marital Confiding
Breast cancer patients were asked to list the first names of supportive relatives and friends in response to the following written question: "In the event of domestic or emotional problems or other stressful situations, who are the people you could call for support or help?" The number of dependable, nonhousehold relationships (SUPP) refers to the total number of relatives and friends listed by the patient as people she could call on for support or help.

The following script, adapted from the Close Persons Questionnaire (35), was used to assess marital confiding during an interview with each subject: "In the event of a serious problem or stressful experience, who is the first person you would want to speak to about the situation? Who would be your second choice?" If the subject had not mentioned her spouse or long-term partner, the interviewer directly asked the subject whether she would choose to speak with her partner about a serious problem or stressful experience. This resulted in a dichotomous variable indicating the presence or absence of confiding marriage/intimate partnership (CONF). Patients without spouse/partners were coded as nonconfiding.

Married status was defined as legal marriage or a committed, cohabiting relationship of at least 6 months duration.

Biomedical Risk Status
The Nottingham Prognostic Indicator (NPI) was used as a measure of biomedical risk. The NPI is calculated as 0.2 x tumor size (cm) + lymph node status (1 = node-negative; 2 = 1–3 positive nodes; 3 = ≥4 positive nodes) + histological grade (1 = good; 2 = moderate; 3 = poor). This is a well-established measure of risk for recurrence and mortality in patients with breast carcinoma (36).

Treatment Aggressiveness
An oncologist rated each participant’s treatment regimen on a scale of 1 to 3 for least to most aggressive adjuvant intervention. Treatments were consistent with the recommended standards of care for women with Stage II and III breast cancer at the time of study enrollment, 1992 to 1993 (37).

Disease Outcome
Medical charts of the subjects were reviewed annually until the end of data collection (2000) to identify breast cancer recurrences and deaths from breast cancer. Dates of recurrence and/or death, as well as cause of death, were determined for all participants through medical record information and/or from contact with medical providers, patients, and/or their family members.

Missing Data and Methods of Data Analysis
Data for emotional acceptance (EA), POMS-TOT, and marital confiding were missing from six, four, and two subjects, respectively. One subject did not complete the assessment of dependable, nonhousehold supports. The technique of multiple imputation (38) was employed to fully utilize the collected information and potentially reduce statistical bias based on the usual Missing at Random assumption. The set of all analysis variables in this report was used to create 20 multiply imputed datasets with the default MCMC algorithm in SAS 9.1.3 (39). Statistical estimates and p values for linear and proportional hazards regression models were then generated through the SAS MIANALYZE module.

Cox proportional hazards regression was used to predict the time from study entry to recurrence or mortality, using biomedical and psychological variables. Data were censored at the time of last available follow-up information for purposes of the survival analyses (40). Each of the psychosocial variables was standardized and analyzed separately to predict time to recurrence and mortality, controlling for NPI. We limited the number of independent variables in the multivariate analyses to three, in order to have at least five deaths per predictor variable in the model, a minimum recommended for Cox proportional hazards regression analyses (40). Pearson correlations were used to assess the associations among predictor and demographic variables.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
DISCUSSION
 NOTES
 REFERENCES
 
Sample Characteristics
The mean interval from breast cancer diagnosis to assessment was 14 ± 5 months. Ninety-five percent of the subjects receiving adjuvant radiation or chemotherapy (other than antihormonal treatment) had completed treatment before data collection. The mean disease severity score, NPI, was 4.9 ± 1 (CI = 2.5–7.2).

At the end of the study period, 21 subjects had been diagnosed with recurrent disease. Sixteen subjects died from breast cancer before the end of data collection. Two subjects died from nonbreast cancer-related causes and their data were censored at their dates of death. Two subjects were lost to follow-up or withdrew consent before the end of data collection and their data were also censored at the last date of contact.

Subjects’ mean age was 51 ± 10 years; self-reported ethnicity in the sample was 45% European American, 37% African American, and 8% other ethnicity. Sixty-eight percent of subjects were married or cohabiting, 23% were divorced, 6% were widowed, and 3% were never married. The married subjects had been with their spouses for a mean of 22 years. Eighty percent had more than high school education. There were no demographic or disease stage differences between participants and eligible nonparticipants, except a larger percentage of participants had >12 years of education (p = .03) (2).

Psychosocial Variables
The mean POMS-TOT scores at T1 was 19.1 ± 31.6 (CI = –24–106). This is higher than POMS-TOT scores from community samples but lower than psychiatric outpatient samples (33). A study of women within 48 hours of breast cancer surgery (41) revealed higher POMS-TOT scores than this sample (59 ± 46), but another study of patients with breast cancer at 1 year post diagnosis (42) revealed similar scores (23 ± 33). The mean score on AE was 59.9 ± 17.6 (CI = 15–95). Normative data for this scale is not available, as it was used for the first time in this study (31).

Forty-one (46%) subjects were in confiding marriage/partnerships, 20 (22%) were in nonconfiding marriage/partnerships, and 28 (32%) were not in marriage/partnerships. The mean number of dependable nonhousehold supports was 6 ± 4 (CI = 1–16).

Correlations Between Demographic, Disease, Treatment, and Psychosocial Characteristics
"Psychosocial variables" are defined here as marital status (MS), SUPP, marital confiding (CONF), EA, and distress (POMS-TOT). Neither disease severity (NPI) nor treatment aggressiveness was related to age, education, or ethnicity. In addition, there were no significant relationships between psychosocial variables and education, or ethnicity. MS, SUPP, CONF, and AE were not related to age. However, younger age was correlated with higher distress (POMS-TOT) (r = –.26, p = .02). Psychosocial variables were not correlated with disease severity (NPI) or treatment aggressiveness. The correlation of NPI and treatment aggressiveness was r = .15, p = .17.

Table 1 provides bivariate correlations as well as mean values for the psychosocial variables.


View this table:
[in this window]
[in a new window]

  		TABLE 1. Sample Size, Means, and Correlations for Psychosocial Study Variables

 

Disease Severity, Treatment Aggressiveness, and Age as Predictors of Disease Outcome
A test of the proportionality assumption of the Cox proportional hazards model (30) indicated acceptable parameters for this data set. The relative risk associated with a 1-standard deviation (SD) increase in disease severity (NPI) was 1.37 (CI = 0.92–2.04) for recurrence and 1.60 (CI = 1.05–2.41) for mortality. The risk ratios (RR) for NPI indicate that, for every increase of 1 SD for NPI, the risk for recurrence increases by 37% and the risk for mortality increases by 60%. Neither treatment aggressiveness score nor estrogen receptor status or age predicted recurrence or mortality.

Psychosocial Variables as Predictors of Disease Outcome
Disease severity, measured by NPI, was included in the analysis of each individual psychosocial variable. Table 2 contains information on emotional processing variables predicting disease outcomes.


View this table:
[in this window]
[in a new window]

  		TABLE 2. Prediction of Recurrence/Mortality From Univariate and Multivariate Analyses of Emotional Processing Variables

 

Neither emotional distress (POMS-TOT) nor acceptance of emotion (AE) was a significant predictor of recurrence or mortality when analyzed separately, although the results were in the predicted direction of influence. In multivariable analysis including AE and POMS-TOT, AE became a statistically significant predictor of both recurrence (RR = 0.55 (CI = 0.23–0.92), p = .02) and survival (RR = 0.46 (CI = 0.24–0.86), p = .01), whereas there were trends for prediction of recurrence (RR = 0.55 (CI = 0.29–1.09), p = .09) and survival (RR = 0.37 (CI = 0.12–1.09), p = .07) by level of POMS-TOT.

AE and distress (POMS-TOT) were negatively correlated (r = –.49), creating a "reciprocal suppressor" effect in which both independent variables (AE and POMS-TOT) end up with higher correlations with survival time after each is adjusted for the other (43). To illustrate this suppressor effect, we created a diagram of subgroup differences in mortality by level of AE and POMS-TOT (Figure 1). High versus low levels of AE were defined by a median split and three levels of POMS-TOT were defined by >1/2 SD below the mean, within 1/2 SD of the mean, and >1/2 SD above the mean. The percent of subjects alive at the end of the study were nearly the same for those with high (85%) versus low AE (83%). However, when these two groups were divided into subgroups of high, medium, or low POMS-TOT, the survival rates when AE is high were 100%, 92%, and 78%, respectively, and 92%, 82%, and 67% when AE is low. The negative correlation of AE and POMS-TOT results in the unequal distribution of cases by subgroup as seen in Figure 1. A {chi}2 analysis for differences between these six groups showed a trend for differences (p = .10) in mortality rate. This {chi}2 test is less sensitive than the Cox proportional hazard analysis for differences in mortality based on AE and POMS-TOT because the length of survival time is not taken into account in the {chi}2.


Figure 116
View larger version (9K):
[in this window]
[in a new window]

  		Figure 1. Suppressor effect: survival rates for combination of high versus low acceptance1 of emotion and high, moderate and low distress.2 1High versus low acceptance of emotion based on median split. 2Definition of levels of distress: high distress = >1/2 SD above the mean; moderate distress = within 1 SD of the mean; low distress = >1/2 SD below the mean. Number of subjects in each group listed with data points. SD = standard deviation.

 

Table 3 contains information on relationship variables predicting mortality. Marital confiding predicted decreased recurrence (RR = 0.27 (CI = 0.10–0.77), p = .01) and mortality (RR = 0.31 (CI = 0.10–0.99), p = .05). Figure 2 shows a Kaplan Meier plot of differences in mortality over time in those with (1) and without (0) a confiding marital relationship. As previously reported, the number of dependable nonhousehold supports predicted decreased mortality (RR = 0.41 (CI = 0.21–0.80), p = .01) with a trend toward decreased recurrence (RR = 0.68 (CI = 0.42–1.10), p =.11). Marital status did not predict differences in disease outcome. We created a composite indicator of close relationships (SUPPCONF) by adding the standardized scores for marital confiding and the number of dependable, nonhousehold supports. These variables were not correlated with one another (r = 0.02). In a survival analysis including NPI, SUPPCONF had a protective effect against recurrence (RR = 0.36 (CI = 0.18–0.71), p = .003) and mortality (RR = 0.30 (CI = 0.13–0.69), p = .005).


View this table:
[in this window]
[in a new window]

  		TABLE 3. Prediction of Recurrence/Mortality From Univariate Relationship Variables and Multivariate Analyses of Close Relationships and Emotional Processing Variables

 

Figure 216
View larger version (11K):
[in this window]
[in a new window]

  		Figure 2. Kaplan-Meier survival estimates among 88 newly diagnosed breast cancer patients by presence versus absence of a confiding marital relationship at study enrollment (14 ± 5 months post diagnosis). Group 1: Patients with a confiding marital relationship: 40 patients, 4 dead (10%), 36 censored (90%); mean (restricted) survival time: 6.86 years (SE 0.24). Group 2: Patients with no confiding marital relationship: 48 patients, 11 dead (33%), 37 censored (77%); mean (restricted) survival time: 6.14 years (SE 0.14): Test statistic (null hypothesis: survival distributions are equal): log rank test, p = .05. SE = standard error.

 

A multivariable survival analysis was conducted using SUPPCONF, AE, and POMS-TOT (Table 3). SUPPCONF predicted disease outcome with RR = 0.57 (CI = 0.35–0.94), p = .03 for recurrence and RR = 0.55 (CI = 0.30–1.00), p = .05 for mortality. The multivariable analysis showed that 1-SD increase in AE reduced the risk for recurrence, RR = 0.60 (CI = 0.36–1.00), p = .05, and for mortality, RR = 0.48 (CI = 0.25–0.91), p = .02. The RRs for 1-SD increase in POMS-TOT were 0.61 (CI = 0.31–1.21), p = .16 for recurrence and 0.40 (CI = 0.14–1.19), p = .10 for mortality in the multivariable analysis.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
NOTES
 REFERENCES
 
This study provides preliminary support for our hypotheses that close relationships and emotional processing, measured near the end of the first year after breast cancer diagnosis, are protective against disease progression. The multivariate analysis suggests unique contributions for predicting mortality from both sets of variables. Women who acknowledged more distress reported having more dependable, nonhousehold supports. This may mean that subjects who had more support felt safer in acknowledging their distress, as documented in research by Lepore and associates (44). Distress was not associated with differences in confiding marriage, however, suggesting a more varied or complex interplay between a woman’s distress and confiding in her spouse. Acceptance of emotion (AE) was not correlated with either marital confiding or number of supports, consistent with our finding of their unique effects on mortality.

This study was the first to investigate the interplay of AE and distress as they affect breast cancer outcome. Our data indicate that the prognostic significance of a given level of emotional distress or AE can best be understood in the context of the other emotion processing variable. The study and interpretation of the effects of distress on disease outcome are complex, at least partially because self-reported measures of distress reflect the end product of several emotion generation and processing activities. After controlling for the expected effect of AE to facilitate processing and reduce distress (negative correlation), we found a trend for the remaining distress to be associated with improved disease outcome. This suggests that acknowledging the distress that cannot be modulated by AE—possibly related to real life, uncontrollable circumstances—may have health protective effects. Similarly, acceptance of the emotions that are not modulated by coping through acceptance seems to have salutary health effects.

The protective effect of AE in this study is consistent with Reynolds’ finding of decreased mortality in breast cancer patients who did not suppress their emotions (27). It is also consistent with the results of a psychological intervention study by Cunningham et al. (45), in which breast cancer patients who actively processed their emotions survived longer. AE supports the "working through" of distress and may prolong survival by facilitating the return to emotional and physiological homeostasis.

This is the first report that marital confiding is protective against breast cancer progression. It supports our hypothesis that it is the way women interact with their spouses, rather than just the presence of a marital relationship, that has salutary effects on health. Self-disclosure has been linked to better overall health, healthier immune function, and lower mortality rates (45,46). It has also been shown to modulate physiologic stress responses (47). Research has consistently shown confiding relationships to correspond with the perception of having adequate social and emotional support (48). Our results are consistent with reports that women who perceive they have adequate support have better prognoses after breast cancer diagnosis (5,10).

We found some support for unique effects of close relationship support and emotion processing on disease outcome in this sample. The RRs associated with AE and distress did not change with the addition of close relationships to the survival analyses. Close relationships were less strongly predictive of improved survival in the multivariate analysis (RR = 0.55 (CI = 0.30–1.00)) than in the univariate survival analysis (RR = 0.30 (CI = 0.13–0.69)). Although this suggests possible mediation of the effect of close relationships on mortality through differences in distress and AE, our data did not meet the criteria for testing mediation (49). The level of close relationship support was not correlated with distress or with AE, although women with higher distress reported more dependable supports. A larger study of the dynamic interplay of supportive relationships with distress and coping with distress would be useful for clarifying their shared and unique contributions to disease outcome.

The major limitation of this study is its small sample size and the low rate of disease progression. The latter is, however, in range for breast cancer patients with Stage II disease treated in the 1990s (50). The study was powered to detect a moderate effect size for the impact of psychosocial variables on disease outcome, with projection of 35% mortality after 5 years in a sample of 95 women with Stage II/III breast cancer. The much lower mortality rate of 16% in our sample after 8 years is the result of a larger than anticipated number of patients with Stage IIa, rather than IIb and III disease, as well as improved efficacy of treatments from the mid-1980s when the study was planned to the early 1990s when the subjects were enrolled. Despite this limitation, the results were statistically significant and consistent with our hypotheses, which were based on a clear theoretical framework. A larger study would be needed to draw firm conclusions, however.

The generalizability of these results to women with Stage 0 or Stage I breast cancer cannot be assumed and it would be difficult to conduct a large enough study to test this effect, given the ≥90% survival of this group of patients with modern treatments. The women in this study were all mothers of at least one child. If childbearing and parenting (51) affect the disease process and its relationship to distress and emotion coping, these results may not generalize to women who have not had children.

Our results predicting time to recurrence were not as strong as for predicting mortality. This may be due to the date of recurrence being a less exact measure of outcome than date of death. Women varied in the time required to have recurrent lesions discovered and confirmed as recurrent disease by a physician, making this measure less consistent across subjects. We, therefore, take mortality as the firmer indication of the association of predictor variables to disease outcome.

This study does not provide information about the influence of psychosocial functioning after the second year post diagnosis on disease progression thereafter. Other studies indicate that marital quality is relatively stable over the course of 5 years post diagnosis, whereas emotional distress is known to decrease with length of time post diagnosis in most but not all patients (52,53). Future research on the stability of emotion processing during survivorship could provide important insights into its longer term effects on biological processes that affect cancer progression.

The strengths of this study include its clear conceptual and measurement framework with hypotheses determined before data collection, its prospective design, relatively high uniformity of diagnostic and treatment characteristics, and control for disease severity in each statistical model of individual psychosocial variables predicting recurrence and mortality.

In summary, these results add to existing evidence that close, confiding relationships may improve disease outcome in women with breast cancer. They also indicate that emotional processing through acceptance and acknowledgment of distress during the recovery from breast cancer treatment may have salutary health effects that are unique from the effects of close relationships.

We wish to acknowledge the contributions of David Reiss, MD, Professor of Psychiatry, George Washington University Medical Center; Robert Siegel, MD, Professor of Medicine, Division of Hematology and Oncology, George Washington University Medical Center; Martha Hunt, PhD, MPH; Carlene Sipma-Dysico, MA; Elise Spertus, PhD; and the breast cancer patients whose participation made this study possible.


   NOTES
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 NOTES
REFERENCES
 
This research was supported by Grant K20 MH00906 from the National Institute of Mental Health (K.W.). The study was presented at the 65th Annual Meeting of the American Psychosomatic Society, March 9, 2007, in Budapest, Hungary.

This research was conducted at the Department of Psychiatry and Behavioral Sciences, George Washington University Medical Center, Washington, DC, 1992–2000.

Received for publication November 3, 2005; revision received August 27, 2007.

DOI:10.1097/PSY.0b013e31815c25cf


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 NOTES
 REFERENCES
 

  1. Antoni MH, Lutgendorf SK, Cole SW, Dhabhar FS, Sephton SE, McDonald PG, Stefanek M, Sood AK. The influence of bio-behavioural factors on tumour biology: pathways and mechanisms. Nature Reviews Cancer 2006;6:240–8.[CrossRef][Medline]
  2. Weihs KL, Simmens SJ, Mizrahi J, Enright TM, Hunt ME, Siegel RS. Dependable social relationships predict overall survival in stages II and III breast carcinoma patients. J Psychosom Res 2005;59:299–306.[CrossRef][Medline]
  3. Kroenke CH, Kubzansky LD, Schernhammer ES, Holmes MD, Kawachi I. Social networks, social support, and survival after breast cancer diagnosis. J Clin Oncol 2006;24:1105–11.[Abstract/Free Full Text]
  4. Sandgren AK, Mullens AB, Erickson SC, Romanek KM, McCaul KD. Confidant and breast cancer patient reports of quality of life. Qual Life Res 2004;13:155–60.[CrossRef][Medline]
  5. Maunsell E, Brisson J, Deschenes L. Social support and survival among women with breast cancer. Cancer 1995;76:631–7.[CrossRef][Medline]
  6. Funch DP, Marshall J. The role of stress, social support and age in survival from breast cancer. J Psychosom Res 1983;27:77–83.[CrossRef][Medline]
  7. Neale AV. Racial and marital status influences on 10 year survival from breast cancer. J Clin Epidemiol 1994;47:475–83.[CrossRef][Medline]
  8. Vachon M. A comparison of the impact of breast cancer and bereavement: personality, social support and adaptation. In: Hobfall S, editor. Stress, Social Support and Women. Washington, DC: Hemisphere Publishing Corp; 1986.
  9. Waxler-Morrison N, Hislop GT, Mears B, Kan L. Effects of social relationships on survival for women with breast cancer: a prospective study. Soc Sci Med 1991;33:177–83.[CrossRef][Medline]
  10. Ell K, Nishimoto R, Mediansky L, Mantell J, Hamovitch M. Social relations, social support and survival among patients with cancer. J Psychosom Res 1992;36:531–41.[CrossRef][Medline]
  11. Hibbard JH, Pope CR. The quality of social roles as predictors of morbidity and mortality. Soc Sci Med 1993;36:217–25.[CrossRef][Medline]
  12. Rohrbaugh MJ, Shoham V, Coyne JC. Effect of marital quality on eight-year survival of patients with heart failure. Am J Cardiol 2006;98:1069–72.[CrossRef][Medline]
  13. Orth-Gomer K, Wamala SP, Horsten M, Schenck-Gustafsson K, Schneiderman N, Mittleman MA. Marital stress worsens prognosis in women with coronary heart disease: the Stockholm female coronary risk study. JAMA 2000;284:3008–14. [See comment in JAMA 2001;285:1289–90.].
  14. Kimmel PL, Peterson RA, Weihs KL, Shidler N, Simmens SJ, Alleyne S, Cruz I, Yanovski JA, Veis JH, Phillips TM. Dyadic relationship conflict, gender, and mortality in urban hemodialysis patients. J Am Soc Nephrol 2000;11:1518–25.[Abstract/Free Full Text]
  15. Hayes AM, Beevers CG, Feldman GC, Laurenceau JP, Perlman C. Avoidance and processing as predictors of symptom change and positive growth in an integrative therapy for depression. Int J Behav Med 2005;12:111–22.[CrossRef][Medline]
  16. Greenberg L. Emotion in psychotherapy: a practice friendly research review. J Clin Psychol 2006;62:611–30.[CrossRef][Medline]
  17. Kashdan T, Barrios V, Forsyth J, Steger M. Experiential avoidance as a generalized psychological vulnerability: Comparisons with coping and emotion regulation strategies. Behav Res Ther 2006;44:1301–20.[CrossRef][Medline]
  18. Ganz PA, Kwan L, Stanton AL, Krupnick JL, Rowland JH, Meyerowitz BE, Bower JE, Belin TR. Quality of life at the end of primary treatment of breast cancer: first results from the moving beyond cancer randomized trial. J Natl Cancer Inst 2004;96:376–87.[Abstract/Free Full Text]
  19. Tross S, Herndon JN, Korzun A, Kornblith A, Cella D, Holland JF, Raich P, Johnson A, Kiang D, Perloff M, Norton L, Wood L, Holland JC. Psychological symptoms and disease-free and overall survival in women with stage II breast cancer. Cancer and leukemia group B. J Natl Cancer Inst 1996;88:661–7.[Abstract/Free Full Text]
  20. Hislop TG, Waxler NE, Coldman AJ, Elwood JM, Kan L. The prognostic significance of psychosocial factors in women with breast cancer. J Chronic Dis 1987;40:729–35.[CrossRef][Medline]
  21. Giraldi T, Rodani MG, Cartei G, Grassi L. Psychosocial factors and breast cancer: a 6-year Italian follow-up study. Psychother Psychosom 1997;66:229–36.[Medline]
  22. Brown KW, Levy AR, Rosberger Z, Edgar L. Psychological distress and cancer survival: a follow-up 10 years after diagnosis. Psychosom Med 2003;65:636–43.[Abstract/Free Full Text]
  23. Fawzy FI, Fawzy NW, Hyun CS, Elashoff R, Guthrie D, Fahey JL, Morton DL. Malignant melanoma: effects of an early structured psychiatric intervention, coping, and affective state on recurrence and survival 6 years later. Arch Gen Psychiatry 1993;50:681–89.[Abstract/Free Full Text]
  24. Dean C, Surtees PG. Do psychological factors predict survival in breast cancer? J Psychosom Res 1989;33:561–9.[CrossRef][Medline]
  25. Cassileth BR, Walsh WP, Lusk EJ. Psychosocial correlates of cancer survival: a subsequent report 3 to 8 years after cancer diagnosis. J Clin Oncol 1988;6:1753–9. [See comment in J Clin Onctol 1989; 7:541–2].
  26. Jensen MR. Psychobiological factors predicting the course of breast cancer. J Personality 1987;55:317–42.[CrossRef][Medline]
  27. Reynolds P, Hurley S, Torres M, Jackson J, Boyd P, Chen VW. Use of coping strategies and breast cancer survival: results from the black/white cancer survival study. Am J Epidemiol 2000;152:940–9.[Abstract/Free Full Text]
  28. Wortman C, Dunkel-Schetter C. The interpersonal dynamics of cancer: problems in social relationships and their impact on the patient. In: Friedman HS, editor. Interpersonal Issues in Health Care. New York: Academic Press; 1982.
  29. Ell K, Nishimoto R, Morvay T, Mantell J, Hamovitch M. A longitudinal analysis of psychological adaptation among survivors of cancer. Cancer 1989;63:406–13.[CrossRef][Medline]
  30. Sobbing L, Wittkind C, editors. UICC TNM Classification of Malignant Tumors. New York: John Willey & Sons, Inc; 1997.
  31. Politi MC, Enright TM, Weihs KL. The effects of age and emotional acceptance on distress among breast cancer patients. Support Care Cancer 2006;15:73–9.[CrossRef][Medline]
  32. McNair PM, Lorr M, Droppleman L. Profile of Mood States Manual. San Diego, CA: Educational and Industrial Testing Services; 1971.
  33. Roberts CS, Elkins NW, Baile WF Jr, Cox CE. Integrating research with practice: the psychosocial impact of breast cancer. Health Soc Work 1989;14:261–8.[Medline]
  34. Curran SL, Andrykowski MA, Studts JL. Short form of the profile of mood states (POMS-SF): psychometric information. Psychological Assessment 1995;7:80–3.[CrossRef]
  35. Stansfeld S, Marmot M. Deriving a survey measure of social support: the reliability and validity of the close persons questionnaire. Soc Sci Med 1992;35:1027–35.[CrossRef][Medline]
  36. Balslev I, Axelsson CK, Zedeler K, Rasmussen BB, Carstensen B, Mouridsen HT. The Nottingham prognostic index applied to 9,149 patients from the studies of the Danish breast cancer cooperative group (DBCG). Breast Cancer Res Treat 1994;32:281–90.[CrossRef][Medline]
  37. Polychemotherapy for early breast cancer: an overview of the randomized trials. Early breast cancer trialists’ collaborative group. Lancet 1998;352:930–42.[CrossRef][Medline]
  38. Allison P. Missing Data. Thousand Oaks, CA: Sage Publications; 2002.
  39. SAS/STAT 9.1 User’s Guide. Cary, NC; SAS Institute Inc.; 2004.
  40. Cantor A. Extending SAS Survival Analysis Techniques for Medical Research. Cary, NC: SAS Institute; 1997.
  41. Romsaas E, Malec J, Javenkoski B, Trump D, Wolberg W. Psychological distress among women with breast problems. Cancer 1986;57:890–5.[CrossRef][Medline]
  42. Carlson LE, Speca M, Patel KD, Goodey E. Mindfulness-based stress reduction in relation to quality of life, mood, symptoms of stress and levels of cortisol, dehydroepiandrosterone sulfate (DHEAS) and melatonin in breast and prostate cancer outpatients. Psychoneuroendocrinology 2004;29:448–74.[CrossRef][Medline]
  43. Tabachnick B, Fidell L. Using Multivariate Statistics, 5/E. Boston: Allyn & Bacon; 2007.
  44. Lepore SJ, Silver RC, Wortman CB, Wayment HA. Social constraints, intrusive thoughts, and depressive symptoms among bereaved mothers. J Pers Soc Psychol 1996;70:271–82.[CrossRef][Medline]
  45. Cunningham AJ, Edmonds CV, Phillips C, Soots KI, Hedley D, Lockwood GA. A prospective, longitudinal study of the relationship of psychological work to duration of survival in patients with metastatic cancer. Psycho-Oncology 2000;9:323–39.[CrossRef][Medline]
  46. Tardy CH. Self-disclosure and health: revisiting Sidney Jourard’s hypothesis. In: Petronio S, editor. Balancing the Secrets of Private Disclosure. New Jersey and London: Laurence Erlbaum Assoc; 2000.
  47. Lutgendorf SK, Ullrich P. Cognitive processing, disclosure, and health: psychological and physiological mechanisms. In: Lepore S, Smyth J, editors. The Writing Cure. Washington, DC: American Psychological Association Press; 2002.
  48. Mark EW, Alper TG. Women, men and intimacy motivation. Psychology of Women Quarterly 1985;9:81–8.[CrossRef]
  49. Baron RM, Kenny DA. The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. J Pers Soc Psychol 1986;51:1173–82.[CrossRef][Medline]
  50. Wood WC, Budman DR, Korzun AH, Cooper MR, Younger J, Hart RD, Moore A, Ellerton JA, Norton L, Ferree CR, Ballow AC, Frei E, Henderson IC. Dose and dose intensity of adjuvant chemotherapy for stage II, node positive breast carcinoma. N Engl J Med 1994;330(18):1253–9.[Abstract/Free Full Text]
  51. Kravdal O. Children, family and cancer survival in Norway. Int J Cancer 2003;105:261–6.[CrossRef][Medline]
  52. Weihs K, Enright TM, Howe G. Marital Satisfaction and emotional adjustment to breast cancer. J Psychosoc Oncol 1999;17:33–49.[CrossRef]
  53. Helgeson VS, Snyder P, Seltman H. Psychological and physical adjustment to breast cancer over 4 years: identifying distinct trajectories of change. Health Psychol 2004;23:3–15.[CrossRef][Medline]



This article has been cited by other articles:


Home page
 The GerontologistHome page
S. Ashida, A. E. L. Palmquist, K. Basen-Engquist, S. E. Singletary, and L. M. Koehly
Changes in Female Support Network Systems and Adaptation After Breast Cancer Diagnosis: Differences Between Older and Younger Patients
Gerontologist, August 1, 2009; 49(4): 549 - 559.
[Abstract] [Full Text] [PDF]

Home page
 JAMAHome page
J. C. Harris and C. D. DeAngelis
The Power of Hope
JAMA, December 24, 2008; 300(24): 2919 - 2920.
[Full Text] [PDF]

Home page
 Psychosom. Med.Home page
J. C. Coyne and B. D. Thombs
WAS IT SHOWN THAT "CLOSE RELATIONSHIPS AND EMOTIONAL PROCESSING PREDICT DECREASED MORTALITY IN WOMEN WITH BREAST CANCER?" A CRITIQUE OF WEIHS ET AL. (2008)
Psychosom Med, July 1, 2008; 70(6): 737 - 738.
[Full Text] [PDF]

Home page
 Psychosom. Med.Home page
K. L. Weihs, T. M. Enright, and S. J. Simmens
RESPONSE
Psychosom Med, July 1, 2008; 70(6): 738 - 739.
[Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Weihs, K. L.
Right arrow Articles by Simmens, S. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Weihs, K. L.
Right arrow Articles by Simmens, S. J.
Related Collections
Right arrow Social Support
Right arrow Cancer

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS