Psychosomatics 49:407-412, September-October 2008
doi: 10.1176/appi.psy.49.5.407
© 2008 Academy of Psychosomatic Medicine
Breast Cancer and Psychosocial Factors: Early Stressful Life Events, Social Support, and Well-Being
Karni Ginzburg, Ph.D.,
Margaret Wrensch, Ph.D., M.P.H.,
Terri Rice, M.P.H.,
Georgianna Farren, M.D., and
David Spiegel, M.D.
Received July 18, 2006; revised February 5, 2007; accepted February 9, 2007. From the Bob Shapell School of Social Work, Tel Aviv University, Tel Aviv, Israel; Depts. of Neurological Surgery and Epidemiology and Biostatistics, Univ. of California, San Francisco; the Dept. of Psychiatry and Behavioral Sciences, Stanford Univ. School of Medicine, Palo Alto, CA. Send correspondence and reprint requests to Karni Ginzburg, Ph.D., Bob Shapell School of Social Work, Tel Aviv University, Tel Aviv, Israel. e-mail: karnig{at}post.tau.ac.il
© 2008 The Academy of Psychosomatic Medicine
|
ABSTRACT
|
BACKGROUND: The allostasis theory postulates that stress causes the body to activate physiologic systems in order to maintain stability. OBJECTIVE: The authors sought to examine the relationship between earlier stress and later development of breast cancer (BC). METHODS: Authors correlated discrete and interactive relationships of stressful life events, social support, and well-being during childhood and adolescence with the occurrence of BC in adulthood among 300 women with primary BC and 305 matched control subjects. RESULTS: BC patients and control subjects reported similar childhood experiences. Yet, although childhood stressful life events were associated with reports of less family support and well being among the controls, those in the BC group who experienced high stress in early childhood actually expressed higher levels of family support and well-being than did those who had experienced lower levels of stress. CONCLUSION: These findings may reflect a tendency toward a repressive coping style among the BC group, which may be either a risk factor for the disease or a result of having it.
|
INTRODUCTION
|
According to the allostasis theory,1,2 exposure to stress causes the body to activate physiologic, allostatic systems in order to maintain stability through change and adapt to the threat. Through allostasis, various physiological systems, for example, the autonomic nervous system, the hypothalamic-pituitary-adrenal (HPA) axis, and the cardiovascular, metabolic, and immune systems, react to the stress in order to facilitate individual response and adaptation to the stressors. Experience of chronic stress may result in increased allostatic load, which refers to the physiological price of repeated or prolonged activation of the allostatic systems. It has been suggested that this price may be implicated in the acceleration of disease processes.3
The implications of exposure to stress,4 presence of social support,5 the stress reaction,6 and the development of breast cancer (BC) have been extensively studied, yielding conflicting findings. One potential contributor to the mixed findings is the time-range of the studies. Breast cancer is a slow-growing tumor with a long subclinical phase that, in some women, extends to 18 years or more.7 Thus, Gertis8 notes that most studies aiming to examine stress, social support, and distress as risk factors for BC actually assess the contribution of these psychosocial factors to cancer-progression during the subclinical stage. Indeed, studies of environmental risk factors for BC have pointed to the need to examine the important influence of childhood or adolescent exposure or experience.9
According to McEwen,2 early experiences of chronic stress are risk factors for allostatic load later in life, mainly through alteration in HPA-axis and autonomic nervous system functioning, thus rendering the individual susceptible to the development of various diseases; this is consistent with animal studies, as well.10
Indeed, a number of studies have found associations between early negative life experiences and the development of BC in adulthood. More specifically, maternal death in childhood5,11 and traumatic childhood experiences of World War II5 were found to be more prevalent among women with BC than among control subjects. Jacobs and Bovasso11 report that an early depressive episode at least 20 years prediagnosis is associated with increased risk for BC, unrelated to the impact of maternal loss. Grassi and Molinari12 found that BC patients tended to describe their childhood relationship with their mothers as less close and less demonstrative than did controls. Muslin et al.13 found no association between BC incidence and early experiences of loss and separation. On the whole, however, the studies suggest that early experiences may play a role in the development of cancer in adulthood.
The present study compares the early experiences of women with BC and matched control subjects. More specifically, it compares their exposure to stressful life events during the first 6 years of life, the perceived supportiveness of their nuclear family, and their well-being in childhood and adolescence, in order to examine the association of each of these factors with BC incidence, both singly and in interaction. On the basis of the few studies that pointed to the relevance of early experiences and environment to the development of adult cancer, we hypothesized that women with BC would report higher levels of stress, lower levels of support, and lower levels of well-being in their youth than the comparable control subjects.
|
METHODS
|
Sample and Procedure
Breast Cancer (BC) Group
Eligible cases included any female resident of a single Northern California county who had a diagnosis of primary BC between September 1997 and June 1999 if under age 50, and between July 1997 and March 1999 if age 50 or older at diagnosis (N=401); 15 eligible women had died by the time of the study; 50 (13%) were not interviewed either because they refused to participate in the study (N=36) or because of other circumstances (N=14); 36 women completed only the telephone interview, which consisted of a subset of the full interview; 300 women completed the full interview. Details of subject ascertainment, recruitment, and interview were described by Wrensch et al.14
Control Sample
Women who were never diagnosed with BC, from the same county, frequency-matched to the BC group for age and ethnicity, were recruited through random-digit dialing;14 3,945 random-digit numbers were called: 1,640 (41%) either were not in service, were a business, fax, or modem number, or were from another county; 920 (23%) were reached, but no resident was eligible; 289 (7%) refused with no further information; 733 phone numbers (19%) were in service but were either busy or did not answer; and 347 (9%) yielded an eligible subject. Of the 347 women identified as eligible, 328 were contacted; of these, 7 women (2%) refused to participate in the study, and 305 women completed the full interview.
The data were collected, on average, 2 years after the BC diagnosis (mean: 2.28; standard deviation [SD]: 0.45). The age of the control group was similar to the age of the BC group at the time of diagnosis (mean: 55.95; SD: 9.50); thus, at the time of the interview, the BC group was somewhat older than the controls (mean: 58.27; [9.55] versus 55.69 [9.69]; t[603]=3.30; p<0.001). The BC and the control groups did not differ in race, level of education, current socioeconomic status, religious affiliation, or number of births.14
Some data collected among the women from the BC group who refused to participate in the study (N=36) indicated that they were similar to the participants from the BC group in ethnic background and in age at diagnosis. Those who refused to participate in the study were, however, somewhat less educated: 3 women (8%) did not finish high school (versus 1% of the participants from the BC group), and 11 (31%) had a high school education (versus 14% of BC participants).
Interviews
The data were gathered by in-person interviews. Because many questions required the recall of early experiences, the interview used three memory tools: a visualization exercise, photographs of girls at various stages of development in a variety of social settings, and a magnetic memory board.14
The University of California, San Francisco Committee on Human Research (CHR) approved the original study; and they, along with Stanford University’s Institutional Review Board for Human Studies, approved the analytic methods for this report. Informed consent was obtained from all participants.
Measures
The questionnaire, based on published sources,15–22 asked detailed questions about subjects’ early experiences.
Stressful Life Events
The scale comprised 24 items, each describing a major life event (e.g., death of a sibling, parents’ divorce or separation). Participants were asked to indicate whether they experienced this event and to specify the age at which they experienced it. The number of stressful events experienced between birth and age 6 reflected the accumulation of life events.
Family Support
This scale comprised 16 items, tapping various aspects of family support and sense of belonging (e.g., "Did you have intense sibling rivalry? Was there someone who hugged you?"). Participants were asked to answer the questions with respect to their childhood and adolescence on a 5-point scale. A mean score was calculated, such that a higher score indicates that the participant reported a more supportive family. The high Cronbach alpha (0.89) indicated high internal consistency.
Family Environment
This scale consisted of six items describing the family environment (e.g., critical; supporting and encouraging). The respondent was asked to rate, on a 5-point scale, the extent to which each item applied to her experiences in childhood and adolescence. A mean score was calculated, such that a higher score indicates the woman reported a more positive the family environment. The high Cronbach alpha (0.90) supported the internal consistency of this measure.
Other Social Support Networks
The availability of other social support networks in adolescence was reflected by involvement in church or other religious activities, organized youth groups (such as cheerleading or Girl Scouts), and community volunteer organizations. Respondents were asked to rate, on 5-point scale, the extent to which they were involved in each of these activities during adolescence. A mean score was calculated such that a higher score indicated that the participant reported more involvement in social support networks.
Well-Being
The scale for well-being consisted of 12 questions (e.g., "Were you happy? Did you feel lonely or remote from other people?"). Each question was answered three times, once with respect to junior high school, once with respect to high school, and once with respect to the years after high school up to age 21. The respondent was asked to answer the questions on a 5-point scale for each period of time. A mean score was calculated such that a higher score indicated greater well-being. The high Cronbach alpha (0.90) supported the internal consistency of this measure.
|
RESULTS
|
Stressful Life Events, Social Support, and Well-Being
The t-tests for independent samples indicated that the two groups did not differ in the number of childhood stressful life events (p=0.68) The two groups were also similar in their reports of family supportiveness (p=0.83), family environment (p=0.99), and involvement in social support networks (p=0.43). Finally, the two study groups did not differ in their reports of well-being during adolescence (p=0.40).
Relationship Between Stressful Life Events, Social Support, and Well-Being
Correlations between stressful life events, the availability of family and other support resources, and well-being in the two study groups revealed the following patterns:
1. The two groups show differential patterns of associations between level of exposure to childhood life events and ratings of social support. Major life events experienced between birth and age 6 were inversely associated with measures of family supportiveness, family environment, and the availability of other social networks in the control group (r = –0.27, p<0.001; r = –0.24, p<0.001; r = –0.13, p<0.05; respectively) but not in the BC group (r = –0.11, r = –0.08, r=0.00, respectively). This differential pattern of correlations was supported by Fisher’s z tests of differences between independent-sample correlations, indicating that the correlations of childhood life events with family support (F[1, 604]=3.92; p<0.05) and with family environment (F[1, 604]=3.63; p=0.057), are significantly different for the two groups.
2. In both study groups, there are strong positive significant correlations (0.54–0.69, p<0.001) between both family supportiveness and family environment with well-being. Similarly, in both study groups, there are moderate positive associations between involvement in social networks and well-being (0.26–0.27, p<0.001). That is, the more intense the respondent’s involvement in social support networks, the higher was her perceived well-being during adolescence.
3. BC patients and controls show differential patterns of correlations between childhood life events and well-being. Among the controls, well-being is inversely associated with exposure to stressful life events (r = –0.29, p<0.001). In the BC group, however, this association was not significant (r=0.07; Fisher’s z tests of differences between independent-sample correlations: F[1, 604]=7.28; p<0.01).
To examine whether the source of the differential patterns of associations among the study variables in the BC and control groups can be attributed to the level of childhood exposure to stressful events, the subjects were divided into three subgroups according to their level of exposure: The low-stress group consisted of respondents who reported none of the major stressful life events between ages 0 and 6 (N=324); the moderate-stress group consisted of the respondents who reported 1–2 stressful life events during those years (N=243); and the high-stress group consisted of those who reported experiencing 3 or more stressful life events (N=38).
A MANOVA with Group (BC, control) and Stress subgroup (low-stress, moderate-stress, high-stress) as the independent variables, and the Support measures (family support, family environment, social networks) and well-being as the dependent variables, yielded a significant overall interaction (F[12, 1,190]=2.87; p<0.01). A series of two-way ANOVAs revealed significant interactions between group and stress-level during childhood on family support (p<0.001; Figure 1), family environment (p<0.05; Figure 2), and well-being (p<0.001; Figure 3).
View larger version (26K):
[in this window]
[in a new window]
|
FIGURE 1. Association of Family Support (y axis) and Childhood Stressful Life Events in Breast Cancer Patients and Control Subjects
Scale comprised 16 items, on a 5-point scale.
|
View larger version (26K):
[in this window]
[in a new window]
|
FIGURE 2. Association of Family Environment (y axis) and Childhood Stressful Life Events in Breast Cancer Patients and Control Subjects
Scale comprised 6 items, on a 5-point scale.
|
View larger version (30K):
[in this window]
[in a new window]
|
FIGURE 3. Association of Well-Being (y axis) and Childhood Stressful Life Events in Breast Cancer Patients and Control Subjects
Scale comprised 12 items, on a 5-point scale.
|
As shown, the source of the differences between the BC patients and the controls is in the high-stress subgroup among the BC patients. In the control group, as the level of reported stressful events during childhood increased, reported family support and well-being decreased. This pattern is only partially observed in the BC group. The BC patients with moderate levels of stressful events reported lower levels of support and well-being than those who reported fewer stressful events. However, those with the highest number of childhood life events reported higher, rather than lower, levels of support and well-being than the low- and moderate-stress subgroups.
|
DISCUSSION
|
Contrary to the hypothesis inferred from the theory of allostasis, neither stress, nor social support, nor well-being in childhood and adolescence differentiated the women who developed breast cancer from those who did not. These findings are not consistent with the reports of association between childhood adversity and increased risk of breast cancer.5,11,12 The lack of support for the hypothesis may be attributed to the limitations of the study, especially its retrospective design. It may, however, imply a more complex association between allostatic load and the development of breast cancer. According to Nielsen et al.,6 although allostatic load may accelerate other disease processes, the persistent activation of the HPA axis and the sympathetic nervous system may specifically protect from breast cancer, through the suppression of estrogen secretion. This speculation is contradicted, however, by the finding that an abnormal diurnal pattern of cortisol predicts earlier mortality with breast cancer, independent of other known risk factors.23
The study did, however, find significant interactions between some of these variables and case–control status. Control subjects who reported more childhood stressful life events also reported lower levels of social support and well-being, whereas the cancer patients who reported more stressful life events did not. Moreover, the BC patients with the highest level of childhood stressful events reported greater well-being and more social support than the other two subgroups.
The inverse relationship between well-being and stressful life events found among the control subjects makes intuitive sense; the decline in social support as negative life events increase might be explained by the fact that early stressful events typically affect the entire family and thereby reduce the resources available for social support.24 The pattern evident in the responses of the BC patients is open to interpretation. One possibility is that the women with BC were particularly resilient—they did not appear to allow stressful events to impair their well-being—and had especially supportive family systems that were not affected by the stressful experiences. Alternatively, this pattern may reflect a stress-specific repression response. This interpretation is consistent with previous findings that women with BC tend to exhibit higher levels of emotional repression than healthy-controls and greater control over emotional expression in general and the expression of anger, in particular.4,25,26
The hypothesized repressive response seems to be specific to high levels of stress. On the whole, the women with BC drew a picture of their early life that was similar to that of the control subjects. The unique pattern of relations between social support and well-being was observed only among those who experienced high levels of early stress. This hypothesis is supported by Pettingale et al.’s26 findings that women with BC tend to restrict their emotional expression after a stress manipulation, but not after an emotionally neutral task.
Because of the study’s retrospective design, we cannot know whether this pattern of response is a risk factor for BC or a result of having the disease. There is evidence that there is a cancer-prone personality.27,28 Alternatively, it may be that the repressive response observed among the high-stress BC subjects stems from their cancer. These women, preserving their emotional resources for the challenge of coping with and adjusting to their current cancer-related difficulties, either repress their emotions or adopt a positive perspective on their early experiences. This hypothesis is supported by reports of increased repression after receiving a diagnosis of breast cancer.29 The two causal hypotheses are not mutually exclusive, because the same repressive tendency that may be a risk factor for BC may also be strengthened and solidified after the highly stressful experience of cancer diagnosis and treatment.
The study findings raise two key questions: One concerns the exact nature of the observed reaction of the high-stress patient subgroup. Since we did not examine defenses directly, we cannot know whether their reactions reflect their tendency to suppress negative affects of which they are consciously aware or the more fundamental mechanism of repression, in which one is unaware of the existing affect. In a study of metastatic BC patients, Giese-Davis and Spiegel30 demonstrated that these are separate mechanisms.
The other question is whether the reaction to stress found in this study is unique to BC patients, or extends to cancer patients in general, or to anyone with a life-threatening illness. Although most of the research on psychosocial factors in cancer focuses on breast cancer, there is evidence of the tendency to repression in patients with other types of cancer, as well.31 Other studies suggest that the tendency to react to stress with repression is a cancer-specific characteristic, not found in other illnesses.32
Although they require further validation, the findings of this study have clinical implications. Repressive behavior can have both positive and negative implications. On the positive side, studies have associated repression with heightened well-being under trying circumstances.33 On the negative side, findings suggest that persons who make more use of repression are less engaged with health-promoting behaviors and more prone to delay medical help-seeking than non-repressors.34 Findings that repressors tend to prefer problem-focused to emotion-focused coping mechanisms35 raise questions about whether they would avail themselves of supportive interventions, which have been shown to have positive effects on cancer patients’ well-being and psychosocial functioning.36 Clinicians who treat BC patients should be aware of the range of effects that repressive tendencies can have and should proceed accordingly.
|
ACKNOWLEDGMENTS
|
Data collection was funded by grant 5BB-1201 from the California Breast Cancer Research Program.
|
REFERENCES
|
- McEwen BS: Protective and damaging effects of stress mediators. New Engl J Med 1998; 338:171–179[Free Full Text]
- McEwen BS: The neurobiology of stress: from serendipity to clinical relevance. Brain Res 2000; 886:172–189[CrossRef][Medline]
- Johnston-Brooks CH, Lewis MA, Evans GW, et al: Chronic stress in childhood: the role of allostatic load. Psychosom Med 1998; 60:597–603[Abstract/Free Full Text]
- McKenna MC, Zevon MA, Corn B, et al: Psychosocial factors and the development of breast cancer: a meta-analysis. Health Psychol 1999; 18:520–531[CrossRef][Medline]
- Scherg H, Blohmke M: Associations between selected life events and cancer. Behav Med 1988; 14:119–124[Medline]
- Nielsen NR, Zhang Z, Kristensen TS, et al: Self-report stress and risk of breast cancer: prospective cohort study. BMJ 2005; 331:548[Abstract/Free Full Text]
- Friberg S, Mattson S: On the growth rate of human malignant tumors: implications for medical decision-making. J Surg Oncol 1997; 65:284–297[CrossRef][Medline]
- Gertis P: Life events, coping, and breast cancer: state of the art. Biomed Pharmacother 2000; 54:229–233[CrossRef][Medline]
- Henderson BE, Pike MC, Bernstein L, et al: Breast cancer, in Cancer Epidemiology and Prevention. Edited by Schottenfeld D, Faumeni JF. New York, Oxford Univ. Press, 1996, pp 1022-1039
- Weaver IC, Riorio J, Seckl JR, et al: Early environmental regulation of hippocampal glucocorticoid receptor gene expression: characterization of intracellular mediators and potential genomic target sites. Ann N Y Acad Sci 2004; 1024:182–212[CrossRef][Medline]
- Jacobs JR, Bovasso GB: Early and chronic stress and their relation to breast cancer. Psychol Med 2000; 30:669–678[CrossRef][Medline]
- Grassi L, Molinari S: Family affective climate during the childhood of adult cancer patients. J Psychosocial Oncol 1986; 4:53–62
- Muslin HL, Gagrfas K, Pieper WJ: Separation experience and cancer of the breast. Ann N Y Acad Sci 1966; 125:802–806[Medline]
- Wrensch M, Chew R, Farren G, et al: Risk factors for breast cancer in a population with high incidence rates. Breast Cancer Res 2003; 5:R88-R102
- Holmes TH, Rahe RH: The Social Readjustment Rating Scale. J Psychosom Res 1967; 11:213–218[CrossRef][Medline]
- Yeaworth RC, York J, Hussey MA, et al: The development of an adolescent life-change event scale. Adolescence 1980; 15:91–97[Medline]
- Burnett PC, Fanshawe JP: Measuring school-related stressors in adolescents. J Youth Adolesc 1997; 26:415–428[CrossRef]
- Compas BE, Davis GE, Forsythe CJ, et al: Assessment of major and daily stressful events during adolescence: The Adolescent Perceived Events Scale. J Consult Clin Psychol 1987; 55:534–541[CrossRef][Medline]
- Mullis RL, Youngs A Jr, Mullis AK, et al: Adolescent stress: issues of measurement. Adolescence 1993; 28:267–279[Medline]
- Reynolds P, Boyd PT, Blacklow RS: The relationship between social ties and survival among black and white breast cancer patients. Cancer Epidemiol Biomarkers Prev 1994; 3:253–259[Abstract]
- Berkman LF, Syme SL: Social networks, host resistance, and mortality: a nine-year follow-up study of Alameda County residents. Am J Epidemiol 1979; 109:186–204[Abstract/Free Full Text]
- Seeman TE, Syme SL: Social networks and coronary artery disease: a comparison of the structure and function of social relations as predictors of disease. Psychosom Med 1987; 49:341–354[Abstract/Free Full Text]
- Sephton SE, Sapolsky R, Kraemer HC, et al: Diurnal cortisol rhythm as a predictor of breast cancer survival. J Nat Cancer Inst 2000; 92:994–1000[Abstract/Free Full Text]
- Gilmer WS, McKinney WT: Early experience and depressive disorders: human and non-human primate studies. J Affect Disord 2003; 75:97–113[CrossRef][Medline]
- Cooper CL, Faragher EB: Psychosocial stress and breast cancer: the interrelationship between stress events, coping strategies, and personality. Psychol Med 1993; 23:653–662[Medline]
- Pettingale KW, Watson M, Greer S: The validity of emotional control as a trait in breast cancer patients. J Psychosocial Oncol 1984; 2:21–30[CrossRef]
- Servaes P, Vingerhoets AJJM, Vreugdenhil G, et al: Inhibition of emotional expression in breast cancer patients. Behav Med 1999; 25:23–27[Medline]
- Bleiker EM, van der Ploeg HM, Hendriks JH, et al: Personality factors and cancer development: a prospective longitudinal study. J Nat Cancer Inst 1996; 88:1478–1482[Abstract/Free Full Text]
- Kreitler S, Chaitchik S, Kreitler H: Repressiveness: cause or result of cancer? Psychooncology 1993; 2:43–54[Medline]
- Giese-Davis J, Spiegel D: Suppression, repressive-defensiveness, restraint, and distress in metastatic breast cancer: separable or inseparable constructs? J Pers 2001; 69:417–449[CrossRef][Medline]
- Phipps S, Srivastava DK: Repressive adaptation in children with cancer. Health Psychol 1997; 16:521–528[CrossRef][Medline]
- Kneier AW, Temoshok L: Repressive coping reactions in patients with malignant melanoma as compared to cardiovascular disease patients. J Psychosom Res 1984; 28:145–155[CrossRef][Medline]
- Ginzburg K, Solomon Z, Bleich A: Repressive coping style, acute stress disorder, and posttraumatic stress disorder after myocardial infarction. Psychosom Med 2002; 64:748–757[Abstract/Free Full Text]
- Weinberger DA: The construct validity of repressive coping style, in Repression and Dissociation: Implications for Personality Theory, Psychopathology, and Health. Edited by Singer JL. Chicago, IL, Univ. of Chicago Press, 1990, pp 337-386
- Myers LB, Derakshan N: Monitoring and blunting and an assessment of different coping styles. Pers Indiv Diff 2000; 28:111–121[CrossRef]
- Spiegel D, Bloom JR, Kraemer HC, et al: Effect of psychosocial treatment on survival of patients with metastatic breast cancer. Lancet 1989; 2:888–891[CrossRef][Medline]
Get information about faster international access.
Privacy Policy
Copyright © 2008
Academy of Psychosomatic Medicine.
All rights reserved.
Home
| Search
| Current Issue
| Past Issues
| Subscribe
| All APPI Journals
| Help
| Contact Us
|
TOP
ABSTRACT
INTRODUCTION
