Stress reactivity, distress and attachment in newly diagnosed breast cancer patients

Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=rhpb20 Download by: [b-on: Biblioteca do conhecimento online UL] Date: 22 January 2016, At: 05:36 Health Psychology and Behavioral Medicine An Open Access Journal ISSN: (Print) 2164-2850 (Online) Journal homepage: http://www.tandfonline.com/loi/rhpb20 Stress reactivity, distress and attachment in newly diagnosed breast cancer patients Silvia Ouakinin, Susana Eusebio, Marco Torrado, Hugo Silva, Isabel Nabais, Graça Gonçalves & Leonor Bacelar-Nicolau To cite this article: Silvia Ouakinin, Susana Eusebio, Marco Torrado, Hugo Silva, Isabel Nabais, Graça Gonçalves & Leonor Bacelar-Nicolau (2015) Stress reactivity, distress and attachment in newly diagnosed breast cancer patients, Health Psychology and Behavioral Medicine, 3:1, 424-438, DOI: 10.1080/21642850.2015.1121491 To link to this article: http://dx.doi.org/10.1080/21642850.2015.1121491 © 2015 The Author(s). Published by Taylor & Francis. Published online: 31 Dec 2015. Submit your article to this journal Article views: 42 View related articles View Crossmark data Stress reactivity, distress and attachment in newly diagnosed breast cancer patients Silvia Ouakinin a *, Susana Eusebio a , Marco Torrado a , Hugo Silva b , Isabel Nabais c , Graça Gonçalves c and Leonor Bacelar-Nicolau a a Faculty of Medicine, University of Lisbon, Lisbon, Portugal; b Institute of Telecommunications - Technical Institute, University of Lisbon, Lisbon, Portugal; c Santa Maria Hospital, Lisbon, Portugal (Received 15 June 2015; accepted 14 November 2015) Research on psycho-oncology increased across literature during the last decades, pointing to links between biological, psychosocial and behavioural factors in cancer beginning and progression. This study aimed to characterize a sample of recently breast-cancer-diagnosed women, compared to a control group, regarding their stress reactivity at a psychological and autonomic levels, anger regulation and attachment styles. Eighty-seven females (52 breast cancer patients and 33 controls) respectively from Portuguese public hospitals and general population were recruited. They were assessed through psychometric measures (distress, attachment styles and anger regulation) and psychophysiological parameters of reactivity were collected. The breast cancer patients studied seem to be less anxious in their attachment patterns compared to healthy people, but they report significant distress while facing a threatening situation. This clinical group also shows lower psychophysiological reactivity, both at the baseline and confronted with different emotional eliciting stimuli. Self-directed anger was associated with the presence of anxious attachment schemes and strictly linked to the appraised distress. A predictive model suggests the impact of this pattern of anger management and an anxious attachment style in the emotional disturbance reported by these patients. Findings suggest that insecure attachment schemes, dysfunctional anger regulation strategies and a lack of psychophysiological activation may be discussed as relevant factors that modulate emotional distress associated with the diagnosis of breast cancer. Keywords: distress; attachment; emotional regulation; psychophysiological reactivity; breast cancer Introduction A cancer diagnosis can be conceptualized as a major threat, affecting survival expectancy along with social and personal identity. Emotional distress, such as anxiety, depression and other emotional regulation disorders arising from illness experience, has been increasingly associated to disease outcomes, as well as to patients’quality of life. Moreover, psychological aspects © 2015 The Author(s). Published by Taylor & Francis. *Corresponding author. Email: [email protected] This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/ licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Health Psychology and Behavioral Medicine, 2015 Vol. 3, No. 1, 424–438, http://dx.doi.org/10.1080/21642850.2015.1121491 Downloaded by [b-on: Biblioteca do conhecimento online UL] at 05:36 22 January 2016 seem to be essential, not only in the referred sense, but as an undeniable part of the disease, understanding from a pathophysiological level to a personal and to a bioecological level (Linden & Girgis, 2012; Lutgendorf, Costanzo, & Siegel, 2007). Research on psycho-oncology increased in the last decades, pointing to links between biological, psychosocial and behavioural factors in cancer initiation and progression (Lillberg et al., 2003; Spiegel, 2012; Stommel, Given, & Given, 2002). Supporting data from several theoretical frameworks, including stress and distress investigations, development, emotion regulation and epigenetics (Cole, 2013; McEwen, 2008), are unveiling the possible impact of distress, negative emotions, life events and stress response, on macro and microenvironmental aspects at a cellular and systemic level (McDonald, O’Connell, & Lutgendorf, 2013). In cancer research, Psychoneuroimmunology advances play an important role, contributing to clarify the relations between tumoural cells’development or survival and individual determinants of vulnerability. Such a model highlights the stressful nature of social, interpersonal or intraindividual threats, personal resources and the activation of communication mediators between central nervous system and the immune system (Kiecolt-Glaser, Robles, Heffner, Loving, & Glaser, 2002). Neuroendocrine organs and pathways, including the autonomic nervous system and hypothalamic–pituitary– adrenal axis, are responsible for the production of inflammatory mediators and stress hormones which, acting in the tumour microenvironment, can influence tumour growth and disease progression (Armaiz-Pena, Cole, Lutgendorf, & Sood, 2013). Stress, loneliness and social support have a strong impact on health and disease, determining morbidity and mortality risks, probably through the complex interplay of mind–body connections across development (Lutgendorf & Sood, 2011). At an individual level, attachment styles, relatively stable throughout life, seem to influence stress responses in adulthood and how emotion regulation strategies are acquired through development (Cassidy, 1994). The first internal working models, as implicit memories, modulate future attachment styles. Although genetic issues influence these schemes, they are experience-dependent and express complex ways of learning through relations across development, which modulate neural networks with strong connections to orbitofrontal cortex, limbic system and other circuits involved in arousal and emotion processing (Schore, 2000,2001). Schore and Schore (2008) have described, along an extensive framework, how these networks are regulated by early attachment relations in which feelings of comfort and safeness, or anxiety and fear, become strongly related to internal working models of self and others. The secure attachment schemes, associated with positive emotions, provide a biochemical milieu that supplies a good self-regulation and autonomy. Inversely, insecure attachment schemes related to negative emotional experiences seem to be associated with physical and mental disorders (Cozolino, 2006). The pathways connecting early stressful adversity to the quality of health mechanisms in adulthood include many-sided components, from physiological to behavioural dimensions, and growing evidences suggest that immune dysregulation is one of eventual areas linking early life experiences to physical and mental disorders (Fagundes, Glaser, & Kiecolt-Glaser, 2013; Maunder & Hunter, 2001,2008). These overall aspects have psychophysiological correlates and support the mechanisms by which individuals regulate their emotions in adulthood and deal with stressful events. Thus, whereas secure attachment sponsor a good integration of the physiological and cognitive dimensions of arousal, increasing the quality of the coping mechanisms in response to stress, insecure attachment styles or even disorganized ones seem related to behavioural, emotional and neurophysiological dysregulation that may establish future vulnerabilities for managing stress (Mikulincer & Florian, 1998; Mikulincer, Shaver, & Pereg, 2003). Despite some inconsistent evidences, current investigation supports that, in breast cancer patients, patterns of stress reactivity (autonomic activation and abnormal cortisol profiles) are Health Psychology and Behavioral Medicine 425 Downloaded by [b-on: Biblioteca do conhecimento online UL] at 05:36 22 January 2016 linked to depression, emotional distress and a worse prognosis of the disease (Wu, Yang, Thayer, & Andersen, 2014). Childhood adversity, determining a lack of personal resources and psychophysiological dysfunction, can be a major source of vulnerability in patients (Crosswell, Bower, & Ganz, 2014) as it has been evident in different populations and diseases (Shonkoff, Boyce, & McEwen, 2009; Slopen, Non, Williams, Roberts, & Albert, 2014). Regarding the above-mentioned links, this research aimed to characterize autonomic reactivity and emotional distress, as well as attachment styles, in recently diagnosed breast-cancer patients compared to a healthy control group. It was hypothesized that distress levels will be higher in patients, compared to controls. The quality of attachment styles and emotion regulation strategies, such as dysfunctional internal working models and anger expression or control modalities, may determine the experience of distress and psychophysiological activation facing diagnosis. Methods Participants The research was carried out in two public hospitals in Lisbon, Portugal, and both Ethics committees from the Centro Hospitalar Lisboa Norte and Centro Hospitalar Lisboa Ocidental approved it. The sample included recently diagnosed patients with breast cancer (BCP) and control subjects from public institutions of health and educational sectors, both recruited in a voluntary modus, after explaining the objectives and procedures. Patients were evaluated after contacting the surgeon/gynaecologist that informed the patient of the diagnostic of breast cancer and confirmed the inclusion criteria. The participants were invited to take part in the study complying with the following inclusion criteria: (a) diagnosed within the previous 6 months, (b) sufficient language abilities, which was made explicit along with the diagnostic criteria in the short telephone contact, (c) aged between 30 and 55 years, (d) being premenopausal and (e) no mental disorder or other present physical disease. Regarding the previous psychopathology, it was assessed by self-report and participants were asked if they were ever followed by a psychiatrist or clinical psychologist. It was also asked if they had a psychiatric diagnosis and were medicated with psychotropics. Controls were sampled in a convenience design, reporting the absence of any physical or mental disease and declaring no current medication intake. All participants gave their written informed consent before being admitted to the study. Data collection and measures Patients and controls were assessed through a psychopathology inventory (Brief Symptoms Inventory (BSI)), the Distress Thermometer, the Adult Attachment Scale and the State-Trait Anger Expression Inventory (STAXI). In patients, the evaluation was planned for the week before treatment beginning. Evaluation included the completion of psychological questionnaires for about 40 minutes and a structured interview for 20 minutes using a Socio-demographic and Medical Questionnaire, while psychophysiological data were recorded. Brief Symptom Inventory:BSI (Derogatis, 1993) is an inventory of psychopathological symptoms with 53 items and is derived from the short form of the Symptom Checklist-90 (SCL-90) from the same author, in which the subject is asked to rate the degree in which every problem mentioned affected him during past week, using a Likert scale which classifies the responses from “never”(0) to “Very often”(4). A Global Symptoms Index (GSI) is provided by this 426 S. Ouakinin et al. Downloaded by [b-on: Biblioteca do conhecimento online UL] at 05:36 22 January 2016 instrument and it is a well-established score of reported distress. It is extracted as the average of distress ratings assigned to each symptom and assessment of nine subscales –Somatization, Obsessive-compulsive, Interpersonal Sensitivity, Phobic Anxiety, Paranoid Ideation, Psychotic Ideation, Depression, Hostility and Additional Subscales. For this study only, the GSI was considered. In this research, a Portuguese version with adequate psychometric properties was used (Canavarro, 1997,1999). The Distress Thermometer,DT (Roth et al., 1998), is a visual analogue scale –ranging from 0 (no distress) to 10 (extreme distress) –created to provide self-report of emotional distress in cancer patients. A Portuguese translation of the scale was carried out in the present study, duly allowed by the National Comprehensive Cancer Network. The participants were invited to select the number which described better the overall level of distress they experienced in the past week. For assessing attachment schemes, the Portuguese version of the Revised Adult Attachment Scale was used (Canavarro, 1999). This scale is composed of 18 items and 3 subscales (close, dependent and anxiety). The ‘close’scale measures the extent to which a person is comfortable with closeness and intimacy. The ‘dependent’scale measures the extent to which a person feels that he/she can depend on others to be available when needed. The ‘anxiety’subscale measures the extent to which a person is worried about being abandoned or unloved, reflecting the internal working model of self. Following Collins recommendations (Brennan, Clark, & Shaver, 1998; Collins, 2008), we used an alternative score system, measuring two attachment dimensions –attachment anxiety (model of self) and attachment avoidance (model of other). The STAXI (Spielberger, 1999) is a self-assessment scale, coded on a 4-point scale (0–3). The STAXI is a self-assessment scale, coded on a 4-point scale (0–3) to assess anger experience and expression. The Portuguese-adapted version used (Silva, Campos, & Prazeres, 1999) included 44 items, divided in six scales measuring different construct dimensions and an Anger Expression Index. The scale State Anger dimension reflects the intensity of angry feelings in the last week. Trait Anger measures a disposition to experience anger and it is divided into two subscales – Anger Reaction and Anger Temperament. Those subscales measure the propensity to react under situations which involve frustration or feelings of unfairness, or without such circumstances, respectively. Anger Expression-Out is related to the expression of angry feelings towards others or objects and Anger Expression-In to the suppression and self-directedness of those feelings. Anger Control reflects the extension of control or suppression of anger expression, and the Anger Expression Index is an overall index of the frequency of anger expression, regardless of its direction. The Socio-demographic and Medical Questionnaire included questions intended to provoke different levels of arousal, during which skin conductance and heart rate (HR) were measured. In Moment 1, socio-demographic questions were asked, and, assuming their neutrality, basal activity was assessed. Moment 2 corresponded to an Arithmetic question for eliciting arousal with a neutral emotional valence. Moment 3 addressed the medical history and BC diagnosis, which presumably elicit negative emotional arousal, as for Moment 4, when asked for negative life events (NLEs) in the past year. Autonomic measures were recorded using biosignalsPlux wireless (Plux Wireless Biosignals S.A., Portugal). Participants were connected to an IT physiological data system and electrodermal and cardiovascular activity were continuously collected between the baseline and moment 4. Two Ag/AgCl-electrodes from edaPLUX sensor were placed on the thenar and hypothenar eminence of the non-dominant hand for measuring the electrodermal activity. Blood volume pressure (BVP) was measured based on photoplethismography technique, applied on the middle finger of the dominant hand. Initially, the experimenter demonstrated at the computer screen how participants’movements produce artefacts and then gave instructions to sit still during the interview. The four interview moments were marked with a switch button (syncPLUX sensor) ensuring temporal synchrony, with millisecond accuracy, between the recording of psychophysiological parameters and the questions asked. Health Psychology and Behavioral Medicine 427 Downloaded by [b-on: Biblioteca do conhecimento online UL] at 05:36 22 January 2016 Data analysis Skin conductance tonic level and HR were extracted from the psychophysiological collecting data system. For feature selection, the raw data were processed using the Python programming language, and the SciPy scientific computing library. The raw data were first filtered to eliminate artefacts such as powerline interference; the BVP signal is bandpass filtered with a [1–8] Hz passing band, using a 4th order filter and the raw EDA signal is lowpass filtered with a 0.25 Hz cut-off frequency, using a 2nd order filter. For both signals, a Butterworth filter design was used since it is acknowledged to provide the best impulse response in the cut-off frequency boundaries, and a zero-phase forward and reverse digital filtering process was adopted (Silva, Fred, Eusebio, Torrado, & Ouakinin, 2012). Statistical analyses were generally conducted using the Statistical Package for Social Sciences (SPSS), version 19 for Macintosh. All statistical tests were two-tailed, with significance levels of 5%. Comparisons between independent samples regarding ordinal and continuous variables and associations between continuous variables were performed through Mann–Whitney non-parametric tests and Spearman correlations and associated significance tests, respectively, since assumptions of normality of variables were not met. Independency between ordinal and/or categorical variables was assessed through Fisher Exact tests either since 20% or more of the tables’ cells presented expected counts below 5 or 2 × 2 tables were analysed. Multiple regression linear models were carried out for characterizing eventual predictive effects of independent variables such as self-directed anger and anxious attachment style on emotional disturbance in the studied patients. Potentially confounding variables were systematically included in the model to control for such effects, namely Age, Education (5 points ordinal variable) and Employment Situation (dichotomized by Active-Working versus Other –Sick leave, Unemployed, Domestic). These results are presented in the next section. An alternative model using Education (dichotomized by Professional or Higher Education versus Secondary School or lower) was also built, as well as a model with GSI as dependent variables and only using the three potential confounders as independent variables as validation. Since these results were coherent with the ones of previous models, they were not included here. Results The studied sample included 85 females, 52 BCP and 33 controls, with a mean age of 41.87 and 37.42 years, respectively. Most of the BCP participants were married (46.2%), followed by the divorced (21.2%) and the civil union categories (15.4%). Forty-six percent of these participants have the high-school level of education and 67.3% were employed. Regarding their health status, most of them reported regular check-ups and blood tests (61.5%), breast palpation (67.3%) and did not show gynaecologic diseases history (73.1%). The majority of participants did not state familial breast cancer history (67.3%) or antecedent breast pathology (55.8%). A relevant part of the participants reported no physical diseases history (33%) nor previous psychopathology (65.4%) and 50% stated no current use of medication. The BCP participants initiated a treatment modality for cancer for overall 1 month at the date of the interview. Comparison of groups Excluding age (U= 501.00, p= .001), education level (p= .028, Fisher’s exact test), working status (p= .001, Fisher’s exact test) and regularity of check-ups reported (p= .048, Fisher’s exact test), the studied groups were homogenous in many socio-demographic variables such as 428 S. Ouakinin et al. Downloaded by [b-on: Biblioteca do conhecimento online UL] at 05:36 22 January 2016 marital status or family composition (Table 1) and in other variables considered (gynaecologic diseases history, familial breast cancer history, antecedent breast pathology, previous psychopathology, and breast palpation habits). In what concerns the occurrence of NLEs in the previous Table 1. Socio-demographic and clinical characterization of BCP and control group. BCP (n= 52) Controls (n= 33) UpMean value (SD) Age 41.87 (6.90) 37.42 (4.22) 501.000 .001** Menarche age 12 (1.49) 13 (1.42) 807.000 .638 N p-value (Fisher Exact test) Education level Basic school 5 0 .028* Junior high school 12 3 Secondary school 24 14 Professional school 1 1 Graduation 10 14 Working status Employed 46 33 .001** Unemployed 6 0 Marital status Single 6 10 .129 Married 24 12 Divorced 11 4 Civil union 8 7 Civil separated 3 0 Household Husband/partner 4 5 .103 Partner and son/daughter 31 16 Son/daughter 11 2 Parents 2 4 Others 2 2 Alone 2 4 Gynaecologic diseases history Yes 14 9 .582 No 38 24 Familial breast cancer history Yes 17 6 .111 No 35 27 Antecedent breast pathology Yes 23 14 .525 No 29 19 Breast palpation habits Yes 35 21 .453 No 17 12 Regular check-up and blood tests Yes 32 23 .048* No 20 5 Previous psychopathology None 34 20 .496 Depression 14 12 Anxiety 4 1 Note: *p< .05. **p< .01. Health Psychology and Behavioral Medicine 429 Downloaded by [b-on: Biblioteca do conhecimento online UL] at 05:36 22 January 2016 year (Table 2), such as deaths of close people or close relatives’diseases, marital and relational problems or financial problems, groups did not differ in a significant way. Concerning the attachment styles, BCP described themselves as significantly less anxious than controls (p< .05), and similar to controls in what regards to avoidant patterns of attachment. In terms of reactivity, BPC showed significantly higher rated distress compared to controls, but a lower electrodermal reactivity both at the baseline (U= 422.00, p< .001) and confronted with different emotional eliciting stimuli: arithmetic question (U=453.00, p= .001), recalling the medical history and BC diagnosis (U= 428.00, p< .001), as well as when they were asked for NLEs in the past year and (U= 346.00, p< .001). BCP also showed less cardiovascular reactivity at the four moments of assessment, even though not significantly different than controls. Statistical analysis between groups did not show any significant differences regarding reported psychopathological symptoms, anger states and traits, and anger regulation strategies (Table 3). Correlations between several variables studied in BCP group point relevant associations between an anxious pattern of attachment, anger regulation strategies (r= .445; p< .01), specially self-directedness ones (STAXI –anger expression in) and reported distress from GSI scores (r = .580; p< .01). Results showed a negative correlation between the expression of angry feelings towards others or objects (r=−.320; p< .05) and anger control and positive associations between distress measured by BSI and anger expression-in (r= .393; p< .01) and also with perceived distress rated by the distress thermometer (r= .415; p< .01). Anger expression-in was especially associated with the presence of an anxious attachment and, simultaneously, strictly linked to the reported level of psychopathological symptoms. No significant associations between psychophysiological parameters and psychological measures of attachment styles were verified in both groups (Table 4). Although the clinical sample is composed of a relatively small number of subjects, which provides a modest support to the development of robust statistics models of prediction, a multiple regression model was carried out using the General Symptoms Index as dependent variable and AAS-Anxiety (Anxiety factor extracted from the Adult Attachment Scale) and Anger Expression In (extracted from the STAXI) as independent variables (Model 1 in Table 5). These results pointed to a pattern of self-directed anger expression and particularly an anxious Table 2. NLEs in the previous year. N BCP (n= 52) Controls (n= 33) p-value (Fisher Exact test) Deaths of close people in the last year Yes 9 4 0.549 No 43 29 Close relatives diseases Yes 12 8 1.000 No 40 25 Marital problems Yes 13 4 0.286 No 39 29 Relational problems Yes 3 3 0.681 No 49 30 Financial problems Yes 5 3 1.000 No 47 30 430 S. Ouakinin et al. Downloaded by [b-on: Biblioteca do conhecimento online UL] at 05:36 22 January 2016 attachment style as variables with impact on the general psychopathological symptoms report (R 2 = .294; regression: ANOVA, F= 9.974, p= .000). The confidence intervals support the tests regarding the regression coefficients as showed in Table 5. Multiple regression models were additionally used with potential confounders’age, education and employment status, but did not indicate any biases in findings (Models 2–4inTable 5). Discussion Results found in this study are in line with literature that points to the relevance of distress screening and the need of being aware of patients’emotional suffering when they are faced with a severe disease diagnosis, even when they do not report it spontaneously (Carlson, Waller, & Mitchell, 2012; Linden & Girgis, 2012). Despite rating a level of distress that can be considered significant (DT > 4), BCP seem to cope with negative emotions by controlling its expression or turning it Table 3. Differences between BCP and healthy participants on psychological measures and psychophysiological parameters. Mean values for BCP and controls and Mann–Whitney test BCP (n= 52) Controls (n= 33) Mann–Whitney Up AAS-Anxiety 2.05 2.34 605.500* .020 AAS-Avoidance 2.60 2.66 794.000 .663 BSI-Gen Sympt Ind 0.67 .70 792.000 .650 Distress Thermometer 6.16 4.51 546.000* .011 STAXI –state anger 11.23 11.67 841.000 .996 STAXI –trait anger 15.47 16.39 738.500 .342 STAXI –angry temperament 5.67 5.67 830.000 .914 STAXI –angry reaction 7.18 7.85 734.500 .321 STAXI –anger expression in 14.73 14.88 828.500 .905 STAXI –anger expression out 12.47 13.12 753.000 .413 STAXI –anger control 23.08 22.40 772.000 .523 STAXI –anger expression 20.10 21.60 783.000 .591 EDA Baseline 8.41 13.45 422.000** .000 EDA Moment 1 10.04 14.74 453.000** .001 EDA Moment 2 9.39 14.17 428.000** .000 EDA Moment 3 7.52 13.45 346.000** .000 HR Baseline 80.16 80.85 726.000 .435 HR Moment 1 75.50 82.03 654.000 .144 HR Moment 2 73.92 80.20 679.000 .221 HR Moment 3 74.10 79.60 721.000 .408 Notes: AAS-Anxiety –anxiety factor extracted from the Adult Attachment Scale; AAS-Avoidance –avoidance factor extracted from the Adult Attachment Scale; BSI-Gen Sympt Ind –General Symptoms Index extracted from the Brief Symptoms Inventory; EDA Baseline –skin conductance tonic level at the baseline measured by the bioPLUX research device; EDA Moment 1 –skin conductance tonic level at the moment 1 measured by the bioPLUX research device; EDA Moment 2 –skin conductance tonic level at the moment 2 measured by the bioPLUX research device; EDA Moment 3 –skin conductance tonic level at the moment 3 measured by the bioPLUX research device; HR Baseline –heart rate in beats per minute at the baseline measured by the bioPLUX research device; HR Moment 1 –heart rate in beats per minute at the moment 1 measured by the bioPLUX research device; HR Moment 2 –heart rate in beats per minute at the moment 2 measured by the bioPLUX research device; HR Moment 3 –heart rate in beats per minute at the moment 3 measured by the bioPLUX research device. *p< .05. **p< .01. Health Psychology and Behavioral Medicine 431 Downloaded by [b-on: Biblioteca do conhecimento online UL] at 05:36 22 January 2016 Silva, H., Fred, A., Eusebio, S., Torrado, M., & Ouakinin, S. (2012). Feature extraction for psychophysiological load assessment in unconstrained scenarios. Conference Proceedings IEEE Engineering in Medicine and Biology,2012, 4784–4787. Slopen, N., Non, A., Williams, D. R., Roberts, A. L., & Albert, M. A. (2014). Childhood adversity, adult neighborhood context, and cumulative biological risk for chronic diseases in adulthood. Psychosomatic Medicine,76(7), 481–489. Spiegel, D. (2012). Mind matters in cancer survival. Psycho-Oncology,21, 588–593. Spielberger, C. D. (1999). Manual for the state-trait anger expression inventory-2. Odessa, FL: Psychological Assessment Resources. Stommel, M., Given, B. A., & Given, C. W. (2002). Depression and functional status as predictors of death among cancer patients. Cancer,94, 2719–2727. Tamagawa, R., Giese-Davis, J., Speca, M., Doll, R., Stephen, J., & Carlson, L. E. (2013). Trait mindfulness, repression, suppression, and self-reported mood and stress symptoms among women with breast cancer. Clinical Psychology,69(3), 264–277. Torrado, M., Silva, H., Eusébio, S., Fred, A., & Ouakinin, S. (2015). Alexithymia, physiological reactivity and cognitive appraisals of emotional stimuli in opiate dependents: A pilot study. Journal of Neurology and Neurophysiology,6(263). doi:10.4172/2155-9562.1000263 Watson, M., & Greer, S. (1983). Development of a questionnaire measure of emotional control. Journal of Psychosomatic Research,27(4), 299–305. Wu, S. M., Yang, H. C., Thayer, J. F., & Andersen, B. L. (2014). Association of the physiological stress response with depressive symptoms in patients with breast cancer. Psychosomatic Medicine,76(4), 252–256. 438 S. Ouakinin et al. Downloaded by [b-on: Biblioteca do conhecimento online UL] at 05:36 22 January 2016