Attachment styles and risk of getting cancer

Results of a retrospective study

Claudio Lallaa*, Alessandra Bottaa & Laura Corvagliaa.

a2th Center of Cognitive Psychotherapy, Rome, 00192 Italy
*Corresponding author. Email:

Lo stile affettivo può essere valutato sulla base della Teoria dell’Attaccamento. E’ stata riscontrata un’importante  correlazione statistica fra attaccamento evitante e patologie somatiche. Lo scopo del presente studio è di conoscere se lo stile distanziante è più frequente nelle persone ammalate di cancro. Lo studio ha esaminato un campione di 125 adulti, 58 dei quali affetti da cancro e 67 senza alcuna malattia manifesta. Non era presente nessuna persona affetta da melanoma. A fine di valutare lo stile di attaccamento si è fatto ricorso al questionario Experiences in Close Relationships (ECR). E’ emersa una proporzione significativamente più alta di donne con stile preoccupato fra i controlli e una proporzione  significativamente più alta di donne  con stile distanziante fra  i casi. Lo stress causato da eventi affettivi, se associato a un senso di impotenza, che è tipico dello stile distanziante, sarà vissuto dalle donne con pessimismo e rassegnazione. Ne deriverà una maggiore vulnerabilità alla malattia oncologica. All’opposto, le donne preoccupate, sebbene  tendano ad avere una visione del partner sentimentale come imprevedibile rispetto alla disponibilità affettiva e siano quindi esposte più facilmente a una condizione di stress, sono però costantemente impegnate, grazie a un senso di autoefficacia nella gestione delle relazioni sentimentali,  a mantenere il partner legato a sé stesse. Ne conseguirà una maggiore resistenza all’insorgere della malattia oncologica. 


Background and objective

The affective style can be assessed on the basis of the Attachment Theory. An important statistical correlation was found between avoidant attachment style and somatic pathologies. The purpose of this study is to know  if the dismissing style is more frequent in people with cancer.


The study examined a sample composed of 125 adults, 58 of which affected by cancer and 67 without any evident disease. No people affected by melanoma were present. In order to assess the kind of attachment style, the questionnaire Experiences in Close Relationships (ECR) was used.


A significantly higher proportion of women with preoccupied style among the controls and a significantly higher proportion of women with dismissing style among the cases emerged.


The stress caused by affective events, if associated with a sense of helplessness, which is typical of the dismissing style, is experienced by women with pessimism and resignation. The result is a greater vulnerability to cancer. On the contrary, preoccupied women, although they tend to have a vision of romantic partner as unpredictable with respect to affective availability and are therefore more easily exposed to a condition of stress, are nevertheless constantly engaged, thanks to a sense of self-efficacy in the management of romantic relationships, to keep the partner tied to themselves. The  result is a greater resistance to cancer developing.

Background and objective

The researches investigating psychological processes involved in the cancer pathogenesis have increased over the last few years. With this regard, both the role of some specific personality traits and the role of specific stressful events have been studied. In the new model of integrated medicine, the cancer initiation phase begins regardless of psychological factors. These last, however, together with food [1] influence the promotion phase (cancer growth) as well as the progression phase (invasion on the surrounding tissues and metastasis). In fact, through the central nervous system, psychological factors can depress the immune processes by which our body defends itself from the cancer aggression. Therefore, psychological factors act as risk or protection factors which, inside of a model more complex of the traditional one, shift the balance at stake towards the disease or the healing.

This article shows a research carried out in order to verify whether or not people with cancer have personality traits making them more susceptible to this disease. In particular, the study tried to establish whether or not the dismissing style is more frequent among people with cancer.

The formulation of such hypothesis came as a result of some important researches.

Spangler and Grossman’s study [2] highlighted how during the Strange Situation, despite their apparent calm, avoidant children showed an increase in heart rate higher than secure children’s one, as well as a growing increase in the salivary cortisol level. This last survey, carried out 15 and then 30 minutes after the conclusion of the procedure, was revealing the perpetuation of a stressful condition that hadn’t been efficiently faced.

Also more recently [3] it was found that in the Strange Situation children with avoidant attachment show less behavioral indicators of discomfort compared to children with secure and anxious-ambivalent attachments, while like the others, they show tachycardia during the mother’s absence and therefore present a contrast between behavior and  physiological reactions to the stress of separation.

Temoshock [4] revealed that normal subjects show a prompt heart activation at the moment of the stressful event which then quickly comes back to normal values, while denial-oriented subjects (similar to the avoidant children of the above-mentioned case) show an increase in heart rate much further the conclusion of the stressful event.

Furthermore, a meta-analysis [5] highlighted an important statistical correlation between avoidant attachment style and somatic pathologies.

The influence of attachment style on romantic relationships can be nowadays assessed. According to attachment theory and the studies it gave rise to [6-7], it is possible to divide adults’ attachment styles into 4 types: secure (the person is at ease with both physical and emotional intimacy established with the partner and furthermore he/she trusts the partner’s affective availability within the relationship); preoccupied (the person can approach the partner but is strongly anxious, at the same time, about the evolution of the relationship); dismissing (the person is barely concerned with the interest showed by the partner and is not at ease with the physical and emotional intimacy); fearful (the person is not at ease with the intimacy and tends to be anxious as a result of the love and attachment showed by the partner). This fourfold division of attachment styles reflects that one of the attachment patterns (secure, anxious-ambivalent, avoidant and disorganized) as widely studied in children thanks to the Strange Situation [8-9-10-11-12]. The typology of attachment styles in romantic relationships is closely connected also to the classification obtained with the Adult Attachment Interview [9], a semi-structured interview by which the state of mind with respect to attachment can be defined as autonomous (corresponding to secure attachment in infants), dismissing (corresponding to insecure-avoidant attachment in infants), preoccupied (corresponding to insecure-ambivalent attachment in infants and unresolved with respect to loss or trauma (corresponding to disoriented/disorganized attachment in infants).

The purpose of this study is to know  if the dismissing style is more frequent in people with cancer.


The study examined a sample composed of 125 people, 58 of which affected by neoplasia and 67 without any evident disease. The two groups were similar at socio-cultural level (age, sex and education). People with cancer were looked after in the medical oncological unit of the University General Hospital “Agostino Gemelli” in Rome and suffered from several kinds of solid cancers. The most frequent types were: colon cancer (41%), breast cancer (27%) and lung cancer (19%) The remaining 13% of patients had other types of solid tumors. No persons affected by melanoma were present. Informed consent was obtained from all individual partecipants included in the study.

In order to assess the kind of emotional style, the Experiences in Close Relationships (ECR) was used. It is  a questionnaire formulated by Brennan, Clark and Shaver [7]. of which there is a validated version in Italy, [13-14]. This instrument is composed of two 18-item sub-scales. The first one refers to the avoidant dimension, while the second one reflects anxiety for relationships. Scores are assigned along a Likert scale going from 1 (completely false) to 7 (completely true). On the basis of the achieved results in the two scales four types of attachment style appear: secure (low levels achieved both in the avoidance dimension and in the anxiety one); preoccupied (low levels registered in avoidance and high levels in anxiety); dismissing (low levels showed in anxiety and high levels in avoidance); fearful (high levels both in anxiety and in avoidance). The suggested cut-off is the value resulting from the average score, obtained in the avoidance scale or in anxiety scale by the sample of the validation study of the Italian version, added to the related standard deviation. Such cut-off has however led to attachment style classification percentages in the validation sample, which above all for the secure one, appear very distant from those found in other studies, such as those carried out by several authors [15-16-17-18-6]. In fact, the massive percentage of 69.7% of subjects categorized as secure in the validation study reveals, in the light of the comparison with other epidemiological studies, a lack of sensitivity at identifying insecure subjects. On the other hand, the same authors of the Italian validation suggest the possibility of lowering the cut-off according to the specific objectives of the research carried out.

A new cut-off was chosen. It is a cut-off able to bring the percentage of secure subjects of the control group to the same level defined by the previous researches, so  as to make in such group the framework of the percentage distributions among the different affective  styles as much more similar as possible to that one already known. The cut-off useful to achieve this objective is the proportion of 0,62 of the standard deviation (SD). Such a calibration allowed in fact to get in the control group a percentage of secure subjects of 46,27%, that is extremely close to that one of  47% by Bartholomew and Horowitz  [6].

The normative scores followed by this study are  indicated below. To a purpose of comparison, the cut-off of the average + SD used in the validation study of the Italian version was pointed out as well.

Normative scores for avoidance scale

Women and men

Age        AverageSDAverage + SDAverage +0,62 SD

Normative scores for anxiety scale


Age        AverageSDAverage + SDAverage +0,62 SD


Age        AverageSDAverage + SDAverage +0,62 SD


Table 1. Proportion of attachment styles (cut-off = average + 0,62 SD), women and men

Attachment stylesWomen and men: casesWomen and men: controlsChi-squaredpOdds ratio
Other styles2336
Other styles4153
Other styles5549
Other styles5563

* Significant difference (prob. 1%)

Table 2. Proportion of attachment styles (cut-off = average + 0,62 SD), only women

Attachment stylesWomen: casesWomen: controlsChi-squaredpOdds ratio
Other styles1628
Other styles2335
Other styles3323
Other styles3337

* Significant difference (prob. 1%)

** Significant difference (prob. 5%)

Table 3. Proportion of attachment styles (cut-off = average + 0,62 SD), only men

Attachment stylesMen: casesMen: controlsChi-squaredpOdds ratio
Other styles714
Other styles1618
Other styles2132
Other styles2132

Within controls and cases, the proportion of women showing each of the four affective styles was compared to the men’s one. (tables 4 and 5).

Table 4. Proportion of women and men showing the four attachment styles in the group of controls

Attachment stylesWomen: controls Men: controlsChi-squaredpOdds ratio
Other styles2814
Other styles3518
Other styles2332
Other styles3732

* Significant difference (prob. 1%)

** Significant difference (prob. 5%)

Table 5. Proportion of women and men showing the four attachment styles in the group of cases

Attachment stylesWomen: casesMen: casesChi-squaredpOdds ratio
Other styles167
Other styles2316
Other styles3321
Other styles3321


By using the average  + 0,62 DS  as cut-off (see table 1) in the group of controls a percentage of secure very close  (46%) to that one of 47% by Bartolomew and Horowitz was obtained.

We can see significant differences between cases and controls in the percentages of dismissing style and preoccupied style among women (table 2), but not among men  (table 3). The significantly higher proportion of preoccupied style among controls when  both women and men are considered together (table 1)  is due only to the contribution of women, as shown by the table 2, 3 and 4. Considering the higher proportion among cases of women with dismissing style, the critical observation a reader may make could be: is the dismissing style that can contribute among  women to the risk of getting cancer or is the condition suffered by these cancer patients that drives them to develop a greater extent of dismissing style?

The first observation to make is that a case-control study cannot by its nature assure the absolute certainty of identifying a risk factor. In the field of science, the absolute certainty does not actually exist, but it is anyway possible to assess what kind of studies allow us to get closer to it. A longitudinal study, assessing first the attachment styles of a sample of people and only later recording the number of people getting cancer among those who showed one or the other of four attachment styles, would be undoubtedly a more reliable method compared to the case-control study. But everyone knows that a longitudinal study is much longer and much more expensive, and just for this reason it calls for being preceded by case-control studies, which are shorter and cheap. So, before other researchers make a longitudinal study, the problem fundamentally is that of stability of attachment styles and their respective working models, with related thoughts and expectations. In this respect, around 70–80% of people experience no significant changes in attachment styles over time [19-20-21-22-23].

The fact that attachment styles do not change for a majority of people indicates that working models are relatively stable. Only around 20–30% of people do experience changes in  their attachment styles. These changes primarily occur toward non-secure forms of attachment, because of negative life-events that are bad for secure attachment, directly by altering the child-parent relationship, and indirectly by increasing life stress for the parents. These events include the death of a parent, foster care, parental divorce, chronic and severe illness of parent, parental psychiatric disorder, parental drug and alcohol abuse, and child experience of physical or sexual abuse. The aforesaid negative life events were identified by means of AAI and checklists   [24-25]  or otherwise assessed (like  child maltreatment, maternal depression, and family malfunctioning in early adolescence) during longitudinal studies [26-27]. Like the other affective styles, the dismissing style of  women suffering from cancer represents a datum belonging to their personality. Affective style changes when relationship with significant figures changes too, above all relationship with parents. Therefore dismissing style comes before the disease onset and it is for women a psychological factor of cancer risk.

Moreover we think that the data of the study corroborate their reading in terms of risk and protective factors by means of their own internal organization and their connection with other studies done on the relationship between stress and the onset of cancer.

In the control group (see table 4) there is a higher proportion of dismissing men compared to dismissing women and a higher proportion of preoccupied women compared to preoccupied men, which in our sample drops to zero. The measure achieved by preoccupied category of attachment style in women of the controls is so high that leads in such group to a significant difference between the proportion of secure women and proportion of secure men, for the clear advantage of the second ones.

It is interesting to notice that some previous researches formerly carried out with non-clinical populations let emerge data consistent with those observed in the control group of this study. In particular, Bartholomew & Horowitz [6] by their interview noticed “preoccupied” style scores in women much higher than those obtained by men. On the other hand, the same authors have also obtained in men “dismissing” style scores significantly higher than those obtained in women.

Similar results were obtained also by Del Giudice [28] in a meta-analysis which took into account gender differences as to avoidance and anxiety dimensions of adults’ attachment styles. The research was based on 113 samples got from 100 studies, for a total of 66.132 examined subjects. Also this study let emerge an avoidance level significantly higher in men and anxiety scores significantly higher in women.

A proportion of preoccupied subjects significantly higher in controls than in cases has emerged (see table 1). But, as we said, such datum completely depends on women because among men of the control group, no preoccupied subject was  found (table 4).

The significantly higher proportion of women with preoccupied style among the controls is complementary to the significantly higher proportion of women with dismissing style among the cases (see table 2). Preoccupied style is the opposite of the dismissing one. The differences between the proportions of preoccupied and dismissing women, that have been noticed between cases and controls should be set within those ones, previously mentioned, outlined in their turn between women and men inside controls (table 4).

The emerged data show that these last differences fall down among the oncological subjects of this research, who don’t show inside them any significant difference among the proportions achieved by men and women  in the different affective styles (table 5).

A significantly higher proportion of secure women among the cases compared to the controls appears (see table 2). This result is only apparently paradoxical. In fact it forms a whole with the big presence of preoccupied women among the controls and the consequent decrease among them of the number of secure women.

Why do just women show such differences? The reason could be connected to the fact that they are more sensitive to the course of affective events. As a result, the stress caused by these events is higher in women. Parkes [29] in a retrospective study carried out on adults undergoing psychotherapeutic treatment because of psychiatric disorders arisen after a bereavement commented the clear prevalence of women characterizing his sample, hypothesizing in the attachment models the presence of aspects making women more vulnerable than men to the event of  loss. We are, in turn, inclined to believe women give more importance to affective bonds, considering them as crucial in life, much more than men do.

If the stress caused by  affective events is associated with a sense of helplessness (which, according to the attachment theory, comes from a mental scheme of the attachment figure that is characterized by a  poor affective availability and is typical of the dismissing style), it will be experienced most dramatically and with a sense of deep pessimism.

Conversely, although preoccupied women experience the affective stress coming from a mental scheme that represents the attachment figure as unpredictable with respect to affective availability, they are constantly engaged in controlling such figure and keeping her linked to themselves. This behavior reflects a sense of self-efficacy rather than helplessness in the management of romantic relationships.

Secure affective style seems to be for women a factor playing an intermediate role between the protection towards cancer offered by preoccupied style and the higher exposure to the risk of it that is caused by the dismissing style. Odds ratios confirm such reading (see table 4). Cancer shows a negative and strong association with preoccupied style (odds ratio = 0, 08), a positive but quite moderate association with secure style (odds ratio = 2, 56), a positive and moderate association with dismissing style (odds ratio = 3, 04).

The deservedly well-known experimental study carried out by Visintainer, Volpicelli & Seligman [30] confirms the aforesaid reading of results. It considered three groups of rats: the first one experienced an escapable electric shock, the second one an inescapable shock and the third, no electric shock. Before doing this, all rats were implanted with a tumor preparation.  In natural conditions, such preparation would have caused the death of 50% of animals, while the remaining 50% would have survived thanks to the immune response. The results were the following: the survivors were 27% of the animals receiving inescapable stressor, 54% of the animals receiving no stressor and 63% of the animals receiving escapable stressor.

It is also evident, as demonstrated by a further experiment [31] , that what is important is not the actual inescapability of the stressor, rather the believe it can be manageable or not, that is the mental scheme through which life events are read.


If dismissing style is a psychological cancer risk factor for women, what comes to light is the need to prevent the establishing of its forerunner in the development of the person: avoiding attachment of children towards parents and particularly towards the mother. Such objective would be naturally achieved by means of a no more dismissing parental caregiving. In order to decrease the frequency of such caregiving style and promote parental sensitivity to children’s attachment signals, two strategies appear crucial: a work of social education carried out through mass-media and a clinical intervention on the single families through the health care.

A second line of action involves the integration of psychotherapy into cancer treatment. As demonstrated by previous researches [32-33] enriching the treatment in this way would produce much more than a palliative effect.


  • [1]   Servan-Schreiber D. Anticancer, Paris: Robert Laffont, 2007.
  • [2] Spangler, G. & Grossmann, K.E.: Biobehavioral organization in securely and insecurely attached infants. Child Development, 1993;64: 1439-1450.
  • [3] Zelenko M, Kraemer H, Huffman L, Gschwendt M, Pageler N, Steiner H.: Heart rate correlates of attachment status in young mothers and their infants. Journal of the American Academy of Child & Adolescent Psychiatry, 2005; 244:470–476.
  • [4]  Temoshock L.R.: On assessing such complex construct as Type C coping. Paper presented at the 1st International  Conference on The (non)Expression of Emotions in Health and Disease, 1996, Tilburg, The Netherlands.
  • [5] van IJzendoorn M.H., Bakermans-Kranenburg M.J.: Attachment representations in mothers, fathers, adolescents, and clinical groups: a meta-analytic search for normative data. J Consult Clin Psychol. 1996.;64(1):8-21.
  • [6] Bartholomew K. & Horowitz L.M.: Attachment Styles Among Young Adults: A Test of a Four-Category Model. Journal of Personality and Social Psychology, 1991;61, (2): 226-244.
  • [7] Brennan K.A., Clark C.L. & Shaver P.R.: Self-report measurement of adult attachment. In J.A. Simpson & W.S. Rholes (Ed), Attachment Theory and Close Relationships (pp. 155-178). New York NY: Guilford Press, 1998.
  • [8] Ainsworth M.D.S., Blehar M.C., Waters E. & Wall S.: Patterns of Attachment: A Psychological Study of the Strange Situation. New York: Erlbaum, Hillsdale, 1978.
  • [9] Main, M., Goldwyn R., & Hesse E.: Adult Attachment Scoring and Classification Systems. Unpublished Manuscript, University of California at Berkeley, 2003.
  • [10] Ainsworth, M.D.S., & Bowlby, J.:An Ethological Approach to Personality Development. American Psychologist, 1991; 46(4), 333-341.
  • [11] Crittenden P.: Quality of Attachment in the preschool years. Development and Psychopathology, 1992; 4, 209-241.
  • [12] Waters E.: Appendix A: The Attachment Q Set (version 3.0). In E. Waters, B.E. Vaughn, G. Posada & K. Kondo-Ikemura, Caregiving, Cultural and Cognitive Perspectives on secure-base Behaviour and Working Models: New Growing Points of attachment Theory and Research. Monographs of the Society for Research in Child development,1995; 60, 236-246.
  • [13] Picardi A., Bitetti D., Puddu P. & Pasquini P.: La scala ‘Experiences in Close Relationships’, un nuovo strumento per la valutazione dell’attaccamento negli adulti: traduzione, adattamento e validazione della versione italiana [The scale ‘Experiences in Close Relationships’, a new tool for the assessment of attachment in adults: translation, adaptation and validation of the Italian version]. Rivista di Psichiatria, 2000; 35, 114-120.
  • [14]  Picardi A. et al.: Il questionario ‘Experiences in Close relationships’ (ECR) per la valutazione dell’attaccamento negli adulti: ampliamento delle evidenze di validità per la versione italiana [The questionnaire ‘Experiences in close relationships’ (ECR) for the assessment of attachment in adults: expanding the evidences of validity for the Italian version]. Giornale Italiano di Psicopatologia, 2002; 8, 282-294.
  • [15]  Ainsworth, M. D. S., & Bell, S. M.:Attachment, exploration, and separation: Illustrated by the behavior of one-year-old in a strange situation. Child Development, 1970; 41, 49-67.
  • [16]  Ainsworth, M. D. S., Bell, S. M., & Stayton, D.: Individual differences in Strange Situation behavior of one-year-old. In H. R. Schaffer (Ed,), The origins of human social relations (pp. 17-57). London: Academic Press, 1971.
  • [17]  George C., Kaplan N. & Main M.: Adult Attachment Interview Protocol. Berkeley: University of California, 3th ed. 1996.
  • [18]   Hazan, C. & Shaver, P.: Romantic Love conceptualized as an attachment process. Journal of Personality and Social Psychology, 1987; 52(3), 511-524.
  • [19]   Baldwin, M.W., & Fehr, B: On the Instability of Attachment Style Ratings. Personal Relationships, 1995; 2, 247–261.
  • [20]   Waters, E., Merrick, S., Treboux, D., Crowell, J., & Albersheim, L.: Attachment security in infancy and early adulthood: A twenty-year longitudinal study. Child Development, 2000; 71, 684–689
  • [21]    Kirkpatrick, L., & Hazen, C.: Attachment styles and close relationships: A four year prospective study. Personal Relationships, 1994; 1, 123–142
  • [22]   Feeney, J., & Noller, P.: Attachment style and romantic love: Relationship dissolution. Australian Journal of Psychology, 1992; 44, 69–74.
  • [23]   Scharfe, E., & Bartholomew, K.: Reliability and stability of adult attachment patterns. Personal Relationships, 1994; 1, 23–43.
  • [24]   Waters, E., Weinfield, N.S., & Hamilton, C.E.: The stability of attachment styles from infancy to adolescence and early adulthood: General discussion. Child Development, 2000; 71, 703–706.
  • [25]  Hamilton, C.E.: Continuity and discontinuity of attachment from infancy through adolescence. Child Development, 2000; 71, 690–694.
  • [26]  Hamilton, C.E.: Continuity and discontinuity of attachment from infancy through adolescence. Child Development, 2000; 71, 690–694.
  • [27]  Weinfield, N. S., Sroufe, L.A., & Egelund, B.: Attachment from infancy to early adulthood in a high-risk sample: Continuity, discontinuity, and their correlates. Child Development, 2000; 71, 695–702.
  • [28]   Del Giudice M.:  Sex Differences in Romantic Attachment: A Meta-Analysis. Personality and Social Psychology Bulletin 2011; 37(2): 193–214.
  • [29]    Parkes C.M. (1991). Attachment, Bonding and Psychiatric Problems after Bereavement in Adult Life, in C.M. Parkes, J. Stevenson-Hinde & P. Marris, Attachment Across the Life Cycle (pp. 268-292), New York, NY: Routledge.
  • [30]  Visintainer M.A., Volpicelli J., Seligman M.E.: Tumor rejection in rats after inescapable or escapable shock. Science, 1982; 216: 437-439.
  • [31]    Seligman, M.E.P. and Visintainer, M.A.: Tumor rejection and early experience of uncontrollable shock in the rat.  In F.R. Brush and J.B. Overmier  (Eds.), Affect, conditioning, and cognition:  Essays on the determinants of behavior. Hillsdale, N.J.: Erlbaum,  203-210, 1985.
  • [32]  Grossarth-Maticek R. & Eisenck H.J.: Lenght of survival and Lymphocyte percentage in Women with Mammary Cancer as a Function of Psychotherapy. Psychol Rep.,1989; 65: 315-321.
  • [33]    Spiegel D. et al.:Effects of Supportive-expressive Group Therapy on Survival of Patients with Metastatic Breast Cancer: A Randomized Prospective trial, Cancer, 2007; 110(5), 1130-1138.

Leggi anche

Iscriviti alla Newsletter

Iscriviti alla Newsletter

Iscriviti alla newsletter per ricevere le nostre novità direttamente nella tua casella di posta elettronica.

Ah... gratis!

Fantastico! Ti sei appena iscritto

Pin It on Pinterest

Share This