Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Email this article to a Colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to My Articles & Searches Download to citation manager Citing Articles via Google Scholar Articles by Beresford, T. P. Articles by Martin, B. Search for Related Content PubMed Citation Articles by Beresford, T. P. Articles by Martin, B. Other Patient Groups/Issues Primary Care

Cancer Survival Probability as a Function of Ego Defense (Adaptive) Mechanisms Versus Depressive Symptoms

Received December 3, 2004; revised August 31, 2005; accepted September 29, 2005. From the VA Medical Center (Research 151) and the Dept. of Psychiatry, Univ. of Colorado Health Sciences Center, Denver, CO. Address correspondence and reprint requests to Dr. Beresford, VA Medical Center (116), 1055 Clermont St., Denver, CO 80220-0116. e-mail: thomas.beresford{at}uchsc.edu

ABSTRACT

TOP ABSTRACT INTRODUCTION METHOD RESULTS DISCUSSION REFERENCES

 
Psychological treatment studies, uncontrolled for ego defense (adaptive) styles, report conflicting survival results. The authors hypothesized that "immature" adaptive styles and frequent depression symptoms would independently predict lower survival rates. This study followed 86 consecutive, mostly late-stage, cancer outpatients for up to 5 years; their survival data were analyzed in relation to the Beck Depression Inventory and the Defense Style Questionnaire scores at study entry. Cumulative survival probability curves contrasted the extreme cases: the most (N=15) to the least (N=21) depressed, and the "immature" (N=14) to the "mature" (N=16) adaptors. Depression did not separate the groups until 30 months after diagnosis. Ego defense style separated them at 8 months; by 18 months, the "immature" survival probability had dropped to 50%, versus 87% for the "mature." At 36 months, survival probabilities were 19% and 57%, respectively. These data direct clinical attention toward ego defense mechanisms as indicators of distress and lowered survival in cancer patients. They further suggest that the maturity of adaptive mechanisms must be controlled for in behavioral-treatment trials of cancer patients.

INTRODUCTION

TOP ABSTRACT INTRODUCTION METHOD RESULTS DISCUSSION REFERENCES

 
Widely publicized studies ^1,2 noted increased survival rates among cancer patients who underwent psychotherapeutic treatment. However, a recent multicenter study³ could not replicate this improved survival after behavioral treatment, and other studies have reported similarly conflicting results.⁴ Since published reports suggest that patients likely to benefit from psychotherapies are generally those with the most psychological maturity,^5,6 it seems possible that the underlying health of ego-defense mechanisms may be related to cancer survival. To our knowledge, the maturity of ego-defense mechanisms⁷ has not been considered as a potential mediating or moderating variable either in behavioral-treatment studies or in trials of antidepressants in cancer patients. We therefore used measures of both depression symptom frequency and ego-defense style maturity to assess the relationship each bore on survival probability in cancer patients. On the basis of previous studies^8–10 of depression and ego-adaptation,^11–13 we hypothesized that 1) increased depression symptom frequency and 2) immature defense-mechanism endorsement would, independently, predict lower cumulative survival probability. We tested this hypothesis in the following comparisons of survival probability with respect to both depression symptoms and ego-defense styles.

METHOD

TOP ABSTRACT INTRODUCTION METHOD RESULTS DISCUSSION REFERENCES

 
Subjects
Consecutive cancer outpatients treated at the University of Colorado Cancer Center made up the sample group. Data were collected during the summer months of 1999, 2000, and 2002. These dates of study reflect the funding for this exploratory investigation, which was supported in part by summer fellowship grants from the National Cancer Institute through the University of Colorado Cancer Center. Summer research assistants were trained in interviewing with each of the study instruments and were certified in the institutional requirements with respect to confidentiality, research ethics, and good clinical practice. Their ongoing work was supervised by the principal investigator (TPB) and by permanent research staff members. Subjects were entered consecutively during the respective time-periods. The data in this report were gathered prospectively in accord with the study hypotheses and are not the product of post-hoc analyses, as might be seen in a convenience sample. At the same time, however, funding limitations prevented more systematic procedures, such as random sampling.

All subjects were adults who consented to this Institutional Review Board-approved study. Subjects were included if they 1) presented a diagnosis of cancer; 2) were free of delirium; and 3) were either pain-free or presented with minimal pain, as measured below; and 4) agreed to participate. Twenty potential subjects declined participation (18.9%; N=20 of 106): 14 gave no reason, 4 said the process was too long, and 2 reported feeling too tired. Consistent with privacy regulations, no further data were collected on those who declined consent. Subjects suffering from pancreatic neoplasm or neuro-humoral secreting tumors, such as carcinoid or pituitary neoplasm, were excluded from the study because of the known effects of these neoplasms on mood and affect. Before entering the study, subjects were given a delirium screening examination¹⁴ and were asked to rate their pain, if they had pain, on a linear-analog scale.¹⁵ No patients reported pain of sufficient magnitude as to be subjectively distracting or greater than the first quartile of the analog scale; two patients failed the delirium screening. In this way, we accrued a sample who then provided responses to the study instruments (N=86). With the assistance of the University of Colorado Cancer Center, we were able to gather independently recorded survival data for all 86 subjects, the final sample for this report.

All subjects were adults whose ages ranged from 32 to 84 years, with a mean age of 55.5 years (standard deviation [SD]: 13.2); there were 47 women and 39 men, with no statistical separation by mean age. Neoplastic conditions were classed as Stage I (8%), Stage 2 (6%), Stage 3 (23%), Stage 4 (40%), and missing or not applicable (23%); that is, the majority of the subjects (63%) suffered from late-stage (Stage 3 or 4) neoplasms. Subject frequencies of cancer diagnoses by type included: genito-urinary (including prostate; N=14), breast (N=14), colon and rectum (N=13), lung (N=12), skin and melanoma (N=8), Hodgkin’s disease and leukemia (N=7), digestive tract, including liver (N=7), ovarian and cervical (N=4), brain and spinal cord (N=3), and endocrine, bone, pharynx, and heart (N=1 each).

Study Instruments
Each subject was given printed versions of the original 21-item Beck Depression Inventory (BDI)¹⁶ and the 40-item Defense Style Questionnaire (DSQ)¹⁷ and asked to fill them out. The BDI presents questions on specific depressive symptoms and asks respondents to rate their occurrence, using four alternatives varying from "rarely" through "often." Each item is scored, yielding a total BDI score. In this version of the BDI, mood symptoms, Items 1–14, can be separated from vegetative symptoms, Items 15–21, and analyzed separately.

The DSQ was designed for recognition and quantitative measurement of ego-defense mechanisms—that is, behavioral strategies by which individuals adapt to the stresses they encounter in their lives. The DSQ was first developed by Bond^18,19 and was validated by Andrews and others.^17,20,21 It contains 40 itemized statements about what a person does in a difficult situation. The subject is asked to agree or disagree with each specific statement on a 9-point Likert-type scale. The 40 items reflect 20 adaptive styles, with two items for each ego-defense mechanism. The two respective item scores are averaged, resulting in a mean score for each mechanism.

Based on Vaillant’s empirical work,^7,12,13,22 the DSQ scores for specific mechanisms are then grouped in an a-priori fashion into three domains of psychological adaptive styles: immature, neurotic, and mature. Combined in this way, investigators may then compute category means for each of the three domains for each subject.

Data Analysis With Respect to Survival Probability
The demographic and statistical core of the University of Colorado Cancer Center recorded data on date of diagnosis, date of death, and date of the last clinic appointment attended for each subject. With Institutional Review Board approval, we were allowed access to these data. Using date of diagnosis as a baseline, we calculated months of survival for the follow-up sample, recording deaths only as verified by the University of Colorado Cancer Center.

Using length of time since diagnosis either to death or to the date of our statistical analysis as length of survival, we sought statistical relationships between survival and age, gender, cancer diagnosis (type), and cancer stage. Regression analysis showed none of these to be significantly related to length of survival (p>0.05 for each).

Given the distribution of cancer stage and survival probability, we asked whether either depression symptom frequencies or maturity of adaptive mechanisms would affect probable longevity. To assess this, we stratified the sample in two ways before any statistical analysis. First, a BDI score >15 characterized the most depressed (N=15; 17.4% of the sample), and a BDI score <6 identified the least depressed (N=21; 24.4%). These cutoff points fell approximately at the uppermost and lowermost quartiles and were chosen according to quartiles as well as clinical significance (Table 1).

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

TABLE 1. Distribution of Main-Effect Variables (N=86)

Second, we calculated a simple "mature-minus-immature" score for each subject by subtracting the respective subject’s mean "immature" DSQ score from his or her mean "mature" DSQ score. This allowed a "mature-minus-immature" adaptation score for each individual. Those whose mature-minus-immature adaptation scores were >2 constituted the Mature Adaptor group (N=16; 18.6% of the total sample), and those with a negative mature-minus-immature adaptation score (that is, where the immature score exceeded the mature score) were identified as the Immature Adaptor sample group (N=14; 16.3% of the total sample). These cutoffs were also chosen according to quartile and clinical significance (Table 1).

The follow-up data were then arranged in a Kaplan-Meier life-table analysis, with cumulative survival probability calculations made for each of four subgroups stratified by depression symptom frequency and by mature-minus-immature adaptation scores, respectively.²³ Probability calculations were computed as percent-survival at each 3-month interval after baseline diagnosis, as above. Potential covariates were considered by use of Cox regression (Table 2).

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

TABLE 2. Distribution of Potential Confounding Variables by Main Effect Variables (N=86)

RESULTS

TOP ABSTRACT INTRODUCTION METHOD RESULTS DISCUSSION REFERENCES

 
Subgroup Characteristics
Each mature and immature ego-defense subgroup resembled its counterpart on demographic and clinical descriptors (Table 2). There were no significant statistical differences with respect to mean age, gender distribution, gender-specific cancer types, or cancer stage; nor did either of the ego-defense subgroups differ significantly on these measures from the whole sample group or from those excluded in this comparison of extreme responders.

By contrast, the depression symptom subgroups differed significantly by age (Table 2): the mean age of the most depressed patients was 49.4 years (SD: 6.9) and, for the least depressed, was 60.1 years (SD: 12.8; Student’s t-test, p=0.002). The young mean age of the most depressed subgroup separated it significantly from the mean for the whole sample (t-test, p<0.02), and from those who did not reside at either extreme (t-test, p<0.04); the least depressed subgroup did not differ in these ways. There was a nonsignificant trend toward more women and fewer men in the most depressed subgroup (Fisher’s exact test, p=0.09). Gender-specific cancer types did not separate the two groups statistically, however, nor did cancer stage.

Depression and Survival Probability
As shown in Figure 1, the most depressed and the least depressed appeared to share the same characteristics until about the 30th month after cancer diagnosis. At that point, survival probability among the most depressed began a precipitous drop, from 91% to 53% by the 36th month, and then appeared to stabilize through 60 months. By contrast, the cumulative survival probability among the least depressed dropped only slightly at the 30th month, from 88% to 82%, and remained constant through 48 months. At 52 months, their survival probability decreased to approximate that of the most depressed by 60 months. Survival probability between the two depression symptom subgroups differed significantly only when the curves separated, at 36 months (p<0.01). When covariates of age, stage, and gender were examined in a Cox regression analysis, none of these were found to significantly influence the relationship (p>0.05 each), and they were dropped from the model.

View larger version (15K): [in this window] [in a new window]

FIGURE 1.  Five-Year Cumulative Survival Probability by Depression Scores

Adaptation and Survival Probability
The cumulative survival probability curves for the Immature and Mature subject groups were very different. The survival probability for the Immature declined steadily and rapidly to 50% after 18 months and then to 19% at 3 years. The Mature subjects’ cumulative survival probability declined only to 90%, and then to 79%, at the same time-points. At 5 years, the Immature subjects’ survival probability remained at 19%, whereas that for Mature style declined only to 54%. The survival probabilities at each measured time-point—18 months, 3 years, and 5 years—were significantly different between the two subgroups stratified by adaptive maturity (p<0.001; Figure 2). Again, when covariates of age, stage, and gender were considered in a Cox regression model, none of these significantly influenced the relationship (p>0.05 each), and they were dropped from the model.

View larger version (16K): [in this window] [in a new window]

FIGURE 2.  Five-Year Cumulative Survival Probability by Maturity of Adaptive Styles

Finally, because we found no interaction effect when considering adaptation and depression at the same time, we did not stratify the sample any further.

DISCUSSION

TOP ABSTRACT INTRODUCTION METHOD RESULTS DISCUSSION REFERENCES

 
The data from this exploratory study suggest that both 1) maturity of ego-defense adaptation, and 2) frequency of depression symptoms may have predictive value with respect to cumulative survival probability in patients suffering from cancer. If so, this suggests a model in which adaptation and depression are inversely related.

The apparent ability of the Defense Style Questionnaire (DSQ) "Immature" items to predict the most precipitous declines in survival probability was striking: these data suggest that about 1 in every 6 patients may suffer early mortality related to failures in ego-defense adaptation and that these may be treatable. This possibility calls out for better understanding of psychological adaptation and survival—especially potential pathophysiological interactions between the two—and for better clinical tools in recognition and treatment.

This study has significant limitations, including its relatively small sample size, the lack of a randomized sampling procedure, and its seasonal funding support. Stable funding for a significant period of time would allow for more systematic sampling and for the collection of larger sample groups. It may also be true that the inclusion of subjects with more than one cancer type or with variable cancer prognoses may create a bias in data collection toward more severe emotional responses to illness, possibly reflected in depression scores. Further, Beck Depression Inventory (BDI) scores tell only of symptoms and do not diagnose clinical depressive disorders, and the Defense Style Questionnaire is a relatively crude, if valid,²⁰ method of assessing ego-adaptation mechanisms. Although there is a wide representation of ego-defense mechanism types in the whole sample, the data here do not answer whether the level of ego-defense maturity may relate to cancer type—such as neoplasms associated with heavy nicotine or alcohol use that result in "excess mortality"—or to cancer stage—when more primitive ego-defense mechanisms may obstruct early recognition of signs or symptoms or may result in avoiding early medical attention. Effects such as these on survival probability will require systematic controls and future study.

An important variable that was not addressed in this investigation is the time-proximity of Beck Depression Inventory or Defense Style Questionnaire measurement to the cancer diagnosis itself in each subject. Theory suggests that ego-defense mechanisms will be engaged in response to any setting of stress or psychological conflict and thereby represent a ubiquitous phenomenon necessary for psychological adaptation. Longitudinal reports strongly suggest that specific adaptive styles are relatively stable over time.²⁴ For the purposes of this study, therefore, we considered ego-defense mechanism responses on the Defense Style Questionnaire as indications of characteristic responses from each subject, rather than as indications of acute stress—trait, rather than state, responses, so to speak. This question must be addressed in future studies by examining both the timing of Defense Style Questionnaire measurement after diagnosis and the longitudinal change in the Defense Style Questionnaire response over the course of the illness.

Despite these concerns, the data raise clinically important questions that would be fruitful topics of further research. Are the immature adaptors, because of problematic coping styles, simply less cooperative with their caregivers or less adherent to their treatment regimens? Are they more likely to engage in high-risk behaviors after receipt of a cancer diagnosis? Is the cancer simply a last stop in a life-course of poor concern for their physical well-being? Should clinical recognition and care be focused more on psychological adaptation and less on depressive symptoms?

Alternatively, clinicians may ask whether the most depressed patients care for themselves less well than the least depressed? Do the former "run out of gas" after managing bravely for a long while? Is the "fighting spirit" mentioned by other observers²⁵ a finite source of hope that can be exhausted in the most depressed patients? To what extent are the most depressed—and the immature adaptors, for that matter—likely to benefit from antidepressant medicines that might improve either mood or adaptation? Are there subgroups to be identified who might be helped in this way?

Despite the fact that the likelihood of recovering from various forms of cancer has improved significantly over recent years, studies suggest that mental adjustment in cancer patients remains one of the important factors correlating with quality of life and the degree of psychological distress.^25,26 Because previous treatment trials, including medication trials,^27–31 for depression in cancer patients present widely differing study designs, clinicians may reasonably ask whether measures of ego-defense adaptation might make better treatment targets for behavioral treatment or for psychoactive medications.

The data presented here point up the necessity of considering commonly observed clinical symptoms in relation to the underlying complexities of the mind/brain adaptive process that all humans must utilize as they meet the conditions of their being. Measures of ego-defense adaptation, such as that provided by the Defense Style Questionnaire, will require further development if they are to be brought into clinical use and offer a productive way of enhancing quality of life in the face of a diagnosis eliciting severe forms of distress. Systematic clinical measurement of ego-defense mechanisms may allow differentiation of depressive presentations most amenable to medication, to behavioral therapies, or to some combination of both. In this way, improved behavioral diagnosis can improve treatment specificity and may affect survival probability in those with the greatest difficulty in adapting to cancer.

ACKNOWLEDGMENTS

 
This study was selected as an oral presentation at the annual meeting of the Society for Biological Psychiatry, 2004. The authors thank Leah Widger, Allyson Miller, Heather Donn Romero, Amie Hull, Jody Combs, Kathleen Frechetti, and Elizabeth Major, all of whom worked with us on this project.

This study was funded in part by the VA (Dr. Beresford) and by the University of Colorado Cancer Center student fellowship program.

REFERENCES

TOP ABSTRACT INTRODUCTION METHOD RESULTS DISCUSSION REFERENCES

 

1. Fawzy FI, Cousins N, Fawzy NW, et al: Structured psychiatric intervention for cancer patients, I: changes over time in methods of coping and affective disturbance. Arch Gen Psychiatry 1990; 47:720–725[Abstract/Free Full Text]
2. Spiegel D, Bloom JR, Kraemer HC, et al: Effect of psychosocial treatment on survival of patients with metastatic breast cancer (comment). Lancet 1989; 2:888–891[CrossRef][Medline]
3. Goodwin PJ, Leszcz M, Ennis M, et al: The effect of group psychosocial support on survival in metastatic breast cancer (comment). N Engl J Med 2001; 345:1719–1726[Abstract/Free Full Text]
4. Spiegel D: Effects of psychotherapy on cancer survival. Nat Rev Cancer 2002; 2:383–389[CrossRef][Medline]
5. Buckley P, Conte HR, Plutchik R, et al: Psychodynamic variables as predictors of psychotherapy outcome. Am J Psychiatry 1984; 141:742–748[Abstract/Free Full Text]
6. Conte HR, Plutchik R, Buck L, et al: Interrelations between ego functions and personality traits: their relation to psychotherapy outcome. Am J Psychotherapy 1991; 45:69–77[Medline]
7. Vaillant GE: An empirically-derived hierarchy of adaptive mechanisms and its usefulness as a potential diagnostic axis. Acta Psychiatr Scand Suppl 1985; 319:171–180[Medline]
8. Perry J: Psychological defense mechanisms in the study of affective and anxiety disorders, in Comorbidity in Anxiety and Mood Disorders. Edited by Maser J, Cloninger C. Washington, DC, American Psychiatric Press, 1990, pp 545-562
9. Akkerman K, Carr V, Lewin T: Changes in ego defenses with recovery from depression. J Nerv Ment Dis 1992; 180:634–638[CrossRef][Medline]
10. Akkerman K, Lewin TJ, Carr VJ: Long-term changes in defense style among patients recovering from major depression. J Nerv Ment Dis 1999; 187:80–87[CrossRef][Medline]
11. Perry J, Hoglend P, Shear K, et al: Field trial of a diagnostic axis for defense mechanisms for DSM-IV. J Personal Disord 1998; 12:1–13[Medline]
12. Soldz S, Vaillant GE: A 50-year longitudinal study of defense use among inner-city men: a validation of DSM-IV defense axis. J Nerv Ment Dis 1998; 186:104–111[CrossRef][Medline]
13. Vaillant GE: Theoretical hierarchy of adaptive ego mechanisms: a 30-year follow-up of 30 men selected for psychological health. Arch Gen Psychiatry 1971; 24:107–118[Abstract/Free Full Text]
14. Trzepacz PT, Baker RW, Greenhouse J: A symptom rating scale for delirium. Psychiatry Res 1988; 23:89–97[CrossRef][Medline]
15. Melzack R: The McGill Pain Questionnaire: major properties and scoring methods. Pain 1975; 1:277–299[CrossRef][Medline]
16. Beck AT, Steer RA: Internal consistencies of the original and revised Beck Depression Inventory. J Clin Psychol 1984; 40:1365–1367[Medline]
17. Andrews G, Singh M, Bond M: The Defense Style Questionnaire. J Nerv Ment Dis 1993; 181:246–256[CrossRef][Medline]
18. Bond M, Gardner ST, Christian J, et al: Empirical study of self-rated defense styles. Arch Gen Psychiatry 1983; 40:333–338[Abstract/Free Full Text]
19. Bond M, Perry J, Gautier M, et al: Validating the self-report of defense styles. J Personal Disord 1989; 3:101–112
20. Perry JC, Hoglend P: Convergent and discriminant validity of overall defensive functioning. J Nerv Ment Dis 1998; 186:529–535[CrossRef][Medline]
21. Andrews G, Pollock C, Stewart G: The determination of defense style by questionnaire. Arch Gen Psychiatry 1989; 46:455–460[Abstract/Free Full Text]
22. Perry JC, Hoglend P, Shear K, et al: Field trial of a diagnostic axis for defense mechanisms for DSM-IV. J Personal Disord 1998; 12:56–68[Medline]
23. Fleming TR, Lin DY: Survival analysis in clinical trials: past developments and future directions. Biometrics 2000; 56:971–983[CrossRef][Medline]
24. Vaillant GE: The Wisdom of the Ego. Cambridge, MA, Harvard University Press, 1995
25. Watson M, Greer S, Rowden L, et al: Relationships between emotional control, adjustment to cancer, and depression and anxiety in breast cancer patients. Psychol Med 1991; 21:51–57[Medline]
26. Lampic C, Wennberg A, Schill JE, et al: Coping, psychosocial well-being, and anxiety in cancer patients at follow-up visits. Acta Oncol 1994; 33:887–894[Medline]
27. Holland JC, Romano SJ, Heiligenstein JH, et al: A controlled trial of fluoxetine and desipramine in depressed women with advanced cancer. Psycho-Oncology 1998; 7:291–300[CrossRef][Medline]
28. Costa D, Mogos I, Toma T: Efficacy and safety of mianserin in the treatment of depression in women with cancer. Acta Psychiatr Scand Suppl 1985; 320:85–92[Medline]
29. Pezzella G, Moslinger-Gehmayr R, Contu A: Treatment of depression in patients with breast cancer: a comparison between paroxetine and amitriptyline. Breast Cancer Res Treat 2001; 70:1–10[CrossRef][Medline]
30. Razavi D, Allilaire JF, Smith M, et al: The effect of fluoxetine on anxiety and depression symptoms in cancer patients. Acta Psychiatr Scand 1996; 94:205–210[Medline]
31. van Heeringen K, Zivkov M: Pharmacological treatment of depression in cancer patients: a placebo-controlled study of mianserin. Br J Psychiatry 1996; 169:440–443[Abstract/Free Full Text]

Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Email this article to a Colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to My Articles & Searches Download to citation manager Citing Articles via Google Scholar Articles by Beresford, T. P. Articles by Martin, B. Search for Related Content PubMed Citation Articles by Beresford, T. P. Articles by Martin, B. Other Patient Groups/Issues Primary Care

Get information about faster international access.

Privacy Policy

Copyright © 2006 Academy of Psychosomatic Medicine. All rights reserved.

Home | Search | Current Issue | Past Issues | Subscribe | All APPI Journals | Help | Contact Us