A study of stability and change in defense mechanisms during a brief psychodynamic investigation

This study investigated the stability of defensive functioning over the course of a 4-session Brief Psychodynamic Investigation (BPI). The sample included 61 outpatients from the Adult Psychiatry Department of the University of Lausanne, Switzerland.
of 7
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  O RIGINAL  A RTICLES  A Study of Stability and Change in Defense MechanismsDuring a Brief Psychodynamic Investigation  Martin Drapeau, PhD,*† Yves De Roten, PhD,* J. Christopher Perry, MPH, MD,† and  Jean-Nicolas Despland, MD* Abstract:  This study investigated the stability of defensive func-tioning over the course of a 4-session Brief Psychodynamic Inves-tigation (BPI). The sample included 61 outpatients from the AdultPsychiatry Department of the University of Lausanne, Switzerland.Defenses were measured from session transcripts using the DMRSquantitative method. Specific changes in defenses were found over the course of the ultrabrief investigation. First, the overall defensivefunctioning (ODF) score and the proportion of obsessional leveldefenses increased significantly, with a significant increase in intel-lectualization. Second, the number of defenses used and the propor-tion of narcissistic level defenses decreased, with a decreasing prevalence of devaluation and idealization. Third, high adaptive(mature) level defenses increased then decreased over the course of BPI, returning to their level at intake by the 4th session. Relief fromdistress and attending to the tasks of BPI tends to improve defensivefunctioning, but more likely returns it to usual levels rather than producing permanent change. Future studies will need to use designsthat allow estimation of state changes while taking sufficient mea-surements to estimate potential changes in trait levels of defenses. Key Words:  Defense mechanisms, Defense Manual Rating Scale,change, stability, brief investigation(  J Nerv Ment Dis  2003;191: 496–502) I n 1894, Freud published an article in which he presented theconcept of defense mechanisms for the first time and defined it as a central aspect of neurotic functioning. In later correspondence with Fleiss, he further addressed the questionof defense mechanisms by stating that they are aimed atmanaging libido and the pressing requests of the Id and theSuperego. As of 1926, Freud had already listed such defensesas repression, isolation, reaction formation, undoing, introjec-tion, and projection.In the decades that followed, this list was expanded  by A. Freud and later object-theory psychoanalysts ( e.g. ,Kernberg, 1967, 1980, 1992; Klein, 1928) to include subli-mation, displacement, various forms of denial, identificationwith the aggressor, different major and minor image distort-ing mechanisms, projective identification, and psychotic de-nial (Cooper, 1989; Despland et al., 2001a). Since these earlyworks of Freud and other pioneer psychoanalysts, manyclinicians and researchers have pointed to the necessity of  better understanding defense mechanisms. Once exclusivelyassociated with psychoanalytic psychotherapy, the concept of defenses has now been addressed in other fields (Caspar,1995), to the point where adding a 6th Axis to the psychiatricDSM is considered feasible, reliable, and valid (Perry and Hoglend, 1998; Perry et al., 1998).However, as Vaillant suggested 10 years ago (1992a),this shift in status from the realm of folk belief to that of scientific psychology requires addressing three crucial ques-tions. First, can defenses truly and reliably be identified?Second, if defense mechanisms can indeed be reliably iden-tified, can they be considered to have any predictive validity?Third, can the maturity level of the defensive functioning bedemonstrated to be independent of environmental factors?The first two questions have been abundantly documented inrecent years. Much research has demonstrated that defensemechanisms can be reliably identified by independent raters(Perry, 2001; Perry and Cooper, 1989; Perry et al., 1993;Vaillant, 1992b) with acceptable reliability. As for the second question raised by Vaillant, most of the studies in the nine-teen eighties were aimed at differentiating and predictingdifferent symptoms and Axis I and II disorders on the basis of defensive functioning. Although the theoretical conception of defenses was originally elaborated in relation to hysteria,Freud later considered it to be crucial in differentiatingdifferent disorders (Despland et al., 2001a; Freud, 1894).Many researchers have pursued this same line of thought and have shown that certain defense mechanisms can differentiateand predict various Axis I and Axis II pathologies as well asgeneral indexes of severity of psychopathology or of impair- *Unite´ de Recherche en Psychothe´rapie, Adult Psychiatry Department of theUniversity of Lausanne, Lausanne, Switzerland; and †The Institute of Community and Family Psychiatry, Sir Mortimer B. Davis JewishGeneral Hospital & McGill University, Montreal, Canada.This research was supported by the Swiss National Fund for ScientificResearch (FNRS), Grant no. 3200–05901.98.Reprints: Dr. Drapeau, Department of Counselling Psychology, McGillUniversity, 3700 McTavish, Montreal, Quebec, H3A 1Y2 Canada.E-mail: martin_drapeau@hotmail.comCopyright © 2003 by Lippincott Williams & Wilkins0022-3018/03/19108-0496DOI: 10.1097/01.nmd.0000082210.76762.ec The Journal of Nervous and Mental Disease   • Volume 191, Number 8, August 2003 496  ment (Maffei et al., 1995; Noam and Recklitis, 1990; Perry,2001; Vaillant and Drake, 1985).By contrast, the third question raised by Vaillant(1992a) regarding the independence of defensive functioningfrom the environment has rarely been addressed. AlthoughVaillant’s work did somewhat support the hypothesis of independence between defenses and environment, it alsoraised another crucial question: just how stable are an indi-vidual’s defenses after all?The stability of defenses has been considered by theo-reticians, many of whom posit that an individual’s defensivefunctioning is relatively stable and that defensive functioningis closely related to his personality structure or organization(Bergeret, 1985; Kernberg, 1984). As a central aspect of  personality, defenses are hence considered stable and changewould only be possible through intensive long-term psycho-therapy. Furthermore, the few existing studies on the subjecthave suggested that many aspects of defensive functioningdemonstrate moderate stability over time (Perry, 2001; Perryand Cooper, 1992). Recent longitudinal and follow-alongstudies have been designed to test the hypothesis that dy-namic factors such as defensive functioning are putativefactors that underlie personality functioning and hence should change more slowly than descriptive or symptomatic phe-nomena. Preliminary reports have indicated that defensivefunctioning indeed appears to be more stable than variousdescriptive measures ( e.g. , GAF), reporting that Overall De-fensive Functioning (ODF) improves slowly, demonstrating amoderate effect size only by 3 to 5 years of follow-along intreated samples (Perry et al., 2001).An interesting consequence of considering defenses asa basic psychological mechanism is that they can then beaddressed as both outcome and process phenomena in psy-chotherapy. As such, researchers have examined change indefensive functioning in long term (Perry et al., 2000) and short-to-medium term psychotherapy (Winston et al., 1994).Recent studies have given increasing attention to the interplay between defensive functioning and therapist interventions(Despland et al., 2001b). However, few studies, if any, haveexamined changes in defensive functioning in ultrabrief forms of therapy. Such investigations are important to pro-vide further information on the microchanges in defensive func-tioning and information on the stability of defenses during veryshort periods of time. Hence, this study aimed at examiningthe evolution of defenses in an ultrabrief form of therapyknown as the Brief Psychodynamic Investigation (BPI). METHODSPatients The final sample (  N   61) included 24 men (39%) and 37 women (61%) with a mean age of 29.13 ( SD    9.18).DSM-IV Axis I and II assessment was done by the assigned clinician at the end of the 4-session investigation. Most of the patients were seeking help for mood (62.3%) or anxietydisorders (36.1%), and occasionally for eating (4.9%), sexual(4.9%), or substance abuse related (3.3%) disorders. Themean number of Axis I diagnoses was 2 disorders. Thirty-eight percent presented a Cluster C personality disorder onAxis II. For a subsample of 36 of the patients, the reliabilityof the clinical diagnoses was established using the Guided Clinical Interview (GCI; Perry, 1993; Perry et al., 1999). Thekappa chance-corrected agreement between the GCI and thestructured assessment was acceptable, with an overall    0.65for the DSM-IV Axis I, and     0.54 for the DSM-IV Axis II.Finally, all patients were assessed using the SCL-90R.The mean scores at the first session were 0.90 ( SD  0.48) for the Global Symptom Index (GSI), 41.70 for the PositiveSymptom Total (PST;  SD  15.15), and 1.84 for the PositiveSymptom Distress Index (PSDI;  SD  0.47). Brief Psychodynamic Investigation Brief Psychodynamic Investigation (BPI) was devel-oped in Lausanne, Switzerland, by Gillie´ron (1989, 1996). Arecent manual guides this 4-session method of assessment for  psychodynamic psychotherapy, which focuses on the pa-tient’s reasons or motives for consultation and on the earlyinteraction between the therapist and the patient (Tadic et al.,2003). Brief Psychodynamic Investigation aims at: a) devel-oping an optimal plan to resolve the patient’s crisis situationthrough the use of an initial dynamic interpretation and itsworking through, b) establishing a psychodynamic and psy-chiatric diagnosis, c) providing information on indications for further therapeutic interventions, and d) furthering the devel-opment of early alliance. At the end of the first or the beginning of the second session, the therapist offers an initialinterpretation built on his understanding of the patient’s pretransference and Core Conflictual Relationship Themes(CCRT; Luborsky, 1998). The 2 or 3 following sessions arethen used to work through this hypothesis. The fourth and lastsession involves summarizing the 3 previous sessions and discussing future therapeutic support or interventions. Therapists The BPIs were completed by 3 female and 7 maletherapists affiliated with the Adult Psychiatry Department of the University of Lausanne (Switzerland). One therapist wasa licensed psychologist while the other 9 were licensed  psychiatrists. All had had prior training in BPI and an averageof 19 years of psychodynamic therapy practice (range 8–38years). The basic training in BPI involves 2 years of weeklyclinical supervision as well as theory seminars. Five thera- pists were considered to be senior therapists with a minimumof 3 years of experience in BPI after the basic training, whilethe remaining 5 were considered to be junior therapists withless than 3 years of experience in BPI after the training. The Journal of Nervous and Mental Disease   • Volume 191, Number 8, August 2003  Short-Term Change in Defenses  © 2003 Lippincott Williams & Wilkins  497  Instrument The Defense Manual Rating Scale (DMRS) was used toassess defense mechanisms (Perry, 1990; see Table 1). TheDMRS is an observer-rated method that can be applied to theaudio or video recording or the written transcription of various forms of interviews or of therapy sessions. We used the DMRS quantitative directions. Many studies have sup- ported the validity and reliability of the method (Guldberg etal., 1993; Perry, 2001; Perry et al., 1993; Skodol and Perry, 1993). The defenses described by the DMRS are virtuallycomparable to those listed and described in the DSM-IV(American Psychiatric Association, 1994). The instrument’slatest edition includes a total of 30 defense mechanismsassigned to 7 hierarchical levels of defensive functioning:high adaptive (mature), obsessional, other neurotic, minor image-distorting (narcissistic), disavowal, major image-dis-torting (borderline), and action defenses. Each of these levelsincludes 3 to 8 individual defenses. Hence, researchers and clinicians can examine a patient or sample in regards to the 7levels of defensive functioning or to the 30 individual de-fenses. Furthermore, as each level of defensive functioning isweighted according to its level of maturity, an overall defen-sive functioning score (ODF) can be computed. In this study,we rated defenses in all 244 BPI sessions (61 patients, 4sessions) employing 4 well-trained raters using standardized transcripts (Mergenthaler and Stigler, 1997; Mergenthaler and Stinson, 1992). Consensus ratings and reliability assess-ment were completed in more than 20% of the cases. Data Analysis All 4 sessions of the BPI were examined. Change in thenumber of defenses and in the ODF score from the first to thelast BPI session were first examined using repeated measureANOVA procedures. Paired-sample  t   tests examined differ-ences in ODF and in the number of defenses between specificsessions. The within-subjects effect size of change in ODF between sessions 1 and 4 was calculated using Glass’ delta ( d  ).Regarding the distribution of the defenses into the 7levels of defensive functioning, a doubly multivariate re- peated measures model (SPSS 11.0) was first used to deter-mine if the general distribution of the defenses differed over the 4 BPI sessions. The proportion of each of the 7 levels of defensive functioning was then considered individually. Afirst step involved the use of a repeated measure ANOVA todetermine if any change could be found in any one givenlevel over the 4 sessions. Students  t   tests were again used toexamine between which sessions any change occurs. Within-subjects effect sizes comparing sessions 1 and 4 were alsocomputed. Finally, when a significant change could be found for any given defensive level, we proceeded to examine the proportion of individual defenses. This was done using  t   testscomparing individual defenses in session 1 with those of session 4. Bonferroni corrections for     0.05 were made inall necessary cases.To further verify the stability of defenses, Pearsoncorrelations were used to examine the relation between thenumber of defenses, the ODF score, and each level of defensive functioning comparing the first session and the last. RESULTS Reliability of the 7 defensive levels was considered to be good with a median intra-class coefficient of 0.79; MeanICC (2,1)  0.75. Change in the Number of Defenses and ODF The repeated measure ANOVA procedure indicated that a significant difference could be found in the number of defenses over the 4 BPI sessions;  F  (1, 60)    52.15,  p   TABLE 1.  Levels of Defensive Functioning and Individual Defenses According to the DMRS a 7. Mature LevelAffiliation, Altruism, Anticipation, Humor, Self-assertion, Self-observation, Sublimation, Suppression6. Obsessional LevelIsolation, Intellectualization, Undoing5. Other Neurotic LevelRepression, Dissociation, Reaction Formation, Displacement4. Narcissistic LevelOmnipotence, Idealization (of Self or Object), Devaluation (of Self or Object)3. Disavowal Level Neurotic Denial, Projection, Rationalisation2. Borderline LevelSplitting of Other, Splitting of Self, Projective-Identification1. Action LevelActing out, Passive Aggression, Help-Rejecting Complaining a Perry, 1990.  Drapeau et al   The Journal of Nervous and Mental Disease   • Volume 191, Number 8, August 2003 © 2003 Lippincott Williams & Wilkins 498  0.001. The following  t   tests revealed that differences could befound between sessions 1 and 2,  t  (60)  2.33,  p  0.05; 1 and 3,  t  (60)    3.31,  p    0.01; and 1 and 4,  t  (60)    7.21,  p   0.001 (Table 2), with the mean number of defenses decreas-ing from 1 session to the next. A significant difference wasalso found between session 2 and session 4,  t  (60)  6.81,  p  0.001, and sessions 3 and 4,  t  (60)    5.68,  p    0.001. Theonly nonsignificant result was found between session 2 and session 3,  t  (60)  1.18, n.s.Pearson correlations indicated that the number of de-fenses used in the first session is correlated with the number of defenses used in the second ( r   0.38,  p  0.003) and third session ( r   0.31,  p  0.012), but not the fourth ( r   0.23,  p  0.06).Repeated measure univariate analysis of variance onthe ODF score indicated that a significant difference in ODFcould be found over the 4 BPI sessions,  F  (1, 60)  5.85,  p  0.05. Subsequent  t   tests indicated that the ODF score in-creased significantly from the 1st to the 4th session,  t  (60)  2.65,  p    0.01,  d     0.47. Significant improvements inODF were also found between the second and fourth ses-sions,  t  (60)  2.86,  p  0.01, and between the third and thefourth sessions,  t  (60)  3.00,  p  0.01. Change in the Levels of Defensive Functioning For the 7 defensive levels over the course of the BPI, adoubly multivariate repeated measures design indicated thatthe general distribution of the defensive functioning changed significantly over the 4 sessions,  F  (21, 40)  2.65,  p  0.01.Repeated measure univariate analysis of variance indicated that the use of mature level defenses was significantly differ-ent over the 4 sessions,  F  (1, 60)    13.45,  p    0.001, and followed a quadratic distribution (Fig. 1). Additional  t   testsshowed that the use of mature defenses decreased betweensessions 1 and 2,  t  (60)    2.79,  p    0.01 (Table 2), and sessions 1 and 3,  t  (60)    2.28,  p    0.05, but increased  between sessions 2 and 4,  t  (60)  2.35,  p  0.05. Hence, nosignificant difference could be found between sessions 1 and 4,  t  (60)  0.04, n.s.,  d   0.007.A significant linear change over the 4 BPI sessions wasalso found for the obsessional level defenses,  F  (1, 60)   13.68,  p    0.001 (Table 2). The patients’ use of suchdefenses increased significantly between sessions 1 and 2, t  (60)  2.17,  p  0.05, and 1 and 4,  t  (60)  3.79,  p  0.001, d   0.88. The last session was also characterized by a greater use of obsessional defenses than in sessions 2,  t  (60)  3.01,  p  0.05, and 3,  t  (60)  2.88,  p  0.05.Significant linear change was found for the narcissisticdefenses,  F  (1, 60)  14.83,  p  0.001. A  t   test revealed thatthe only session with a significant drop in narcissistic de-fenses was session 4. The last session of the BPI showed significantly less use of these defenses than sessions,  t  (60)  TABLE 2.  Defensive Level Means and Standard Deviations for All BPI Sessions and t-test Comparisons (N  61) Mean Proportion (SD) by BPI Sessiont-test Comparisonsfor Sessions 1 and 41 2 3 4  Number of defenses 31.03 (13.01) 26.97 (10.66) 25.21 (9.09) 17.74 (9.28) 7.21***ODF 4.37 (0.57) 4.38 (0.53) 4.32 (0.59) 4.65 (0.74) 2.65**Mature 9.75 (9.64) 6.46 (6.84) 6.75 (7.60) 9.82 (11.12) 0.04Obsessional 23.13 (11.69) 27.14 (14.18) 26.69 (14.26) 33.54 (18.88) 3.79***Other neurotic 15.50 (10.15) 16.80 (10.66) 14.92 (10.53) 14.88 (12.76) 0.32 Narcissistic 15.34 (10.77) 13.03 (9.40) 12.55 (8.87) 8.92 (10.41) 4.04***Disavowal 26.82 (13.28) 27.52 (14.37) 29.43 (14.67) 25.43 (16.95) 0.57Borderline 2.10 (4.01) 2.10 (4.50) 2.20 (5.15) 1.24 (3.09) 0.14Action 7.36 (7.52) 6.91 (7.33) 7.43 (9.97) 6.12 (7.73) 1.06 *  p  0.05; **  p  0.01; ***  p  0.001 FIGURE 1.  Evolution of defensive levels over the 4 BPI sessions. The Journal of Nervous and Mental Disease   • Volume 191, Number 8, August 2003  Short-Term Change in Defenses  © 2003 Lippincott Williams & Wilkins  499  4.04,  p  0.001,  d   0.59, 2,  t  (60)  2.71,  p  0.01, and 3,  t  (60)  2.34,  p  0.05.However, no significant change over the 4 BPI sessionswas found for the neurotic defenses,  F  (1, 60)  0.42, n.s. Thestability of neurotic defenses was also confirmed from onesession to another, including comparing the first and lastsessions of the BPI. Additionally, no significant change over the 4 BPI sessions was found for the disavowal,  F  (1, 60)  0.08, n.s., borderline,  F  (1, 60)    1.57, n.s., and actiondefenses,  F  (1, 60)    0.69, n.s. The stability of the patients’use of disavowal, borderline, and action level defenses fromone session to another was also confirmed by the  t   tests.Significant correlations were found between sessions 1and 4 for use of narcissistic ( r   0.31,  p  0.01) and actionlevel defenses ( r   0.28,  p  0.02). However, no significantcorrelation was found between session 1 and session 4 for useof mature ( r   0.05,  p  0.65), obsessional ( r   0.08,  p  0.53), neurotic ( r     0.20,  p    0.11), disavowal ( r     0.07,  p  0.55), and borderline defenses ( r   0.02,  p  0.98). Change in the Individual Defenses The results indicated that obsessional and narcissisticlevel defenses change significantly between the first and fourth BPI sessions. Hence, we further examined change inthe three obsessional, and three narcissistic level defenses.For the obsessional level defenses, a significant differencecould be found in session 1 and session 4 for use of intellec-tualization,  t  (60)  4.28,  p  0.001. The use of intellectual-ization almost doubles between these two sessions. No sig-nificant difference was found in the use of undoing whencomparing the first and the last sessions,  t  (60)  1.06, n.s. Nosignificant difference was found in sessions 1 and 4 for use of isolation after the Bonferroni corrections,  t  (60)  2.31, n.s.For the minor image-distorting (narcissistic) level de-fenses,  t   tests showed that the use of devaluation,  t  (60)   3.24,  p    0.002, and idealization,  t  (60)    2.44,  p    0.01,decreased sharply in the fourth session. Of these two de-fenses, devaluation is more prevalent at intake and session 4. No significant difference was found in the patients’ use of omnipotence,  t  (60)  0.71, n.s. DISCUSSION These findings suggest that specific changes in defensescan be found over the course of a 4-session Brief Psychody-namic Investigation. First, the overall defensive functioningscore and the proportion of obsessional-level defenses in-crease significantly. Second, the number of defenses and the proportion of narcissistic-level defenses decrease. More spe-cifically, intellectualization is the obsessional defense thatgained in overall proportion, and devaluation and idealizationare the 2 narcissistic defenses that decreased. Third, althoughno significant change was found between sessions 1 and 4 for the high adaptive (mature) level defenses, these defensesdecreased significantly during the BPI to finally return to their level at the first session. Finally, no significant change wasfound in other neurotic, disavowal, borderline, and actionlevel defenses.Such important changes in defensive functioning in avery short time period, with within-subject effect sizes reach-ing 0.88 for the obsessional defenses, are certainly not inagreement with findings in other studies. According to Perryet al. (2000) and Hersoug et al. (2003), similar improvementin defensive functioning requires between 1 and 2 years of  psychotherapy. Two important questions must be addressed,including determining whether change in defenses in such ashort period of time can be considered as an outcome measurereflecting trait change or whether it reflects a state-dependentimprovement, and determining if these same changes arerelated to clinical processes. TABLE 3.  Individual Obsessional and Narcissistic Defense Means and StandardDeviations for the First and Last BPI Sessions and t-test Comparisons (N  61) BPI SESSIONSM% (S.D.)t-test Comparisonsfor Sessions 1 and 41 4 ObsessionalIsolation 2.07 (3.64) 9.24 (2.53) 2.31Intellectualization 14.32 (11.14) 25.05 (17.94) 4.28***Undoing 6.40 (6.07) 7.51 (8.54) 1.06 NarcissisticOmnipotence 1.38 (2.61) 1.06 (2.81) 0.71Idealisation 2.93 (3.82) 1.46 (3.92) 2.44**Devaluation 11.02 (8.39) 6.40 (8.31) 3.24** *  p  0.05; **  p  0.01; ***  p  0.001  Drapeau et al   The Journal of Nervous and Mental Disease   • Volume 191, Number 8, August 2003 © 2003 Lippincott Williams & Wilkins 500
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks