1. Gender differences in response to transference interpretations Randi Ulberg October 2009 UNIVERSITY OF OSLO, NORWAY

2. Per Høglend, research director Svein Amlo, clinical director Alice Marble Kjell-Petter Bøgwald Øystein Sørbye Mary Cosgrove Sjaastad Oscar Heyerdahl Paul Johansson Randi Ulberg Martin M. Nilsen Martin Furan Knut Hagtvet Jan Ivar Røssberg FEST - First Experimental Study of Transference-interpretations Department of Psychiatry Diakonhjemmet Hospital, Oslo Blakstad Hospital, Akershus Ullevål University Hospital, Division of Psychiatry Ullevål University Hospital, Centre for Child and Adolescent Psychiatry Department of Psychology, University of Oslo Dep. of Child and Adolecent Psychiatry, Vestfold Per Høglend, research director Svein Amlo, clinical director Alice Marble Kjell-Petter Bøgwald Øystein Sørbye Mary Cosgrove Sjaastad Oscar Heyerdahl Paul Johansson Randi Ulberg Martin M. Nilsen Martin Furan Knut Hagtvet Jan Ivar Røssberg Hanne –Sofie Dahl FEST - First Experimental Study of Transference-interpretations D Department of Psychiatry

3. Gender differences in response to transference interpretations Høglend, P., Bøgwald, K.P., Amlo, S., Marble, A., Ulberg, R., Sjaastad, M.C., Sørbye, Ø., Heyerdahl, O. & Johansson, P. (2008). Transference interpretations in dynamic psychotherapy: do they really yield sustained effects? Am. J. Psychiatry, 165: 763-771. Ulberg, R., Johansson, P., Marble. A. & Høglend, P. (2009). Patient sex as moderator of effects of transference interpretation in a randomized controlled study of dynamic psychotherapy. Can. J. Psychiatry, 54: 78-86. Ulberg, R., Høglend, P., Marble, A. & Johansson, P. (Submitted). Women respond favourably to transference interpretation, men do not: a randomized controlled study of long-term effects of dynamic psychotherapy Ulberg, R., Marble. A. & Høglend. P. (2009). Do gender and level of relational functioning influence the long-term treatment response in dynamic psychotherapy? Nordic. J. Psychiatry, 63: 412-419. Ulberg, R., Høglend, P., Marble, A., & Sørbye, Ø. (2009).From submission to autonomy; approaching independent decision making; A single case study in a randomized controlled study of long-term effects of dynamic psychotherapy American Journal of Psychotherapy, 63: 227-243.

4. FEST Patients referred from general practitioners, outpatient departments and private practice (N =100) Patients referred from general practitioners, outpatient departments and private practice (N =100) Randomisation: Transference/not transference Randomisation: Transference/not transference Dynamic Psychotherapy (1 weekly session;max 40 sessions) Dynamic Psychotherapy (1 weekly session;max 40 sessions) 7 therapists (5 men and 2 women) 7 therapists (5 men and 2 women) Same gender therapist – patient 56 % Same gender therapist – patient 56 % Different gender therapist – patient 44 % Different gender therapist – patient 44 % UNIVERSITY OF OSLO, NORWAY

5. Outcome measures Psychodynamic Functioning Scales (PFS) Psychodynamic Functioning Scales (PFS) Inventory of Interpersonal Problems (IIP-C) Inventory of Interpersonal Problems (IIP-C) (Dynamic Change) Global Assessment of Functioning (GAF) Global Assessment of Functioning (GAF) Global Severity Index - SCL-90 (GSI) Global Severity Index - SCL-90 (GSI) (Symptom Change) ) Global Severity Index - SCL-90 (GSI) Global Severity Index - SCL-90 (GSI) UNIVERSITY OF OSLO, NORWAY

6. Gender in Psychotherapy   In clinical theory it is recognized that patient gender contributes to the transference   Search: “gender OR sex AND psychotherapy” PubMed –result; 118 209 International Journal of Psychoanalysis – result: 2008; 41   Erotic transference UNIVERSITY OF OSLO, NORWAY

7. Erotic transference John William Waterhouse 1849 - 1917

8. Erotic transference - in treatment Opdal LC: Erotization of the analytic situation. Scandinavian Psychoanalytic Review. 2007;30:2-12. Opdal LC: Erotization of the analytic situation. Scandinavian Psychoanalytic Review. 2007;30:2-12. Gabbard G: Gabbard G: Boundaries, technique, and self-deception: a discussion (2008) Lessons to be learned from the study of sexual boundary violation (1996) Sexual excitement and countertransference love in the analyst (1994) Joy Scaverien : Gender, Countertransference and the Erotic Transference. (2006) Joy Scaverien : Gender, Countertransference and the Erotic Transference. (2006)

9. GENDER VS SEX Distinction: Distinction: GENDER (determined by psychological and cultural conditions) GENDER (determined by psychological and cultural conditions) SEX (determined by physical conditions) Stoller (1968). SEX (determined by physical conditions) Stoller (1968). GENDER is used because it is obvious that the gender-variable not only encloses the patient’s biological sex. GENDER is used because it is obvious that the gender-variable not only encloses the patient’s biological sex.

10. Quality of Object Relations (QOR) Lifelong pattern of relationships from primitive to mature Lifelong pattern of relationships from primitive to mature 7-8: Mature equitable relationships 7-8: Mature equitable relationships 5-6: Recent relationships may be difficult, but there is evidence of at least one mature relationship in the patient’s history 5-6: Recent relationships may be difficult, but there is evidence of at least one mature relationship in the patient’s history 3-4:Need of dependency or overcontrol in most relationships 3-4:Need of dependency or overcontrol in most relationships 1-2: Unstable, less gratifying relationships 1-2: Unstable, less gratifying relationships UNIVERSITY OF OSLO, NORWAY

11. Pretreatment characteristics of patients who received dynamic psycho-therapy of one year duration with (N=52) and without (N=48) transference interpretations. Age Age Global optimism ** Global optimism ** Expectancy Expectancy Motivation Motivation Quality of Object relations Quality of Object relations Female sex Female sex Single Single Depressive disorders Depressive disorders Anxiety disorders Anxiety disorders Other Other No diagnosis No diagnosis Personality disorders Personality disorders More than one pers.dis. More than one pers.dis. Transference Comparison 38 (9) 36 (10) 61 (14) 69 (13) 8 (2) 8 (2) 5.4 (0.6) 5.4 (0.6) 5.1 (0.8) 5.1 (0.8) 50 % 63 % 38 % 54 % 50 % 52 % 26 % 23 % 14 % 17 % 17 % 19 % 44 % 46 % 19 % 19 %

12. 101 pre-treatment characteristics Women and Men GSI PSYCHOFARMAKA SELF ATTACK HOSTILE MOTHER CARING INTRUSIVE DIVORCED AGE GAF EDUCATION CARING DOMINATING PAID WORK LIFE QUALITY PFS LOVE SELF DEPRESSION SED. MED. QOR INSIGHT ANTIDEPR. MED IIP BLAME SELF SENSITIVITY ANXIETY EXPECT MUTUALITY PREVIOUS TREATMENT FATHER CARING EXPECT. TREATMENT MARRIED DEAD PARENTS PARENTS PSYCH. ILL PROTECT SELF FRIENDSHIP SOMATIC D. MOTHER CONTROL PD SICK LEAVE SOLVE PROBLEM

13. QOR Men: N = 44 QOR 5.0 Men: N = 44 QOR 5.0 Women: N = 56 QOR 5.1 Women: N = 56 QOR 5.1 UNIVERSITY OF OSLO, NORWAY

14. Patient Gender as General Predictor in Individual Psychotherapy The majority of studies have found that improvement in therapy is independent of patient gender, but a small number of studies found women more responsive to psychotherapy than men. The majority of studies have found that improvement in therapy is independent of patient gender, but a small number of studies found women more responsive to psychotherapy than men. Sue and Lam (2001,2002),Bergin and Garfield (2004), Zlotnic et al (1996), Ogrodniczuk et al (2004) Sue and Lam (2001,2002),Bergin and Garfield (2004), Zlotnic et al (1996), Ogrodniczuk et al (2004) UNIVERSITY OF OSLO, NORWAY

15. Patient Gender as Moderator in Individual Psychotherapy : Patient Gender as Moderator in Individual Psychotherapy :  Male patients had better outcome in interpretive therapy than in supportive therapy  Female patients had better outcome in supportive therapy than in interpretive therapy  Moderator effect during therapy. No significant long-term moderator effect of gender Ogrodniczuk & al 2001 Ogrodniczuk & al 2001  Patients with Bipolar I Disorder after treatment with interpersonal and social rhythm therapy and/or intensive clinical management:  Women who received interpersonal and social rhythm therapy showed more marked and rapid improvement than men Frank et al Am J Psychiatry December 2008 Frank et al Am J Psychiatry December 2008 UNIVERSITY OF OSLO, NORWAY

16. Research questions; Research questions; during therapy Will patient gender predict outcome across treatments (No)? Will patient gender predict outcome across treatments (No)? Will there be an interaction effect ; Will there be an interaction effect ; gender /treatment with or without transference interpretations ? gender /treatment with or without transference interpretations ? Is patient gender a moderator over and above QOR ? Is patient gender a moderator over and above QOR ? UNIVERSITY OF OSLO, NORWAY

17. Gender in FEST at post-treatment No predictor effect of gender No predictor effect of gender UNIVERSITY OF OSLO, NORWAY

18. Recovered cases post-treatment Gender as moderator N = 18 25 a 30 26 a One outlier deleted from the transference group Proportion of recovered cases 0 10 20 30 40 50 60 70 80 90 100 MenWomen comparison transference GAF UNIVERSITY OF OSLO, NORWAY

19. Recovered cases post-treatment Combined QOR and gender 0 10 20 30 40 50 60 70 80 90 100 High QOR Men a Low QOR Women b comparison transference GAF Proportion of recovered cases N = 10 11 11 11 a Significant difference between treatments (x 2 =3.9, df=1, p=0.05) b Significant difference between treatments (x 2 =4.7, df=1, p=0.03) UNIVERSITY OF OSLO, NORWAY

20. PFS 0 10 20 30 40 50 60 70 80 90 100 High QOR MenLow QOR Women a Proportion of Recovered Cases comparison transference Combined QOR and gender N = 10 11 11 11 a Significant difference between treatments ( Fisher exact test: p < 0.032) UNIVERSITY OF OSLO, NORWAY a

21. Gender in FEST at post- treatment Significant gender differences in symptom change GAF and GSI at post-treatment (linear mixed model).Gender was a moderator of the effects of transference interpretations: Significant gender differences in symptom change GAF and GSI at post-treatment (linear mixed model).Gender was a moderator of the effects of transference interpretations: Women improve more with transference interpretations than men Women improve more with transference interpretations than men Men improve more without transference interpretations than women Men improve more without transference interpretations than women Pre-treatment variations between women and men (hostile and dominant) could not explain the outcome differences. Pre-treatment variations between women and men (hostile and dominant) could not explain the outcome differences.

22. Long -Term Responses We wanted to explore whether the average QOR women and men responded differently to transference interpretation We wanted to explore whether the average QOR women and men responded differently to transference interpretation Outcome Measures: PFS and IIP-C (dynamic, interpersonal change) Outcome Measures: PFS and IIP-C (dynamic, interpersonal change) UNIVERSITY OF OSLO, NORWAY

23. Model predicted “true” trajectories of the PFS for the transference group and comparison group within the subsamples of female patients. Controlled for the effects of QOR. Effect size 0.49 0.59 0.69 0.77 p=0.08 p=0.03 p=0.01 p=0.01

24. Model predicted “true” trajectories of the PFS for the transference group and comparison group within the subsamples of male patients. Controlled for the effects of QOR.

25. Model predicted “true” trajectories of the PFS for the transference group and comparison group within the subsamples of male and female patients. Controlled for the effects of QOR. Between groups effect sizes (Cohen’s d) for Women : 0.59 0.69 0.77 p=0.03 p=0.01 p=0.01

26. Match and mismatch therapist-patient gender There were 56 matched and 43 mismatched therapy dyads. There were 56 matched and 43 mismatched therapy dyads. For women treated by women the treatment effect of transference interpretations became stronger than for women in general. (The estimate increased from 1.16 to 1.56) For women treated by women the treatment effect of transference interpretations became stronger than for women in general. (The estimate increased from 1.16 to 1.56)

27. Gender as moderator of long-term treatment effects, controlled for Quality of Object Relations Scale (QOR); Low QOR women Dependent variable and parameter Estimate 90% CItdfp Psychodynamic Functioning Scales: Women coded 0 and men coded 1 Intercept 60.4759.48 to 61.45101.9120,5200.000 Time (log)3.753.18 to 4.3211.0117,0010.000 Time x Treatment1.870.79 to 2.942.999,7120.005 QOR( - 4.41)4.373.60 to 5.139.5113,5450.000 Gender1.360.18 to 2.531.9112,5070.059 Time x Treatment x (QOR - 4.41) -1.20-1.94 to -0.47-2.7109,4650.008 Time x Treatment x Gender -1.28-2.39 to –0.17-1.9110,4460.059

28. Trajectories of the Psychodynamic Functioning Scales (PFS) for the transference group and comparison group within the sub samples of female patients with low scores and male patients with high scores on the Quality of Object Relations Scale (QOR)

29. Summary gender-findings FEST  Women respond significantly better than men to transference interpretation during therapy and during long-term follow-up.  Poor relational functioning women benefit much more from transference interpretations than good relational functioning men.  During therapy: Gender differences measured with GAF and GSI (symptomatic change)  During follow-up: Gender differences measured with GSI (symptomatic change) and PFS (dynamic change)  When including the two pre-treatment differences (dominant and hostile) between men and women as covariates in the model, the moderator findings became stronger for both outcome measures.  Match and mismatch therapist-patient gender; women treated by women profited even more from transference interpretations

31. Poor relational functioning women is the sub group of patients in FEST which improve most from dynamic psychotherapy with transference interpretations

32. A Single Case Study: From Submission to Autonomy: Approaching Independent Decision Making. Explore one single case to shed light on and illustrate a good outcome therapy process in one woman with poor relational functioning. Explore one single case to shed light on and illustrate a good outcome therapy process in one woman with poor relational functioning. Hypothesize that changes in symptoms, self-image, and changes in insight and interpersonal functioning would coincide well with the changes in the process between therapist and patient and the therapist’s countertransference feelings. Hypothesize that changes in symptoms, self-image, and changes in insight and interpersonal functioning would coincide well with the changes in the process between therapist and patient and the therapist’s countertransference feelings.

33. Anonymous patient and therapist Miklós Ligeti1871-1944

34. From Submission to Autonomy Method Case formulations Case formulations Repeated applications of self–reports and observer–rated measures, (during and after therapy) Repeated applications of self–reports and observer–rated measures, (during and after therapy) Transcription of sessions; detailed observer ratings of the therapist–patient interaction, using Structural Analysis of Social Behaviour (SASB) Transcription of sessions; detailed observer ratings of the therapist–patient interaction, using Structural Analysis of Social Behaviour (SASB)

35. Egon Schiele 1890-1918

36. Munch 1863-1944

41. Patient’s Expectations TREATMENT HELPS Insight Independen t Less exploitable

42. Example transference interpretation: Therapist: You experienced that you became attracted to that handyman because he seemed to be somehow attentive and caring towards you. You fell in love very fast like you have done since you were a teen-ager Patient: Yes, ehm Therapist: I don’t know, but when I was attentive towards you and linked your relation to me to your relation to other people, you didn’t fall in love with me, but it seems you became more attentive towards yourself Patient: Yes, that is absolutely right. I really wondered what happened with me when I met that guy. I find it easier to reflect here with you…….I’m not hiding ….my feelings…I don’t fall in love to escape….

43. Egon Schiele 1890-1918

44. Marc Chagall 1887-1985

45. Low QOR woman; transference therapy session 07 (SASB)

46. Low QOR woman; transference therapy session 16 (SASB)

47. Low QOR woman; transference therapy session 29 (SASB)

48. Changes in process and outcome measures in a good outcome therapy during 1 year dynamic psychotherapy and up to 3 year follow-up Process Weighted autonomy a Therapist Weighted autonomy a Patient Counter transference; Positive b Counter transference; Negative b Counter Transference; Paternalistic b Pre - treatment-16.25-4.32.70.23.0 Mid-treatment9.2718.12.80.02.3 Late in treatment9.9715.43.20.02.3 1 year post- treatment ---- 3 year post- treatment ---- Outcome Depression c Insight d Internal dialogue with therapist e Interpersonal functioning d Exploitable f Pre - treatment1.965.7-65.02.00 Mid-treatment0.3---0.69 Treatment termination 0.080.07.273.70.79 1 year post- treatment 0.081.02.081.10.50 3 year post- treatment 0.080.32.079.80.50

49. At 3-year follow-up: “I am more in contact with myself and take myself more seriously and see my own worth. I can forgive myself and ask for help if I need it”. “I am more in contact with myself and take myself more seriously and see my own worth. I can forgive myself and ask for help if I need it”.

50. Summary single- case The process between the therapist and the patient showed friendly complementarity The process between the therapist and the patient showed friendly complementarity The autonomy–encouraging interaction coincided with the positive changes in The autonomy–encouraging interaction coincided with the positive changes in depressive feelings depressive feelings insight insight self–image self–image interpersonal functioning. interpersonal functioning. The patient became less exploitable and more autonomous, independent, and capable of choosing what was best for her The patient became less exploitable and more autonomous, independent, and capable of choosing what was best for her

51. Discussion; Findings in brain function 1. 1. Gender differences in brain activation-pattern (fMRI) between men and women during processing of negatively valenced words versus non-words Women showed greater activation in areas participating in understanding and expressing language and processing emotions and memories. Men showed greater activation in areas important for executive functions and behavior. Transference interpretations often spot difficult conflict areas. Gender-related neural responses to emotional stimuli might be connected to different response between men and women to transference interpretations. Hofer et al; Sex differences in brain activation patterns during processing of positively and negatively valenced emotional words. Psychological Medicine 2007 (pp109-119) 2. 2. Oxytocine (higher in women than in men) partly determines interpersonal affiliation during loss of social contact. Mechanisms distinctly different in women and men. Taylor SE and Gonzaga GC; Affiliative responses to stress. In Harmon-Jones E, Winkielman P (Eds.) Social Neuroscience. NY Guilford Press. 2007 (pp 454-473) UNIVERSITY OF OSLO, NORWAY

52. Summary Women responded significantly better than men to transference interpretation: Women responded significantly better than men to transference interpretation: Post-treatment; a difference in symptom change between women and men Post-treatment; a difference in symptom change between women and men Long-term; different symptom and dynamic change between women and men Long-term; different symptom and dynamic change between women and men Female patients, who have difficult relationships to other people (low QOR) was the sub group of patients in FEST showing the best treatment effects from dynamic psychotherapy with transference interpretations. Female patients, who have difficult relationships to other people (low QOR) was the sub group of patients in FEST showing the best treatment effects from dynamic psychotherapy with transference interpretations. UNIVERSITY OF OSLO, NORWAY

54. Methode Psychodynamic Functioning Scales (PFS) Inventory of Interpersonal Problems - Circumplex version (IIP-C) Global Assessment of Functioning (GAF) Symptom Checklist-90 (SCL-90) Feeling Word Checklist (FWC–58) Countertransference Therapist Representation Inventory (TRI) Structural Analysis of Social Behaviour (SASB)

55. Hypotheses – during therapy (short-term resposes) Patient gender will not predict outcome across treatments Men and women may respond differently to dynamic psychotherapy with and without transference interpretation at post- treatment, but we make no prediction with regard to the direction of an interaction effect UNIVERSITY OF OSLO, NORWAY

56. Gender as a Moderator of Treatment Effects GAF dependent variable ParameterEstimateDft-valuep-value90%CI Intercept 59,44104,770,560,001 58,04 to 60,84 Time 5,00102,19,790,0014,16 to5,86 Time x Treatment 0,9093,61,150,25-0,40to2,21 Gender 3,44109,02,790,011,40to5,49 Time x Treatment x Gender -2,78102,12,970,01 -4,33to-1,23 UNIVERSITY OF OSLO, NORWAY

57. Gender as Moderator of Treatment Effects over and above QOR GAF dependent variable ParameterEstimate 90% CI tdf p Intercept 57.4555,84 – 59,0559,46105,80,001 Time 4.934,09 – 5,779,7199,70,001 Time x treatment 2.110,56 – 3,672,2595,40,03 QOR – 4,41 2.801,54 – 4,4073,69106,60,001 Gender 3.671,73 – 5,623,14106,60,001 Time x Treatment x QOR(-4,41) -1.49-2,53 - -0,452,3799,60,02 Time x treatment x Gender -3,07-4,63 – 1,503,2499,60,001 UNIVERSITY OF OSLO, NORWAY

58. Munch 1863-1944

59. So we see that you avoid talking about this because you feel anxious and uncomfortable discussing your sadness and anger over your fathers death. You did the same after your divorce and again now that we are ending therapy. Psychodynamic interpretation Defence Anxiety Impulse Parents Others Therapist DEFENCE ANXIETY IMPULSE PARENTS OTHERS THERAPIST

60. Level of transference interventions Degrees of comprehension from superficial to profoundanalysis of the emotions and behavioral patterns Level 1: the therapist addressed transactions in the patient–therapist relationship. Level 2: the therapist encouraged exploration of thoughts and feelings about the therapy and the therapist. Level 3: the therapist was to encourage the patient to discuss how the patient believed the therapist might feel or think about the patient. Level 4: the therapist includes himself explicitly in interpretive linking of dynamic elements (conflicts), direct manifestations of transference, and allusions to the transference. Level 5: the therapist interprets repetitive interpersonal patterns and links these patterns to transaction between the patient and the therapist.

61. Gabriel Byrne; In Treatment Anonymous therapist