Therapeutic Attachment and Outcome In Adult Psychotherapy

Jacobsen, C. F., Falkenström, F., Castonguay, L., Nielsen, J., Lunn, S., Lauritzen, L., & Poulsen, S. (2024). The relationship between attachment needs, earned secure therapeutic attachment and outcome in adult psychotherapy. Journal of Consulting and Clinical Psychology, 92(7), 410–421. https://doi.org/10.1037/ccp0000900

Rationale

Attachment theory has increasingly been applied to psychotherapy research to understand therapeutic processes and outcomes (Mikulincer et al., 2013).

The therapeutic relationship may function similarly to early attachment bonds by providing a “secure base” for clients (Bowlby, 1988).

Recent research has explored how secure therapeutic attachment may catalyze change (Mallinckrodt et al., 2015, 2017) and how therapists’ management of therapeutic distance may facilitate corrective experiences for clients with different attachment needs (Daly & Mallinckrodt, 2009).

However, the full sequence of associations between pretreatment attachment, development of secure therapeutic attachment, and client experiences of growing engagement or autonomy has not been comprehensively investigated in a large outpatient sample.

• Growing engagement measures the client’s experience of decreased anxiety about and in therapy, and a growing comfort with emotional disclosure. It reflects the client becoming more relaxed and comfortable discussing personal topics in therapy.
• Growing autonomy measures the client’s experience of increased agency and self-reliance in therapy. It reflects the client’s perception of becoming more independent and self-reliant through the therapeutic process.

Additionally, the role of these dynamics as potential change mechanisms impacting interpersonal outcomes remains unclear.

This study aimed to address these gaps by testing an extended version of the T-GRAF model (Mallinckrodt et al., 2015) to elucidate attachment processes in therapy and their relationship to treatment outcomes.

Method

The study used a naturalistic prospective cohort design, collecting data from clients in individual psychotherapy with Danish practicing psychologists.

Multilevel modeling was used to analyze associations between attachment variables and outcomes.

Procedure:

Clients completed background questionnaires prior to therapy, including measures of adult attachment style and interpersonal problems.

During therapy, they completed measures of therapeutic attachment and therapeutic distance at regular intervals.

Post-treatment, clients again completed the measure of interpersonal problems.

Sample:

330 adult clients (75% female, mean age 40.6 years) seen by 44 therapists participated.

Clients presented with various mental health issues, with interpersonal difficulties, self-esteem problems, and burnout being most common.

Measures:

• Experiences in Close Relationships-Revised (ECR-R): Measured adult attachment insecurity
• Client Attachment to Therapist Scale (CATS): Assessed client-therapist attachment
• Therapeutic Distance Scale (TDS): Measured client’s experience of engagement/autonomy in therapy
• Inventory of Interpersonal Problems-32 (IIP): Assessed interpersonal difficulties

Controls:

By controlling for pseudosecurity in their analyses (through the ESAT measures), the researchers attempted to identify more genuine, “individuated-secure” therapeutic attachments that are not biased by these underlying insecure attachment patterns.

Pseudosecurity refers to characteristics or behaviors in the therapeutic relationship that may appear similar to secure attachment but are actually rooted in maladaptive attachment patterns.

The concept of pseudosecurity suggests that some clients may present as securely attached to their therapist, but this apparent security masks underlying attachment insecurities.

Statistical measures:

Multilevel growth models were used to analyze associations between attachment variables over time. Two-level models examined associations with treatment outcomes.

Results

Hypothesis I: Higher levels of individuated-secure (as opposed to anxious or avoidant pseudosecure) therapeutic attachment developed during therapy would be associated with higher levels of growing engagement and growing autonomy in therapy.

The results did not support the prediction as originally stated. The study found no significant positive associations between individuated-secure therapeutic attachment and growing engagement.

For growing autonomy, there was a significant association with ESAT_avoid, but it was in the opposite direction of what was predicted.

Higher levels of individuated-secure attachment (controlling for avoidant characteristics) were associated with decreases in growing autonomy over time, rather than increases.

Hypothesis II: Clients with higher pretreatment attachment avoidance, who developed a more individuated-secure attachment to the therapist during therapy, would experience
higher levels of growing engagement in therapy.

The study did not find support for this prediction.

Clients with higher pretreatment attachment avoidance who developed more individuated-secure attachment to the therapist did not show significantly higher levels of growing engagement in therapy.

The hypothesized relationship between attachment avoidance, secure therapeutic attachment, and growing engagement was not observed in this sample.

Hypothesis III: Due to the inconsistent findings in previous research, we explored the associations between clients with higher levels of pretreatment attachment anxiety, the development of individuated-secure attachment to the therapist during therapy, and the experience of growing autonomy without any a priori hypotheses.

Direct association:

• The preliminary model examining pretreatment attachment anxiety as an independent predictor found no significant associations between higher levels of attachment anxiety and the client’s experience of growing autonomy in therapy.

Interaction effect:

• The moderator model, which examined the interaction between pretreatment attachment anxiety and individuated-secure therapeutic attachment (ESAT_anx), found no significant interaction effect on growing autonomy in therapy.

Hypothesis IV: We explored a possible association between higher levels of growing engagement or growing autonomy in therapy, either as independent predictors or in interaction with higher levels of individuated-secure therapeutic attachment developed during therapy, and a decrease in
interpersonal problems at the end of therapy.

The study found that both higher levels of growing engagement and higher levels of individuated-secure attachment controlling for anxious characteristics (ESAT_anx) independently predicted decreases in interpersonal problems at the end of therapy.

However, growing autonomy and individuated-secure attachment controlling for avoidant characteristics (ESAT_avoid) did not show significant associations with therapy outcomes.

Additionally, the hypothesized interaction effects between these variables were not observed.

Insight

The study did not fully support the proposed ET-GRAF model, suggesting attachment dynamics in therapy may be more complex or context-dependent than theorized.

The finding that more secure therapeutic attachment (controlling for avoidant characteristics) predicted decreased growing autonomy could indicate corrective experiences for avoidantly attached clients, shifting them away from maladaptive self-reliance.

The association between growing engagement and improved interpersonal outcomes aligns with the model’s emphasis on engagement as a key process for avoidantly attached clients.

However, the lack of significant findings for many hypothesized relationships highlights the need for further research on how client characteristics, therapeutic approach, and treatment duration may moderate these dynamics.

Future research could benefit from more frequent repeated measures, investigation of therapist behaviors related to managing therapeutic distance, and examination of how attachment dynamics may differ across various client populations or treatment modalities.

Strengths

• Large, diverse sample of clients with various presenting problems
• Use of multilevel modeling to account for nested data structure
• Investigation of both anxious and avoidant attachment characteristics
• Examination of therapeutic attachment controlling for pseudosecurity
• Naturalistic design enhancing external validity

Limitations

• Reliance on self-report measures potentially influenced by social desirability or attachment biases
• Lack of repeated outcome measures limiting analysis of change processes
• Potential selection bias in client recruitment
• Predominantly Caucasian Danish sample limiting generalizability
• Heterogeneity in treatment approaches and durations complicating interpretation
• Missing data due to naturalistic design

Implications

The findings suggest that therapists should be attuned to clients’ attachment needs, which may vary based on pretreatment attachment styles and level of interpersonal impairment.

For some clients, particularly those with avoidant tendencies, fostering engagement rather than autonomy may be more beneficial.

The study highlights the potential importance of pseudosecurity in therapeutic attachment, with different types (anxious vs. avoidant) potentially having distinct clinical implications.

Clinicians may benefit from training to recognize signs of pseudosecurity and tailor interventions accordingly.

The results also suggest that brief, low-intensity treatments may be sufficient for some clients to establish secure therapeutic attachment, while others may require more prolonged or intensive interventions to modify attachment patterns.

The study underscores the complexity of attachment dynamics in therapy and the need for nuanced, individualized approaches to maximize interpersonal outcomes.

References

Primary reference

Jacobsen, C. F., Falkenström, F., Castonguay, L., Nielsen, J., Lunn, S., Lauritzen, L., & Poulsen, S. (2024). The relationship between attachment needs, earned secure therapeutic attachment and outcome in adult psychotherapy.Journal of Consulting and Clinical Psychology, 92(7), 410–421. https://doi.org/10.1037/ccp0000900

Other reference

Bowlby, J. (1988). A secure base. Basic Books.

Daly, K. D., & Mallinckrodt, B. (2009). Experienced therapists’ approach to psychotherapy for adults with attachment avoidance or attachment anxiety. Journal of Counseling Psychology, 56(4), 549–563.

Mallinckrodt, B., Anderson, M. Z., Choi, G., Levy, K. N., Petrowski, K., Sauer, E. M., Tishby, O., & Wiseman, H. (2017). Pseudosecure vs. individuated-secure client attachment to therapist: Implications for therapy process and outcome. Psychotherapy Research, 27(6), 677–691.

Mallinckrodt, B., Choi, G., & Daly, K. D. (2015). Pilot test of a measure to assess therapeutic distance and its association with client attachment and corrective experience in therapy. Psychotherapy Research, 25(5), 505–517.

Mikulincer, M., Shaver, P. R., & Berant, E. (2013). An attachment perspective on therapeutic processes and outcomes. Journal of Personality, 81(6), 606–616.

Keep Learning

Socratic questions for a college class to discuss this paper:

1. How might cultural factors influence the attachment dynamics observed in therapy? How could future research address this?
2. What are the ethical implications of therapists deliberately managing therapeutic distance to facilitate change? How can this be balanced with client autonomy?
3. How might the findings differ if the study had focused on a specific clinical population (e.g., individuals with personality disorders) rather than a general outpatient sample?
4. In what ways might technology-mediated therapy (e.g., teletherapy) impact the development of therapeutic attachment and the dynamics described in the ET-GRAF model?
5. How could attachment-informed interventions be integrated into evidence-based treatment protocols for specific disorders?