Attachment theory, developed by Bowlby to explain human bonding, has profound implications for conducting and adapting psychotherapy. We summarize the prevailing definitions and measures of attachment style. We review the results of three meta-analyses examining the association between attachment anxiety, avoidance, and security and psychotherapy outcome. Fourteen studies were synthesized, which included 19 separate therapy cohorts with a combined sample size of 1,467. Attachment anxiety showed a d of -.46 with posttherapy outcome, while attachment security showed a d of.37 association with outcome. Attachment avoidance was uncorrelated with outcome. The age and gender composition of the samples moderated the relation between attachment security and outcome: samples with a higher proportion of female clients and a higher mean age showed a smaller relation between security and outcome. We discuss the practice implications of these findings and related research on the link between attachment and the therapy relationship.

No abstract

This meta-analysis synthesizes research on the relation between patient adult attachment style and patient-rated working alliance. A random-effects model was used to calculate the mean weighted product-moment correlation (r) for 24 studies (12 published in peer-reviewed journals and 12 unpublished doctoral dissertations) of individual outpatient therapy with adults. The mean weighted r for attachment avoidance and alliance was -.137, p<.001, and the mean weighted r for attachment anxiety and alliance was -.121, p<.001. These findings suggest that therapists should attend to attachment in order to foster alliance and have additional implications for theory and future research.

Patients' psychotherapy outcome expectations and credibility beliefs are associated with adaptive treatment processes and outcomes. However, little is known about correlates of such beliefs. We explored how participant characteristics relate to patients' baseline and early treatment beliefs. Data for this preliminary correlational study derived from 65 adult outpatients in psychotherapy-as-usual. Self-report measures assessed patient (demographics, general and mental health history, global and specific symptom severity, and diagnostic comorbidity), therapist (demographics and theoretical orientation), and dyad (demographic match) characteristics, and patients' baseline treatment outcome expectations and their post-Session 1 outcome expectations and credibility beliefs. Bivariate correlations indicated that patients' manic symptoms were positively associated with baseline expectations ( ϭ .292, p ϭ .002), while substance abuse symptoms ( ϭ Ϫ.209, p ϭ .038) and Axis II comorbidity (r ϭ Ϫ.257, p ϭ .045) were negatively associated with such expectations. Patients' psychotic symptoms were positively associated with post-Session 1 expectations ( ϭ .221, p ϭ .035) and global functioning was positively associated with post-Session 1 credibility beliefs (r ϭ .343, p ϭ .016). Multivariate regression indicated a positive and negative association, respectively, between manic symptoms and baseline expectations (␤ ϭ .28, p ϭ .027) and substance abuse symptoms and baseline expectations (␤ ϭ Ϫ.33, p ϭ .032) when accounting for patient global symptom severity and Axis II comorbidity. Results highlight several associations between participant characteristics and patients' presenting and early treatment beliefs, suggesting the importance of attending to such characteristics in case formulation, treatment planning, and treatment implementation.

No abstract