L’attachement du client au thérapeute en psychothérapie : une revue systématique des études empiriques

Abstract

Attachment theory describes characteristic patterns of relating to close others and has important implications for psychotherapy. Consistent with Bowlby's (1988) secure base conception of attachment in psychotherapy, several instruments have been developed to measure client attachment to therapist. Despite a large number of studies, no review takes into account all published work. The purpose of this systematic review was to compare different measures of client's attachment to therapist, with different groups of client and in different therapeutic contexts. Medline, Embase, Pubmed, PsycInfo, and Web Of Science were searched to identify studies published between 1995 and 2019 reporting on help-seeking client's attachment to therapist. Twenty-five empirical papers met the inclusion criteria. Meta-analyses were conducted for studies that examined client attachment to therapist subscales (Secure, Avoidant–Fearful, Preoccupied–Merger) as correlates of client-rated working alliance (K=11, 892 clients), and general adult attachment (K=11, 752 clients). The results show that the client's secure attachment to therapist is strongly correlated with the therapeutic alliance (mean weighted R=0,71 [95% CI=0.62–0.79]), moderately correlated with the results, and negatively correlated with the avoidance (mean weighted r=−0,12 [IC 95%=−0.06–−0.21]) and anxiety dimensions (mean weighted R=−0,11 [IC 95%=−0.03–−0.17]) of adult attachment. With the Avoidant–Fearful style, results go in the opposite direction, negatively correlated with the therapeutic alliance (mean weighted R=−0.55 [IC 95%=−0.59–−0.50]), negatively correlated with the results, and correlated with the avoidance dimension of adult attachment (mean weighted R=0,16 [IC 95%=0.09–0.23]). The Preoccupied–Merger style shows weak (positive or negative) to zero correlations. In addition, nine studies looked at some dimensions of the psychotherapy process (e.g., resistance, transference, emotions, self-disclosure, and attitude toward psychotherapy), showing that insecurity of attachment to the therapist (both Avoidant–Fearful style and Preoccupied–Merger style) seem to interfere in one way or another with the development of a productive psychotherapy process. Overall, these results are in line with what was expected. They are consistent with Bowlby's conception of a secure base of attachment in psychotherapy. They showed that dysfunction and maladaptive developmental experience interfere with adult's ability needed to establish secure attachments and that psychotherapy may play the role of what has been previously defined as corrective emotional–or intersubjective–experience. In addition, these results do not seem to vary according to the instruments used. Together, the high correlation with therapeutic alliance and the weak correlation with pretherapy adult attachment confirm the relevance of the client's attachment to therapist as a specific variable related to the process of change in psychotherapy.