Background and objectives: Attachment theory suggests that internal working models developed from early experiences with attachment figures biases cognitive appraisals a person makes of themselves and others. The current paper investigates whether attachment-related interpretative biases can be altered using Cognitive Bias Modification (CBM-I).

Methods: Eighty anxiously attached participants were randomly assigned to receive either secure or insecure CBM-I training. To measure training effects on attachment-related interpretation bias, participants read scenarios involving attachment figures whereby the availability of the attachment figure was undetermined, followed by test sentences that assigned an attachment-secure or -insecure interpretation to the situations. Participants rated the similarity of these sentences to the previously read ambiguous situations.

Results: Participants who received secure CBM-I training ascribed higher similarity ratings to secure sentence interpretations of ambiguous scenarios compared to insecure sentence interpretations. Attachment anxiety increased after CBM-I training for those who received insecure training, but did not differ for those who received secure training.

Limitations: This study was limited to healthy participants and did not include clinical participants. These findings need to be replicated by assessing the effects of CBM-I over an extended period.

Conclusions: CBM-I training may provide a viable means of modulating attachment anxiety. If validated with more potent strategies for secure attachment training, this approach could have significant implications for the treatment of affective disorders characterized by insecure attachments.