Could guided affective imagery support emotional access in eating-disorder care? A hypothesis-generating reading of a pilot study

Preprint. This Spotlight commentary manuscript has not been peer reviewed. Objective: Many people with eating disorders find emotions difficult to notice, distinguish, tolerate, accept, or understand. This common but non-universal challenge may complicate both care and recovery. In the multicomponent intervention described by Savidaki et al., the reported session structure placed guided imagery before movement. This Spotlight asks whether participant-controlled imagery may support emotionally meaningful access, and what, if anything, optional movement might contribute next. Method: A hypothesis-generating reading examined the published intervention description and anonymised diary extracts concerning guided imagery, the Inner Safe Place exercise, and imagery-to-movement sequences. Therapist observations and aggregate findings provided package-level context. Intervention-development and harms-reporting guidance informed a staged research agenda. Results: The intervention-group diaries provide a coherent signal of contact with emotion during complete sessions, including pleasant and unpleasant experiences. They cannot show improvement or identify imagery as the cause: no pre-movement emotional-access measure or equivalent control-group qualitative data were collected, and imagery was combined with movement, writing, group and therapist contact, and treatment as usual. Discussion: Future studies should assess emotional state before imagery, participant-defined emotional access after imagery and before movement, and participant-rated change after optional movement, followed by recovery, delayed effects, and harms. A participant-valued shift from unwanted numbness or disconnection towards tolerable, meaningful emotion, with choice and control preserved, may be an immediate process benefit. Participants may also value pleasant emotion or felt safety that lasts longer or feels stronger than at baseline; valence, intensity, or duration alone does not establish benefit or clinical improvement. The manuscript contains the full conflict-of-interest, data-availability, ethics, and artificial-intelligence assistance statements.