Experimentally Manipulating Dream Content by Reactivating a Remote Emotional Memory

Abstract Sleep and dreaming have long been hypothesized to serve a functional role in emotion regulation. Here, we investigated how reactivating highly emotional autobiographical memories prior to sleep affects dream content and subsequent emotional appraisal of these memories. Participants (N = 42) completed the Autobiographical Emotional Memory Task (AEMT) prior to bed for three consecutive nights, providing up to 3 verbal dream reports each night, with sleep monitored via wrist actigraphy. Consistent with emotion regulation theories of dreaming, we hypothesized that reactivating negative emotional memories before sleep would affect the content and emotion of dreams more than reactivating positive memories, and that participants who dreamed about negative memories would have greater overnight reductions in emotional reactivity toward them than those who did not dream about them. As hypothesized, the valence of reactivated memories strongly influenced participants’ dream emotions. Contrary to our hypotheses, positive memories were descriptively more likely to be incorporated into dream content than negative memories, although this difference did not reach significance. Furthermore, overnight changes in emotional appraisal of memories were unrelated to dream emotion or content. These observations demonstrate that it is possible to experimentally manipulate dreaming by reactivating highly emotional autobiographical memories before sleep. However, we find no support for preferential incorporation of negative emotional experiences into dreams, and no evidence that dreaming about emotional memories is associated with overnight changes in reactivity toward them.