A meta-analysis of the relation between dream content and memory consolidation

Lauren Hudachek,Erin J Wamsley
DOI: https://doi.org/10.1093/sleep/zsad111
IF: 6.313
2023-12-11
SLEEP
Abstract:The frequent appearance of newly learned information in dreams suggests that dream content is influenced by memory consolidation. Many studies have tested this hypothesis by asking whether dreaming about a learning task is associated with improved memory, but results have been inconsistent. We conducted a meta-analysis to determine the strength of the association between learning-related dreams and post-sleep memory improvement. We searched the literature for studies that (1) trained participants on a pre-sleep learning task and then tested their memory after sleep, and (2) associated post-sleep memory improvement with the extent to which dreams incorporated learning task content. Sixteen studies qualified for inclusion, which together reported 45 effects. Integrating across effects, we report a strong and statistically significant association between task-related dreaming and memory performance (SMD = 0.51 [95% CI 0.28, 0.74], p < 0.001). Among studies using polysomnography, this relationship was statistically significant for dreams collected from non-rapid eye movement (NREM) sleep (n = 10) but not for dreams collected from rapid eye movement (REM) sleep (n = 12). There was a significant association between dreaming and memory for all types of learning tasks studied. This meta-analysis provides further evidence that dreaming about a learning task is associated with improved memory performance, suggesting that dream content may be an indication of memory consolidation. Furthermore, we report preliminary evidence that the relationship between dreaming and memory may be stronger in NREM sleep compared to REM.
What problem does this paper attempt to address?