The impact of sleep on true and false memory across long delays

• True and false memory performance in the DRM task was probed 24 or 48 h after encoding.
• Subjects who slept soon after encoding had better true memory.
• To a lesser extent, sleep also increased false memory.
• Slow-wave sleep was negatively correlated with false recognition.
• SWS may be detrimental for gist-based false memory processing.

While the influence of sleep on memory has a long history, sleep’s role in the formation of false memories is less clear. Moreover, virtually nothing is known about the development of false memories beyond delays of about 12 h. Here, for the first time, we assess post-sleep development of true and false memories across longer delay intervals of 24 and 48 h. Although technically a false memory, remembering information that is related to the theme, or gist, of an experience can be considered an adaptive process. Some evidence suggests that sleep, compared to a wake period, increases both true and gist-based false memories in the Deese-Roediger-McDermott (DRM) task, but not all studies have returned this result, and most studies cannot rule out the possibility that sleep is merely protecting the information from interference, as opposed to actively aiding its consolidation. Here, to equate amount of time spent awake and asleep across groups, we assess how the positioning of sleep relative to memory encoding impacts retention across longer delays of 24 and 48 h. Participants encoded 16 DRM lists in the morning (WAKE 1st Groups) or evening (SLEEP 1st Groups), and were tested either 24 or 48 h later at the same time of day. Results demonstrate that true memory is better when participants sleep soon after learning. Sleeping first also increased false memory, but only in low performers. Importantly, and similar to previous studies, we found a negative correlation between slow-wave sleep (SWS) and false memory, suggesting that SWS may be detrimental for semantic/gist processing.