• In (A), the authors found that spindle power after memory cues predicted better memory retention, but spindle power before cues predicted worse memory retention. They reasoned that post-cue spindle power may benefit memory because they are involved in reactivation, but spindle power before cues may hurt memory because spindles have a refractory period, which limits how often they can occur, which makes spindles less likely to occur post-cue. In (B), they followed up on this idea with a new experiment where they developed a real-time algorithm and delivered cues either immediately after spindles (Early cues) or a few seconds after spindles (Late cues). In (C), they found less forgetting for memories associated with late cues, confirming that reactivation is unlikely to occur within the spindle refractory period, when spindles are unlikely.

  • Bursts of brain activity linked to memory reactivation

    A team of scientists from Princeton and Northwestern universities set out to find more direct and precisely timed evidence for the involvement of one particular sleep wave -- known as the “sleep spindle.” 
    In the study, sleep spindles, described as bursts of brain activity typically lasting around one second, were linked to memory reactivation, and that sounds during sleep can cue sleep spindles that improve memory, but can be blocked if a spindle has just occurred. The paper, “Sleep spindle refractoriness segregates periods of memory reactivation,” appears in the journal Current Biology.
    “The most novel aspect of our study is that we found these spindles occur rhythmically -- about every three to six seconds -- and this rhythm is related to memory,” said James W. Antony, first author of the study and a postdoctoral fellow in Princeton’s Computational Memory Lab.
    Three experiments explored how recent memories are reactivated during sleep. While volunteers took an afternoon nap, sound cues were surreptitiously played. Each was linked to a specific memory. The researchers’ final experiment showed that if cues were presented at opportune times such that spindles could follow them, the linked memories were more likely to be retained. If they were presented when a spindle was unlikely to follow, the linked memories were more likely to be forgotten. 
    Read More: