Electronic Thesis and Dissertation Repository

Thesis Format

Integrated Article


Master of Science




Grahn, Jessica A.

2nd Supervisor

Batterink, Laura J.



In ‘targeted memory reactivation’ (TMR) paradigms, information learned during wakefulness is paired with a cue, and reactivated during sleep by presenting that same cue. TMR improves memory. In a prior study (Antony et al., 2012), participants learned two melodies. One melody was cued during a nap, and performance was better than for the uncued melody. The current study reanalyzed these data to characterize sleep spindle density during TMR cue-periods relative to non-cued periods, and whether spindle density correlated with performance. During TMR stimulation, spindle density was significantly higher than during non-stimulation in four time windows. Compared to the non-TMR group, higher spindle density occurred in two windows in the TMR group. Within-subject, spindle density was not correlated with accuracy, while between-subjects, spindle quantity correlated with post-nap accuracy improvements (r = .507). Thus, spindle density is altered at specific times by TMR, but TMR-specific density changes may not predict performance.

Summary for Lay Audience

‘Targeted memory reactivation’ (TMR) is a technique that uses a stimulus like sound or smell associated with prior learning to boost the memory-bolstering processes that happen during sleep. One of these processes is called sleep spindles, which are fast brainwaves that burst in a rhythmic, repetitive manner. In this study, a non-invasive electrophysiological monitoring method (elecroencephalography; EEG) was used to record electrical activity on the scalp that been shown to represent the activity of the surface layer of the brain underneath. Using data collected first reported in a previous, auditory TMR study (Antony et al., 2012) we investigated the occurrence of sleep spindles relative in time to the sounds during the sleep recordings. This was contrasted to silent periods within the same EEG recordings, as well as against a different group of nappers who were presented no sounds during sleep.

In the original study, participants produced a melody with button presses before napping, and after napping, with the melody presented to them during their nap. They found that participants were better at producing the that they had been presented during the nap when compared to a control melody.

We found higher concentrations of sleep spindles in four small time windows within the periods of sound stimulation when comparing to periods of no sound stimulation of the same participants’ recordings. When we compared to the group that received no sounds at all, we found that the TMR group had significantly higher concentrations of sleep spindles in two small time windows. This indicated that while TMR alters sleep spindle concentration, this may not alone correspond to better performance in cued tasks.