Schedule of Events | Symposia

Stability of working memory related theta synchrony over time

Poster Session B - Sunday, March 30, 2025, 8:00 – 10:00 am EDT, Back Bay Ballroom/Republic Ballroom

Jenna N. Pablo1 (jpablo@unr.edu), Beau Oster1,2, Marian E. Berryhill1; 1University of Nevada, Reno, 2Johns Hopkins University School of Medicine

There is tremendous interest in developing protocols to preserve or enhance working memory (WM) performance. One emerging target is strengthening underlying neural oscillations, including theta-gamma coupling to improve WM. Recent noninvasive brain stimulation (NIBS) studies tailor interventions to an individual’s peak phase-locking value (PLV) theta frequency over anterior-posterior sites. An assumption of this approach is that this peak value is stable across multi-session training paradigms occurring over time. To test this assumption, we examined (1) intra-individual and (2) inter-individual stability of peak PLV theta frequencies across time. We recorded EEG (BioSemi, 32-channel cap) from n=30 participants during three sessions (separated by >48 hours) as they performed a visual WM change detection task. Reaction times significantly improved across sessions (p<.001), but WM accuracy did not change. We conducted single-trial time-frequency decomposition with Morlet wavelets to extract the peak PLV (Fz to Pz electrodes) theta frequency over the delay period of correct trials. A mixed-effects model revealed no main effect of session on peak PLV frequencies (p=0.34). However, the low intraclass correlation coefficient (ICC=0.14) indicated substantial intra-individual variability (accounts for 86.3% of the variability) relative to inter-individual variability (13.7%). This suggests individualized peak PLV theta frequencies should be reassessed before each neurostimulation session to optimize tuning. These findings challenge the assumption of stable oscillations over time and raise questions regarding the mechanism by which NIBS benefits WM.

Topic Area: EXECUTIVE PROCESSES: Working memory

CNS Account Login

CNS2025-Logo_FNL_HZ-150_REV

March 29–April 1  |  2025

Latest from Twitter