Better with age: Developmental changes in oscillatory activity during verbal working memory encoding and maintenance

Abstract

Numerous investigations have characterized the oscillatory dynamics serving working memory in adults, but few have probed its relationship with chronological age in developing youth. We recorded magnetoencephalography during a modified Sternberg verbal working memory task in 82 youth participants aged 6-14 years old. Significant oscillatory responses were identified and imaged using a beamforming approach and the resulting whole-brain maps were probed for developmental effects during the encoding and maintenance phases. Our results indicated robust oscillatory responses in the theta (4-7 Hz) and alpha (8-14 Hz) range, with older participants exhibiting stronger alpha oscillations in left-hemispheric language regions. Older participants also had greater occipital theta power during encoding. Interestingly, there were sex-by-age interaction effects in cerebellar cortices during encoding and in the right superior temporal region during maintenance. These results extend the existing literature on working memory development by showing strong associations between age and oscillatory dynamics across a distributed network. To our knowledge, these findings are the first to link chronological age to alpha and theta oscillatory responses serving working memory encoding and maintenance, both across and between male and female youth; they reveal robust developmental effects in crucial brain regions serving higher order functions.

Fig. 1

Modified Sternberg verbal working memory task and behavior. (Top) Example trial of the modified Sternberg working memory task along with the corresponding correct response. (Bottom) Accuracy and reaction time were significantly correlated with chronological age. ₸ p < .001.

Fig. 2

Grand-averaged MEG sensor-level spectrogram. Grand-averaged time-frequency spectrogram taken from a representative posterior sensor (MEG2043) near parieto-occipital cortices. Time is denoted on the x-axis (0.0 s = encoding onset) and frequency (Hz) is shown on the y-axis. The time-frequency windows used for subsequent beamforming are denoted by the dashed black boxes. The spectrogram is shown in percent power change from baseline, with the color scale shown to the right of the spectrogram.

Fig. 3

Developmental alterations in the neural oscillations serving working memory encoding. Whole-brain correlation maps and associated scatterplots showing the relationship at the peak voxel, with age in years on the x-axes and oscillatory amplitude in pseudo-t units on the y-axes. (A-B) Theta oscillations became stronger with increasing age in the bilateral occipital cortices. (C-E) Alpha oscillations became stronger with increasing age in the left dorsal prefrontal cortices, left supramarginal, and left occipito-temporal cortices, while the opposite pattern was observed in the right anterior prefrontal cortices. **p < .005; ^{₸ ₸}p < .001.

Fig. 4

Developmental alterations in the neural oscillations serving working memory maintenance. Whole-brain correlation maps and associated scatterplots showing the relationship at the peak voxel, with age in years on the x-axes and oscillatory amplitude in pseudo-t units on the y-axes. (A–C) Alpha oscillations became stronger with increasing age in the right occipital cortices, right parietal, left prefrontal, and left occipito-temporal cortices. **p < .005; ^{₸ ₸}p < .001.

Fig. 5

Sexually divergent developmental patterns of neural oscillations serving working memory encoding and maintenance. Whole-brain Fisher-r-to-Z correlation maps and associated scatterplots showing the relationship at the peak voxel, with age in years on the x-axes and oscillatory amplitude in pseudo-t units on the y-axes. (A-B) Theta oscillations became weaker with increasing age in posterior cerebellar cortices among females, and weaker with increasing age in anterior cerebellar cortices among males. (C) Alpha oscillations became stronger with increasing age in right superior temporal cortices among females, but weaker with increasing age among males. *p < .05; ^₸p < .01; **p < .005; ^{₸ ₸}p < .001.