The effect of cathodal transcranial direct current stimulation during rapid eye-movement sleep on neutral and emotional memory

Abstract

Sleep-dependent memory consolidation has been extensively studied. Neutral declarative memories and serial reaction time task (SRTT) performance can benefit from slow-wave activity, characterized by less than 1 Hz frequency cortical slow oscillations (SO). Emotional memories can benefit from theta activity, characterized by 4-8 Hz frequency cortical oscillations. Applying transcranial direct current stimulation (tDCS) during sleep entrains specific frequencies to alter sleep architecture. When applying cathodal tDCS (CtDCS), neural inhibition or excitation may depend on the waveform at the applied frequency. A double dissociation was predicted, with CtDCS at SO frequency improving neutral declarative memory and SRTT performance, and theta frequency CtDCS inhibiting negative emotional memory. Participants completed three CtDCS conditions (Theta: 5 Hz, SO: 0.75 Hz and control: sham) and completed an SRTT and word recognition task pre- and post-sleep, comprising emotional and neutral words to assess memory. In line with predictions, CtDCS improved neutral declarative memory when applied at SO frequency. When applied at theta frequency, no negative emotional word memory impairment was found but a positive association was found between post-stimulation theta power and emotional word recognition. SRTT performance was also not altered by either CtDCS frequency. Future studies should investigate overnight theta CtDCS and examine the effects of CtDCS during and after stimulation.

Keywords: emotional; memory; procedural; sleep-dependent consolidation; tDCS.

Figure 1.

Example of a retrieval trial. Participants rated valence and arousal of the presented word on a nine-point scale as well as an ‘RKN’ memory judgement, with reaction times and accuracy being recorded.

Figure 2.

Example of a pre-sleep SRTT session. Participants pressed the key corresponding to the filled square onscreen, with accuracy and reaction times being recorded.

Figure 3.

Depiction of experimental protocol. W, wake; REM, rapid eye-movement sleep; N1, N2: stages of non-rapid eye-movement sleep (NREM); SWS, slow-wave sleep.

Figure 4.

Alertness, measured by the SSS, shown before and after sleep in each CtDCS condition.

Figure 5.

Performance on the word recognition task separately for each CtDCS condition and emotion category. (a) The number of correct trials; (b) Remember recognition performance measured by d’ (**p < 0.01).

Figure 6.

Valence and arousal ratings for encoding and retrieval separately for each CtDCS condition and emotion category. (a) Valence ratings during encoding; (b) valence ratings during retrieval; (c) arousal ratings during encoding and (d) arousal ratings during retrieval (**p < 0.01).

Figure 7.

SRTT accuracy, response times and amount of learning separately for each CtDCS condition before and after sleep. (a) Accuracy in control trials; (b) accuracy in transfer trials; (c) response times in control trials; (d) response times in transfer trials and (e) amount of learning (*p < 0.05,+p < 0.01).

Figure 8.

Relative spectral power during SWS and REM sleep separately for each CtDCS condition. (a) Relative power in the SWA (0.5–2 Hz) band; (b) relative power in the theta (4–8 Hz) band (**p < 0.01).