Episodic memory improvements due to noninvasive stimulation targeting the cortical–hippocampal network: A replication and extension experiment

Abstract

Introduction The distributed cortical network of the human hippocampus is important for episodic memory. In a previous experiment, noninvasive stimulation of the hippocampal‐cortical network applied for five consecutive days improved paired‐associate learning measured after the stimulation regimen via cued recall (Wang et al., Science, 2014, 345, 1054). This finding has not yet been directly replicated. Furthermore, evidence for long‐lasting effects of stimulation on paired‐associate learning was obtained by analyzing relatively small subsamples (Wang & Voss, Hippocampus, 2015, 25, 877) and requires further evaluation. Methods Sixteen healthy young adults participated in this replication study using the same experimental design as the original study. Participants received 1 week of active stimulation and 1 week of sham stimulation, with memory assessments occurring at the beginning (pre) and end (post) of each week. Assessments included the paired‐associate task used in the original study, as well as a long‐term episodic memory retention task in order to test the hypothesis that increased paired‐associate learning could come at the cost of accelerated long‐term forgetting. Change in memory scores was evaluated within (pre vs. post) and across (active vs. sham) weeks. Results Similar to Wang et al., paired‐associate learning was significantly improved after 1 week of active stimulation but not after 1 week of sham stimulation. We found no evidence that stimulation increased long‐term forgetting for either week. Conclusion These findings confirm the beneficial effects of stimulation on episodic memory that were reported previously and indicate that stimulation‐related gains in new learning ability do not come at the price of accelerated long‐term forgetting.

Multi-day rTMS exerts site-specific effects on functional connectivity but does not influence associative memory performance

“…The absence of this mid-week task exposure in our study may have attenuated this effect. However, a recent study has successfully replicated these associative memory effects without a mid-week assessment in a sample of 16 healthy adults (Hermiller et al, 2019), suggesting that there may also be other factors influencing our observed outcomes. It is possible that inconsistent effects of rTMS on associative memory may more broadly reflect a variable response to the stimulation.…”

“…retrieving an online link to a questionnaire from a personal email inbox). We acknowledge that this filled delay protocol may differ from past studies (Wang et al, 2014;Hermiller et al, 2019). However, this task minimised the potential for participants to employ mnemonic rehearsal strategies during the delay period, which may have led to biased memory assessments.…”

“…Impressively, a subsequent follow-up study conducted on a smaller sub-sample of the same experiment (N = 8) revealed that these effects persisted for ~15 days following stimulation (Wang & Voss, 2015). Both the improvements in associative memory (Hermiller et al, 2019) and the increases in parieto-hippocampal connectivity (Freedberg et al, 2019b) following multi-day rTMS have recently been replicated in independent samples, although the sample sizes in these studies were moderate.…”

“…'The Winner's Curse', see Button et al, 2013). Encouragingly, the declarative memory and FC effects initially reported by Wang et al (2014) following multiday 20 Hz to the parietal-hippocampal network have since been replicated in similar moderate sample sizes (N = 15 -16) (Freedberg et al, 2019b;Hermiller et al, 2019). In light of these findings, investigating the reproducibility (i.e.…”

“…Past studies examining the effects of multi-day parietal-hippocampal rTMS on associative memory have been conducted in young-adult samples (Hermiller et al, 2019;Wang et al, 2014). Our sample was predominantly comprised of young adults (35 out of 40 aged between 18-31 years), while five participants were aged between 40-55 years.…”