How Can Transcranial Magnetic Stimulation Be Used to Modulate Episodic Memory?: A Systematic Review and Meta-Analysis

Abstract

A systematic review and meta-analysis were conducted to synthesize the existing literature on how transcranial magnetic stimulation (TMS) has been used to modulate episodic memory. Given the numerous parameters of TMS protocols and experimental design characteristics that can be manipulated, a mechanistic understanding of how changes in the combination of parameters (e.g., frequency, timing, intensity, targeted brain region, memory task) modulate episodic memory is needed. To address this, we reviewed 59 studies and conducted a meta-analysis on 245 effect sizes from 37 articles on healthy younger adults (N = 1,061). Analyses revealed generally more beneficial effects of 1-Hz rTMS vs. other frequencies on episodic memory. Moderation analyses revealed complex interactions as online 20-Hz rTMS protocols led to negative effects, while offline 20-Hz rTMS led to enhancing effects. There was also an interaction between stimulation intensity and frequency as 20-Hz rTMS had more negative effects when applied below- vs. at-motor threshold. Conversely, 1-Hz rTMS had more beneficial effects than other frequencies when applied below- vs. at- or above-motor threshold. No reliable aggregate or hypothesized interactions were found when assessing stimulation site (frontal vs. parietal cortex, left vs. right hemisphere), stimulated memory process (during encoding vs. retrieval), the type of retrieval (associative/recollection vs. item/familiarity), or the type of control comparison (active vs. sham or no TMS) on episodic memory. However, there is insufficient data to make strong inference based on the lack of aggregate or two-way interactions between these factors, or to assess more complex (e.g., 3-way) interactions. We reviewed the effects on other populations (healthy older adults and clinical populations), but systematic comparison of parameters was also prevented due to insufficient data. A database of parameters and effects sizes is available as an open source repository so that data from studies can be continuously accumulated in order to facilitate future meta-analysis. In conclusion, modulating episodic memory relies on complex interactions among the numerous moderator variables that can be manipulated. Therefore, rigorous, systematic comparisons need to be further investigated as the body of literature grows in order to fully understand the combination of parameters that lead to enhancing, detrimental or null effects on episodic memory.

Keywords: episodic memory; non-invasive brain stimulation; recall; recognition; repetitive TMS; theta-burst stimulation; transcranial magnetic stimulation.

Figure 1

A total of 59 articles were included in the systematic review. The meta-analysis focused on healthy younger adults and rTMS in episodic memory resulting in 37 articles with 245 effect sizes.

Figure 2

Quality assessment findings for the 59 articles included in the systematic review. Blinding of outcome assessment reflects blinding of the experimenter in coding memory outcome measures (e.g., recall tasks).

Figure 3

Contour-Enhanced funnel plot. Reference line is set to g = −0.06. Contour lines represent p-value significance (i.e., <0.05, < 0.01) with the white area representing p-values > 0.05.

Figure 4

Residual funnel plots for moderator variables. The residuals are on the x-axis with the standard error on the y-axis. Contour lines represent p-value significant (i.e., <0.05, < 0.01) with the white area representing p-values > 0.05. Reference line set to g = −0.06.

Figure 5

Overview of effect sizes included in meta-analysis. Reference line set to the average effect size (g = −0.06). Legend represents study sample size ranges: 10–19, 20–29, 30–39, 40–49, 50–59, 60–69.