Targets and methods used in HITS experiments.

(A) The general logic of HITS experiments, whereby hippocampal indirect targets are used to define connectivity networks from which stimulation-accessible locations are selected and stimulated with TMS. Red coloration indicates an example group-level fMRI connectivity network of the hippocampal indirect target, from which a stimulation location is selected. A representative electrical field (60) induced by TMS of this location at a typical intensity (estimate of 100% MT) is shown. The field is thresholded at 66% of its maximum intensity, as stimulation of lower intensity should have negligible effects on neuronal activity (all analyzed studies used stimulation intensity of ~70% MT or above). (B) Hippocampal indirect targets and neocortical stimulation locations in the analyzed studies are shown overlaid on a template brain. An example resting-state fMRI connectivity network of the hippocampal indirect targets is displayed in red, as in (A). Stimulation locations are colorized separately for studies that used individualized targeting versus those that used group/atlas-based targets. For studies using individualized targeting, the average (centroid) location of the targets for all subjects in the study is shown. The hippocampal indirect targets are shown for those studies using individualized targeting. Note that not all stimulation locations fall within the highlighted red hippocampal network, as that specific network is shown for illustrative purposes to highlight proximity of all stimulation targets to a typical group-defined hippocampal network. (C) Coronal slices for the indicated positions show the hippocampal indirect targets and stimulation locations in greater detail. K-means clustering indicated that the majority of stimulation locations (84%) comprised a cluster within left parietal cortex, as indicated.

Studies identified by systematic review.

Study information is listed separately for studies that were included versus excluded from meta-analyses. Sample sizes listed are for the number of subjects contributing to statistical effects that were analyzed or described. *Studies that re-analyzed data from another study and therefore did not contribute independent data are indicated.

HITS selectively improved episodic memory overall.

(A) Forest plot of all effects of HITS on episodic-memory task outcomes, ordered by size. (B) Forest plot of all effects of HITS on non-memory task outcomes. The pooled effect is shown for both plots. Red circles indicate outliers that were excluded in sensitivity analyses. Circle size indicates the study’s weighted contribution to the meta-regression model.

Study factors that modulated effects of HITS on episodic memory.

(A) Factors that significantly modulated HITS effects on episodic memory. The pie chart indicates the percentage of effects for each category of each factor. Individual effects are shown in the box plots, grouped by factor levels. (B) Factors that did not significantly modulate HITS effects on episodic memory, plotted in the same format. *Indicates an effect modification that was significant in the main analysis but not in the sensitivity analysis.

Greater effects of HITS on memory in studies having optimized designs.

Box plots of expected effect sizes projected from the meta-regression model for studies in which HITS was applied pre-task and measured using recollection format tests, versus all other study factors. Circles are individual projected effects, with circle size indicating standard error of the prediction, as indicated.

Histogram of sample sizes per analyzed effect, plotted separately for episodic-memory and non-memory effects.

Relationships among factors in episodic memory studies.

Funnel plot of episodic memory HITS effects.

Outliers are indicated in red. The I2 statistic is indicated.

Funnel plot of non-memory HITS effects.

Outliers are indicated in red. The I2 statistic is indicated.

Description of study factors.