Modulation of premotor cortex response to sequence motor learning during escitalopram intake

Abstract

The contribution of selective serotonin reuptake inhibitors to motor learning by inducing motor cortical plasticity remains controversial given diverse findings from positive preclinical data to negative findings in recent clinical trials. To empirically address this translational disparity, we use functional magnetic resonance imaging in a double-blind, randomized controlled study to assess whether 20 mg escitalopram improves sequence-specific motor performance and modulates cortical motor response in 64 healthy female participants. We found decreased left premotor cortex responses during sequence-specific learning performance comparing single dose and steady escitalopram state. Escitalopram plasma levels negatively correlated with the premotor cortex response. We did not find evidence in support of improved motor performance after a week of escitalopram intake. These findings do not support the conclusion that one week escitalopram intake increases motor performance but could reflect early adaptive plasticity with improved neural processing underlying similar task performance when steady peripheral escitalopram levels are reached.

Keywords: Functional magnetic resonance imaging; neural plasticity; post-stroke motor dysfunction; selective serotonin reuptake inhibitors; sequential motor learning.

Figure 1.

Study design and task: following baseline, escitalopram, or placebo administration took place for seven consecutive days. Post-baseline, motor training took place at single dose (first day), days 5 and 6, and at steady state (day 7). Motor training on days 5 and 6 was completed outside the scanner. fMRI data were acquired at baseline, single dose, and steady state. fMRI: functional magnetic resonance imaging. Notes: Task: sequential pinch force task; Force: the yellow bar controlled by the participants, the rise and fall of which was required to match the rise and fall of the blue (reference bar, i.e. the bar controlled by a computer).

Figure 2.

Sequential motor learning. Left: significant improvements in lag scores over five days of sequential motor training across both escitalopram and placebo. However, despite a significant learning effect, we observed no significant group differences in performance, nor did we observe an interaction effect. Right: comparison of the rate of change between baseline and steady state yield no significant group differences. Bold fonts indicate training completed in the scanner.

Figure 3.

Escitalopram-induced decreases in whole-brain cortical motor responses during sequential motor learning: Orthogonal brain slices showing group-dependent changes in the learning contrast over time. Mean functional group response (red) of the escitalopram group (top) and placebo (bottom) at each baseline, single dose, and steady-state measurements, as computed by a series of one-sample t-tests in SPM12. Single Dose > Steady State: brain regions in the escitalopram group with significant decreases in the learning contrast (blue) between single dose and steady state (top row) show decreases in bilateral premotor and temporal–parietal regions (Table 3). Comparisons between single dose with steady state in the placebo group do not yield any significant changes across time (bottom). Interaction: comparisons of groups over time reveal decreases in the whole-brain learning contrast in the left premotor cortex of the escitalopram group that are not observed in placebo (violet). Consideration of behavioral performance as a variable of interest shows brain regions where changes in the learning contrast positively correlate with improvement in motor performance, also with a peak in the left premotor cortex (overlaid in yellow). All results are shown for sequence-specific learning with p < 0.05 family-wise error (FWE) correction at a cluster forming threshold of p < 0.001. All orthogonal planes presented are the same. β = beta value at global maximum coordinate. See supplementary Table 3 for an overview of significant brain regions corresponding to correlation analyses.

Figure 4.

Correlations between escitalopram plasma levels and whole-brain cortical premotor response during sequence-specific learning from single dose to steady state: (a) Escitalopram plasma concentrations negatively correlate with changes in the whole-brain learning contrast in bilateral cortical motor regions, including the premotor cortex (premotor cortex from significant 2 × 2 interaction overlaid in yellow), with a peak in the left supramarginal gyrus. (b) Betas containing parameter estimates for error from the left premotor cortex plotted against escitalopram plasma levels at single dose and steady state, respectively. Results refer to the sequence-specific learning contrast and are shown with p < 0.05 family-wise error (FWE) correction at a cluster forming threshold of p < 0.001. ng/ml: nanograms/milliliters; β: beta value at premotor MNI coordinates. Note: see supplementary Table 4 for an overview of significant brain regions.