Motor Cortex Excitability Reflects the Subjective Value of Reward and Mediates its Effects on Incentive Motivated Performance

Abstract

1Performance-based incentives tend to increase an individual's motivation, 2 resulting in enhancements in behavioral output. While much work has focused on 3 understanding how the brain's reward circuitry influences incentive motivated 4 performance, fewer studies have investigated how such reward representations act on 5 the motor system. Here we measured motor cortical excitability with transcranial magnetic 6 stimulation (TMS) while female and male human participants performed a motoric 7 incentive motivation task for prospective monetary gains and losses. We found that 8 individuals' performance increased for increasing prospective gains and losses. While 9 motor cortical excitability appeared insensitive to prospective loss, temporal features of 10 motor cortical excitability for prospective gains were modulated by an independent 11 measure of an individual's subjective preferences for incentive (i.e., loss aversion). Those 12 individuals that were more loss averse had a greater motor cortical sensitivity to 13 prospective gain, closer to movement onset. Critically, behavioral sensitivity to incentive 14 and motor cortical sensitivity to prospective gains were both predicted by loss aversion. 15 Furthermore, causal modeling indicated that motor cortical sensitivity to incentive 16 mediated the relationship between subjective preferences for incentive and behavioral 17 sensitivity to incentive. Together our findings suggest that motor cortical activity integrates 18 information about the subjective value of reward to invigorate incentive motivated 19 performance. 20 SIGNIFICANCE STATEMENT 21Increasing incentives tend to increase motivation and effort. Using a motoric 22 incentive motivation task and transcranial magnetic stimulation, we studied the motor 23 cortical mechanisms responsible for incentive motivated motor performance. We provide 24 experimental evidence that motor cortical sensitivity to incentive mediates the relationship 25 between subjective preferences for incentive and incentive motivated performance. 26These results indicate that, rather than simply being a reflection of motor output, motor 27 cortical physiology integrates information about reward value to motivate performance. 28

Reward-Based Improvements in Motor Control Are Driven by Multiple Error-Reducing Mechanisms

“…While the spinal stretch reflex is extremely fast, it is difficult to assume an effect of reward or motivation occurring at the spinal level. On the other hand, transcortical feedback includes primary motor cortex processing (Pruszynski et al, 2011), a structure that shows sensitivity to reward (Bundt et al, 2016;Galaro et al, 2019;Thabit et al, 2011). Consequently, an exciting possibility for future research is that transcortical feedback gain is directly enhanced by the presence of reward.…”

“…Similarly, in saccades, reward reduces reaction times and sensitivity to distractors (Manohar et al, 2015). Reports also indicate that reward invigorates movement execution by increasing peak velocity and accuracy during saccades (Takikawa et al, 2002) and reaching movements (Carroll et al, 2019;Galaro et al, 2019;Summerside et al, 2018). Together, these studies suggest that reward can shift the speed-accuracy function, at least in isolation, of both selection and execution.…”

Reward-driven enhancements in motor control are robust to TMS manipulation

“…The premotor area is central to movement planning and several studies have shown its sensitivity to reward (Ramkumar et al 2016;Olson 2003, 2004). Regarding M1, a large literature demonstrates effects of reward on various aspects of M1 processing (Bundt et al 2016;Galaro et al 2019;Kapogiannis et al 2008;Mawase et al 2016Mawase et al , 2017Ramkumar et al 2016;Thabit et al 2011), making it a suitable candidate for mediating the reward-driven effects observed in our study. Furthermore, we show in Codol et al (2020) that some execution improvements may be due to an increase in feedback control, likely transcortical (Omrani et al 2016;Pruszynski et al 2011) and visuomotor feedback (Carroll et al 2019).…”

“…However, this would not explain previous reports of reward-driven increase in feedback control (Carroll et al 2019;Manohar et al 2019) and end-point stiffness (Codol et al 2020) during reaching tasks. Rather, reward could directly modulate M1, as M1 activity has been shown to be highly sensitive to reward (Bundt et al 2016;Galaro et al 2019;Kapogiannis et al 2008;Mawase et al 2016Mawase et al , 2017Ramkumar et al 2016;Thabit et al 2011;Zhao et al 2018), shaping processing near the end of the sensorimotor arc. Another reasonable hypothesis is that reward information is integrated earlier on, with M1 being merely the final recipient.…”

Rat Paraventricular Neurons Encode Predictive and Incentive Information of Reward Cues

“…Furthermore, late responses might also reflect motor preparation process that might reduce lick latencies. Human behavioral and EEG studies reported that reward affects this process ( Mir et al, 2011 ; Schevernels et al, 2014 ), which might be mediated through the basal ganglia including the ventral striatum ( Pasquereau et al, 2007 ; Galaro et al, 2019 ). The PVT might affect lick latency through its projections to the nucleus accumbens (see below).…”