Without Blinking an Eye: Proactive Motor Control Enhancement

The process of stopping a planned action occurs in different behavioral modes quite frequently in everyday life. The cognitive mechanism of canceling an intended action has been termed “response inhibition” (Aron et al. 2004; Logan et al. 1984) or “reactive control” (Aron 2011). Many tasks are aimed at measuring the ability to control prepotent response tendencies. Among the most popular are the Go/No-Go and stop signal tasks (SST) (Aron et al. 2014). Numerous research studies have endeavored to enhance this ability by improving its performance in normal and clinical populations. For example, in an ADHD sample, pharmacological interventions such as methylphenidate improved SST performances (Devito et al. 2009). Cognitive training presented mixed results (Koster et al. 2017; Maraver et al. 2016; Spierer et al. 2013). Non-invasive brain stimulation proved to be a useful tool in improving inhibition-based performances (Jacobson et al. 2011; Kwon and Kwon 2013).

While reactive control refers to the process of stopping an intended action, another type of inhibition mechanism refers to a preparatory process occurring before a response tendency is triggered. This process is described as a proactive inhibition or proactive inhibitory control (Aron 2011; Verbruggen and Logan 2009). Selective proactive control impairments have been observed in various clinical populations, i.e., patients and their siblings diagnosed with schizophrenia showed reduced proactive control compared to normal subjects while their reactive inhibition remained intact. The behavioral results were supported by fMRI data showing weaker activation in the right striatum, right inferior frontal cortex, and left and right temporoparietal junction in the schizophrenia group compared to those of the normal controls during the SST (Zandbelt et al. 2011). Pani et al. (2013) demonstrated an opposite pattern in children with ADHD showing an impaired reactive control measured by the stop signal reaction time (SSRT) during the SST and an unimpaired proactive control measured by a general slowdown in reaction time as an adaptive strategy (Pani et al. 2013).

The majority of studies investigating impulsivity and self-control in normal and clinical populations utilized the reactive inhibition paradigm’s tasks and based on the results formulated conclusions as to the general inhibitory mechanisms (Barkley 1997; Corbetta and Shulman 2002; Lipszyc and Schachar 2010; Monsell 1996; Thomalla et al. 2013). Clearly, the data is of great importance, however, it only depicts part of the ability to control cognitive and behavioral processes. To overcome this lacuna, it is important to employ a task which measures the specific proactive inhibitory control. For this purpose, we chose to recruit a simple urge-driven motor action involving automatic and controlled processes—the eye blink (EB). EB is a natural motor behavior that can spontaneously occur or manifest as a reflex but can also be voluntarily controlled. Blinking can occur intentional or can be partially suppressed, thus, creating an uncomfortable sensation and an urge which in turn promotes the execution of the blink (Berman et al. 2012).

The most obvious function of EB is the lubrication of the eye (Doane 1980; Karson et al. 1990). Another role is the stabilization of the tear film (Korb et al. 1994). Beyond its protective properties, EB encompasses different cognitive processes associated with attention (Corbetta and Shulman 2002; Nakano and Kitazawa 2010; Nakano et al. 2013). In contrast to reflexive EB, spontaneous blinks do not require sensory inputs. It appears that their underlying mechanism is an internal blink generator (Kaminer et al. 2011), dopamine-mediated to some extent (Colzato et al. 2008; Ladas et al. 2014; Slagter et al. 2015). Indeed, the spontaneous EB rate in some dopamine-related psychiatric conditions (e.g., Parkinson’s disease and schizophrenia) was found to vary when compared to that of the general population (Karson et al. 1990).

On average, people blink 15–17 times a minute (Bentivoglio et al. 1997; Karson et al. 1990) with the rate changing according to different mental states such as reading, talking, daydreaming, and resting (Bentivoglio et al. 1997; Doughty 2001). When asked, subjects can inhibit their EB to some extent for a limited amount of time (Mazzone et al., 2010; Moraitis and Ghosh 2014; Schmeichel and Zell 2007). The ability to inhibit EB can be moderately predicted by a self-control report, i.e., people who reported themselves high on self-control inhibited EB better that those who reported themselves low on the trait scale (Schmeichel and Zell 2007). Several previous studies have affirmed that a few brain regions are associated with blink inhibition, i.e., the insular cortex, the supplementary motor areas (SMA), the primary facial motor areas, and the prefrontal regions such as the right inferior frontal gyrus (rIFG, Berman et al. 2012; Lerner et al. 2008).

Herein, we focused on the rIFG due to this region’s hypothesized involvement in proactive and reactive inhibitory control processes (Aron 2011; Aron et al. 2014; Jaffard et al. 2008). We designed a simple blink inhibition task and attempted to improve the ability to suppress blinking by a non-invasive and safe brain stimulation method, transcranial direct current stimulation (tDCS), over the rIFG. tDCS is a non-invasive, painless cortical stimulation technique which can increase or decrease brain excitability by adjusting the polarity of a weak current flow (Nitsche and Paulus 2000). Various tDCS studies have applied different stimulation protocols manipulating a variety of cognitive abilities such as language (e.g., Cerruti and Schlaug 2009; Flöel et al. 2008), attention (e.g., Bolognini et al. 2010; Kraft et al. 2010; Sparing et al. 2009), memory (e.g., Berryhill et al. 2010; Chi et al. 2010), and executive functions (e.g., Boggio et al. 2010; Fregni et al. 2005; Jacobson et al. 2011). The aim of the present study was therefore to investigate whether modulating activity in the rIFG by using tDCS can enhance proactive inhibitory control of an urge-driven behavior—EB.

We based our stimulation protocol on studies which had demonstrated an increased inhibition performance (both reactive and proactive) after anodal stimulation over the rIFG (Cunillera et al. 2014; Cunillera et al. 2016; Jacobson et al. 2011; Juan and Muggleton 2012). Based on these studies, we hypothesized that applying anodal stimulation over the rIFG will improve the subjects’ ability to suppress eye blinks in our novel task as well their performance in a classic response inhibition task—the SST.

None of the participants experienced adverse effects during or after stimulation. The results of four subjects (two from the rIFG anodal group and another two from the sham group) were excluded as they apparently did not understand the instructions and blinked substantially more during the stop phase compared to the baseline and resting phases.

Stop Signal Task

To investigate the effect of the tDCS condition, a MANOVA was conducted. The stimulation conditions (rIFG, rOFC, and sham) served as independent between-subject variables. The SSRT and response time in the go trials (go RT) were classified as dependent variables. An analysis revealed a significant general effect for all variables F(3, 45) = 6.52, p < .001, eta² = .3.

A significant difference was found in the go RT variables in the SST, F(2, 46) = 3.61, p < .04, eta² = .13. While the sham group responded with a mean go RT of 596 ms, the rOFC group responded with a mean go RT of 682 ms and the rIFG group responded slower with a mean go RT of 775 ms. Post hoc comparisons revealed a significant difference in this measure only between the rIFG and the sham groups (p < .02). Lastly, no significant difference was found between the stimulation groups for the SSRT variable, F(2, 46) = 2.26, p > .1, eta² = .09. Results are shown in Table 2. Post hoc comparisons are shown in Figs. 2 and 3.

Table 2

Mean (and SE) performance scores of the EB task and the SST as a function of group

	Sham		rIFG		rOFC
Variable	M	SE	M	SE	M	SE
EB baseline	34.13	6.95	38.5	6.65	31.72	4.43
EB stop %**	50.68	7.68	78.87	4.81	63.19	3.48
Go RT (ms)*	596	46	775	41	682	47
SSRT (ms)	249	9	247	13	280	14

rIFG right inferior frontal gyrus, rOFC right orbitofrontal cortex, ms milliseconds, M mean, SE standard error, EB eye blink, RT response time, SST stop signal task, SSRT stop signal reaction time

*p < .05; **p < .01

Open image in new window

Open image in new window

Herein, we applied anodal tDCS to the rIFG and measured its ability to suppress EBs compared to sham and active control conditions. To the best of our knowledge, this is the first attempt at manipulating the inhibition of urge-driven behavior by using tDCS. We demonstrated that anodal stimulation of the rIFG resulted in a marked improvement in the ability to suppress EBs as well as inducing a longer general RT during the response inhibition task. These findings are in line with other studies which have investigated the enhanced proactive inhibition as indicated by the prolonged RT during the go/no-go task (Cunillera et al. 2014) and during the SST (Cunillera et al. 2016).

The exact mechanism of proactive inhibition is still unclear. Different brain regions have been linked with this ability (Aron 2011). The rIFG seems to mediate cognitive processes associated with executive control, both in reactive and proactive inhibition (Aron 2011; Chikazoe et al. 2009; Jacobson et al. 2011). Moreover, longer go trial RTs observed in the active stimulation group support the assumption that this region serves as a “brake mechanism” by generally slowing down the response execution (Jahfari et al. 2010). Indeed, an earlier study demonstrated that stimulation of the rIFG slowed down responses (Wessel et al. 2013).

A recent review concluded that response inhibition is not a core function of the OFC (Stalnaker et al. 2015). However, although the rIFG group showed longer go RTs than the rOFC group, the difference was not significant (p = .15). A possible explanation of this outcome can arise from the anticipation function of the OFC, which is typically related to a reward or punishment signaling the expected outcome of an action (Schoenbaum et al. 2011). Taken together with findings regarding the neuronal connectivity of the OFC and IFG (Frühholz and Grandjean 2012), we speculated that enhancement of the anticipation function will result in the minor elevation of the go RT in the rOFC group compared to the sham group.

Previous studies have found that anodal stimulation of the rIFG resulted in a shorter SSRT, an index of efficient response inhibition (Campanella et al. 2017; Jacobson et al. 2011; Jacobson et al. 2012a, b; Stramaccia et al. 2015). The current study did not replicate this finding. A possible explanation could be the stimulation method. In the studies mentioned earlier, 25-cm² saline-soaked surface sponge electrodes were used whereas in the current experiment, a chloride gelled electrode with a surface contact area of 3.14 cm² was used. Clearly, this electrode allows more focal stimulation by targeting more specific regions in the rIFG.

In line with this hypothesis, a previous fMRI study found different activations in the posterior inferior frontal gyrus (pIFG) and inferior frontal junction (IFJ) regions. The authors concluded that these regions are associated with different cognitive control processes (Chikazoe et al. 2008). Similarly, findings from a transcranial magnetic stimulation (TMS) study suggested that the right IFJ was associated with attentional detection, whereas the pIFG was found associated with the inhibitory control process (Verbruggen et al. 2010). Overall, we can speculate that our stimulation protocol affected the “brake mechanism” resulting in a RT slowdown but did not reduce the SSRT. However, in order to confirm this hypothesis, a study combining this stimulation protocol with a neuroimaging technique is needed. Another limitation is the positioning of the reference electrode over the vertex. We cannot rule out the claim that the tDCS effect is unrelated to the rIFG due to a varied distribution of electrical current across different regions (De Berker, Bikson and Bestmann 2013). Using an extracephalic reference, electrode might rule out this possibility.

Our finding of tDCS enhancement of proactive motor control ability can expand the concept of non-invasive interventions in various movement disorders such as Tourette syndrome (TS) or chronic tic syndrome and might assist as well in treating addictions or other urge-driven behaviors. Tic movement in TS is a semi-involuntary behavior, phenomenologically similar to EB. One of the earliest tic symptoms in childhood is the pattern changes in EB (Jung et al. 2004). The temporal dynamics of tics and EBs are similar (Peterson and Leckman 1998).

In an fMRI study measuring brain activation during inhibition of EB in TS and healthy control subjects, inverse correlations between the current severity of tic symptoms and the magnitude of activation were found in the IFG as well as other regions. Individuals experiencing more severe tics activate these regions less than those with fewer severe symptoms. The authors suggested that greater functional disturbances in the frontostriatal systems may therefore contribute to more severe symptoms (Mazzone et al. 2010). Another study found inverse correlations between bilateral IFG gray matter volumes and clinical severity scores in patients with TS (Müller-Vahl et al. 2009). Lastly, Worbe and colleagues found a negative correlation between TS patients’ clinical severity scores and cortical thickness in the posterior part of the IFG (Worbe et al. 2010). All these findings suggest that the rIFG plays a role in tic suppression in TS. We suggest that TS patients should enhance activity in this region by using tDCS, thus, possibly reducing tic severity.

To conclude, the tDCS-induced enhancement of proactive motor control of urge-driven behaviors observed in the current study confirms the beneficial properties of our stimulation protocol. Future studies applying this protocol on clinical populations such as TS patients or other urge/impulse control disorders would test its effectiveness in these populations. Moreover, while considering the limited therapeutic options for TS, our protocol might serve as a useful therapeutic tool.