Secondary Logo

Journal Logo

Neural Correlates of Expert Visuomotor Performance in Badminton Players

HÜLSDÜNKER, THORBEN; STRÜDER, HEIKO K.; MIERAU, ANDREAS

Medicine & Science in Sports & Exercise: November 2016 - Volume 48 - Issue 11 - p 2125–2134
doi: 10.1249/MSS.0000000000001010
BASIC SCIENCES
Free

Introduction Elite/skilled athletes participating in sports that require the initiation of targeted movements in response to visual cues under critical time pressure typically outperform nonathletes in a visuomotor reaction task. However, the exact physiological mechanisms of this advantage remain unclear. Therefore, this study aimed to determine the neurophysiological processes contributing to superior visuomotor performance in athletes using visual evoked potential (VEP).

Methods Central and peripheral determinants of visuomotor reaction time were investigated in 15 skilled badminton players and 28 age-matched nonathletic controls. To determine the speed of visual signal perception in the cortex, chromatic and achromatic pattern reversal stimuli were presented, and VEP values were recorded with a 64-channel EEG system. Further, a simple visuomotor reaction task was performed to investigate the transformation of the visual into a motor signal in the brain as well as the timing of muscular activation.

Results Amplitude and latency of VEP (N75, P100, and N145) revealed that the athletes did not significantly differ from the nonathletes. However, visuomotor reaction time was significantly reduced in the athletes compared with nonathletes (athletes = 234.9 ms, nonathletes = 260.3 ms, P = 0.015). This was accompanied by an earlier activation of the premotor and supplementary motor areas (athletes = 163.9 ms, nonathletes = 199.1 ms, P = 0.015) as well as an earlier EMG onset (athletes = 167.5 ms, nonathletes = 206.5 ms, P < 0.001). The latency of premotor and supplementary motor area activation was correlated with EMG onset (r = 0.41) and visuomotor reaction time (r = 0.43).

Conclusion The results of this study indicate that superior visuomotor performance in athletes originates from faster visuomotor transformation in the premotor and supplementary motor cortical regions rather than from earlier perception of visual signals in the visual cortex.

Institute of Movement and Neurosciences, German Sport University Cologne, Cologne, GERMANY

Address for correspondence: Andreas Mierau, Ph.D., Institute of Movement and Neurosciences, German Sport University Cologne, Am Sportpark Müngersdorf 6, 50933 Cologne, Germany; E-mail: mierau@dshs-koeln.de.

Submitted for publication April 2016.

Accepted for publication May 2016.

The ability to rapidly perceive visual cues in the visual field and to initiate a targeted motor response is essential in many sports when athletes have to perform visuomotor tasks under critical time pressure. Within a visuomotor reaction, several consecutive and parallel processing stages on the central and peripheral levels contribute to the overall visuomotor performance. These include the perception of a visual stimulus in retinal photoreceptors and the primary visual cortex; visual processing in higher visual areas and transformation into a motor command as well as finally movement execution. A simplified model of the proposed neural pathway is presented in Figures 1A and 1B.

FIGURE 1

FIGURE 1

On a behavioral level, considering the start and end point of a visuomotor reaction (Fig. 1C) (1), numerous studies have demonstrated superior visuomotor reaction time (VMRT) during simple and sport-specific visuomotor tasks in athletes when compared with nonathletes (1,3,17,29). Specifically, studies differentiating between premotor and motor time (Fig. 1C) (2) have indicated superior VMRT in athletes being primarily attributable to a reduction of premotor reaction time (1,29) suggesting functional differences at the neuronal level. The premotor reaction time can be further subdivided into the perception and processing of visual information as well as the visuomotor transformation necessary for motor command generation (Fig. 1C) (3). Previous studies have thus far not differentiated between these premotor stages.

Neural correlates of visual perception and processing can be studied using the pattern reversal visual evoked potential (VEP). The VEP consists of three prominent waves, an initial negative potential arising approximately 75 ms after stimulus onset (N75) followed by a positive peak around 100 ms (P100) and a second negative potential peaking approximately 145 ms (N145) (28). Because of their origin in the primary visual cortex, the N75 and the P100 potentials are suggested to subserve visual perception (11,18), whereas the N145 is also generated by extrastriate areas and, thus, may be attributable to visual processing (2).

In sports, reduced P100 latencies have been shown for tennis (8) and volleyball players (31) as well as fencers (26). Importantly, latencies in tennis players were reduced when compared with not only nonathletes but also endurance athletes (8,16). This suggests adaptations in early visual perception to be attributable to the sport-specific requirements on the visuomotor system rather than participation in elite sports or to physical performance in general. However, reductions in P100 latency were in the range of a few milliseconds, which cannot fully account for the observed differences in VMRT.

After visual perception and processing, visuomotor transformation in the premotor areas (Brodmann area 6 [BA6]) and primary motor areas (M1 [BA4]) subserves the translation of a visual signal into a motor command. In this context, SMA activation and connectivity have previously been shown to be markedly different between athletes and nonathletes (5,9). These differences directly affect visuomotor performance as SMA activation is closely related to VMRT (4), and thus, it may be causal to superior performance in athletes. Moreover, athletes exhibited higher cortical activation of BA4 during a visuomotor reaction task (10), indicating that the primary motor cortex may also contribute to the differences in VMRT between athletes and nonathletes.

Despite existing evidence for shorter VMRT in athletes, the exact central and peripheral mechanisms remain unclear because of insufficient differentiation between processing stages illustrated in Figure 1. On the group level, such an approach will provide specific information on the neural/and or peripheral basis of superior athletes VMRT. In addition, on an individual level, the identification of the athlete’s strength and weaknesses during a visuomotor reaction will help to individualize visual/visuomotor training approaches.

In this study, we aimed to identify the determinants of VMRT in badminton athletes and nonathletes by investigating the determinants of visual perception, visual processing, visuomotor transformation, and movement execution (Fig. 1C) (4). We chose badminton because of the robust visuomotor demands of this sport (for a review, see [23]).

Chromatic and achromatic pattern reversal VEP amplitude and latency were analyzed to identify visual perception (N75 and P100) and processing (N145) of luminance and color contrast stimuli. In addition, subjects completed a simple visuomotor reaction task using a luminance contrast pattern reversal stimulus to determine visuomotor transformation in BA6/BA4 as well as EMG onset and VMRT.

On the behavioral level, we expected shorter VMRT in athletes compared with nonathletes. It was further hypothesized that the athletes’ faster VMRT results primarily from differences in visuomotor transformation (BA6/BA4 activation) rather than faster visual perception and processing.

Back to Top | Article Outline

METHODS

Subjects and Ethics

A total of 47 healthy subjects participated: 16 athletes (8 males and 8 females) and 31 nonathletes (13 males and 18 females). Four subjects (one athlete and three nonathletes) did not meet the requirements for visual acuity and/or color vision (see next section) and were excluded from analyses. Anthropometric measures of athletes and nonathletes are summarized in Table 1. Badminton athletes were recruited from a national youth training center in Bonn, Germany. All athletes had participated in regular badminton training for more than 9 yr (9.5 ± 3.0) with at least three training sessions and a total number of >8 h (12.5 ± 4.7) of training per week. All badminton players compete at high national level in their respective age-group during the regular season and participate in professional international and/or national tournaments. The experiment was performed during a training workshop in preparation for the season.

TABLE 1

TABLE 1

Age-matched control subjects did not participate in badminton and/or other racket or team sports. All subjects confirmed being free of injury for at least 6 month, having no pain and/or discomfort, and experiencing no limitation during their daily routines and exercise. Participants were provided the experimental protocol, and their written consent was obtained. In case of participants under the legal age, the consent form was signed by a parent or guardian. The study was approved by the local research ethics committee of the university in accordance with the Declaration of Helsinki.

Back to Top | Article Outline

Experimental Protocol

Before the experiment, the subject’s dominant hand was determined by applying the Edinburgh Handedness Inventory test (21). On the test day, subjects were instructed not to drink caffeinated drinks or alcohol. Questionnaires were used to ensure normal sleep quality and exclude depression (Beck Depression Inventory [BDI]). Ophthalmologic tests were performed to ensure visual acuity (Landolt test) of at least 20/20 as well as color vision (Ishihara color charts). The room was completely darkened and illuminated by an 8-W LED panel placed 10 cm behind the screen to keep lightning conditions constant. To ensure the subject’s eyes were level with the center of the screen during testing, the head was placed on a stepless high-adjustable chin rest. During testing, subjects were seated in a comfortable chair, and they placed their hands on the table with an angle of approximately 90° in the elbow joint. Subjects were instructed to avoid head movement and blinking during the experiment. Earplugs were used to avoid disturbance from surrounding noise. Before starting the experiment, all visual stimuli were presented during a practice period. Overall, the experimental protocol lasted for approximately 15 min. All visual stimuli were presented binocularly, and subjects were instructed to keep their gaze on a fixation point presented in the center of the screen.

The experimental protocol was subdivided into two experiments to optimize the investigation of visual perception/processing and visuomotor processing/movement execution, respectively. Specifically, valid VEP acquisition requires a large number of trials and a constant reversal frequency (cf. Experiment 1). However, such a VEP paradigm is not suitable for the determination of VMRT because the temporally invariant presentation of the stimuli would allow the subject to anticipate the upcoming stimulus already after a few trials. Therefore, we conducted a second experiment where interstimulus intervals were randomized in the range of a few seconds (average: 4 s) to avoid anticipation. The protocol used in Experiment 2 allowed for the investigation of movement-related BA6 activation, EMG onset, and VMRT.

Back to Top | Article Outline

Experiment 1: VEP

Experiment 1 investigated VEP using three different modalities of pattern reversal chessboard stimuli of luminance (black–white) and chromatic (red–green/blue–yellow) contrast. The sequence of pattern reversal stimuli was counterbalanced across subjects. For each stimulus, 200 sweeps were recorded, subdivided into two blocks of 100 stimuli. Visual stimuli were displayed with a reversal frequency of 1 Hz (two reversals per second). The pause between the two blocks was 15 s. Between two pattern modalities, there was a pause of 1 min.

Back to Top | Article Outline

Experiment 2: visuomotor reaction task

The second experiment contained a simple visuomotor reaction task to determine visuomotor transformation in the premotor and supplementary motor regions (BA6) as well as movement execution. The visual cue given the subjects was a reversal of the black–white pattern stimulus also used in Experiment 1. Subjects were instructed to press a button as fast as possible with the index finger of their dominant hand whenever they perceived a pattern reversal. Interstimulus intervals were randomized between 2 and 6 s during the visuomotor reaction task so that subjects were not able to anticipate the pattern reversal. Sixty reactions were performed, subdivided into three blocks of 20 reactions with a pause of 15 s between two blocks.

Back to Top | Article Outline

Visual stimuli

Visual stimuli were programmed with the CRS toolbox (Cambridge Research System, Rochester, UK), implemented in Matlab (The Mathworks, Natick, MA) and presented using the ViSaGe MKII system (Cambridge Research System) on a 22-inch CRT Monitor (HP P1230 Colour CRT) with a refresh rate of 120 Hz. Viewing distance was set to 750 mm. Single pattern reversal checks subtended a visual angle of 1°4′ within a field of view of 27.7° × 21.3°. Black–white pattern reversal stimuli had a mean luminance of 60 cd·m−2 and a contrast of 98%. Equiluminant red–green and blue–yellow pattern reversal stimuli had a mean luminance of 15 and 28 cd·m−2, respectively. The following CIE1931 coordinates were used: red, x = 0.591, y = 0.333; green, x = 0.288, y = 0.588; yellow, x = 0.394, y = 0.499; blue, x = 0.155, y = 0.073. Before the experiment, luminance and color coordinates were calibrated using a ColorCalII Colorimeter (Cambridge Research System). During the tests, all recording systems were synchronized by an electrical trigger signal frame-synchronously generated by the ViSaGe MKII.

Back to Top | Article Outline

Data Recording

EEG recording

EEG was recorded from 64 scalp locations overlying the whole scalp and equally distributed across both hemispheres according to the international 10:10 system (15). One additional electrode was used to measure electrooculographic signals. The electrical reference and the ground electrode were located on positions FCz and AFz, respectively. The sampling rate was 1000 Hz. Electrode impedances were kept at values less than 15 kΩ.

Back to Top | Article Outline

EMG recording

EMG activity was recorded from the musculus flexor carpi radialis of the dominant hand. A DE-2.1 Ag double differential surface sensor (Bagnoli; Delsys, Natick, MA) with a contact spacing of 10 mm and an input impedance >1015 Ω was used and amplified (1000×) by a BagnoliTM amplifier (Bagnoli, Delsys). The skin was shaved, cleaned, and abraded before the sensor was attached using a sensor-tailored adhesive interface. Data were sampled at 1000 Hz.

Back to Top | Article Outline

VMRT

In response to a pattern reversal, VMRT was determined by pressing a button on a Cedrus RB-530 response box (Cedrus, San Pedro, CA). Each button press was registered by the EEG and EMG recording systems via an electrical trigger pulse. The response box signal was sampled with a frequency of 1000 Hz.

Back to Top | Article Outline

Data Analysis

Experiment 1: VEP

EEG data analysis

EEG data were analyzed using the Brain Vision Analyzer 2 software (Brain Products, Munich, Germany) and scripts of the EEGlab toolbox (7) implemented in Matlab (The Mathworks).

For the analysis of chromatic and achromatic VEP, raw data were first band-passed filtered (0.3–40 Hz) (25). The window of analysis was defined from 50 ms before the stimulus to 250 ms after the stimulus. For each epoch, the prestimulus activity was subtracted from the data (baseline correction). A semiautomatic artifact rejection procedure was conducted to identify prominent artifacts (mainly movement and blink artifacts). Segments containing voltage steps >50 μV and amplitudes >100 μV were excluded from further analysis. Subsequently, all remaining segments were visually inspected. An extended runica algorithm implemented in EEGlab (7) was applied for independent component analysis (ICA) decomposition. On the basis of cortical mapping, frequency spectrum, and time course, components reflecting muscular activity were identified by visual inspection and removed from the data. After ICA back transform, current source density(CSD) was calculated (order of splines = 4, maximal degree of Legendre polynomials = 10, lambda = 1e−5), and data were averaged over segments.

For the achromatic (black–white) pattern reversal stimuli, amplitude and latency of VEP components (N75, P100, and N145) were identified based on the time course of electrode position Oz. The N75, P100, and N145 components were defined as the maximal negative potential in the interval from 50 to 100 ms, the maximal positive value between 75 and 125 ms, and the maximal negative value between 110 and 160 ms, respectively. VEP latencies were expressed as the difference between pattern reversal (time 0) and peak potential. VEP amplitudes were defined as the difference between baseline (zero) and peak amplitude, respectively. For chromatic (red–green; blue–yellow) pattern reversal stimuli, only the P100 was investigated because N75 and N145 were not validly identifiable in all subjects. Cortical sources of N75, P100, and N145 potentials were localized within a 10-ms peri-peak window using the LORETA module implemented in the Analyzer 2 software.

Back to Top | Article Outline

Experiment 2: visuomotor reaction task

EEG data analysis

For the analysis of BA6/BA4 activation, raw data were first band-passed filtered (0.3–100 Hz). The window of analysis was defined from 500 ms before the stimulus to 1000 ms after the stimulus. Baseline correction was performed for each epoch by subtracting prestimulus activity from the data. An ocular correction ICA procedure implemented in the Analyzer 2 software was used to identify eye blinks. Only segments containing blinks in the interval between −500 and 200 ms relative to reversal onset were excluded from the data as blinks occurring in this period may affect visual perception and/or visuomotor reaction. A semiautomatic artifact rejection algorithm (voltage step = 50 μV, min/max amplitude = ±150 μV) including the same time interval was applied, and all segments were visually inspected afterward. Similar to the VEP data analysis, the remaining segments were decomposed into independent components, and those reflecting artifacts were excluded from the data. CSD was calculated, and data were averaged across segments. In addition, we calculated the activation of premotor or supplementary motor regions (BA6) as well as M1 (BA4) during the visuomotor reaction task in the time domain by using the LORETA transformation module. On the basis of averaged LORETA time traces across trials, the amplitude and the latency of maximal BA6 positivity and negativity within the interval from 75 to 400 ms after pattern reversal were determined.

Back to Top | Article Outline
EMG data analysis

EMG data were investigated during the visuomotor reaction experiment. Data were first band-pass filtered (5–450 Hz) and rectified. Similar to the EEG data, the window of analysis was defined from 500 ms before the stimulus to 1000 ms after the stimulus. Baseline correction was performed for each epoch by subtracting prestimulus activity from the data. EMG onset was defined in a single-trial approach for each reaction trial individually according to the suggestions of Hodges and Bui (12). In particular, EMG data were first low-pass filtered (50 Hz). Then a 25-ms moving average window was applied, and the EMG onset was defined when the signal exceeds three standard deviations from baseline. Subsequently, EMG onset in each trial was visually examined and adjusted if necessary.

Back to Top | Article Outline
VMRT

VMRT was defined as the time between the visual cue (pattern reversal) and the movement execution (button press). Trials indicating reaction times faster than 100 ms or slower than 500 ms were excluded from further analysis.

Back to Top | Article Outline

Statistical Analysis

All statistical comparisons were conducted in Statistica 7.1 (StatSoft, Tulsa, OK).

In a preliminary analysis, we used unpaired t-tests to determine whether there were any differences in anthropometric parameters between groups. As the distribution of males and females was different between groups (athletes: 7 males, 8 females; nonathletes: 11 males, 17 females), we also analyzed the effect of sex on neurophysiological (N75, P100, N145, and BA6 peak activation), muscular (EMG onset), and behavioral (button press) parameters. For neurophysiological measures, unpaired t-tests revealed significantly reduced latency in females for the black white pattern reversal N145 (P = 0.002) as well as for the P100 potentials during red–green (P = 0.006) and blue–yellow (P = 0.009) conditions. There were no differences for EMG onset, BA6 peak activation, and button press. To account for the differences in sex distribution between groups, VEP amplitude and latency values of male and female subjects in the control group were weighted during t-tests and ANOVA by the factors 0.88 and 1.18, respectively.

Back to Top | Article Outline

Experiment 1: VEP

Amplitude and latency of black–white (N75, P100, and N145), blue–yellow (P100), and red–green (P100) VEP components were compared between athletes and nonathletes using unpaired t-tests. To further determine differences in visual perception between chromatic (blue–yellow and red–green) and achromatic (black–white) contrast conditions, an ANOVA with the within-subject factor modality (black–white, red–green, and blue–yellow) and the between-subject factor group was calculated for P100 amplitude and latency. As described previously, weighted values were used for all VEP analyses.

Back to Top | Article Outline

Experiment 2: visuomotor reaction task

Parameters investigated during the visuomotor reaction task were the amplitude and latency of BA6 activation as well as latencies of EMG onset and VMRT (button press). Unpaired t-tests were used for statistical comparisons between athletes and nonathletes.

The normal distribution was investigated using the Kolmogorov–Smirnov test. The sphericity assumption was evaluated according to the Mauchly test, and the Greenhouse–Geisser correction of degrees of freedom was applied in case of nonsphericity. Effect sizes were calculated using partial eta square (η2) and Cohen’s d for ANOVA and t-tests, respectively.

Significance levels were defined as follows: *P < 0.05, **P < 0.01, and ***P < 0.001.

Back to Top | Article Outline

RESULTS

Visual perception (VEP) experiment

Average VEP time courses are presented in Figures 2A and 2B. LORETA localization revealed the N75 and P100 of the black–white pattern-reversal stimulus located in the primary visual cortex (BA17), whereas the N145 component also includes activation of BA18. P100 potentials for the chromatic (red–green/blue–yellow) pattern reversal stimuli were located in BA18. No differences in localization were observed between athletes and nonathletes. LORETA results as well as mappings of cortical activity are presented in Figures 2C and 2D.

FIGURE 2

FIGURE 2

Table 2A presents the mean values of the VEP experiment for both groups. Unpaired t-tests for VEP amplitude and latency revealed similar amplitude and latency values for achromatic (black–white) VEP amplitude (N75: t15,28 = −0.85, P = 0.403, d = 0.27; P100: t15,28 = 0.82, P = 0.418, d = 0.38; N145: t15,28 = 1.40, P = 0.169, d = 0.45) and latency (N75: t15,28 = −0.56, P = 0.552, d = 0.19; P100: t15,28 = −0.50, P = 0.623, d = 0.16; N145: t15,28 = −1.61, P = 0.115, d = 0.52).

TABLE 2

TABLE 2

The ANOVA for P100 latency and amplitude of black–white, blue–yellow, and red–green pattern reversal stimuli yielded a significant main effect for modality in both analyses (latency: F2,60 = 30.80, P < 0.001, η2 = 0.51; amplitude: F2,56 = 34.96, P < 0.001, η2 = 0.54). Post hoc tests indicated longer P100 latencies for blue–yellow (P < 0.001) and red–green (P < 0.001) when compared with black–white pattern reversal stimuli. Further, P100 amplitude was significantly higher during black–white pattern reversal when compared with blue–yellow (P < 0.001) and red–green (P < 0.001) conditions.

Back to Top | Article Outline

Visuomotor reaction experiment

Figure 3 presents the example data of one subject during the visuomotor reaction. On the group level, outcome values of the visuomotor reaction experiment are presented in Table 2B. Unpaired t-tests for BA6 peak activation indicated an earlier positive (t13,22 = −3.38, P = 0.002, d = 1.18) as well as negative (t15,27 = −2.52, P = 0.015, d = 0.813) BA6 peak activation in the athlete’s group when compared with nonathletes. This was accompanied by a faster EMG onset (t15,28 = −4.06, P < 0.001, d = 1.3) and VMRT (t15,28 = −2.53, P = 0.015, d = 0.81) in athletes.

FIGURE 3

FIGURE 3

Back to Top | Article Outline

LORETA localization

In addition to the previously described group differences about the time course of BA6 activation, we were also interested whether groups differed about the amount of BA6 activation during the visuomotor transformation period. To this end, we focused on the neural processes in the time interval 75–400 ms, which is essentially the window immediately between the initial V1 activation (N75) and the button press (including EMG onset). The average cortical activity of athletes and nonathletes was contrasted for all electrodes, and the resulting differential signal was localized using LORETA. This analysis revealed that the most striking difference between athletes and nonathletes was the activation of BA6 in the time interval 130–210 ms. This time interval encompasses a large part of the visuomotor transformation stage between N75/P100 and EMG onset. To determine whether these differences between groups can be considered as statistically significant, t-tests for BA6 amplitude were performed for each time point between 75 and 400 ms. The results yielded higher BA6 activation in athletes compared with nonathletes in the interval 116–166 ms. The t-tests for BA4 activation further revealed a higher BA4 activation in athletes between 144 and 340 ms. The time courses of BA6 and BA4 activation as well as the LORETA localization of differential cortical activity in the time intervals of significant group differences are presented in Figure 4.

FIGURE 4

FIGURE 4

Back to Top | Article Outline

Correlation analysis

On the basis of the previously mentioned findings, earlier BA6 peak activation was accompanied by faster EMG onset and VMRT in athletes. To determine whether there was a direct relationship between these parameters, we calculated Pearson correlation coefficients. This analysis yielded a significant correlation only between negative BA6 peak activation and EMG onset (r = 0.41, P = 0.007) as well as VMRT (r = 0.43, P = 0.005).

Back to Top | Article Outline

DISCUSSION

In this study, determinants of visuomotor performance were investigated in badminton athletes and nonathletes by differentiating between different consecutive and parallel processing stages, including visual perception, visual processing, visuomotor transformation, and movement execution. Athletes showed faster VMRT that was accompanied by an earlier activation of premotor and supplementary motor regions as well as an earlier EMG onset. No differences between groups were observed during visual perception. These results suggest superior visuomotor performance in athletes being mainly attributable to faster visuomotor processing rather than to visual perception.

Back to Top | Article Outline

VMRT

Badminton players were characterized by significantly shorter VMRT, which agrees with numerous previous findings (for a review, see Phomsoupha and Laffaye [23]).

Back to Top | Article Outline

Visual perception (N75 and P100)

N75 and P100 potentials were located in the primary visual cortex for both groups, which is line with previous findings suggesting that these potentials reflect visual perception (11,18,28).

More importantly, the N75 and the P100 potentials of the black–white pattern-reversal VEP as well as the chromatic P100 potentials obtained during Experiment 1 did not differ in latency and amplitude between athletes and nonathletes. This indicates similar speed of visual perception in the primary visual cortex for both the achromatic as well as the chromatic contrast conditions.

These results are in line with several previous studies on volleyball and badminton players that did not report faster visual perception using pattern reversal (22) and sport-specific visual stimuli (6,14). However, it is in contrast to some VEP studies reporting reduced P100 latency in fencers (26) as well as in tennis (8) and volleyball (31) players. This heterogeneity of results may be attributable to differences in visual stimulus parameters, the type of sport, and the skill level of participants. However, it also raises the question of whether there is evidence of structural and/or functional adaptations in the visual pathway of athletes that could account for earlier visual perception in athletes.

In an electroretinography study, Zwierko et al. (30) did not observe differences in implicit b-wave time between volleyball athletes and nonathletes, indicating similar speed of visual perception between groups on the retinal level. Furthermore, although Thomas et al. (27) reported a slightly reduced N75 in cricketers (P = 0.03), this result could not be confirmed in several subsequent studies (8,22,26,31). Therefore, there is currently no convincing evidence for potential differences between athletes and nonathletes during visual signal transmission via the nervus opticus, lateral geniculate nucleus, and optic tract to layer IV of the visual cortex. Hence, the previously reported reduction of P100 latencies in athletes is probably related to differences in signal transmission between the generation of the N75 and P100 potentials.

In sum, the results of this study clearly indicate that there are no differences between groups during visual perception for chromatic as well as for achromatic stimuli. When combined with the current literature, this argues against differences during signal transmission via the visual pathway between athletes and nonathletes. Importantly, even if there are differences in some sports or with specific stimulus parameters, these are in the range of a few milliseconds, which is not sufficient to fully explain faster VMRT in athletes.

Back to Top | Article Outline

Visual processing (N145)

LORETA analysis revealed that the N145 is located in the visual area V2 (BA18), which is in line with previous research suggesting extrastriate regions as also contributing to the N145 (19,24). Similar to the N75 and P100 potentials, no differences in amplitude and latency were observed between athletes and nonathletes. This is accordance with previous findings in volleyball players (31) and fencers (26). However, Ozmerdivenli et al. (22) reported reduced N145 latency in volleyball players, although this difference was only observed when stimulating the left eye and in the range of approximately 5 ms. As previously argued for the N75 and P100 VEP components, this difference could not fully explain faster VMRT in athletes. In addition, BA6 peak activation preceded the N145, indicating visual processing (N145) runs in parallel rather than in series to visuomotor transformation (BA6 peak activation). The absence of correlations between N145 latency and BA6 negativity peak (r = 0.18) as well as EMG onset (r = 0.18) and VMRT (r = 0.14) further supports the assumption that the N145 is not directly related to the faster VMRT in athletes.

Back to Top | Article Outline

Visuomotor transformation (BA6)

During visuomotor transformation, athletes revealed a significantly stronger BA6 activation already 116 ms after the visual cue. This was further accompanied by a significantly earlier negative peak activation of BA6 (approximately 36 ms). This matched well and was directly related to the observed differences in EMG onset (approximately 39 ms). These results suggest faster visuomotor transformation as reflected by premotor activation, which is the main determinant of superior visuomotor performance in athletes.

These findings support previous studies suggesting activation of premotor and supplementary motor regions being crucially related to reaction time. Specifically, when artificially modulating the excitability state of the SMA by transcranial direct current stimulation, reaction times were accelerated or delayed in a polarity specific manner (4). On the basis of these results, the authors suggested a threshold model where SMA activation has to exceed a certain level before the movement is initiated. With regard to the results of this study, such a movement initiation threshold would be reached earlier in athletes, as reflected by earlier BA6 activation, thus leading to an earlier EMG onset and movement execution.

The neurophysiological basis for an earlier BA6 activation in athletes may result from functional differences in visuomotor network activation (5,9). In resting state connectivity measurements using functional magnetic resonance imaging, badminton athletes exhibited stronger connectivity in a frontoparietal network (9) that subserves visuomotor transformation. Importantly, only in the athletes group were parietal regions functionally connected to BA6, whereas the control group showed parietal connectivity only to BA9. This may suggest a more direct transmission of visual signals to executive motor regions in athletes, whereas the visual information has to pass a more “indirect” and thus more time-consuming pathway in nonathletes. This is further supported by motor imagery experiments where nonathletes exhibited widespread neuronal activity in a distributed network in contrast to focal activation especially of the SMA in athletes (5). Therefore, earlier BA6 activation and, thus, earlier movement initiation may result from efficient visuomotor networks in athletes that provide straightforward and thus time-efficient visuomotor signal transmission. By contrast, indirect visuomotor connections and the activation of a distributed network in nonathletes may slow down the visuomotor transformation process and delay premotor activation.

In addition to the negative peak in BA6 activation, also the peak positivity occurred earlier in athletes (approximately 93 ms) when compared with nonathletes (approximately 127 ms). As this peak is observed after the N75, it may potentially be related to visuomotor transformation. However, multiple t-tests revealed no significant differences in BA6 activation between groups in the time interval 75–116 ms. Therefore, our results indicate that early BA6 positivity peak is not related to differential visuomotor processing between groups. In line with this, the latency of the positive BA6 peak was also not correlated to EMG onset or button press. Further, when compared with the robust BA6 peak negativity, the positive peak was less pronounced or not detectable in approximately 20% of the subjects. In sum, this pattern of results suggests the early positive BA6 peak does not explain superior visuomotor performance in athletes. Of note, a previous study reported V1 activation already approximately 27.5 ms following the visual stimulus (13). However, flash stimuli that naturally exhibit V1 activation around 30 ms (20) were used in this study. By contrast, the first V1 activation after pattern reversal is found after approximately 75 ms (20). Therefore, we do not expect visuomotor transformation to start before 75 ms. Future studies using different visual stimuli such as light flashes that are accompanied by a significantly earlier V1 activation (13) may provide further insights into the timing of visual and motor cortex activation and its relation to visuomotor performance.

Back to Top | Article Outline

Movement execution (BA4)

Athletes and nonathletes also differed with regard to BA4 activation. Specifically, athletes showed a stronger BA4 activation in the time interval between 144 and 340 ms. This is in accordance with the results of Endo et al. (10) likewise reporting higher M1 activation in athletes compared with nonathletes during a simple visuomotor reaction task. The timing of group differences in BA6 (start at 116 ms) and BA4 (start at 144 ms) may further suggest a serial dependence of visuomotor processing in BA6 and BA4. Therefore, our results indicate not only faster visuomotor transformation in the athletes’ supplementary and premotor region (BA6) but also an earlier activation of M1 (BA4), which is responsible for movement execution.

Back to Top | Article Outline

CONCLUSION

The results of this study confirm superior visuomotor performance in badminton athletes. This is related to faster visuomotor transformation processes as reflected by an earlier activation of premotor and primary motor regions rather than to a faster perception of visual signals in the visual cortex. The combined pattern of results is of particular importance for training and diagnostic purposes in visuomotor demanding sports as this approach provides detailed insights into specific central and peripheral processes contributing to visuomotor reactions. Identification of the athlete’s strengths and weaknesses on different levels during the visuomotor task could help to individualize training and to improve visuomotor performance more effectively. Further, the efficiency of training approaches such as visual training, and its impact on different visuomotor processing stages could be systematically evaluated. These questions should be addressed in future studies.

The authors thank Martin Lemke from the national youth training center in Bonn for supporting this study and all badminton athletes for participation. The study was supported by DFG (INST 229/2-1 FUGG) and an internal grant from the German Sport University Cologne (no. 920133). TH is recipient of the graduate school doctoral scholarship at the German Sport University Cologne. The sponsor had no role in designing and conducting the study as well as during data analysis and interpretation of results. The authors have no conflict of interest to declare. The present study does not constitute endorsement by the American College of Sports Medicine.

Back to Top | Article Outline

REFERENCES

1. Ando S, Kida N, Oda S. Central and peripheral visual reaction time of soccer players and nonathletes. Percept Mot Skills. 2001;92(3 Pt 1):786–94.
2. Barnikol UB, Amunts K, Dammers J, et al. Pattern reversal visual evoked responses of V1/V2 and V5/MT as revealed by MEG combined with probabilistic cytoarchitectonic maps. Neuroimage. 2006;31(1):86–108.
3. Bhabhor MK, Vidja K, Bhanderi P, Dodhia S, Kathrotia R, Joshi V. A comparative study of visual reaction time in table tennis players and healthy controls. Indian J Physiol Pharmacol. 2013;57(4):439–42.
4. Carlsen AN, Eagles JS, MacKinnon CD. Transcranial direct current stimulation over the supplementary motor area modulates the preparatory activation level in the human motor system. Behav Brain Res. 2015;279(15):68–75.
5. Chang Y, Lee J, Seo J, et al. Neural correlates of motor imagery for elite archers. NMR Biomed. 2011;24(4):366–72.
6. Del Percio C, Brancucci A, Vecchio F, et al. Visual event-related potentials in elite and amateur athletes. Brain Res Bull. 2007;74(1–3):104–12.
7. Delorme A, Makeig S. EEGLAB: an open source toolbox for analysis of single-trial EEG dynamics including independent component analysis. J Neurosci Methods. 2004;134(1):9–21.
8. Delpont E, Dolisi C, Suisse G, Bodino G, Gastaud M. Visual evoked potentials: differences related to physical activity. Int J Sports Med. 1991;12(3):293–8.
9. Di X, Zhu S, Jin H, et al. Altered resting brain function and structure in professional badminton players. Brain Connect. 2012;2(4):225–33.
10. Endo H, Kato Y, Kizuka T, Takeda T. A comparison of stimulus synchronous activity in the primary motor cortices of athletes and non-athletes. Exp Brain Res. 2006;174(3):426–34.
11. Hashimoto T, Kashii S, Kikuchi M, Honda Y, Nagamine T, Shibasaki H. Temporal profile of visual evoked responses to pattern-reversal stimulation analyzed with a whole-head magnetometer. Exp Brain Res. 1999;125(3):375–82.
12. Hodges PW, Bui BH. A comparison of computer-based methods for the determination of onset of muscle contraction using electromyography. Electroencephalogr Clin Neurophysiol. 1996;101(6):511–9.
13. Inui K, Kakigi R. Temporal analysis of the flow from V1 to the extrastriate cortex in humans. J Neurophysiol. 2006;96(2):775–84.
14. Jin H, Xu G, Zhang JX, et al. Athletic training in badminton players modulates the early C1 component of visual evoked potentials: a preliminary investigation. Int J Psychophysiol. 2010;78(3):308–14.
15. Jurcak V, Tsuzuki D, Dan I. 10/20, 10/10, and 10/5 systems revisited: their validity as relative head-surface-based positioning systems. Neuroimage. 2007;34(4):1600–11.
16. Magnié MN, Bermon S, Martin F, Madany-Lounis M, Gastaud M, Dolisi C. Visual and brainstem auditory evoked potentials and maximal aerobic exercise: does the influence of exercise persist after body temperature recovery? Int J Sports Med. 1998;19(4):255–9.
17. Nakamoto H, Mori S. Sport-specific decision-making in a Go/NoGo reaction task: difference among nonathletes and baseball and basketball players. Percept Mot Skills. 2008;106(1):163–70.
18. Nakamura M, Kakigi R, Okusa T, Hoshiyama M, Watanabe K. Effects of check size on pattern reversal visual evoked magnetic field and potential. Brain Res. 2000;872(1–2):77–86.
19. Noachtar S, Hashimoto T, Lüders H. Pattern visual evoked potentials recorded from human occipital cortex with chronic subdural electrodes. Electroencephalogr Clin Neurophysiol. 1993;88(6):435–46.
20. Odom JV, Bach M, Brigell M, et al. ISCEV standard for clinical visual evoked potentials (2009 update). Doc Ophthalmol. 2010;120(1):111–9.
21. Oldfield RC. The assessment and analysis of handedness: the Edinburgh inventory. Neuropsychologia. 1971;9(1):97–113.
22. Ozmerdivenli R, Bulut S, Bayar H, et al. Effects of exercise on visual evoked potentials. Int J Neurosci. 2005;115(7):1043–50.
23. Phomsoupha M, Laffaye G. The science of badminton: game characteristics, anthropometry, physiology, visual fitness and biomechanics. Sports Med. 2015;45(4):473–95.
24. Schroeder CE, Steinschneider M, Javitt DC, et al. Localization of ERP generators and identification of underlying neural processes. Electroencephalogr Clin Neurophysiol Suppl. 1995;44:55–75.
25. Souza GS, Gomes BD, Lacerda EM, Saito CA, da Silva Filho M, Silveira LC. Amplitude of the transient visual evoked potential (tVEP) as a function of achromatic and chromatic contrast: contribution of different visual pathways. Vis Neurosci. 2008;25(3):317–25.
26. Taddei F, Viggiano MP, Mecacci L. Pattern reversal visual evoked potentials in fencers. Int J Psychophysiol. 1991;11(3):257–60.
27. Thomas NG, Harden LM, Rogers GG. Visual evoked potentials, reaction times and eye dominance in cricketers. J Sports Med Phys Fitness. 2005;45(3):428–33.
28. Tobimatsu S, Celesia GG. Studies of human visual pathophysiology with visual evoked potentials. Clin Neurophysiol. 2006;117(7):1414–33.
29. Zwierko T. Differences in peripheral perception between athletes and nonathletes. J Hum Kinet. 2008;19:53–62.
30. Zwierko T, Czepita DM, Lubinski W, Lubkowska A. Effects of physical effort on neuroretinal function in athletes and non-athletes: an electroretinographic study. Eur J Ophthalmol. 2010;20(2):381–8.
31. Zwierko T, Osinski W, Lubinski W, Czepita D, Florkiewicz B. Speed of visual sensorimotor processes and conductivity of visual pathway in volleyball players. J Hum Kinet. 2010;23:21–7.
Keywords:

VISION; REACTION TIME; EEG; SPORT; ATHLETE; VISUAL EVOKED POTENTIALS

© 2016 American College of Sports Medicine