The study was approved by the local ethics committee of the Medical Faculty at the University of Leipzig (309/17-ek). All participants gave written informed consent to participate in the experiments according to the Declaration of Helsinki.

A total number of 16 healthy participants [mean age: 27.69 ± 0.86 years (mean ± SE); range 20–32 years; three females] were enrolled in the present study. None of the participants had a history of neurological illness, and during the time of the experiment, none of the volunteers was taking any central-acting drugs. All participants were right-handed according to the Edinburgh Handedness Inventory (mean handedness score: 79.41 ± 5.25; Oldfield, 1971). Total hours of sports per week (mean: 7.09 ± 1.87 h), hours of fine-motor training/activity per week (e.g., playing a musical instrument, knitting, doing handcrafts, playing video games, mean: 1.94 ± 0.73 h) as well as the extent of slacklining activities in the past and present were assessed prior to testing. Hence, participants were recruited based on their slacklining experience (mean: 7.25 ± 0.68 years) and their current slacklining activities (mean: 4.09 ± 0.56 h per week) and were therefore considered as advanced slackliners.

The present study on advanced slackliners aimed at quantifying hemodynamic response alterations during standing on a slackline (ST) and walking on a slackline (WA) using a solid slacklining frame (Slackstar, Braun GmbH, Neurmarkt, Germany). The slackline had a length of 4 m and a height of 50 cm with small platforms at each end at the same height (see Figure 1D). Using a within-subject design, both conditions were performed in a block-wise manner for 10 × 30 s (30 s resting phases between trials) in one session (see Figure 1A). The order of conditions was randomized; however, all trials of each condition were performed consecutively in one block.

During WA, participants were asked to walk forward over the slackline at a preferred pace. Platforms were used for direction changes and inter-trial resting phases, respectively. During ST, participants were asked to stand in the middle of the slackline without changing their standing leg. In case of a loss of balance (drops from the slackline), participants were asked to get back on the slackline immediately. Motor performance was videotaped and assessed as the number of drops per trial. Additionally, after each condition participants were asked to rate the perceived level of task difficulty on a scale from 1 (very easy) to 10 (very difficult). Moreover, a chest strap (Polar Electro Oy, Kempele, Finland) was applied to assess HR. Before and after the entire experiment, participants rated their levels of attention, fatigue, and discomfort on a visual analog scale (VAS).

Hemodynamic responses during ST and WA were recorded using the portable NIRSport system (NIRx Medical Technologies, Glen Head, NY) covering sensorimotor brain areas [premotor and supplementary motor cortex (PMC–SMA), primary motor cortex (M1), somatosensory association cortex (SAC), and inferior parietal cortex (IPC)] on both hemispheres (see Figure 1C). Using a total of eight light sources and seven detectors with an inter-optode distance of 30 mm, hemodynamic responses were measured in 22 channels (see Figure 1B). Optodes were placed on an fNIRS cap (with different sizes according to participants' head sizes) including optode distance holders, which ensured fixed source-detector distances and standardized sensor placement according to the international 10–20 system. Infrared light was emitted by sources with wavelengths of 760 and 850 nm. The NIRSport system uses time and frequency multiplexing to minimize crosstalk between wavelengths and optodes and acquires data with a sampling frequency of 7.81 Hz.

The comparison of slacklining conditions on a behavioral level revealed a significant difference (p = 0.019, Z = −2.354) between WA [1.31 ± 0.51 (mean ± SE) drops] and ST (3.63 ± 1.02 drops) (see Figure 2A). However, with regards to perceived level of task difficulty, Wilcoxon test revealed no differences (p = 0.317, Z = −1.000, see Figure 2B) between WA (3.06 ± 0.28) and ST (3.38 ± 0.36).

With regards to HR, rmANOVA revealed a non-significant trial × condition interaction [F_{(1, 9)} = 1.220, p = 0.289, η²_p = 0.080]. Moreover, we found no significant influence of factor condition on HR [main effect of factor condition: F_{(1, 14)} = 1.008, p = 0.332, η²_p = 0.067]. Only factor trial showed a significant influence on HR [main effect of factor trial: F_{(9, 126)} = 170.917, p = 0.001, η²_p = 0.413], indicating alteration in HR over time during both conditions (see Figure 2C). In detail, post-hoc tests revealed a significant decline in HR between trial 1 and trial 3–10 as well as trial 2 and trial 4–10 [t₍₁₅₎ = 3.53–5.39, p_adjusted < 0.024].

We hypothesized to observe an involvement of sensorimotor brain areas such as M1, PMC, and SMA, since both standing and walking on a slackline are considered demanding and represent multisensory balance tasks (hypothesis a). Indeed, we found hemodynamic response alterations in the above-mentioned regions for both chromophores during both slacklining conditions. These results go in line with previous studies investigating the crucial role of sensorimotor brain areas for balance control (Hiyamizu et al., 2014; Herold et al., 2017a; Seidel et al., 2017). For example, similar to our findings, Herold et al. (2017a) recently observed enhanced cortical activity in SMA, precentral gyrus (PrG), and postcentral gyrus (PoG) using fNIRS during the execution of a balance task on a balance board. Similar activation pattern were found in another recent study by Seidel et al. (2017) in endurance athletes and non-athletes. Moreover, Herold et al. (2017a) even found a strong negative correlation between the magnitude of SMA activity and sway in mediolateral direction during task execution using an additional inertial sensor. In another fNIRS study, ΔHbO₂ concentrations in SMA were increased after a single training session on a balance board (Hiyamizu et al., 2014). These findings suggest a crucial involvement of sensorimotor brain areas such as M1, PMC, and SMA in balance control since different sensory information are integrated and processed in these brain regions during the execution of a motor task (Nachev et al., 2008).

Based on the above mentioned results, hemodynamic response alterations in sensorimotor brain areas during balancing on a balance board and also during slacklining can be explained by indirect and direct locomotor pathways as proposed by Herold et al. (2017b). According to the authors, PMC and SMA are part of an indirect locomotor pathway indicating a more controlled task execution. More automatic motor control, in contrast, is realized via a direct locomotor pathway, which includes M1 and other brain regions such as the cerebellum. The fact that we observed an increased brain activity in areas of both the direct and the indirect locomotor pathway indicates that slacklining, even by advanced slackliners, can neither be exclusively assigned to a controlled nor an automatic motor control during task execution. In fact, the demands of slacklining seem to be more complex (higher number of degrees of freedom) that both controlled and automatic components of motor control are required for task execution. Therefore, it remains speculative whether experienced slackliners would show less and/or more focal cortical activity during balancing on a balance board as compared to balance untrained participants, which should be addressed more precisely in future studies.

We further hypothesized to detect task-related differences in hemodynamic response alterations between slacklining conditions (hypothesis b). Indeed, we observed no differential neural effects between ST and WA condition, which might be explained on a behavioral level by the fact that participants did not perceive any differences regarding the task complexity of both conditions. However, the significant differences in motor performance (drops from slackline) in turn point to a differential task complexity, interestingly toward a more challenging ST condition. This in turn might be explained by the fact that ST might be rather untypical for slacklining, requires altered motor control as compared to WA and therefore causes a higher number of drops. On the one hand, the lack of task-related differences in brain processing may be attributed to methodological issues discussed in the limitation section.

On the other hand, to date, investigations of task-related neuronal effects using fNIRS are rather sparse. However, as previously demonstrated by Holper et al. (2009), fNIRS is indeed capable of detecting differences in hemodynamic response alterations as a function of task complexity. In this study, participants performed five different finger tapping tasks with increasing task complexity: unimanual simple and complex tasks (each with left and right hand) as well as bimanual complex tasks. The authors reported significant hemodynamic response alterations in M1 for both ΔHbO₂ and ΔHHb between all conditions (Holper et al., 2009), confirming evidence on hand/finger tapping tasks from previous fMRI studies (Solodkin et al., 2001; Verstynen et al., 2005; Witt et al., 2008). Based on this evidence, it can therefore be assumed that it is crucial for observing differential task-related effects whether the task is executed unilaterally or bilaterally. This is further supported by a recent fNIRS investigation by Carius et al. (2020b), where participants performed a complex basketball slalom dribbling task. Here, the authors found that task-related hemodynamic response alterations in M1 are related not only to laterality but also to the pace of task execution. Hence, the lack of differential hemodynamic response alterations between conditions in our results might be attributed to the fact that both ST and WA were performed bipedally and also the upper extremities each realized comparable support functions. Further explanations, however, remain purely speculative since, to our knowledge, there is no further evidence to date regarding differential task-related neuronal effects in lower limb or whole-body motor tasks.

In addition, we performed a correlation analysis investigating the effect of slacklining experience on hemodynamic response alterations during slacklining (hypothesis c). Here, we found no significant correlation for both chromophores, neither for ST nor for WA. However, without applying FDR-correction, our results revealed that slacklining experience might be associated with lower hemodynamic response alterations during ST. This finding can be considered a tentative indication that more experienced slackliners exhibit less cortical activity, mainly within right PMC–SMA and right M1. This trend would be consistent with previous studies showing that experts have more efficient neural networks while performing motor tasks (Seidel et al., 2017; Carius et al., 2020a) as proposed by the “neural efficiency” hypothesis (Dunst et al., 2014). Moreover, less cortical activity in PMC–SMA might indicate a greater automaticity and fine tuning of the motor performance in higher-skilled individuals, as automaticity in the motor learning process is associated with a decrease in frontal and secondary motor areas (Poldrack et al., 2005). This would also be the case for less cortical activity in M1, which, as part of a direct locomotor pathway, also points to automaticity of a motor/balance task (Herold et al., 2017b). All in all, however, the uncorrected findings of the correlation analysis need to be considered with caution and the potential associations need to be retested using a larger sample size with a bigger range of expertise including also novices and professionals.

From a methodological point of view, another explanation for the absence of significant associations between slacklining experience and hemodynamic response alterations might be attributed to the method used for rating the participants, since only slacklining experience in years was surveyed. More slacklining experience, however, is not necessarily associated with a higher slacklining expertise and/or superior slacklining performance. Hence, future studies on experts should find a solution to operationalize the slacklining skills more precisely. In addition, further behavioral parameters (e.g., body sway, pace of execution, number of steps, vibration of the slackline, etc.) should be recorded during task execution in order to quantify slacklining performance and to explain related hemodynamic response alterations more profoundly. Indeed, this would also be conceivable by using state-of-the-art approaches such as motion capturing, allowing the quantification of body part movements (e.g., arms) and the identification of potential (balance) execution strategies.

In the present study, hemodynamic response alterations were assessed during slacklining, which can be considered a challenging whole-body balance task. Therefore, only advanced slackliners were recruited, who were able to perform both conditions of the task reliably in a block design consisting of multiple trials. Since slacklining novices were expected not to be able to perform this experimental procedure reliably without causing a minor data quality, an additional novice control group was not included. This, however, came with the cost of a missing classification of the results as specific for advanced slackliners. From a methodological perspective it is worth mentioning that hemodynamic response alterations were only captured in sensorimotor brain areas since this was the main focus of the present study. There is no doubt, however, that other brain areas outside the human motor system are also involved in the execution of a whole-body balance tasks such as slacklining. Hence, future studies should address this issue using whole-brain fNIRS montages. Furthermore, we only investigated hemodynamic response alterations within cortical sensorimotor brain areas and not in extra- and/or sub-cortical regions, which is due to the limited penetration depth of fNIRS. Beyond that, it is necessary to mention that whole-body balance tasks such as slacklining might lead to motion artifacts, e.g., caused by head movements, influencing the fNIRS signal. Therefore, we applied state-of-the-art motion artifact and baseline correction methods to minimize the effect of artifacts on the data. In future studies, further influences on the fNIRS signal such as physiological confounders (e.g., scalp blood flow) should be reduced by using multi-distance fNIRS measurements. Furthermore, additional physiological data assessments such as respiratory measurements might help quantifying the physical effort of the slacklining task more precisely and monitoring systematic influences.