Twelve weeks of physical exercise breaks with coordinative exercises at the workplace increase the sulcal depth and decrease gray matter volume in brain structures related to visuomotor processes

Throughout the lifespan, physical exercise, which is a planned, structured, and repetitive form of physical activity (= all muscle induced bodily movements, which leads to an increased energy expenditure) (Herold et al. 2022), is positively linked with physical health, mental health (Hillman et al. 2008), and maintenance of brain health (Cabral et al. 2019). Physical exercise can also prevent the development of neurodegenerative diseases (Kramer and Erickson 2007) and cognitive decline (Ma et al. 2017). Consequently, physical activity has become an important lifestyle factor to prevent cognitive health.

The volume of the gray matter (GM) of the brain changes with age (Fjell and Walhovd 2010), but regular physical activity can prevent those age-related changes. For example, Wittfeld et al. (2020) conducted a cross-sectional study involving adults aged between 21 and 84 years. They found that cardiorespiratory fitness, developed from a regular level of physical activity, is strongly positively related to the gray matter (GM) volume of the frontal lobe, temporal lobe, the hippocampal gyrus, and the cingulate cortex. In older adults, cardiorespiratory fitness and a regular level of physical activity were positively linked to the GM volume of the prefrontal cortex and the hippocampus (Erickson et al. 2014). Furthermore, in older adults, regular level of physical activity correlates positively with the GM volume of the frontal lobe, temporal lobe of the brain, and the hippocampus (Domingos et al. 2021; Erickson et al. 2014). A positive relationship between the GM volume of the hippocampus and a regular level of physical activity was also found for young to middle-aged adults (Killgore et al. 2013). In addition, in a recent cohort study by Fox et al. (2022), an association between a regular level of physical activity with larger total GM volume was found in adults aged between 30 and 94 years.

However, most studies which have investigated the impact of a regular level of physical activity or physical exercise on brain structure and/or function have focused mostly on aerobic, anaerobic, and resistance training (Ai et al. 2021; Alkadhi 2018; Erickson et al. 2014; Gaertner et al. 2018; Herold et al. 2019; Hillman et al. 2008; Sexton et al. 2016; Smith et al. 2010; Voelcker-Rehage and Niemann 2013; Voss et al. 2011; Wilke et al. 2019; Wittfeld et al. 2020). Although the first study that showed structural brain changes due to exercise in young healthy participants was based on juggling (Draganski et al. 2004), there are fewer studies investigating the effects of coordinative training on brain structure (see Voelcker-Rehage and Niemann 2013). For example, Niemann et al. (2014a) examined how coordinative training, which included exercises to improve eye–hand and leg–arm coordination, balance, and spatial orientation and reaction, affects the brain structure in older adults. They found an increase in the GM volume of the hippocampus (Niemann et al. 2014a) and the basal ganglia (Niemann et al. 2014b). Increases in the GM volume of the hippocampus (Rehfeld et al. 2017), and the parahippocampal region (Müller et al. 2017) were found after dancing training in elderly adults. Furthermore, increases after dancing training were found in the right subiculum, the left dentate gyrus (Rehfeld et al. 2017), left precentral gyrus (Müller et al. 2017; Rehfeld et al. 2018), postcentral gyrus, left supplementary motor area, left superior temporal gyrus, medial frontal gyrus, left insula, the anterior, and medial cingulate cortex (Rehfeld et al. 2018). After balance training, increases in GM volume were found in the left supplementary motor area (SMA), superior frontal gyrus, and medial orbitofrontal cortex (Taubert et al. 2010), but decreases in GM volume were found in the putamen (Rogge et al., 2018; Taubert et al., 2010), inferior orbitofrontal cortex, middle temporal gyrus, left inferior occipital gyrus, cerebellum bilaterally (Taubert et al. 2010), right superior temporal gyrus, and left parahippocampus (Weber et al. 2019). Juggling interventions led to increases in the GM volume of the hMT + /V5 (Boyke et al. 2008; Driemeyer et al. 2008), which is an area located at the occipito-temporo-parietal pit (Sousa et al. 2016), with relevance for hand movements (Oreja-Guevara et al. 2004) and processing tactile and visual motion direction information (van Kemenade et al. 2014). Further increases were found in the frontal, temporal (Driemeyer et al. 2008), parietal lobes (Draganski et al. 2004; Driemeyer et al. 2008), the hippocampus, and the nucleus accumbens (Boyke et al. 2008). Although Sampaio-Baptista et al. (2014) found no main effect of time after 6 weeks of juggling training on the GM volume, increases in GM volume in the occipital lobe and parietal lobe, and decreases in the superior temporal gyrus, the insula, and the operculum were found during a follow-up period 4 weeks after the training.

Besides the assessment of volumetric measures, the measurement of sulcal measures can be useful. Lamont et al. (2014) indicated that changes in brain structure due to regular physical activity can be detected earlier in sulcal measures compared to volumetric measures. Of the previous published studies on coordinative exercise, three studies included measures on cortical thickness (Rogge et al. 2018; Taubert et al. 2016; Weber et al. 2019) but, to our knowledge, no one included the variables sulcal width or depth. Especially the intraparietal sulcus has been found as being strongly implicated in visuomotor and cognitive functions (e.g., Capizzi et al. 2023; Davare et al. 2012; Grefkes and Fink 2005; Richter et al. 2019). Therefore, it would be important including these measures to assess the effects of coordinative exercise, which strongly draws on visuomotor task demands.

Nevertheless, the current state of research does not provide a sufficient amount of evidence to assess the influence of physical exercise including coordinative exercises on brain outcomes in young and middle-aged adults (Erickson et al. 2019). Another challenge is to reach this age group since the largest group of adults typically spend about half of their waking hours during weekdays at the workplace (Conn et al. 2009). Therefore, implementing a physical exercise intervention in their spare time might reduce the adherence, due to multiple other distractions and time restrictions. In contrast, the workplace offers a low-threshold opportunity to reach this age group (WHO 2018). Furthermore, office workers are confronted with demanding cognitive tasks, and if they are unable to perform those tasks, it could have a negative influence on mental well-being (Bridger and Brasher 2011). Therefore, it would be helpful to integrate coordinative exercises, like juggling, at the workplace because higher level cognitive processing is required for performing coordinative exercises (Voelcker-Rehage and Niemann 2013) and, compared to other exercise types, coordinative exercises are more beneficial for improving cognitive functions (Ludyga et al. 2020). In addition, performing coordinative exercises could lead to improvements in brain areas, e.g., frontal cortical (Driemeyer et al. 2008), parietal cortical (Draganski et al. 2004; Driemeyer et al. 2008; Sampaio-Baptista et al. 2014), and hippocampal area (Boyke et al. 2008), which are related to (higher) cognitive functions (Agosta et al. 2017; Behrmann et al. 2004; Borders et al. 2022; Frith and Dolan 1996; Helfrich and Knight 2019; Lisman et al. 2017; Opitz 2014; Toichi et al. 2004). Moreover, participating in a physical exerercise break (PEB) with coordinative exercises could be useful in preventing age-related structural (Fjell and Walhovd 2010; Kochunov et al. 2005; Steffener 2021) brain changes, and brain changes due to neurodegenerative diseases (e.g., Parkinson’s and Alzheimer’s disease) (Baghdadi et al. 2022; Chishiki et al. 2020; Pereira et al. 2012; Pettigrew et al. 2017; Ramírez-Ruiz et al. 2005). Furthermore, almost all studies so far implemented coordinative exercises with a higher training duration (> 75 min per week, Sampaio-Baptista et al. 2014), and if juggling training was performed, the training mostly focused on learning the three-ball cascade (Boyke et al. 2008; Draganski et al. 2004; Driemeyer et al. 2008; Sampaio-Baptista et al. 2014). However, since insufficient time (Justine et al. 2013) to exercise and exercise-related boredom (Velasco and Jorda 2020; Wolff et al. 2021) play a role for participating in physical exercise a high training volume combined with monotonous exercise tasks might represent a threshold for average office workers to start physical exercise.

Therefore, the aim of the present study was to investigate how 12 weeks of physical exercise breaks (PEBs) with coordinative exercises with a low training volume and a variety of juggling exercises performed at the workplace affects the brain structure in young and middle-aged adults. We hypothesized that the PEBs will lead to increases in the GM volume, and the surface-based brain metrics, especially in regions related to visuomotor tasks.

During the intervention period, four participants of the IG and three participants of the CG dropped out of the study (Fig. 1). At least 2/3 of the 24 intervention’s sessions was completed by 16 participants of the intervention group. Of those, 12 participants completed all 24 sessions (100.0%), two completed 22 sessions (91.7%), and one participant each completed 21 (87,5%), and 16 sessions (67.0%). The overall attendance adherence was 96.1% (± 8.7%). No adverse events occurred during the intervention. Accordingly, 25 participants were included in the final analysis (Fig. 1). Moreover, we excluded one participant from the CG from the analysis of the juggling performance data who was already a skilled juggler (but was not performing juggling training before or during the intervention).

Regular engagement over 12 weeks in low-volume PEBs with coordinative exercises was associated with an improvement in juggling performance and an increase in vertex-wise cortical depth in a cluster including the right inferior parietal lobe. In addition, a reduction of GM volume in a cluster primarily involving the right insula and the right operculum following the intervention was also observed. Decreases in GM volume of the right insula/operculum were correlated with improvements in juggling performance. This corroborates nicely with the fact that the volumetric decreases in this brain structure were related to the juggling training.

The increase in the sulcal depth in the inferior parietal lobe following juggling training is a novel finding. Sulcal depth is a measure of cortical shape (van Essen 2005), which has been associated with neurodevelopmental disorders (van der Meer et al. 2021), working memory performance (van der Meer et al. 2021; Yao et al. 2022), and changes of the structural characteristics of the aging brain (Jin et al. 2018). In one of the rare studies on sulcal characteristics in the context of physical activity, Lamont et al. (2014) reported evidence that higher physical activity was associated with larger sulcal width, while there was no significant relationship with sulcal depth. The identified sulcal cluster in this study appears to display some overlap with areas of the intraparietal sulcus, which have been referred to as “interfaces between the perceptive and motor systems for controlling arm and eye movements in space” (Grefkes and Fink, 2005, p.3). Grefkes and Fink (2005) summarized evidence from lesion studies and brain imaging studies revealing that this brain structure is implicated in tactile and visual object processing, and especially in cross-modal information integration between the visual and the sensorimotor systems. The medial portion of the intraparietal sulcus is thought to be implicated in tasks requiring visuomotor coordination of hand movements, especially in “transforming visual coordinates into motor programs, and for the online control of goal-directed precision movements” (Grefkes and Fink, 2005, p.9). Hence, the current study provides the first evidence that structural changes of the intraparietal sulcus can constitute an important biomarker for changes following a juggling training that poses relatively high visuomotor demands.

The decrease of GM volume in response to our 12-week long juggling training was restricted to a cluster involving the right insula, the right operculum, and the right inferior frontal gyrus. The right inferior frontal gyrus is a part of the Broca’s region and represents motor functions, including complex hand movements, associative sensorimotor learning, and sensorimotor integration (Binkofski and Buccino 2004). In a quite similar vein, the insula is known to be implicated in hand–eye movements (Fink et al. 1997), proprioceptive functions (Chilvers et al. 2021), and in sensorimotor and vestibular functions (Uddin et al. 2017). The frontal operculum has been found to be specifically involved in monitoring performance during goal-directed hand movements (Quirmbach and Limanowski 2022). Interestingly, when using transcranial magnetic stimulation, Tunik et al. (2008) even demonstrated a possible causal role of the inferior frontal gyrus opercularis in planning hand–object interactions. The observed training-related GM changes in this brain structure provide empirical support for the idea that this structure is of importance for visuomotor processes being involved in the coordination of hand and eyes. The observed association between brain changes in this structure and juggling performance nicely supports this finding. However, a somewhat surprising finding is that, compared to previous juggling studies, we only found decreases of GM volume after the juggling training. In previous research, mostly, though not exclusively, increases of GM volume in different areas were found (Boyke et al. 2008; Draganski et al. 2004; Driemeyer et al. 2008; Sampaio-Baptista et al. 2014). Furthermore, when comparing the brain structure of skilled and non-skilled jugglers, Gerber et al. (2014) found a higher GM density in the hMT + /V5 area, the intraparietal sulcus, and the secondary visual cortex. In particular, the GM density of the hMT + /V5 area was positively related to the performance of the skilled jugglers. However, compared to the juggling skills of the participants in the previous juggling studies, the participants of the IG of the present study were not very skilled jugglers after completing the intervention. The participants were only able to juggle for 4.4 s (SD = 4.4) after 12 weeks of training. In the other studies, participants reached or had to reach juggling times between 20 and 180 s (Boyke et al. 2008; Draganski et al. 2004; Driemeyer et al. 2008; Sampaio-Baptista et al. 2014). Nevertheless, Sampaio-Baptista et al. (2014) also found some decreases of GM volume following juggling, especially in the left superior temporal gyrus, the insula, and operculum. Moreover, in other studies where training with high coordinative demands was included, decreases of the GM volume were also found in different areas of the brain (Rogge et al. 2018; Taubert et al. 2010; Weber et al. 2019). For example, Weber et al. (2019) found that learning to ride a unicycle over a time period of 3 weeks was associated with significant reductions of GM volume in regions supporting visuospatial processes. Similarly, in professional or skilled athletes who are regularly engaged in tasks requiring a high level of coordinative skills, both greater and smaller GM volumes have been reported. Skilled golf players have a higher GM volume for the premotor cortex and parietal areas, compared to less-skilled golf players (Jäncke et al. 2009). Professional ballet dancers have a lower GM volume in the left premotor cortex, SMA, putamen, and superior frontal gyrus, compared to non-skilled dancers (Hänggi et al. 2010). The results of a cross-over study of skilled dancers (ballet and figure skating) and slackliners showed a lower GM volume in the anterior hippocampal formation and parieto-insular vestibular cortex, but a higher GM volume in the posterior hippocampal formation, lingual, and fusiform gyri, compared to a non-trained CG (Hüfner et al. 2011). Taken together, these findings clearly indicate that these decreases in GM volume do not indicate that a brain structure is “deactivated” following the training, but they may rather reflect the reorganization of brain tissue facilitating more automated and efficient task performance (in this case, visuomotor coordination; see Weber et al. 2019).

It is evident that brain changes following physical exercise vary as a function of the practice level. For example, Sampaio-Baptista et al. (2014, 2015) found that a juggling training with a lower duration of the exercise sessions (75 min per week) showed a negative correlation between the brain changes of GM volume of the left motor cortex and dorsolateral prefrontal cortex (DLPFC) with the changes in juggling performance. In contrast, a juggling training with higher duration of the exercise sessions (150 min per week), the same relationship was positive. For the white matter (WM) volume, the relationship was the opposite, i.e., a juggling training with a lower duration of the exercise sessions had increased functional connectivity and a juggling training with higher duration of the exercise sessions had decreased functional connectivity, as indicated by increased and decreased motor resting state structure strength, respectively. It is important to mention that the juggling performance was not significantly different between both these groups after finishing the juggling intervention. Weber et al. (2019) found similar effects after 3 weeks of unicycle training. They reported a decrease in the GM volume and an increase of the WM fractional anisotropy after the training. Both research groups, i.e., Sampaio-Baptista et al. (2014, 2015) and Weber et al. (2019), suggested that the changes rely on the different reorganization processes of brain tissue and are moderated by the participant’s skill or amount of practice. Weber et al. (2019) showed that the GM decreases in task-relevant structures following unicycle training subsequently increased again after a follow-up period during which no unicycle training was performed. This strongly supports the view that brain changes following physical activity are highly dynamic and strongly related to the amount of physical exercise.

This study also has some limitations. Firstly, we did not assess changes in diffusion-weighted MRI and functional characteristics of the brain following the intervention. Recent research has indicated that the combined consideration of these different imaging modalities provides a comprehensive and more holistic view of the manifold brain changes associated with coordinative exercises (e.g., Weber et al. 2019). A further limitation is the higher number of women (n = 19) compared to men (n = 6) participating in the current study. According to the literature, it is more likely that women participate in worksite programs (Beck et al. 2016). Moreover, these results are in line with other studies implementing physical exercise programs at the workplace, where a higher percentage of women (between 64.2 and 90.0%) participated in the programs (Dalager et al. 2017; Gram et al. 2014; Grande et al. 2015; Hartfiel et al. 2011; Hunter et al. 2018; Puig-Ribera et al. 2008). Finally, no a priori sample size calculation was done and the sample size of this study is small, which certainly limits the statistical power. This limitation of the statistical power could lead to a reduction of the likelihood that significant results reflect a true effect (Button et al. 2013). Nevertheless, the fact that we found brain changes in regions strongly implicated in visuomotor processes indicated that the findings are meaningful. This study was hypothesis generating and should serve as basis for future studies in this field where we await replication of these findings in larger samples of participants.

The present study provides the first evidence that 12 weeks of PEBs with coordinative exercises and with a low training volume and a variety of juggling exercises performed at the workplace result in changes in GM volume and sulcal depth. These changes were found in brain structures strongly implicated in visuo-coordinative processes involving hand and arm movements. Future research in this field would benefit by the use of different brain imaging modalities. In particular, the assessment of magnetic resonance spectroscopy in the context of physical exercise could also provide important new insights into the biochemical and cellular mechanisms underlying the observed changes in brain volume.