- Email: [email protected]
Changes in brain cortical activity measured by EEG are related to individual exercise preferences
Physiology & Behavior 98 (2009) 447–452
Contents lists available at ScienceDirect
Physiology & Behavior j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / p h b
Changes in brain cortical activity measured by EEG are related to individual exercise preferences Stefan Schneider a,⁎, Vera Brümmer a, Thomas Abel a, Christopher D. Askew b, Heiko K. Strüder a a b
Institute of Movement and Neurosciences, Dept. of Exercise Neuroscience, German Sport University Cologne, Germany Faculty of Science, Health and Education, University of the Sunshine Coast, Maroochydore, Queensland, Australia
a r t i c l e
i n f o
Article history: Received 7 May 2009 Received in revised form 29 June 2009 Accepted 20 July 2009 Keywords: EEG Physical exercise history Electrotomography Exercise preferences LORETA
a b s t r a c t Exercise is well known to result in changes of brain cortical activity measured by EEG. The aim of this study was (1) to localise exercise induced changes in brain cortical activity using a distributed source localisation algorithm and (2) to show that the effects of exercise are linked to participants' physical exercise preferences. Electrocortical activity (5 min) and metabolical parameters (heart rate, lactate, peak oxygen uptake) of eleven recreational runners were recorded before and after incremental treadmill, arm crank and bicycle ergometry. Electroencephalographic activity was localised using standardised low resolution brain electromagnetic tomography (sLORETA). Results revealed an increase in frontal α activity immediately post exercise whereas increases after bike exercise were found to be localised in parietal regions. All three kinds of exercise resulted in an increase of β activity in Brodmann area 7. Fifteen and thirty minutes post exercise a specific activation pattern (decrease in frontal brain activity–increase in occipital regions) was noticeable for treadmill and bike but not arm crank exercise. We conclude that specific brain activation patterns are linked to different kinds of exercise and participants' physical exercise preferences. © 2009 Elsevier Inc. All rights reserved.
1. Introduction In the last twenty years the promotion of exercise has become an important public health message. This can be attributed primarily to the impact of exercise on cardio-respiratory and metabolic parameters and their influence on physical health [1]. But apart from those peripheral changes recent evidence suggests that there is also a major influence of exercise on brain cortical function which seems to be connected to changes in cognitive performance, general arousal and well being [2,3]. It is generally agreed that there are temporary changes in the EEGs α and β activities caused by exercise [4,5]. These changes were first reported in the 1950s [6] and are considered to reflect a state of accumulated relaxation [4,7,8]. An important aspect when considering effects of exercise on brain cortical activity is to assign effects to specific brain regions. Recent publications have contributed to the hypothesis that especially α activity within frontal brain regions is affected by exercise [9–13]. Although promising a fundamental limitation of these approaches is
⁎ Corresponding author. Institute for Movement and Neurosciences, German Sport University Cologne, Carl-Diem Weg 6, 50933 Köln, Germany. Tel.: +49 173 7078760; fax: +49 221 4973454. E-mail address: [email protected] (S. Schneider). 0031-9384/$ – see front matter © 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.physbeh.2009.07.010
linked to missing possibilities of brain imaging during and immediately after exercise. In the recent years standardised low resolution brain electromagnetic tomography (sLORETA) has become an accepted method to localise changes in brain cortical activity by analysing EEG derivations [14–17] especially in experimental situations that do not allow to use standard brain imaging methods like magnetic resonance imaging (MRI) or positron emission tomography (PET) [13,18]. sLORETA is a source localisation method relying on mathematical models of the bio-electrical generators and the volume conductors within which they lie. It is based on standardised EEG recordings, which are modeled to a probabilistic head model provided by the Montreal Neurological Institute (MNI). Active cortical regions are identified created by allocating the raw sLORETA values of individual voxels to their corresponding Brodmann areas (BA) or cerebral gyri on the basis of the coordinates of the digitized Talairach Atlas [19]. Apart from sLORETA there are similar approaches to overcome the inverse solution like BESA, ST-MAP, MUSIC and others. By comparing these methods it gets clear that each has its own specific advantages but sLORETA was shown to give the most satisfactory results [20]. Just recently we were able to show that major changes in brain cortical activity occurred when individuals were allowed to exercise at their preferred intensity rather than constricting their intensity by heart-rate limits [21]. This is in accordance with previous findings, linking relaxational effects of exercise to individuals' physical activity
448
S. Schneider et al. / Physiology & Behavior 98 (2009) 447–452
history and exercise preferences [8,22]. Accordingly we hypothesized that changes in brain cortical activity are connected not only to preferred exercise intensities but also to participants' exercise history. Within this study we tested a group of regular runners with little or no preferences to bike or arm crank exercise (i.e. none of these participants ever used an arm crank before, none of them was a regular biker) on changes in brain cortical activity after treadmill, bike and arm crank ergometry. We hypothesized that previously reported changes in EEG α and β activities, which were linked to recreational processes (changes in frontal α activity [12] and overall β activity [23]) would be dominant after treadmill but not arm crank and bike exercise. 2. Materials and methods 2.1. Participants and procedure The Institutional Review Committee of the German Sport University approved this study. All procedures were in compliance with the Declaration of Helsinki for human subjects. Twelve adults, either students or academic staff of the German Sport University, aged 26.3 ± 3.8 (male (n = 8): 25.6 ± 4.2; female (n = 4): 27.5 ± 2.89) participated in this study. All participants were regular runners performing a minimum of 2 h weekly training mainly for health reasons. Some of the participants also reported to cycle regularly but only to travel within the city and without any exercise background. None of the participants had previous experience with arm crank ergometry or similar (e.g. handbiking). All participants received a pre exercise medical screening. Incremental peak oxygen uptake (VO2peak) tests were performed either on an arm crank ergometer (Cyklus2 — Richter, D), a bicycle ergometer (Ergoline ER 900 — Ergoline, Bitz, Germany) or a treadmill (Woodway ELG 55 — Woodway, Weil am Rhein, Germany). Each participant performed one VO2peak test a day and testing days were separated by at least three days. In order to prevent the effects of sequence participants were randomly assigned into one of the three groups either (1) arm crank–bike–treadmill, (2) bike–treadmill–arm crank or (3) treadmill– bike–arm crank. Arm crank performance started at 20 W and was increased every 3 min by 20 W [24]. Bike ergometer performance started at 50 W and was increased by 50 W every 3 min [25]. The incremental treadmill test started at 2.0 m/s and was increased every 3 min by 0.5 m/s [26]. All tests ended at volitional exhaustion (RPE scale 19–20, metabolic respiratory quotient ≥ 1). At the end of each stage 20 μl capillary blood from the ear lobe was collected for blood lactate (LAC) analysis (Biosen C_Line Glukose-/Laktat-Messsystem — EKF-diagnostic, Barleben, Germany). Rating of perceived exertion (RPE) was assessed at the end of each stage (Borg-scale) and heart rate (HR) was recorded at 5 s intervals (Polar S810i — Polar Electro, Kimpele, Finland). Expired gasses were continuously collected and analysed at 30 s intervals (ZAN 600 ErgoTest — ZAN, Oberthulba, Germany) for the determination of oxygen uptake, carbon dioxide production and ventilation. Peak oxygen uptake during incremental stage test (VO2peak) was taken as the highest 30 s value throughout the test. EEG activity was recorded for 5 min in a supine rest position with eyes closed prior to each test (PRE), 2 min after exercise (POST2) and again 15 (POST15) and 30 min (POST30) after exercise. To have similar conditions to PRE measurement subjects were asked to sit up and stay seated in between measurements. As exercise intensity was at a maximum level, we chose a supine position to avoid blackouts (unpublished observation) as EEG was recorded immediately after exercise. Moreover we were worried that an acute shift of body fluids to the lower body parts in a seated position after exercise might influence brain cortical activity [18]. 2.2. EEG recording A 64-channel portable EEG-System (IT-Med, Usingen, D) was used for data acquisition. Thirty minutes prior to exercise an EEG-Cap was mounted (Electro-Cap International, Inc., USA). This was adapted to
individual head size and built of 19 electrodes and one reference electrode (mounted in the triangle of FP1, FP2 and FZ) in the 10–20 system [27]. It recorded EEG activity on positions Fp1, Fp2, F3, F4, F7, F8, Fz, C3, C4, Cz, P3, P4, P7, P8, Pz, T7, T8, O1, and O2. The cap was fixed with two strings to a belt around the chest to prevent shifting during the exercise trials. The cap was permeable to air in order to prevent an increase in heat during exercise. Distances between electrodes were approximately 5 cm to prevent possible cross talk after exercise due to salt bridges between electrodes. Each electrode was filled with Electro-Gel™ (Electro-Cap International, Inc., Eaton, USA) for signal transduction. If impedance of an electrode exceeded 10 kΩ this single electrode was excluded from further analysis. The analogue signal of the EEG was amplified and converted to digital signals using Braintronics Iso 1064 CE Box (Braintronics B. V., Hl Almere, Netherlands; hardware filters: high pass — fixed time constant of 1 s; low pass — 97 Hz ± 15% (−3dB)) and stored with a sampling rate of 256 Hz on a hard disk of Neurofile XP EEG-System (IT-Med, Usingen, Germany). 2.3. Data analysis — EEG/LORETA sLORETA enables the spatial identification and analysis of brain cortical activity via traditional EEG recordings [14,15,28,29]. Scalp recorded cranial EEG activity is evoked by synchronised postsynaptic potentials (PSPs), which are located in clusters of pyramidal cells [30]. sLORETA makes it possible to determine the three dimensional orientation of these highly synchronised PSPs. sLORETA software is based on a probabilistic MNI brain volume which was scanned at 5 mm resolution. The MNI coordinates were converted to “corrected” Talairach coordinates, then given to the Talairach Daemon. Voxels that were unambiguously labeled as cortical grey matter, and that fell unambiguously within the brain compartment, were retained. This produced 6239 cortical grey matter voxels at 5 mm resolution [14]. Coordinates given within this study will refer to the MNI152 template. Cortical regions are created by allocating the raw sLORETA values of individual voxels to their corresponding Brodmann areas (BA) or cerebral gyri on the basis of the coordinates of the digitized Talairach Atlas [19]. EEG data were analysed by Brain Vision Analyzer (Brain Products, München, Germany). After manual artefact detection, high and lowpass filters were applied so that a frequency range from 0.5 to 49 Hz remained for analysis (time constant 0.3183 s; 24 dB/octave). Data were then segmented into 4 s sections where an overlap of 10% was accepted. An allowance of automatic artefact rejection was set to: gradient b 50 μV. Additional data were checked for eye movement artefacts manually and by ICA analysis using the semiautomatic ICA application in Analyzer2 (Brain Products, München, Germany). This resulted in a minimum of seventy 4 s segments of artifact-free data. In sLORETA software the coordinates of the 19 electrode positions were applied to a digitized MRI version of the Talairach Atlas (McConnell Brain Imaging Centre, Montréal Neurological Institute, McGill University). These Talairach coordinates were then used to compute the sLORETA transformation matrix. Following transformation to average reference EEG activity the seventy 4 s segments of artifact-free resting EEG were averaged to calculate cross spectra in sLORETA for α (7.5– 12.5 Hz) and β (12.5–35 Hz) bands for each subject and each of the four measurements. Using the sLORETA transformation matrix, cross spectra of each subject and for each frequency band were then transformed to sLORETA files. This resulted in the corresponding 3D cortical distribution of the electrical neuronal generators for each subject. To display differences in EEG spectral activity between PRE and POST2/POST15/POST30 measurements the following tests were performed: POST2 vs. PRE, POST15 vs. PRE and POST30 vs. PRE. Paired sampled t-test was computed for sLORETA power at each voxel. Statistical significance was assessed using a nonparametric
S. Schneider et al. / Physiology & Behavior 98 (2009) 447–452
randomisation test [31]. To correct for multiple comparisons, a nonparametric single-threshold test was assessed defining a critical threshold (tcritical). Voxels with statistic values exceeding this threshold have their null hypothesis, i.e. no difference in EEG power between the two tests (e.g. POST2 vs. PRE), rejected. The omnibus hypothesis (that all the voxel hypotheses are true) is rejected if a voxel value exceeds the critical threshold for p b .05 defined by 5000 randomisations. Voxel-by-voxel t-values in Talairach space are displayed as statistical parametric maps (SPMs). 2.4. Statistics — physiological values Changes in HR, LAC and VO2-uptake were analysed using the factors of exercise (treadmill, arm crank, bike) and time (baseline value, value at maximal exhaustion). Exercise time was compared using repeated measures ANOVA with the factor exercise (treadmill, arm crank, bike). Where ANOVA revealed significance, Fisher LSD was used for post-hoc analysis in all cases. Data were analysed using Statistica 7.1 (StatSoft, Tulsa, USA).
449
3.2. sLORETA pre vs. post Pre vs. post results showed an increase of activity across the whole scalp in the α and β frequency ranges (Fig. 2). This was found to be significant (tcritical for p b .05 = 3.95⁎; p b .01 = 5.11⁎⁎) after treadmill exercise in BA 6, 8, 9 (frontal lobe) and BA 24, 32 (limbic lobe) for α activity (p b .05, t = 4.76⁎) and in BA 7 (parietal lobe) for β activity (p b .01, t = 5.98⁎⁎). A widespread increase of activity, which reached significance (tcritical for p b .05 = 3.52⁎; p b .01 = 4.19⁎⁎) in BA 7 (parietal lobe) and 23, 31 (limbic lobe) within the α (p b .01, t = 4.83⁎⁎) and β band (p b .01, t = 4.91⁎⁎) could be observed after bike exercise (Fig. 2). Immediately after arm crank exercise significant changes (tcritical for p b .05 = 3.16⁎; p b .01 = 3.92⁎⁎) could be obtained for α activity (p b .05, t = 3.17⁎) in one voxel only (MNI = −45/20/15, BA 45) whereas β activity was significantly increased (p b .01, t = 3.93) in BA 7 and 40 (parietal lobe) (Fig. 2).
3.3. sLORETA pre vs. post15 and post30 3. Results 3.1. Physiological values during incremental exercises Mean exercise time showed no differences between the three different kinds of ergometry (treadmill 15.42 ± 2.81 min; arm crank 15.42±4.08 min; bike 15.00±3.19 min; F(2,22) =0.14; p=0.87; Fig. 1A). LAC values increased after exercise (pb 0.001). No significant differences were found in LAC values after arm crank, bike and treadmill performance (F(2, 20) =0.99; p=0.39; Fig. 1C). Heart rate was increased at maximal exhaustion (p b 0.001). Differences in maximal heart rate (HRmax; F(2, 20) = 5.55; p b 0.05) were obtained between treadmill and arm crank as well as bike and arm crank exercises (p b 0.001). Differences in HRmax between bike and treadmill exercises did not reach significance (p = 0.21; Fig. 1B). Participants showed a higher VO2peak (F(2, 20) = 42,49; p b 0.001) after treadmill and bike exercises compared to arm crank exercise (p b 0.001; Fig. 1D). A slightly lower value which marginally missed significance (p = 0.08) was found after bike exercise compared to treadmill exercise. RPE values were at a maximum level (19 to 20) after each kind of exercise.
No significant effects were found for changes in α (post15 t = 4.03; post30 t = 4.00) and β activities (post15 t = 3.69; post30 t = 2.96) after treadmill exercise (post15 tcritical for p b .05 = 4.16; post30 tcritical for p b .05 = 4.44). α activity was found to be significantly increased (post15 t = 3.78⁎, p b 0.05; post30 t = 3.95) after bike exercise in BA 6, 9 (frontal lobe), 24 and 32 (limbic lobe, Fig. 3). No significant changes could be obtained for β activity (post15 t = 3.63; post30 t = 2.77) after bike exercise (post15 tcritical for p b .05 = 3.64; post30 tcritical for p b .05 = 4.15). Changes in α (p b .05; t = 3.94⁎) and β activities (p b .05; t = 3.63⁎) 15 min after arm crank exercise were found significantly increased mainly in the left and right temporal lobes. β activity 30 min post exercise was increased in BA30/31 (limbic lobe; p b .05; t = 4.24) (post15 tcritical for p b .05 = 3.45; post30 tcritical for p b .05 = 3.61). Specific activation patterns could be noticed after treadmill and bike exercises compared to arm crank exercise (Figs. 2 and 3). Whereas an increase of α activity could be noticed in parietal and occipital regions across all three kinds of exercise, a decrease of β
Fig. 1. Illustration of exercise time (A), heart rate (B), blood lactate (C) and VO2peak (D) across all three kinds of exercise (treadmill, arm crank, bike). Rest values and values at volitional exhaustion are shown. Bars show means and .95 confidence intervals. ⁎⁎⁎p b .001.
450
S. Schneider et al. / Physiology & Behavior 98 (2009) 447–452
Fig. 2. Statistical parametric maps (SPM) of sLORETA differences in the α and β frequency bands comparing POST2 and PRE (n = 12). Red and yellow colours indicate increased activity in the POST2 measurement. Orthogonal views of the cortex (unthresholded) are displayed. Images depicting SPMs seen from different perspectives are based on voxel-byvoxel t-values of differences. Structural anatomy is shown in grey scale (L left, R right, A anterior, P posterior). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
activity in frontal brain regions was noticeable after treadmill and bike exercises only.
the scalp. 15 and 30 min post exercise a specific activation/ deactivation pattern in the β frequency range could be recognized for bike and treadmill exercises.
4. Discussion 4.1. Physiological parameters This study examined changes in brain cortical activity after different forms of exercise performed at maximal intensity. We used three different kinds of ergometry (treadmill, bike, arm crank) and recorded EEG activity prior to and up to half an hour post ergometry. A localisation analysis using sLORETA showed an increase of α and β activities immediately post exercise, which was spread widely across
To our knowledge no comparison between treadmill, arm crank and bike ergometry has been conducted so far using the same set of participants. No changes were observed between exercise duration and LAC concentration between the three kinds of ergometry, but HRmax and VO2peak were clearly lower after arm crank ergometry (Fig. 1BD). This
S. Schneider et al. / Physiology & Behavior 98 (2009) 447–452
combination of reduced HRmax and VO2peak let us suggest that final exhaustion and dropout after arm crank exercise are caused by local muscle fatigue rather than cardio-respiratory parameters [32]. This is supported by participant reports of cramped forearm muscles after arm crank exercise. 4.2. EEG data The overall increase in α and β activities observed immediately after exercise is in accordance with prior results. The first findings from Beaussart et al. [6] showed an increase of around 20% in α activity after
451
exercise at a mean maximum heart rate of 200. Subsequent studies [5,22] confirmed that the increase in α activity was strongest during the first 5 to 10 min after exercise and returned to pre exercise levels by about the 30th minute of recovery. Further sLORETA analysis revealed an increase of activity in α power, which was localised in frontal brain areas after treadmill exercise whereas α power after bike exercise was found to be increased in parietal areas and no significant changes could be obtained after arm crank exercise (Fig. 2). Given the fact that participants in this study reported to have a clear preference for running, an increase in frontal α activity (please remember that α activity is inversely related to cerebral
Fig. 3. Statistical parametric maps (SPM) of sLORETA differences in the β frequency band comparing POST15 vs. PRE (left) and POST30 vs. PRE (right, n = 12). Red and yellow colours indicate increased activity, blue colours decreased activity in the POST15 and POST 30 measurement. Orthogonal views of the cortex (unthresholded) are displayed. Images depicting SPMs seen from different perspectives are based on voxel-by-voxel t-values of differences. Structural anatomy is shown in grey scale (L left, R right, A anterior, P posterior). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
452
S. Schneider et al. / Physiology & Behavior 98 (2009) 447–452
activity) let us assume that the reported effects of treadmill exercise are linked to individuals' physical activity preferences as initially hypothesized. Special attention should be paid to the fact that changes were localised within frontal regions, which are well known to be involved in emotional processing [33–35]. As extensive research within the last years has shown that (1) exercise is linked to changes in mood and general well being [23,36,37] and (2) exercise seems to be connected to changes in frontal α activity [9–13], it could be assumed that the results observed here propose a connection between exercise preferences, frontal lobe activity and mood. Unfortunately we did not check for changes in mood. This should be taken into consideration by future research activities. An influence of individuals' physical activity preferences is further more supported by specific changes within the β frequency range 15 and 30 min post exercise (Fig. 3) which shows a reduction of activity in frontal brain areas after treadmill and bike but not arm crank exercise, which was totally unfamiliar to our participants (please be aware that a reduction of β activity is widely regarded as a decrease of arousal [38,39]). Nevertheless a second explanation is conceivable. With regard to the metabolic values presented in Fig. 1, it seems possible, (despite missing statistical significance) that the obtained cortical activation patterns are specific to maximal exercise intensity. Further studies should try to distinguish between different exercise intensities and their impact on cortical activation patterns, especially in a maximal to submaximal range. Furthermore, as final exhaustion and dropout after arm crank exercise were caused by local muscle fatigue rather than cardio-respiratory parameters, it might be interesting to apply the provided ideas and techniques on the model of central vs. peripheral fatigue. Finally it seems noteworthy that immediately after exercise (Fig. 2) all three kinds of exercise led to a significant increase of β activity in BA 7, which is believed to play a role in motor coordination [40,41]. As this increase is strongly connected to a limited time frame (b15 min) one could speculate whether this increase is a post effect of motion itself, i.e. if similar activation patterns could be described during exercise. Acknowledgments We would like to thank our participants for spending some of their valuable study time during the winter semester 2007 assisting us. Thanks to Petra Wollseiffen, Moritz Fölger and Matthias Lohne for their help with data collection. A special thank goes to Dr. Jayne Lucke for proof reading the first version of this manuscript. This study was made possible by a grant from the German Space Agency (DLR) 50WB0519 and a young investigators grant awarded to S. Schneider by the German Sport University. References [1] Karacabey K. Effect of regular exercise on health and disease. Neuro Endocrinol Lett 2005;26:617–23. [2] Martinsen EW. Physical activity in the prevention and treatment of anxiety and depression. Nord J Psychiatry 2008;62(Suppl 47):25–9. [3] Anish EJ. Exercise and its effects on the central nervous system. Curr Sports Med Rep 2005;4:18–23. [4] Crabbe JB, Dishman RK. Brain electrocortical activity during and after exercise: a quantitative synthesis. Psychophysiology 2004;41:563–74. [5] Mechau D, Mucke S, Weiss M, Liesen H. Effect of increasing running velocity on electroencephalogram in a field test. Eur J Appl Physiol Occup Physiol 1998;78:340–5. [6] Beaussart M, Niquet G, Gaudier E, Guislain F. [The EEG of boxers examined immediately after combat. Comparative study with the EEG recorded before combat in 52 cases.]. Rev Obstet Ginecol Venez 1959;101:422–7. [7] Petruzzello SJ, Landers DM, Hatfield BD, Kubitz KA, Salazar W. A meta-analysis on the anxiety-reducing effects of acute and chronic exercise. Outcomes and mechanisms. Sports Med 1991;11:143–82. [8] Shibata M, Shimura M, Shibata S, Wakamura T, Moritani T. Determination of the optimal walking speed for neural relaxation in healthy elderly women using electromyogram and electroencephalogram analyses. Eur J Appl Physiol Occup Physiol 1997;75:206–11.
[9] Hall EE, Ekkekakis P, Van Landuyt LM, Petruzzello SJ. Resting frontal asymmetry predicts self-selected walking speed but not affective responses to a short walk. Res Q Exerc Sport 2000;71:74–9. [10] Petruzzello SJ, Landers DM. State anxiety reduction and exercise: does hemispheric activation reflect such changes? Med Sci Sports Exerc 1994;26:1028–35. [11] Woo M, Kim S, Kim J, Petruzzello SJ, Hatfield BD. Examining the exercise-affect dose–response relationship: does duration influence frontal EEG asymmetry? Int J Psychophysiol 2009 May;72(2):166–72. (Electronic publication ahead of print 2008 Dec 10). [12] Crabbe JB, Smith JC, Dishman RK. Emotional & electroencephalographic responses during affective picture viewing after exercise. Physiol Behav 2007;90:394–404. [13] Schneider S, Askew CD. Localisation of exercise induced changes in brain cortical activity using a distributed source localization algorithm. Med Sci Sports Exerc 2009;41(5):6. [14] Pascual-Marqui RD. Standardized low-resolution brain electromagnetic tomography (sLORETA): technical details. Methods Find Exp Clin Pharmacol 2002;24 (Suppl D):5–12. [15] Pascual-Marqui RD, Esslen M, Kochi K, Lehmann D. Functional imaging with lowresolution brain electromagnetic tomography (LORETA): a review. Methods Find Exp Clin Pharmacol 2002;24(Suppl C):91–5. [16] Horacek J, Brunovsky M, Novak T, Skrdlantova L, Klirova M, Bubenikova-Valesova V, et al. Effect of low-frequency rTMS on electromagnetic tomography (LORETA) and regional brain metabolism (PET) in schizophrenia patients with auditory hallucinations. Neuropsychobiology 2007;55:132–42. [17] Saletu M, Anderer P, Semlitsch HV, Saletu-Zyhlarz GM, Mandl M, Zeitlhofer J, et al. Low-resolution brain electromagnetic tomography (LORETA) identifies brain regions linked to psychometric performance under modafinil in narcolepsy. Psychiatry Res 2007;154:69–84. [18] Schneider S, Brummer V, Carnahan H, Dubrowski A, Askew CD, Struder HK. What happens to the brain in weightlessness? A first approach by EEG tomography. Neuroimage 2008;42:1316–23. [19] Talairach J, Tournoux P. Co-Planar Stereotaxic Atlas of the Human Brain. Stuttgart: Thieme; 1988. [20] Grech R, Cassar T, Muscat J, Camilleri KP, Fabri SG, Zervakis M, et al. Review on solving the inverse problem in EEG source analysis. J Neuroeng Rehabil 2008;5:25. [21] Schneider S, Askew CD, Diehl J, Mierau A, Kleinert J, Abel T, et al. EEG activity and mood in health orientated runners after different exercise intensities. Physiol Behav 2009 Mar 23;96(4–5):709–16. [22] Boutcher SH, Landers DM. The effects of vigorous exercise on anxiety, heart rate, and alpha activity of runners and nonrunners. Psychophysiology 1988;25:696–702. [23] Schneider S, Mierau A, Diehl J, Askew CD, Strüder HK. EEG activity and mood in health orientated runners after different exercise intensities. Physiol Behav 2009;96:706–16. [24] Schmid A. Rollstuhlergometrie. Dtsch Z Sportmed 2002;53:153–4. [25] Heck H, Mader A, Hess G, Mucke S, Muller R, Hollmann W. Justification of the 4-mmol/l lactate threshold. Int J Sports Med 1985;6:117–30. [26] Hollmann W, Hettinger T. Schattauer. Stuttgart: New York; 2000. [27] Jasper HH. The ten-twenty electrode system of the international federation. Electroencephalogr Clin Neurophysiol Suppl 1958;35:371–5. [28] Fuchs M, Kastner J, Wagner M, Hawes S, Ebersole JS. A standardized boundary element method volume conductor model. Clin Neurophysiol 2002;113:702–12. [29] 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;34:1600–11. [30] Martin JH. The collective electrical behavior of cortical neurons: the electroencephalogram and the mechanisms of epilepsy. In: Kandel JH, Schwartz TM Jessel, editors. Principles of Neuroscience. Vol. London: Prentice-Hall; 1991. p. 777–91. [31] Nichols TE, Holmes AP. Nonparametric permutation tests for functional neuroimaging: a primer with examples. Hum Brain Mapp 2002;15:1–25. [32] Smith PM, Price MJ, Doherty M. The influence of crank rate on peak oxygen consumption during arm crank ergometry. J Sports Sci 2001;19:955–60. [33] Faw B. Pre-frontal executive committee for perception, working memory, attention, long-term memory, motor control, and thinking: a tutorial review. Conscious Cogn 2003;12:83–139. [34] Davidson RJ, Schwartz GE, Saron C, Bennett J, Goleman DJ. Frontal versus parietal EEG asymmetry during positive and negative affect. Psychophysiology 1979;16:202–3. [35] Coan JA, Allen JJ. Frontal EEG asymmetry as a moderator and mediator of emotion. Biol Psychol 2004;67:7–49. [36] Ekkekakis P, Petruzzello SJ. Acute aerobic exercise and affect: current status, problems and prospects regarding dose–response. Sports Med 1999;28:337–74. [37] Struder HK, Weicker H. Physiology and pathophysiology of the serotonergic system and its implications on mental and physical performance. Part I. Int J Sports Med 2001;22:467–81. [38] Lindsley DB. Attention, consciousness, sleep and wakefulness. In: Magoun, editor. Handbook of Physiology. Vol. Washington: Am Physiological Society; 1960. p. 1553–93. [39] Bonnet MH, Arand DL. Impact of activity and arousal upon spectral EEG parameters. Physiol Behav 2001;74:291–8. [40] Hyvarinen J, Laakso M, Roine R, Leinonen L. Comparison of effects of pentobarbital and ethanol on the neuronal activity in the posterior parietal association cortex. Acta Physiol Scand 1979;107:219–25. [41] Hanakawa T, Immisch I, Toma K, Dimyan MA, Van Gelderen P, Hallett M. Functional properties of brain areas associated with motor execution and imagery. J Neurophysiol 2003;89:989–1002.
Contents lists available at ScienceDirect
Physiology & Behavior j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / p h b
Changes in brain cortical activity measured by EEG are related to individual exercise preferences Stefan Schneider a,⁎, Vera Brümmer a, Thomas Abel a, Christopher D. Askew b, Heiko K. Strüder a a b
Institute of Movement and Neurosciences, Dept. of Exercise Neuroscience, German Sport University Cologne, Germany Faculty of Science, Health and Education, University of the Sunshine Coast, Maroochydore, Queensland, Australia
a r t i c l e
i n f o
Article history: Received 7 May 2009 Received in revised form 29 June 2009 Accepted 20 July 2009 Keywords: EEG Physical exercise history Electrotomography Exercise preferences LORETA
a b s t r a c t Exercise is well known to result in changes of brain cortical activity measured by EEG. The aim of this study was (1) to localise exercise induced changes in brain cortical activity using a distributed source localisation algorithm and (2) to show that the effects of exercise are linked to participants' physical exercise preferences. Electrocortical activity (5 min) and metabolical parameters (heart rate, lactate, peak oxygen uptake) of eleven recreational runners were recorded before and after incremental treadmill, arm crank and bicycle ergometry. Electroencephalographic activity was localised using standardised low resolution brain electromagnetic tomography (sLORETA). Results revealed an increase in frontal α activity immediately post exercise whereas increases after bike exercise were found to be localised in parietal regions. All three kinds of exercise resulted in an increase of β activity in Brodmann area 7. Fifteen and thirty minutes post exercise a specific activation pattern (decrease in frontal brain activity–increase in occipital regions) was noticeable for treadmill and bike but not arm crank exercise. We conclude that specific brain activation patterns are linked to different kinds of exercise and participants' physical exercise preferences. © 2009 Elsevier Inc. All rights reserved.
1. Introduction In the last twenty years the promotion of exercise has become an important public health message. This can be attributed primarily to the impact of exercise on cardio-respiratory and metabolic parameters and their influence on physical health [1]. But apart from those peripheral changes recent evidence suggests that there is also a major influence of exercise on brain cortical function which seems to be connected to changes in cognitive performance, general arousal and well being [2,3]. It is generally agreed that there are temporary changes in the EEGs α and β activities caused by exercise [4,5]. These changes were first reported in the 1950s [6] and are considered to reflect a state of accumulated relaxation [4,7,8]. An important aspect when considering effects of exercise on brain cortical activity is to assign effects to specific brain regions. Recent publications have contributed to the hypothesis that especially α activity within frontal brain regions is affected by exercise [9–13]. Although promising a fundamental limitation of these approaches is
⁎ Corresponding author. Institute for Movement and Neurosciences, German Sport University Cologne, Carl-Diem Weg 6, 50933 Köln, Germany. Tel.: +49 173 7078760; fax: +49 221 4973454. E-mail address: [email protected] (S. Schneider). 0031-9384/$ – see front matter © 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.physbeh.2009.07.010
linked to missing possibilities of brain imaging during and immediately after exercise. In the recent years standardised low resolution brain electromagnetic tomography (sLORETA) has become an accepted method to localise changes in brain cortical activity by analysing EEG derivations [14–17] especially in experimental situations that do not allow to use standard brain imaging methods like magnetic resonance imaging (MRI) or positron emission tomography (PET) [13,18]. sLORETA is a source localisation method relying on mathematical models of the bio-electrical generators and the volume conductors within which they lie. It is based on standardised EEG recordings, which are modeled to a probabilistic head model provided by the Montreal Neurological Institute (MNI). Active cortical regions are identified created by allocating the raw sLORETA values of individual voxels to their corresponding Brodmann areas (BA) or cerebral gyri on the basis of the coordinates of the digitized Talairach Atlas [19]. Apart from sLORETA there are similar approaches to overcome the inverse solution like BESA, ST-MAP, MUSIC and others. By comparing these methods it gets clear that each has its own specific advantages but sLORETA was shown to give the most satisfactory results [20]. Just recently we were able to show that major changes in brain cortical activity occurred when individuals were allowed to exercise at their preferred intensity rather than constricting their intensity by heart-rate limits [21]. This is in accordance with previous findings, linking relaxational effects of exercise to individuals' physical activity
448
S. Schneider et al. / Physiology & Behavior 98 (2009) 447–452
history and exercise preferences [8,22]. Accordingly we hypothesized that changes in brain cortical activity are connected not only to preferred exercise intensities but also to participants' exercise history. Within this study we tested a group of regular runners with little or no preferences to bike or arm crank exercise (i.e. none of these participants ever used an arm crank before, none of them was a regular biker) on changes in brain cortical activity after treadmill, bike and arm crank ergometry. We hypothesized that previously reported changes in EEG α and β activities, which were linked to recreational processes (changes in frontal α activity [12] and overall β activity [23]) would be dominant after treadmill but not arm crank and bike exercise. 2. Materials and methods 2.1. Participants and procedure The Institutional Review Committee of the German Sport University approved this study. All procedures were in compliance with the Declaration of Helsinki for human subjects. Twelve adults, either students or academic staff of the German Sport University, aged 26.3 ± 3.8 (male (n = 8): 25.6 ± 4.2; female (n = 4): 27.5 ± 2.89) participated in this study. All participants were regular runners performing a minimum of 2 h weekly training mainly for health reasons. Some of the participants also reported to cycle regularly but only to travel within the city and without any exercise background. None of the participants had previous experience with arm crank ergometry or similar (e.g. handbiking). All participants received a pre exercise medical screening. Incremental peak oxygen uptake (VO2peak) tests were performed either on an arm crank ergometer (Cyklus2 — Richter, D), a bicycle ergometer (Ergoline ER 900 — Ergoline, Bitz, Germany) or a treadmill (Woodway ELG 55 — Woodway, Weil am Rhein, Germany). Each participant performed one VO2peak test a day and testing days were separated by at least three days. In order to prevent the effects of sequence participants were randomly assigned into one of the three groups either (1) arm crank–bike–treadmill, (2) bike–treadmill–arm crank or (3) treadmill– bike–arm crank. Arm crank performance started at 20 W and was increased every 3 min by 20 W [24]. Bike ergometer performance started at 50 W and was increased by 50 W every 3 min [25]. The incremental treadmill test started at 2.0 m/s and was increased every 3 min by 0.5 m/s [26]. All tests ended at volitional exhaustion (RPE scale 19–20, metabolic respiratory quotient ≥ 1). At the end of each stage 20 μl capillary blood from the ear lobe was collected for blood lactate (LAC) analysis (Biosen C_Line Glukose-/Laktat-Messsystem — EKF-diagnostic, Barleben, Germany). Rating of perceived exertion (RPE) was assessed at the end of each stage (Borg-scale) and heart rate (HR) was recorded at 5 s intervals (Polar S810i — Polar Electro, Kimpele, Finland). Expired gasses were continuously collected and analysed at 30 s intervals (ZAN 600 ErgoTest — ZAN, Oberthulba, Germany) for the determination of oxygen uptake, carbon dioxide production and ventilation. Peak oxygen uptake during incremental stage test (VO2peak) was taken as the highest 30 s value throughout the test. EEG activity was recorded for 5 min in a supine rest position with eyes closed prior to each test (PRE), 2 min after exercise (POST2) and again 15 (POST15) and 30 min (POST30) after exercise. To have similar conditions to PRE measurement subjects were asked to sit up and stay seated in between measurements. As exercise intensity was at a maximum level, we chose a supine position to avoid blackouts (unpublished observation) as EEG was recorded immediately after exercise. Moreover we were worried that an acute shift of body fluids to the lower body parts in a seated position after exercise might influence brain cortical activity [18]. 2.2. EEG recording A 64-channel portable EEG-System (IT-Med, Usingen, D) was used for data acquisition. Thirty minutes prior to exercise an EEG-Cap was mounted (Electro-Cap International, Inc., USA). This was adapted to
individual head size and built of 19 electrodes and one reference electrode (mounted in the triangle of FP1, FP2 and FZ) in the 10–20 system [27]. It recorded EEG activity on positions Fp1, Fp2, F3, F4, F7, F8, Fz, C3, C4, Cz, P3, P4, P7, P8, Pz, T7, T8, O1, and O2. The cap was fixed with two strings to a belt around the chest to prevent shifting during the exercise trials. The cap was permeable to air in order to prevent an increase in heat during exercise. Distances between electrodes were approximately 5 cm to prevent possible cross talk after exercise due to salt bridges between electrodes. Each electrode was filled with Electro-Gel™ (Electro-Cap International, Inc., Eaton, USA) for signal transduction. If impedance of an electrode exceeded 10 kΩ this single electrode was excluded from further analysis. The analogue signal of the EEG was amplified and converted to digital signals using Braintronics Iso 1064 CE Box (Braintronics B. V., Hl Almere, Netherlands; hardware filters: high pass — fixed time constant of 1 s; low pass — 97 Hz ± 15% (−3dB)) and stored with a sampling rate of 256 Hz on a hard disk of Neurofile XP EEG-System (IT-Med, Usingen, Germany). 2.3. Data analysis — EEG/LORETA sLORETA enables the spatial identification and analysis of brain cortical activity via traditional EEG recordings [14,15,28,29]. Scalp recorded cranial EEG activity is evoked by synchronised postsynaptic potentials (PSPs), which are located in clusters of pyramidal cells [30]. sLORETA makes it possible to determine the three dimensional orientation of these highly synchronised PSPs. sLORETA software is based on a probabilistic MNI brain volume which was scanned at 5 mm resolution. The MNI coordinates were converted to “corrected” Talairach coordinates, then given to the Talairach Daemon. Voxels that were unambiguously labeled as cortical grey matter, and that fell unambiguously within the brain compartment, were retained. This produced 6239 cortical grey matter voxels at 5 mm resolution [14]. Coordinates given within this study will refer to the MNI152 template. Cortical regions are created by allocating the raw sLORETA values of individual voxels to their corresponding Brodmann areas (BA) or cerebral gyri on the basis of the coordinates of the digitized Talairach Atlas [19]. EEG data were analysed by Brain Vision Analyzer (Brain Products, München, Germany). After manual artefact detection, high and lowpass filters were applied so that a frequency range from 0.5 to 49 Hz remained for analysis (time constant 0.3183 s; 24 dB/octave). Data were then segmented into 4 s sections where an overlap of 10% was accepted. An allowance of automatic artefact rejection was set to: gradient b 50 μV. Additional data were checked for eye movement artefacts manually and by ICA analysis using the semiautomatic ICA application in Analyzer2 (Brain Products, München, Germany). This resulted in a minimum of seventy 4 s segments of artifact-free data. In sLORETA software the coordinates of the 19 electrode positions were applied to a digitized MRI version of the Talairach Atlas (McConnell Brain Imaging Centre, Montréal Neurological Institute, McGill University). These Talairach coordinates were then used to compute the sLORETA transformation matrix. Following transformation to average reference EEG activity the seventy 4 s segments of artifact-free resting EEG were averaged to calculate cross spectra in sLORETA for α (7.5– 12.5 Hz) and β (12.5–35 Hz) bands for each subject and each of the four measurements. Using the sLORETA transformation matrix, cross spectra of each subject and for each frequency band were then transformed to sLORETA files. This resulted in the corresponding 3D cortical distribution of the electrical neuronal generators for each subject. To display differences in EEG spectral activity between PRE and POST2/POST15/POST30 measurements the following tests were performed: POST2 vs. PRE, POST15 vs. PRE and POST30 vs. PRE. Paired sampled t-test was computed for sLORETA power at each voxel. Statistical significance was assessed using a nonparametric
S. Schneider et al. / Physiology & Behavior 98 (2009) 447–452
randomisation test [31]. To correct for multiple comparisons, a nonparametric single-threshold test was assessed defining a critical threshold (tcritical). Voxels with statistic values exceeding this threshold have their null hypothesis, i.e. no difference in EEG power between the two tests (e.g. POST2 vs. PRE), rejected. The omnibus hypothesis (that all the voxel hypotheses are true) is rejected if a voxel value exceeds the critical threshold for p b .05 defined by 5000 randomisations. Voxel-by-voxel t-values in Talairach space are displayed as statistical parametric maps (SPMs). 2.4. Statistics — physiological values Changes in HR, LAC and VO2-uptake were analysed using the factors of exercise (treadmill, arm crank, bike) and time (baseline value, value at maximal exhaustion). Exercise time was compared using repeated measures ANOVA with the factor exercise (treadmill, arm crank, bike). Where ANOVA revealed significance, Fisher LSD was used for post-hoc analysis in all cases. Data were analysed using Statistica 7.1 (StatSoft, Tulsa, USA).
449
3.2. sLORETA pre vs. post Pre vs. post results showed an increase of activity across the whole scalp in the α and β frequency ranges (Fig. 2). This was found to be significant (tcritical for p b .05 = 3.95⁎; p b .01 = 5.11⁎⁎) after treadmill exercise in BA 6, 8, 9 (frontal lobe) and BA 24, 32 (limbic lobe) for α activity (p b .05, t = 4.76⁎) and in BA 7 (parietal lobe) for β activity (p b .01, t = 5.98⁎⁎). A widespread increase of activity, which reached significance (tcritical for p b .05 = 3.52⁎; p b .01 = 4.19⁎⁎) in BA 7 (parietal lobe) and 23, 31 (limbic lobe) within the α (p b .01, t = 4.83⁎⁎) and β band (p b .01, t = 4.91⁎⁎) could be observed after bike exercise (Fig. 2). Immediately after arm crank exercise significant changes (tcritical for p b .05 = 3.16⁎; p b .01 = 3.92⁎⁎) could be obtained for α activity (p b .05, t = 3.17⁎) in one voxel only (MNI = −45/20/15, BA 45) whereas β activity was significantly increased (p b .01, t = 3.93) in BA 7 and 40 (parietal lobe) (Fig. 2).
3.3. sLORETA pre vs. post15 and post30 3. Results 3.1. Physiological values during incremental exercises Mean exercise time showed no differences between the three different kinds of ergometry (treadmill 15.42 ± 2.81 min; arm crank 15.42±4.08 min; bike 15.00±3.19 min; F(2,22) =0.14; p=0.87; Fig. 1A). LAC values increased after exercise (pb 0.001). No significant differences were found in LAC values after arm crank, bike and treadmill performance (F(2, 20) =0.99; p=0.39; Fig. 1C). Heart rate was increased at maximal exhaustion (p b 0.001). Differences in maximal heart rate (HRmax; F(2, 20) = 5.55; p b 0.05) were obtained between treadmill and arm crank as well as bike and arm crank exercises (p b 0.001). Differences in HRmax between bike and treadmill exercises did not reach significance (p = 0.21; Fig. 1B). Participants showed a higher VO2peak (F(2, 20) = 42,49; p b 0.001) after treadmill and bike exercises compared to arm crank exercise (p b 0.001; Fig. 1D). A slightly lower value which marginally missed significance (p = 0.08) was found after bike exercise compared to treadmill exercise. RPE values were at a maximum level (19 to 20) after each kind of exercise.
No significant effects were found for changes in α (post15 t = 4.03; post30 t = 4.00) and β activities (post15 t = 3.69; post30 t = 2.96) after treadmill exercise (post15 tcritical for p b .05 = 4.16; post30 tcritical for p b .05 = 4.44). α activity was found to be significantly increased (post15 t = 3.78⁎, p b 0.05; post30 t = 3.95) after bike exercise in BA 6, 9 (frontal lobe), 24 and 32 (limbic lobe, Fig. 3). No significant changes could be obtained for β activity (post15 t = 3.63; post30 t = 2.77) after bike exercise (post15 tcritical for p b .05 = 3.64; post30 tcritical for p b .05 = 4.15). Changes in α (p b .05; t = 3.94⁎) and β activities (p b .05; t = 3.63⁎) 15 min after arm crank exercise were found significantly increased mainly in the left and right temporal lobes. β activity 30 min post exercise was increased in BA30/31 (limbic lobe; p b .05; t = 4.24) (post15 tcritical for p b .05 = 3.45; post30 tcritical for p b .05 = 3.61). Specific activation patterns could be noticed after treadmill and bike exercises compared to arm crank exercise (Figs. 2 and 3). Whereas an increase of α activity could be noticed in parietal and occipital regions across all three kinds of exercise, a decrease of β
Fig. 1. Illustration of exercise time (A), heart rate (B), blood lactate (C) and VO2peak (D) across all three kinds of exercise (treadmill, arm crank, bike). Rest values and values at volitional exhaustion are shown. Bars show means and .95 confidence intervals. ⁎⁎⁎p b .001.
450
S. Schneider et al. / Physiology & Behavior 98 (2009) 447–452
Fig. 2. Statistical parametric maps (SPM) of sLORETA differences in the α and β frequency bands comparing POST2 and PRE (n = 12). Red and yellow colours indicate increased activity in the POST2 measurement. Orthogonal views of the cortex (unthresholded) are displayed. Images depicting SPMs seen from different perspectives are based on voxel-byvoxel t-values of differences. Structural anatomy is shown in grey scale (L left, R right, A anterior, P posterior). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
activity in frontal brain regions was noticeable after treadmill and bike exercises only.
the scalp. 15 and 30 min post exercise a specific activation/ deactivation pattern in the β frequency range could be recognized for bike and treadmill exercises.
4. Discussion 4.1. Physiological parameters This study examined changes in brain cortical activity after different forms of exercise performed at maximal intensity. We used three different kinds of ergometry (treadmill, bike, arm crank) and recorded EEG activity prior to and up to half an hour post ergometry. A localisation analysis using sLORETA showed an increase of α and β activities immediately post exercise, which was spread widely across
To our knowledge no comparison between treadmill, arm crank and bike ergometry has been conducted so far using the same set of participants. No changes were observed between exercise duration and LAC concentration between the three kinds of ergometry, but HRmax and VO2peak were clearly lower after arm crank ergometry (Fig. 1BD). This
S. Schneider et al. / Physiology & Behavior 98 (2009) 447–452
combination of reduced HRmax and VO2peak let us suggest that final exhaustion and dropout after arm crank exercise are caused by local muscle fatigue rather than cardio-respiratory parameters [32]. This is supported by participant reports of cramped forearm muscles after arm crank exercise. 4.2. EEG data The overall increase in α and β activities observed immediately after exercise is in accordance with prior results. The first findings from Beaussart et al. [6] showed an increase of around 20% in α activity after
451
exercise at a mean maximum heart rate of 200. Subsequent studies [5,22] confirmed that the increase in α activity was strongest during the first 5 to 10 min after exercise and returned to pre exercise levels by about the 30th minute of recovery. Further sLORETA analysis revealed an increase of activity in α power, which was localised in frontal brain areas after treadmill exercise whereas α power after bike exercise was found to be increased in parietal areas and no significant changes could be obtained after arm crank exercise (Fig. 2). Given the fact that participants in this study reported to have a clear preference for running, an increase in frontal α activity (please remember that α activity is inversely related to cerebral
Fig. 3. Statistical parametric maps (SPM) of sLORETA differences in the β frequency band comparing POST15 vs. PRE (left) and POST30 vs. PRE (right, n = 12). Red and yellow colours indicate increased activity, blue colours decreased activity in the POST15 and POST 30 measurement. Orthogonal views of the cortex (unthresholded) are displayed. Images depicting SPMs seen from different perspectives are based on voxel-by-voxel t-values of differences. Structural anatomy is shown in grey scale (L left, R right, A anterior, P posterior). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
452
S. Schneider et al. / Physiology & Behavior 98 (2009) 447–452
activity) let us assume that the reported effects of treadmill exercise are linked to individuals' physical activity preferences as initially hypothesized. Special attention should be paid to the fact that changes were localised within frontal regions, which are well known to be involved in emotional processing [33–35]. As extensive research within the last years has shown that (1) exercise is linked to changes in mood and general well being [23,36,37] and (2) exercise seems to be connected to changes in frontal α activity [9–13], it could be assumed that the results observed here propose a connection between exercise preferences, frontal lobe activity and mood. Unfortunately we did not check for changes in mood. This should be taken into consideration by future research activities. An influence of individuals' physical activity preferences is further more supported by specific changes within the β frequency range 15 and 30 min post exercise (Fig. 3) which shows a reduction of activity in frontal brain areas after treadmill and bike but not arm crank exercise, which was totally unfamiliar to our participants (please be aware that a reduction of β activity is widely regarded as a decrease of arousal [38,39]). Nevertheless a second explanation is conceivable. With regard to the metabolic values presented in Fig. 1, it seems possible, (despite missing statistical significance) that the obtained cortical activation patterns are specific to maximal exercise intensity. Further studies should try to distinguish between different exercise intensities and their impact on cortical activation patterns, especially in a maximal to submaximal range. Furthermore, as final exhaustion and dropout after arm crank exercise were caused by local muscle fatigue rather than cardio-respiratory parameters, it might be interesting to apply the provided ideas and techniques on the model of central vs. peripheral fatigue. Finally it seems noteworthy that immediately after exercise (Fig. 2) all three kinds of exercise led to a significant increase of β activity in BA 7, which is believed to play a role in motor coordination [40,41]. As this increase is strongly connected to a limited time frame (b15 min) one could speculate whether this increase is a post effect of motion itself, i.e. if similar activation patterns could be described during exercise. Acknowledgments We would like to thank our participants for spending some of their valuable study time during the winter semester 2007 assisting us. Thanks to Petra Wollseiffen, Moritz Fölger and Matthias Lohne for their help with data collection. A special thank goes to Dr. Jayne Lucke for proof reading the first version of this manuscript. This study was made possible by a grant from the German Space Agency (DLR) 50WB0519 and a young investigators grant awarded to S. Schneider by the German Sport University. References [1] Karacabey K. Effect of regular exercise on health and disease. Neuro Endocrinol Lett 2005;26:617–23. [2] Martinsen EW. Physical activity in the prevention and treatment of anxiety and depression. Nord J Psychiatry 2008;62(Suppl 47):25–9. [3] Anish EJ. Exercise and its effects on the central nervous system. Curr Sports Med Rep 2005;4:18–23. [4] Crabbe JB, Dishman RK. Brain electrocortical activity during and after exercise: a quantitative synthesis. Psychophysiology 2004;41:563–74. [5] Mechau D, Mucke S, Weiss M, Liesen H. Effect of increasing running velocity on electroencephalogram in a field test. Eur J Appl Physiol Occup Physiol 1998;78:340–5. [6] Beaussart M, Niquet G, Gaudier E, Guislain F. [The EEG of boxers examined immediately after combat. Comparative study with the EEG recorded before combat in 52 cases.]. Rev Obstet Ginecol Venez 1959;101:422–7. [7] Petruzzello SJ, Landers DM, Hatfield BD, Kubitz KA, Salazar W. A meta-analysis on the anxiety-reducing effects of acute and chronic exercise. Outcomes and mechanisms. Sports Med 1991;11:143–82. [8] Shibata M, Shimura M, Shibata S, Wakamura T, Moritani T. Determination of the optimal walking speed for neural relaxation in healthy elderly women using electromyogram and electroencephalogram analyses. Eur J Appl Physiol Occup Physiol 1997;75:206–11.
[9] Hall EE, Ekkekakis P, Van Landuyt LM, Petruzzello SJ. Resting frontal asymmetry predicts self-selected walking speed but not affective responses to a short walk. Res Q Exerc Sport 2000;71:74–9. [10] Petruzzello SJ, Landers DM. State anxiety reduction and exercise: does hemispheric activation reflect such changes? Med Sci Sports Exerc 1994;26:1028–35. [11] Woo M, Kim S, Kim J, Petruzzello SJ, Hatfield BD. Examining the exercise-affect dose–response relationship: does duration influence frontal EEG asymmetry? Int J Psychophysiol 2009 May;72(2):166–72. (Electronic publication ahead of print 2008 Dec 10). [12] Crabbe JB, Smith JC, Dishman RK. Emotional & electroencephalographic responses during affective picture viewing after exercise. Physiol Behav 2007;90:394–404. [13] Schneider S, Askew CD. Localisation of exercise induced changes in brain cortical activity using a distributed source localization algorithm. Med Sci Sports Exerc 2009;41(5):6. [14] Pascual-Marqui RD. Standardized low-resolution brain electromagnetic tomography (sLORETA): technical details. Methods Find Exp Clin Pharmacol 2002;24 (Suppl D):5–12. [15] Pascual-Marqui RD, Esslen M, Kochi K, Lehmann D. Functional imaging with lowresolution brain electromagnetic tomography (LORETA): a review. Methods Find Exp Clin Pharmacol 2002;24(Suppl C):91–5. [16] Horacek J, Brunovsky M, Novak T, Skrdlantova L, Klirova M, Bubenikova-Valesova V, et al. Effect of low-frequency rTMS on electromagnetic tomography (LORETA) and regional brain metabolism (PET) in schizophrenia patients with auditory hallucinations. Neuropsychobiology 2007;55:132–42. [17] Saletu M, Anderer P, Semlitsch HV, Saletu-Zyhlarz GM, Mandl M, Zeitlhofer J, et al. Low-resolution brain electromagnetic tomography (LORETA) identifies brain regions linked to psychometric performance under modafinil in narcolepsy. Psychiatry Res 2007;154:69–84. [18] Schneider S, Brummer V, Carnahan H, Dubrowski A, Askew CD, Struder HK. What happens to the brain in weightlessness? A first approach by EEG tomography. Neuroimage 2008;42:1316–23. [19] Talairach J, Tournoux P. Co-Planar Stereotaxic Atlas of the Human Brain. Stuttgart: Thieme; 1988. [20] Grech R, Cassar T, Muscat J, Camilleri KP, Fabri SG, Zervakis M, et al. Review on solving the inverse problem in EEG source analysis. J Neuroeng Rehabil 2008;5:25. [21] Schneider S, Askew CD, Diehl J, Mierau A, Kleinert J, Abel T, et al. EEG activity and mood in health orientated runners after different exercise intensities. Physiol Behav 2009 Mar 23;96(4–5):709–16. [22] Boutcher SH, Landers DM. The effects of vigorous exercise on anxiety, heart rate, and alpha activity of runners and nonrunners. Psychophysiology 1988;25:696–702. [23] Schneider S, Mierau A, Diehl J, Askew CD, Strüder HK. EEG activity and mood in health orientated runners after different exercise intensities. Physiol Behav 2009;96:706–16. [24] Schmid A. Rollstuhlergometrie. Dtsch Z Sportmed 2002;53:153–4. [25] Heck H, Mader A, Hess G, Mucke S, Muller R, Hollmann W. Justification of the 4-mmol/l lactate threshold. Int J Sports Med 1985;6:117–30. [26] Hollmann W, Hettinger T. Schattauer. Stuttgart: New York; 2000. [27] Jasper HH. The ten-twenty electrode system of the international federation. Electroencephalogr Clin Neurophysiol Suppl 1958;35:371–5. [28] Fuchs M, Kastner J, Wagner M, Hawes S, Ebersole JS. A standardized boundary element method volume conductor model. Clin Neurophysiol 2002;113:702–12. [29] 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;34:1600–11. [30] Martin JH. The collective electrical behavior of cortical neurons: the electroencephalogram and the mechanisms of epilepsy. In: Kandel JH, Schwartz TM Jessel, editors. Principles of Neuroscience. Vol. London: Prentice-Hall; 1991. p. 777–91. [31] Nichols TE, Holmes AP. Nonparametric permutation tests for functional neuroimaging: a primer with examples. Hum Brain Mapp 2002;15:1–25. [32] Smith PM, Price MJ, Doherty M. The influence of crank rate on peak oxygen consumption during arm crank ergometry. J Sports Sci 2001;19:955–60. [33] Faw B. Pre-frontal executive committee for perception, working memory, attention, long-term memory, motor control, and thinking: a tutorial review. Conscious Cogn 2003;12:83–139. [34] Davidson RJ, Schwartz GE, Saron C, Bennett J, Goleman DJ. Frontal versus parietal EEG asymmetry during positive and negative affect. Psychophysiology 1979;16:202–3. [35] Coan JA, Allen JJ. Frontal EEG asymmetry as a moderator and mediator of emotion. Biol Psychol 2004;67:7–49. [36] Ekkekakis P, Petruzzello SJ. Acute aerobic exercise and affect: current status, problems and prospects regarding dose–response. Sports Med 1999;28:337–74. [37] Struder HK, Weicker H. Physiology and pathophysiology of the serotonergic system and its implications on mental and physical performance. Part I. Int J Sports Med 2001;22:467–81. [38] Lindsley DB. Attention, consciousness, sleep and wakefulness. In: Magoun, editor. Handbook of Physiology. Vol. Washington: Am Physiological Society; 1960. p. 1553–93. [39] Bonnet MH, Arand DL. Impact of activity and arousal upon spectral EEG parameters. Physiol Behav 2001;74:291–8. [40] Hyvarinen J, Laakso M, Roine R, Leinonen L. Comparison of effects of pentobarbital and ethanol on the neuronal activity in the posterior parietal association cortex. Acta Physiol Scand 1979;107:219–25. [41] Hanakawa T, Immisch I, Toma K, Dimyan MA, Van Gelderen P, Hallett M. Functional properties of brain areas associated with motor execution and imagery. J Neurophysiol 2003;89:989–1002.