Can biofeedback training of psychophysiological

http://informahealthcare.com/psm ISSN: 0091-3847 (print) Phys Sportsmed, 2015; Early Online: 1–13 DOI: 10.1080/00913847.2015.1069169

CLINICAL FEATURE ORIGINAL RESEARCH

Can biofeedback training of psychophysiological responses enhance athletes’ sport performance? A practitioner’s perspective

The Physician and Sportsmedicine Downloaded from informahealthcare.com by 178.58.159.90 on 07/23/15 For personal use only.

Nika Pusenjak1, Anton Grad2, Matej Tusak3, Matevz Leskovsek4 and Romina Schwarzlin5 1

Research Group Larus Inventa, Ljubljana University Incubator, Vojkova 63, Ljubljana, Slovenia, 2University of Primorska Faculty of Mathematics, Natural Sciences and Information Technologies - Department of Psychology, Koper, Slovenia, 3University of Ljubljana, Sport Faculty - Department of Sport Psychology Ljubljana, Slovenia, 4Zdrav Dih d.o.o., Ljubljana, Slovenia, and 5Institute Jozef Stefan, Ljubljana, Slovenia

Abstract

Keywords:

Background: In recent years, biofeedback has become increasingly popular for its proven success in peak performance training – the psychophysiological preparation of athletes for high-stakes sport competitions, such as the Olympic games. The aim of this research was to test whether an 8-week period of exposure to biofeedback training could improve the psychophysiological control over competitive anxiety and enhance athletic performance in participating subjects. Methods: Participants of this study were highly competent athletes, each training in different sport disciplines. The experimental group consisted of 18 athletes (4 women, 14 men), whereas the Control group had 21 athletes (4 women, 17 men). All athletes were between 16 and 34 years old. The biofeedback device, Nexus 10, was used to detect and measure the psychophysiological responses of athletes. Athletes from both groups (control and experimental) were subjected to stress tests at the beginning of the study and once again at its conclusion. In between, the experimental group received training in biofeedback techniques. We then calculated the overall percentage of athletes in the experimental group compared with those in the control group who were able to control respiration, skin conductance, heart rate, blood flow amplitude, heart rate variability, and heart respiration coherence. One year following completion of the initial study, we questioned athletes from the experimental group, to determine whether they continued to use these skills and if they could detect any subsequent enhancement in their athletic performance. Results: We demonstrated that a greater number of participants in the experimental group were able to successfully control their psychophysiological parameters, in comparison to their peers in the control group. Significant results (p < 0.05) were noted in regulation of GSR following short stress test conditions (p = 0.037), in regulation of HR after exposure to STROOP stressor (p = 0.037), in regulation of GSR following the Math and GSR stressors (p = 0.033, p = 0.409) and in achieving HR – breathing coherence following the math stressor (p = 0.042). Conclusion: One year following completion of the training program, all participants from the experimental group indicated that they were still using the biofeedback – psycho-regulation skills. Furthermore, these participants uniformly reported believing that these skills had enhanced their athletic performance and general well-being.

Biofeedback training, competition, CSAI-2, SAS, athletes, pre-competitive anxiety, stress tests.

Introduction The biofeedback method is adopted from the applied psychophysiology science field. It is believed that, if implemented correctly, its techniques can prepare athletes to achieve a distinctive state known as psychophysiological coherence [1]. This state is connected to positive emotions and establishes synchronization in psychological as well as physiological processes. It is manifested through a coherent heart rhythm pattern and optimization of other physiological parameters. This specific heart rhythm is a consequence of a harmonization of sympathetic and parasympathetic branches of the autonomic nervous system, which has been shown to

History Received 27 March 2015 Accepted 1 July 2015 Published online 22 July 2015

lead to both enhanced cognitive abilities and executive performance [2]. Using a sophisticated biofeedback apparatus, the psychophysiological responses of athletes can be measured before, during and after an exercise, providing quantitative data necessary to assess performance. Different physiological monitoring sensors are attached to the body of an athlete (respiratory sensor, thermometer, blood volume pulse sensor, galvanic skin conductance sensor, electromyography, EEG, etc.), whereby the psychophysiological information yielded by the sensors enhances the biological signal in the biofeedback device. The athlete can receive feedback about his/her

Correspondence: Nika Pusenjak, Research Group Larus Inventa d.o.o., Ljubljana University Incubator, Vojkova 63, 1000 Ljubljana, Slovenia. E-mail: [email protected] 2015 Informa UK Ltd.


2

N. Pusenjak et al.

performance in a variety of forms. These include animations, graphs, video games, or simple audio feedback. As the athlete receives this information in real time, he/she can learn to alter different parameters accordingly, creating a biofeedback loop displayed in the biofeedback loop picture. Owing to rapid technological advancements, physiological monitoring sensors have become more sophisticated, allowing biofeedback training to be more enjoyable. Its motivational benefits have become more apparent and, as the program can be adjusted for use in non-laboratory settings, athletes are increasingly interested in participating in training programs that utilize these methods. The key benefit of biofeedback training is that the psychophysiological control results are provided to the athletes in real time, allowing them to observe their responses and to quickly learn to adjust their actions, thereby gaining more control over the autonomic nervous system. In other words, biofeedback enables the athletes to quantify as their progress can observe in real time in a form of graphs, numbers or animations at the screen. [3]. Biofeedback, psychophysiological self-regulation, anger management and precompetitive/competitive anxiety Biofeedback training helps athletes lower self-reported anxiety, relax their muscles and control their autonomic responses [4]. Wilson and Cummings [5] use the term ‘Learned Self Regulation (LSR)’ to describe a similar process, whereby an individual learns to actively influence his/her mind and body states, to decide when to engage emotionally or mentally, and to shift the attention span to where it is needed the most. LSR is a skill that can be learned and is thus linked to performance enhancement. The technique works on the principle that the athlete must first be aware of his/her problem in order to change his state of mind and physical response. Strack in 2003 examined the high school batting performance of baseball players. His study showed the improvement of biofeedback HRV group for 60% in mastering the batting skill then in control group also evident were results describe raise in low frequencies (%) in the heart rate spectrum. Once capable of controlling the psychophysiological responses, he/she can become more selfconfident. This confidence is the key to enhancing performance [6]. In addition, biofeedback training is beneficial in anger management, as was demonstrated in one case study of a young tennis player [7]. The key turning point in learning auto-regulation skills is known colloquially as the “Aha! moment,” when a trainee suddenly recognizes the connection between his/her thoughts and how thinking can modify the body’s physiological responses. In other words, as our physical movements are linked to our attitudes, real-time feedback allows the athlete to modify his/her psychophysiological state, learning to relax effortlessly and to continue to master other self-regulation techniques [8]. Psychological measurement of anxiety Anxiety is an apprehension of cognitively-perceived threat. It manifests in the physical symptoms of stress, and in

Phys Sportsmed, 2015; Early Online:1–13

psychological, cognitive, and emotional symptoms as well. In his book, Lazarus explains that stress is the imbalance between how a person perceives his/her ability to master the demands of the given situation and what the actual demands of the situation are [9]. Stress is harmful for an athlete, as high levels of unpleasant emotions interrupt his/her ability to achieve peak performance [10]. Athletes sometimes experience more anxiety playing at the international level than playing in a club, because they perceive the international competition as being more important. As a result, they experience pre-competitive anxiety and competitive anxiety. These states can produce negative self-talk, and consequently, negative psychophysiological reaction of the body. Collectively, these manifest as stress responses. Since its inception, biofeedback training has typically been performed within a laboratory setting [11]. Consequently, sport biofeedback practitioners have tended to work with athletes that practice sports, which require a controlled, lab-like environment: shooting, archery, and golf, for example [4]. Findings from several such studies indicate that biofeedback can help shooters enhance their performance and gain control over their stress responses. However, such techniques should be combined with other relaxation methods, such as progressive muscle relaxation, autogenic training and visualization [12]. For example, [13], biofeedback training that incorporates EMG, temperature and GSR sensors in combination with relaxation techniques could be very effective in enhancing athletic performance of gymnasts. Biofeedback was administered in a group meeting with 17 gymnasts. One athlete wore a GSR sensor on his fingers, whereas the others provoked him with negative suggestions. As they continued, the pitch of the sound from the device rose. This allowed them to identify which thoughts were negative for him. Identifying this helped them to stop disturbing feelings. They next tried to relax. After that, they visualized their routine and were aware of their arousal. With some practice, they learned to stop producing the sound from the biofeedback device. This meant that they had achieved psychophysiological desensitization. This in turn helped eliminate anxiety and the probability of mistakes during the performance [14]. In a more recent study, Strack [15] conducted experiments to assess whether heart variability biofeedback training (HRV) could enhance school baseball players’ sport performance. The results obtained revealed that the biofeedback trainee group achieved better focus ability, in addition to having other positive effects, such as better relaxation and improvements in performance. In a similar study, Bar-Eli et al. [16] examined whether there was a link between 14-week-long biofeedback training combined with mental training (Wingate five-step approach [17]) and the performance of 11–14 year- old swimmers. The results reported by the authors indicated an increase in the young athletes’ performance upon completion of the program. Galloway [18] investigated whether the Wingate five-step approach, in combination with biofeedback training, could enhance the serving accuracy in tennis. Although the study


DOI: 10.1080/00913847.2015.1069169

participants consisted of only six elite tennis players, the results proved the program’s efficacy, as the trainees’ serving accuracy as well as their mental skill adherence improved. Biofeedback methods have recently become very popular. In particular, it was proven to be successful in the psychophysiological preparation of Canadian athletes for the 2010 Vancouver Winter Olympic Games. The program was 3 years long for the National Short Track Speed skating team. Psychological skills training were integrated with biofeedback training to optimize self-regulation for performance on demand and under pressure. One of the programs the athletes participated in was ‘An Integrated Biofeedback and Psychological Skills Training Program for Canada’s Olympic Short-Track Speed skating Team.’ This is a seven phasemodel designed to improve athletic performance. It includes both psychological and biofeedback training. The success from participating athletes during the Games demonstrated that the approach was successful [19]. The first notable psychophysiological bio/neurofeedback training laboratory in professional soccer was the so-called Mind Room. In order to score game penalty kicks, athletes learned to stay calm and controlled with bio/neurofeedback training [20]. Dupee and Werthner conducted bio-neurofeedback training in order to successfully prepare 15 elite athletes for the Vancouver 2010 Olympics [21]. Pop-Jordanova and Demerdzieva also trained their athlete on Olympic games and used peripheral biofeedback and neurofeedback. Their clients successfully passed Olympic qualifications competition [22]. Along with the increased popularity of biofeedback peak performance, training was also necessary to test the efficacy of different training protocols designed to teach cognitive and emotional self-regulation and to improve athlete performance on sport-specific tasks. Perry and his colleagues performed this evaluation [23]. Many researchers suggested that training of heart rate variability biofeedback (HRV) could improve athletes coping with the competition stress. L. Lagos and colleagues performed a study using a virtual reality golf center to train golfers to do HRV biofeedback training. They recorded physiological measures, noting HRV and respiration rate in the laboratory between the first, fourth, seventh, and tenth weeks of the study. Observation showed: reduction in symptoms of anxiety, stress, and sensation seeking and increases in total HRV, low-frequency HRV, and amplitude of oscillation at 0.1 Hz, as well as improved sport performance. This effect became more evident during the 10 weeks of HRV BFB training [24]. Thomson and his colleagues concluded that HRV biofeedback training is effective and that HRV-derived sympathetic response is significantly tied to sport performance – in their case, in shooting performance [25]. Garg et al. researched the effect of heart rate variability biofeedback training (10 days, 20 min a day) on performance psychology of 30 college basketball players. The group observed lowered anxiety and correlation to better performance [26]. It is important to also mention Hanin’s model IZOF, which describes five basic dimensions describing emotional experiences and their characteristics, implemented in several psych

Can biofeedback enhance sport performance?

3

biosocial states related to performance. These include form (subjectively perceived), content (quality), intensity, time and context. These five descriptors are similar to widely used emotional components: implied form, valence and intensity. The IZOF model focuses on the effect of emotions on athletic performance, as every athlete has his own combination of emotional experiences resulting in different psychophysiological responses. These can be good or bad for athletic performance. Each athlete has his own zone of optimal performance that is connected to his or her individual anxiety level. Some athletes function better when anxiety is high, some when is low or moderate. In light of these findings, Hanin and colleagues proposed the IZOF model (HANIN, 1997, 2000). This is an ‘intraindividual framework that aims to describe, predict, explain, and control an athlete’s optimal and dysfunctional experiences related to individually successful and poor performances’ [27,28]. Tenenbaum et al. developed the Individual Affect Performance Zone (IAPZ). Here, feelings are described as discrete items (e.g., worry, calmness, confidence, etc.). Ordinal logistic regression was performed to estimate the probability of four feeling categories, based on positive-/negative-valence and functional/dysfunctional groupings [29]. Van der Lei H. and Tenenbaum (2012) used IAPZ in their study investigating performance processes within affectrelated performance zones of golfers. Results highlight the associations between the IAPZs. Decision-making or swing/ stroke executions were strong, and most importantly, unique for each golfer [30]. One of the models used in peak performance psychophysiological training is also the MAP model – multi-action plan. This model monitors the whole spectrum of psychophysiological and behavioural features related to the different types of performance. In one study, Bertollo et al. investigate the behavioral and psychophysiological correlations of the athletic performances of a 20-year-old Italian shooting team athlete and a 46-year-old Italian dart-throwing team athlete who participated in the study. On the basis of results of the study, they concluded that if we want to help athletes find their own zone of optimal performance we have to first monitor the entire spectrum of psychophysiological and behavioral features during different types of sport performances, and to then develop and implement bio/neurofeedback training [31]. The aim of our study was to determine exactly how effective biofeedback training is for learning psychophysiological control over the autonomic nervous system, following the completion of an 8-week-long biofeedback training program. Success was manifested through fewer breaths per minute, lower skin conductance, lower heart rate, higher blood flow and higher heart-breathing coherence. We wanted to know whether the percentage of participants in the experimental group who were able to improve their physiological values is higher when compared to the percentage of participants in the control group. The second goal was to determine if athletes were still using the psychophysiological regulation skills they learned during the 8-week-long Biofeedback training program, and whether these skills enhanced their sport performance 1 year after completing the study.

4

N. Pusenjak et al.

Materials and methods


Participants The participants in this study were Slovenian top athletes between 16 and 34 years of age. They were representative athletes of different sport associations, and though none had any prior psychophysiological training, some of the athletes were already familiar with the autogenic training technique. All athletes agreed to participate in the study willingly and provided their explicit consent. The experimental group consisted of athletes who were available to undergo the 8-week-long Biofeedback training program. This included 18 individuals, of whom 4 were female and 14 male. The control group consisted of 4 women and 17 men. The participants still actively competed in different sports (archery, shooting, fencing, wake boarding, athletic, volleyball, basketball, skiing, cross country skiing, swimming, tennis, cycling, Take Qwon Do, table tennis, carting) at the national and international levels. Instruments All study participants were administered identical stress tests (long and short), irrespective of whether they had taken part in biofeedback training. These tests were conducted before and after the 8-week-long training program, and utilized Bio Trace software, which is a part of the Nexus 10 (Mind Media BV, NL) biofeedback device. It consists of three sensors for measuring physiological responses: a breathing sensor (RESP), placed around participant’s waist that measures the number of breaths per minute; a blood volume sensor (BVP), placed on a middle finger, measuring heart rate, blood volume and coherence; and a galvanic skin response (GSR) sensor attached to bent fingers. Biofeedback training was also performed using the Nexus 10 biofeedback system, with Bio Trace software animations (Mind Media BV, NL). These were developed to train control over different physiological parameters. They are equipped with four sensors—GSR, RESP, BVP and temperature sensor. All sensors were connected to the Nexus 10 biofeedback device, and the apparatus transmitted collected data to a computer via Bluetooth. Data pertaining to each athlete was automatically saved at the end of the stress tests. The sampling rate per sensor channel was 1024 samples/sec (max). At the end of each measurement session, the Biofeedback device software graphically displayed the athletes’ stress profiles with essential statistical calculations for every completed stress test (long and short separately). In addition to the above measurements, Biofeedback training also utilized the Wild Divine biofeedback device and the integrated software Wisdom Quest, which incorporated two sensors, namely GSR and BVP. Wisdom Quest software has a high quality graphic interface and is similar to adventure video games. Here, the user has to practice breathing and relaxation exercises to proceed through the game.


their training over time, they received a questionnaire a year after their participation in the study. They were asked whether they believed that this had helped enhance their performance. They were asked the following questions: 1. Are you using the skills you learned during the biofeedback-training program as a part of your precompetitive and competitive preparation? 2. Did psychophysiological control help you to enhance your sport performance?

Procedure Stress tests The athletes who volunteered to participate in the study were divided randomly into experimental and control groups. Prior to starting the training program, in which only the experimental group was to participate, both groups were given short and long biofeedback stress tests. Testing was conducted in the Psychodiagnostic Laboratory, at the Faculty of Sport. The same tests were repeated upon completion of the training program. The stress tests followed the experimental design described below: The participant sits comfortably on a chair in front of a computer and the biofeedback device. He/she is then asked to observe the computer screen, and the sensors are attached to his/her fingers and around his/her waist. The biofeedback practitioner sits beside the participant, in order to provide guidance and instructions for each task. The short stress test consisting of 5-min-long software animation: . . .

The long stress test consisting of 14-min-long animated program incorporating stressors and relaxation phases: . .

Questionnaire As one aim of the study was to establish whether the athletes in the experimental group continued to use biofeedback in

1-min-long baseline interval (relaxing screen photo and calming music) 2-min-long stressor interval (loud noises, catastrophic screen photos) 2-min-long relaxation interval (relaxing screen photo and music)

.

2-min-long baseline interval (relaxing screen photo and calming music) 2-min-long stressor interval (Stroop task [32] – participant is asked to name the color in which the word indicating color, such as red, blue, green, etc.) displayed on the screen is written. The response has to be fast and precise, as the words are changing fast, and the word and the color in which it is displayed do not correspond (i.e., word RED is written in blue, for example, or ORANGE is displayed in green color). As the brain has difficulty responding correctly to the conflicting stimuli, the participants’ cognition is affected, causing stress. 2-min-long relaxation interval (relaxing screen photo and calming music)


DOI: 10.1080/00913847.2015.1069169

.

. .


.

2-min-long stressor interval (math task: the participant is asked to subtract 7 from a large number and to repeat the process with the obtained result. As the responses have to be given correctly and quickly, this task induces stress) 2-min-long relaxation interval (relaxing screen photo and calming music) 2-min-long stressor interval (traumatic story-telling task – the participant is asked to describe a very traumatic event that he/she experienced in the last year) 2-min-long relaxation interval (relaxing screen photo and calming music)

Biofeedback training/therapy At the beginning and at the end of the study, athletes belonging to both experimental and control groups were subjected to the stress test measurements (short and long stress tests together). The first stress test measurement was followed by an 8-week-long biofeedback training period in which only the experimental group took part. The participating athletes attended the Biofeedback training sessions twice a week, each lasting 1-h. In addition, they were instructed to perform breathing exercises at home. At the end of the program, all participants (experimental and control groups) were subjected to the stress tests identical to those conducted at the beginning of the study. The Biofeedback training provided as a part of this study consisted of breathing, muscle, GSR, and temperature biofeedback exercises provided by Bio Trace software. More specifically, protocols for Abdominal breathing, Heart rate variability, Galvanic skin response, Thermal control and Protocols for multi-physiological biofeedback were used, whereby athletes had to control multiple physiological parameters at the same time. Data analysis The overall improvement in each physiological parameter was measured as the movement of the curve in a specific segment of a stress profile before, during and after the stress phases. The objective of this study was to determine how successful each participant was in controlling his/her psychophysiological responses and to establish if the athletes in the experimental group were more successful in this task when compared to those in the control group. More specifically, the question was—by how many percentage points was the participant able to recover the parameter value during the relaxation phase? Moreover, following the stressor phase, was he/she able to return to the baseline level? The parameters measured in order to answer the above questions were: RESP – number of breaths per minute, HR – heartbeats per minute, BVP amplitude – blood volume pulse amplitude (total volume of blood per heartbeat), GSR – galvanic skin response (expressed in micro Siemens), HRV – heart rate variability amplitude (variation in beat-to-beat interval), respiratory-heart rate coherence (described by Kay in 1988 [32] as a method of linear coupling between two signals – respiration and heart rate).

5

In our study, we were concerned with comparing the expected measurement results with the observed results. We could conduct the RM ANOVA in this study, but as it would not give us more than the relationship between the three points in time, we decided to perform just a simple chi-square test. The psychophysiological parameters in all individuals vary, due to the natural variance in biological rhythms. In particular, due to the slight adaptation of laboratory settings, the testing conditions could not be replicated exactly. Once the above measurements were performed for each participant, the values obtained in the first stress test were compared to the corresponding measurements in the second stress test. In this phase, the aim was to establish which percentage of the participants in both experimental and control groups had improved control over a specific parameter over the 8-week period. Any differences between the two groups could thus be attributed to the effectiveness of the biofeedback-training program. The significance of the results was evaluated with chisquare test to determine if there is a significant difference between expected and observed frequencies. With chi-square analysis, we were testing the null hypothesis (i.e., that there is no significant difference between expected and observed results), although this test did not give us more information about the strength of the relationship and its substantial significance regarding the population. Another drawback to this test is its sensitivity to sample size. Questionnaire response analysis Participants in the experimental group were asked about effectiveness of the biofeedback training. They were given two options to choose between: yes or no. The answers were tallied for each question. Once all responses were given, the number of positive responses was displayed as an indication of the program effectiveness.

Results The results of self-regulation from the first stress test to the second stress test are presented in Figures 1, 2, 3 and 4. They are displayed separately for the control (ctrl) and experimental (exp) groups, and are expressed as improvements in the individual parameters: GSR – galvanic skin response, HR – heart rate, RESP – respiration rate per minute, BVP – amplitude of blood flow, and COH – coherence of respiration and heart rate for different stressors. The first stressor was the SHORT STRESS TEST, followed by STROOP, MATH and STORY stressors. In chi-square calculations figures, there is always a control group described as ctrl and experimental groups as exp. The number of athletes who were able to improve psychophysiological control over specific parameter is described with 1, whereas the number of athletes who were not able to improve psychophysiological control over specific parameter is denoted with 0. Figure 1 shows a comparison of the overall improvement (expressed as a percentage of participants in a


6

N. Pusenjak et al.


a

b

c

d

e

Figure 1. Comparison of the overall improvement (expressed as a percentage of participants in a specific group) after a second SHORT STRESS TEST. (a) Chi-square values for SHORT STRESS TEST – GSR. (b) Chi-square values for SHORT STRESS TEST – HR. (c) Chi-square values for SHORT STRESS TEST –RESP. (d) Chi-square values for SHORT STRESS TEST – BVP. (e) Chi-square values for SHORT STRESS TEST – COH. Abbreviations: BVP = Blood volume sensor; COH = Coherence; GSR = Galvanic skin response; HR = Heart rate; RESP = Respiration.

specific group), specifically after a second SHORT STRESS TEST. In Figure 1a, we can see the chi-square statistic for the short stress test galvanic skin response (GSR) measurement, which is 4.310. The p value is 0.037. The result is significant at p < 0.05. In Figure 1b we have chi-square statistic for the short stress test heart rate (HR) measurements. Here, the result is 0.012. The p value is 0.910. The result is not significant at p < 0.05. Figure 1c shows the chi-square statistic for the short stress test: respiration (RESP) measurement calculated as 0.059. The p value is 0.910. The result is not significant at p < 0.05.

In Figure 1d is the chi-square statistic for the short stress test and parameter blood volume pulse amplitude – BVP measurement is 1.806. The p value is 0.178. The result is not significant at p < 0.05. In Figure 1e is the chi-square statistic for short stress test, coherence parameter – COH measurement is 1.430. The p value is 0.231. The result is not significant at p < 0.05. The second short stress test (performed after the experimental group completed the 8-week-long training program) results indicate that a greater number of athletes in the experimental group (EXP group) showed improvement in all


DOI: 10.1080/00913847.2015.1069169


7

a

b

c

d

e

Figure 2. Comparison of the overall improvement (expressed as a percentage of participants in a specific group), based on the second measurement of the stress test, following STROOP stressor. (a) Chi-square values for STROOP stressor – GSR. (b) Chi-square values for STROOP stressor – HR. (c) Chi-square values for STROOP stressor – RESP. (d) Chi-square values for STROOP stressor – BVP. (e) Chi-square values for STROOP stressor – COH. Abbreviations: BVP = Blood volume sensor; COH = Coherence; GSR = Galvanic skin response; HR = Heart rate; RESP = Respiration.

parameters except the heart rate (HR), where the control group (ctrl group) performed better (by 0.2%, as shown in Figure 1). Figure 2 describes the comparison of the overall improvement (expressed as a percentage of participants in a specific group), based on the second measurement of the stress test after STROOP stressor. In Figure 2a, the chi-square statistic for Stroop stressor for galvanic skin response measurement (GSR) 3.121. The p value is 0.077. The result is not significant at p < 0.05. In Figure 2b the chi-square statistic is

displayed for Stroop heart rate (HR) measurement and is 4.310. The p value is 0.037. The result is significant at p < 0.05. In Figure 2c, we can see the chi-square statistic for Stroop – respiration (RESP) measurement: the value 1.812. The p value is 0.178. The result is not significant at p < 0.05. In Figure 2d, we have the chi-square statistic for Stroop stressor for blood volume pulse (BVP) measurement, which is 1.061. The p value is 0.302. The result is not significant at p < 0.05. Figure 2e shows calculations of the chi-square


8

N. Pusenjak et al.


a

b

c

d

e

Figure 3. Comparison of the overall improvement (expressed as a percentage of participants in a specific group), based on the second measurement of the stress test after MATH stressor. (a) Chi-square values for MATH stressor – GSR. (b) Chi-square values for MATH stressor – HR. (c) Chi-square values for MATH stressor – RESP. (d) Chi-square values for MATH stressor – BVP. (e) Chi-square values for MATH stressor – COH. Abbreviations: BVP = Blood volume sensor; COH = Coherence; GSR = Galvanic skin response; HR = Heart rate; RESP = Respiration.

statistic for Stroop stressor. The coherence (COH) measurement is 1.430, and the p value is 0.231. The result is not significant at p < 0.05. In fact, the only statistically significant parameter that the athletes were able to control after the Stroop stressor is heart rate. In Figure 3, we have comparison of the overall improvement (expressed as a percentage of participants in a specific group), based on the second measurement of the stress test after MATH stressor. In Figure 3a we have the chi-square statistic for Math - GSR measurement and its value is 4.178. The p value is 0.040. The result is significant at p < 0.05.

In Figure 3b, we see the chi-square statistic for Math – heart rate (HR) measurement and the calculation is 4.542. The p value is 0.033. The result is significant at p < 0.05. In Figure 3c is the chi-square statistic for Math – respiration (RESP) measurement. The result is 1.292. The p value is 0.255. This result is not significant at p < 0.05. In Figure 3d is the chi-square statistic for Math – blood volume pulse (BVP) measurement and the value is 1.981. The p value is 0.159. The result is not significant at p < 0.05. In Figure 3e is the chi-square statistic for Math stressor and coherence (COH) measurement. The result is 4.127. The p value is


DOI: 10.1080/00913847.2015.1069169


9

a

b

c

d

e

Figure 4. Comparison of the overall improvement (expressed as a percentage of participants in a specific group), based on the second measurement of the stress test after STORY stressor. (a) Chi-square values for STORY stressor – GSR. (b) Chi-square values for STORY stressor – HR. (c) Chi-square values for STORY stressor – RESP. (d) Chi-square values for STORY stressor – BVP. (e) Chi-square values for STORY stressor – COH. Abbreviations: BVP = Blood volume sensor; COH = Coherence; GSR = Galvanic skin response; HR = Heart rate; RESP = Respiration.

0.042. The result is significant at p < 0.05. The results suggest that significant results of regulation following the math stressor occurred in regulation of galvanic skin response, heart rate and coherence. In Figure 4 we present a comparison of the overall improvement (expressed as a percentage of participants in a specific group), based on the second measurement of the stress test after STORY stressor. In Figure 4a is the chisquare statistic for the Story stressor and galvanic skin response (GSR) measurements. The value is 0.140. The p value is 0.707. The result is not significant at p < 0.05.

In Figure 4b, you can see the chi-square statistic for Story stressor and heart rate (HR) measurement - 1.112. The p value is 0.291. The result is not significant at p < 0.05. In Figure 4c is presented the chi-square statistic for the Story stressor and respiration (RESP) measurement – 1.857. The p value is 0.172. The result is not significant at p < 0.05. In Figure 4d, we can see the chi-square statistic for the Story stressor and blood volume (BVP) measurement. Here, the result is 0.199. The p value is 0.655. The result is not significant at p < 0.05. The Figure 4e displays the chi-square statistic for Story stressor and coherence (COH) measurement.

10

N. Pusenjak et al.

The value is 0.002 and the p value is 0.959. The result is not significant at p < 0.05. There are no significant results of regulation of parameters following the story stressor. Calculations ware made if results were significant at p < 0.05. Significant results were seen in regulation of GSR after the Short stress test condition (p = 0.037). In addition, regulation of HR after the STROOP stressor (p = 0.037), regulation of HR after the Math stressor (p = 0.033), regulation of GSR after the Math stressor (p = 0.409) and regulation of coherence after the Math stressor (p = 0.042) all yielded compelling results.


Questionnaire results As previously noted, only the participants assigned to the experimental group and provided with biofeedback training were asked to answer the questions about effectiveness of biofeedback on two questions. Each of these participants answered positive (YES answers).

Discussion The study findings reported in the preceding sections provide support for the initial hypotheses. The study objective was to determine whether athletes assigned to the experimental group would exhibit better psychophysiological control over their autonomic nervous system after completing the 8-weeklong biofeedback training program in comparison to those assigned to the in control group. The study’s findings demonstrated that the athletes that were successful in attaining psychophysiological control were able to reach a state of deeper relaxation (as indicated by fewer breaths per minute, lower skin conductance, lower heart rate, higher blood flow and higher respiration-heart rate coherence) following the intervals of stress in the stress tests administered at the end of the study. In addition, results pertaining to the overall improvement in the physiological parameters after the second measurement short stress test indicate that a greater proportion of athletes in the experimental group were able to control the specific parameters in comparison to those in the control group. However, heart rate measurements revealed that not only did both groups achieve nearly identical results, but also that only 5% of participants in each group were able to improve the control over their heart rate after being subjected to the stressor. Moreover, the proportion of successful participants was similar between the groups, with respect to the improvement in the RESP, BVP and HR parameter (Figure 1). These results are surprising, as the experimental group was trained to control the breathing pattern, increase coherence and control the heart rate and was thus expected to perform better. Moreover, as the tests performed at the beginning and at the end of the study were identical, in the second test, although the participants were familiar with the images of disasters and the sounds of the stress test, they were still not efficient in controlling the responses. The major difference between the two groups was in the skin conductance parameter. Approximately 67% of the participants in the experimental group improved their control over their GSR value, compared to


only 35% participants in the control group. Similar findings relate to the coherence parameter, where the experimental group achieved 17% better results than the control group (Figure 1). These findings suggest that the majority of the participants in the trained group were able to actively manage these two parameters, and that they had successfully learned the self-regulation (LSR) skill, as suggested by Wilson and Cummings [5]. Furthermore, closer examination of the results achieved by both groups after the second measurement of the Stroop stressor interval reveals that the athletes in the experimental group achieved better results than did those in the control group in all parameters, except BVP (Figure 2). In addition, the differences between the two groups were greater when compared to those found in the short stress test. The important alterations by the majority of the participants in the experimental group ware displayed in the ability to control the GSR, HR, and RESP, as well as in the coherence parameter. Although the outcomes of the control group were also very good, the results achieved by the experimental group were evidently better, with the improvements ranging from 12–37%, compared to the control group, where the BVP parameter was excluded. Thus, it can be assumed that similar outcomes would pertain to the LSR [5]. However, as previously noted, the initial and the final tests were identical, implying that both experimental and control groups were familiar with the testing procedures. Hence, familiarity with the stressor protocols might have resulted in improved coping strategies (as indicated by the participants’ comments provided at the end of the study). For example, some participants noted that, in the Stroop test, in order to avoid the distraction, they focused on last letter in the word to articulate the correct word color, rather than looking at the whole word. Although this is clearly the weak point of the adopted testing protocol, the repetition of the same stress tests was chosen: this is the widely used biomedical testing protocol, and our aim was to use it in the field of psychophysiology in order to assess the training program’s efficacy. Neither group was particularly successful in controlling the BVP parameter (Figure 2), although the control group achieved much better results (25%), compared to the experimental group (11%). One explanation of these rather unexpected findings can be seen in the measurement of blood volume pulse amplitude, whereby BVP signal measures the anticipatory anxiety, that is indicates that a person anticipates the tasks to be stressful: in the experiments, all participants knew that the Stroop stressor would require significant cognitive engagement. Thus, it is expected that this parameter would take more time to return to the baseline levels. Furthermore, it signifies the overall emotions exhibited by the participant [31]. Another reason for these values could be the fact that the BVP amplitude is related to muscle relaxation. Thus, the experimental group could have performed better in this respect if the athletes had been coached to relax their muscles using an EMG (electromyography) sensor. Instead of adopting that approach, the participants were simply instructed to relax their muscles using Jacobson’s progressive muscle relaxation technique [32]. No test was performed to establish whether they had been successful in this exercise.


DOI: 10.1080/00913847.2015.1069169

It is also possible that the athletes assigned to the experimental group approached the second measurement as they would a competition or an exam, where they would have to demonstrate what they had learned in an 8-week-long biofeedback-training program. Consequently, they were likely more tense in comparison to the athletes in the control group, who simply repeated the familiar tests without specific expectations regarding their performance. Finally, as noted above, the experimental group was not trained with EMG biofeedback. As a result, their ability to relax the muscles was reduced, as this is a skill best learned through EMG training. Examination of overall improvement following the Math stressor indicates that, whereas both groups were successful in controlling their responses to cognitive stress, a greater number of participants in the experimental group achieved the control in all parameters, compared with those in the control group. The greatest difference was found in the ability to regulate the GSR parameter: nearly 83% participants in the experimental group were successful, compared with only 50% of participants in the control group. These results suggest that athletes in the experimental group were able to shift their concentration and focus on the mathematical task and at the same time remained self-aware of their body and mind reactions. This could be an effect of GSR and RESP biofeedback training [5]. The smallest difference between the two groups was found in the ability to control the BVP parameter, where only 15% of the control group members were able to improve their results, compared to 33% of those in the experimental group. A more detailed examination of the results indicating improvement in individual parameters following the exposure to the STORY stressor reveals that the differences between the two groups are, in fact, very small. The greatest difference is observed in the respiratory parameter, where 78% of the participants in the experimental group successfully improved their control over their respiration after exposure to this stressor, compared with only 55% in the control group. These results indicate that the experimental group members actively used their new breathing skills in psychophysiology regulation, as they had practiced it for 8 weeks. The differences in the results, although in favor of the experimental group, are minimal: only 2% improvement was achieved in GSR, 5% in BVP and just 1% in coherence. Finally, with respect to the HR parameter, the control group was better than the experimental group by 15% [33]. These results could potentially be explained by the fact that, during the 8-week-long training program, the trainee and the practitioner would develop a close, trust-inducing relationship, allowing the trainees to describe their most painful traumatic events more easily than the athletes in the control group could. Moreover, as previously noted, as they did not undergo a training program, the athletes in the control group were not under performance pressure, unlike the experimental group. This was likely to lead to competitive (or at least task performance) pressure when attending the second stress test session. Of course, this group should in theory have been able to control their responses to the stressful situation, as they had been trained to do so. Nonetheless, it is likely that


11

the level of stress at the final stress test was not equivalent between the two groups, making it difficult for the trained individuals to compensate for it. Clearly, as the traumatic story-telling stressor is emotional in nature, it is influenced by our own expectations, fears and even personality traits. These traits may include optimism, doubt, perfectionism, etc., – none of which were controlled for in this study. Thus, in order to conduct a more reliable assessment of the biofeedback training program, it would be beneficial to include all possible mental techniques, as well as personality assessment tests. Once this is implemented, the biofeedback-training interval can be adjusted to the actual needs of each participating athlete and his/her sport. Several authors have previously suggested this type of combined approach. For example, IZOF [27,28] or Bar-Eli’s [16] Wingate five-step approach, which Galloway [18] later used. Beauchamp et al. [24] later developed their own approach, presented as ‘Integrated biofeedback and psychological skills training’ for Canada’s Olympic speed skating team. Twenty-five Chi-square statistics were calculated for all parameters and groups regarding the all four stressors and overall stress. The hypothesis (H0) was that an 8-week long biofeedback training does not influence successful regulation of certain parameters following the stressor. In each of these cases, H0 can be rejected. We can accept the alternative hypothesis, and assume that 8-week long biofeedback training probably does influence the successful regulation of the above-mentioned parameters and stress conditions, and that this could potentially be valid for the overall population, not just our sample. There were no other significant results. As can be seen from the results, the biofeedback training intervention provided to the experimental group as part of this study improved the cognitive psychophysiological control of measured parameters in the participants assigned to this group.

Conclusion The key conclusion arising from the second study is that the 8-week-long biofeedback training provided to study participants could contribute to their improvement of the psychophysiological control over their autonomic nervous system. Moreover, this finding supports the initial hypothesis, as the overall improvement following prolonged stress in all four stressor conditions was higher in the experimental group when compared to the control group in all measured parameters. The least difference was found in the BVP parameter: here, the experimental group was better by only 3%. However, when all parameters were examined, the experimental group was better by >15%. Thus, these results indicate that the 8-week-long biofeedback training program (which athletes attended two times per week in 1-h-long sessions), when combined with visualization, progressive muscle relaxation for faster learning of psychophysiological control, is an effective method for learning psychophysiological selfregulation [3]. In addition, the experimental group responses to the two questions about biofeedback and sport enhancement suggest that athletes gained new and valuable skills by


12

N. Pusenjak et al.

using the biofeedback technique, also known as ‘Learned Self Regulation (LSR)’ [5]. The respondents indicated that they used these skills regularly as a part of their precompetitive and competitive preparation, even 1 year after the training. Moreover, they reported that these skills helped them to enhance their athletic performance. Research conducted by Wilson et al. [5] indicates that, when a person experiences the so-called Aha moment and learns to effortlessly self-regulate and modify his/her physiological reactions, movements and attitudes, implementing these new mastery skills in real sport situations enhances performance. This, in turn, creates the desire to continue using these newly acquired skills, as their effectiveness has been proven. However, in order for athletes to embrace this form of training, it is essential that biofeedback can be utilized in real-time and in a real sports situation, away from the laboratory setting. This is the most vital and difficult component for athletes to achieve if they had previously been trained in the lab setting. The results of this study could thus prompt more athletes to take part in similar training programs where they can learn LSR skills. These skills are proven to remain stable, as the athletes that took part in this study retained them 1 year after completing the training and believed that their application helped them improve their sport performance. The 8-weeklong biofeedback-training program could be developed to specifically target different levels, and thus serve as a foundation for psychophysiological training for all athletes in any competitive sport. It is fun, motivational, and costeffective. But as an effective and quality training plan, it has to be combined first with performance models such as IZOF [29] or Individual Affect Performance zones (IAPZ) to fully determine an athlete’s performance zones and how they are connected to their emotional states [30]. Future studies may explore whether the same length and structure of biofeedback training would be equally successful for younger athletes, or if the duration of training should be further optimized.

Acknowledgements The protocols used in this study are in accordance with the requirements of good clinical practice and all the applicable regulations of the Republic of Slovenia. The testing protocol of this study was approved by The National Medical Ethics Committee of the Republic of Slovenia (NMEC) that operates in compliance with the ICH/GCP requirements of good clinical practice and all the applicable regulations of the Republic of Slovenia. Chi-square statistic calculations were made with Social Science Statistics web site online calculation program derived from SPSS - link http://www.socscistatistics.com/tests/chisquare/Default.aspx.

Declaration of interest This research was supported by the European Union through the European Social Fund and was coordinated by the Public Agency for Technology of Republic of Slovenia (TIA -


SPIRIT) according to the program of Human Resources Development 2007–2013. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

References [1] Schwarz MS, Andrasik F. Biofeedback: A practitioner guide. New York (NY): The Guilford Press; 2003. 27. [2] Bradley RT, McCraty R, Atkinson M, Tomasino D, Daugherty A, Argueles L. Emotion self-regulation, psychophysiological coherence, and test anxiety: results from an experimental using electrophysiological measures. Appl Psychophysiol Biofeedback 2010;35:261–83. [3] Zaichkowsky LD. Biofeedback for self-regulation o competitive stress. In Zaichkowsky LD, editor. Stress management for sport. Reston (VA): American Alliance for Health, Physical Education, Recreation and Dance, National Association for Sport and Physical Education; 1982. p 55–64. [4] Blumenstein B, Bar-Eli M, Thenenbaum G. Brain and body in sport and exercise. West Sussex (UK): John Wiley & Sons, LTD; 2002. 43. [5] Wilson VE, Cummings M. Owner’s manual for self-regulation of your brain and body. YSAM Inc; Toronto: 2004. [6] Strack B. The effect of heart rate variability biofeedback on batting performance in baseball [dissertation]. San Diego (CA): Alienate International University; 2003. 3. [7] Wilson V, Peper E. Athletes are different: factors that differentiate biofeedback/neurofeedback for sport versus clinical practice. Biofeedback 2011;39:27–30. [8] Wilson VE, Peper E, Gibney KH. Using the ’Aha’ experience with biofeedback: enhancing body-mind integration. Biofeedback 2004;32:21–5. [9] Lazarus RS. Emotion and adaptation. New York: Oxford University Press; 1991. 2. [10] Burton D. Measuring competitive state anxiety. In Duda J, Morgantown, Fitness Informational Technology, editor. Advances in sport and exercise psychology 1998;129–48. [11] Andreassi JL. Psychophysiology: human behavior and physiological response. 4th edition. Hillsdale, NJ: Erlbaum; 2000. p 116. [12] Collins D. Psychophysiology and sport performance. In Biddle SJH, Editor. European Perspective in Exercise and Sport. Leeds, UK: Human Kinetic press; 1995. p 154–78. [13] Petruzzelo SJ, Landers DM, Salzar W. Biofeedback and sport/ exercise performance: Application and limitations. Behav Ther 1991;22:379–92. [14] Peper E, Schmid A. The use of electrodermal biofeedback for peak performance training. Somatics 1983;4:16–18. [15] Strack B. The effect of heart rate variability biofeedback on batting performance in baseball. In Edmonds A, Gershon T, editors. West Sussex: John Wiley & sons. Case Studies in Applied Psychophysiology Neurofeedback and Biofeedback treatments for Advance in Human Performance; 2003. p 9. [16] Bar-Eli MD, Dreshman R, Blumenstein B, Weinstein Y. The Effect of Mental Training with Biofeedback on the Performance of Young Swimmers. Appl Psychol 2002;51:567–81. [17] Blumenstein B, Bar-Eli M, Tenenbaum G. A five-step approach to mental training incorporating biofeedback. Sport Psychol 1997;11:440–53. [18] Galloway SM. The Effect of Biofeedback on Tennis Service Efficiency Biofeedback. Int J Sport Exerc Psychol 2011;9:251–66. [19] Beauchamp MK, Harvey RH, Beauchamp PH. Integrated biofeedback and psychological skills training for Canada’s Olympic speed skating team. J Clin Sport Psychol 2012;6:67–84. [20] Wilson V, Peper E, Donald M. “The Mind Room” in Italian Soccer Training: The Use of Biofeedback and Neurofeedback for Optimum Performance. Biofeedback 2006;34:70–81. [21] Dupee M, Werthner P. Managing the Stress Response: The Use of Biofeedback and Neurofeedback with Olympic Athletes. Biofeedback 2011;39:92–4.


DOI: 10.1080/00913847.2015.1069169

[22] Pop –Jordanova N, Demerdzieva A. Basic science biofeedback training for peak performance in sport - case study. Macedon J Med Sci 2010;3:113–18. [23] D Perry F, Shaw L, Zaichkowsky Z. Biofeedback and neurofeedback in sports. Biofeedback 2011;39:95–100. [24] Lagos L, Vaschillo E, Vaschillo B, Lehrer P, Bates M, Pandina R. Virtual reality–assisted heart rate variability biofeedback as a strategy to improve golf performance: a case study. Biofeedback 2011;39:15–20. [25] Thompson AG, Swain DP, Branch JD, Spina RJ, Grieco CR. Autonomic response to tactical pistol performance measured by heart rate variability. J Strength Cond Res 2015;29:926–33. [26] Paul M, Garg K. The effect of heart rate variability biofeedback on performance psychology of basketball players. Appl Psychophysiol Biofeedback 2012;37:131–44. [27] Hanin Y. Emotions in sport: current issues and perspectives. In Tenenbaum Eklund RC, editors. Handbook of Sport Psychology. 3rd edition. Hoboken, NJ: John Wiely & Sons; 2007. p 31–58.


13

[28] Hanin Y. Performance related emotional states in sport: a qualitative analysis. Forum qualitative sozialforshung/forum qualitative social research [online journal]. 2003. Available from http://www.qualitative-research.net/fqs-texte/1-03/1-03hanin-e.htm. [cited 2 October 2014]. [29] Kamata A, Tenenbaum G, Hanin Y. Individual Zone of Optimal Functioning (IZOF): A probabilistic estimation. J Sport Exerc Psychol 2002;24:189–208. [30] Van der Lei H, Tenenbaum G. Performance processes within affectrelated performance zones: a multi-modal investigation of golf performance. Appl Psychophysiol Biofeedback 2012;37:229–40. [31] Bertollo M, Bortoli L, Gramaccioni G, Hanin Y, Comani S, Robazza C. Behavioural and psychophysiological correlates of athletic performance: a test of the multi-action plan model. Appl Psychophysiol Biofeedback 2013;38:91–9. [32] Stroop JR. Studies of interference in serial verbal reactions. J Exp Psychol 1935;18:643–62. [33] Kay SM. Modern spectral estimation: theory and application. Upper Saddle River (NJ): Prentice Hall; 1988. p 470.