1. IntroductionThe term “stress” as it is currently used was conceived in 1936 by Hans Selye, who defined it as “the non-specific response of the body to any demand for change” [1]. The key components of the “stress system” are the hypothalamic–pituitary–adrenal (HPA) axis and the sympathetic nervous system (SNS). Stressors cause the production of corticotrophin-releasing hormone (CRH) in the hypothalamus, inducing the secretion of adrenocorticotropic hormone (ACTH) in the posterior pituitary and the activation of the noradrenergic neurons in the locus coeruleus/norepinephrine (LC/NE) system in the brain. The LC/NE system plays the main role in the immediate “fight or flight” response by releasing epinephrine and norepinephrine, while ACTH elicits the secretion of cortisol in the adrenal cortex. Normally, CRH and ACTH levels vary in a predictable circadian cycle and are suppressed by high concentrations of blood cortisol via negative feedback loop [2]. The acute-stress response is immediate and intense. When acute stress is severe or lasts a longer period of time, it might have deleterious consequences on health [3].Chronic stress can cause many different symptoms: cognitive, which include memory problems, difficulty in concentrating, constantly worrying, racing mind; emotional: depression, anxiety, mood swings, irritation, abandonment; physical: pains and aches, diarrhea or constipations, chest pain, tachycardia, sickness, frequent colds; and behavioral: eating disorders, changes in resting hours, withdrawing from others, neglecting responsibilities, smoking, and alcohol or drug use in order to relax [2,4].Fortunately, there are many evidence-based stress-reduction techniques which are easy to learn and practice. They include biofeedback, autogenic training, progressive muscle relaxation, diaphragmatic breathing, emotional freedom technique, relaxation response, guided imagery, transcendental meditation, and mindfulness-based interventions. They all can diminish bodily and mental tension, causing a decrease in disease symptoms, prevention of disease, and improvement of the patient’s quality of life [5].Biofeedback, which is the main subject of this research, is a process which helps an individual to learn how to change physiological activity with the intention to improve health. Heart rate, blood pressure, and muscle tension are examples of physiological functions that people can learn to control [6]. During biofeedback training, precise instruments measure and give information to the user about distinct physiological stress/relaxation indicators such as brainwave or breathing patterns, heart function, muscle activity, and skin temperature. The presentation of the aforementioned information, often in conjunction with changes in thinking, emotions, and behavior, supports desired physiological changes. Over time, these alterations can be maintained without the continuous use of an instrument [7].Based on the content of biofeedback, it can be categorized into several types: neurofeedback, respiratory, heart rate variability, galvanic skin response, blood pressure or thermal feedback, and electromyography [6]. These techniques have been successfully used for the treatment of headache [8], hypertension and type II diabetes [9], asthma [10], anxiety disorders [11], depression [12], as well as for the reduction of pain and mental stress during an early postpartum period [13].Respiratory biofeedback systems measure and present breathing-related information to help users learn specific breathing skills for relaxation and stress relief [14]. Abdominal or diaphragmatic breathing is considered to be the best technique for subjects beginning to practice breathing exercises. During abdominal breathing, a person inhales through the nose and expands the abdomen slowly by gently pushing out and down as the oxygen fills the lower lung cavity. When the abdomen is full, the individual exhales through the nose and pulls the abdomen back, pressing the lungs from the bottom. Overall, the main purpose of diaphragmatic breathing is to fill up the lungs completely. The major advantage of this breathing is the invigoration of the abdominal muscles, as their persistent movements massage the internal organs and boost blood circulation [15].In the past two decades, respiratory sinus arrhythmia (RSA)—a biomarker of parasympathetic-nervous-system-mediated cardiac control—has proved to present reliable information of emotion regulation [16]. Heart rate varies simultaneously with respiration: the inter-beat-interval (IBI), the time difference between two beat pulses, is shorter during inhalation and longer during exhalation [17]. This physiological phenomenon is called “respiratory sinus arrhythmia” [14]. In some studies, RSA biofeedback is considered to be a certain type of HRV biofeedback (HRV-BF), as the IBI shows heart rate variability (HRV) and assists individuals in controlling their breathing [14]. Previous studies showed that RSA biofeedback interventions effectively reduced resting heart rate, skin conductance level, and systolic blood pressure [18,19], lowered anxiety levels, and improved resistance to stressful situations [20].It has been suggested that the use of virtual reality (VR) can make respiratory-based biofeedback techniques an even more powerful tool for coping with stress. VR can help the user to immerse completely into the biofeedback process by improving the visual attractiveness of biofeedback stimuli. Furthermore, the virtual environment is controllable and can be created both to maintain relaxation and to focus attention by using engaging environments [21,22].Previous RSA and respiratory biofeedback examinations showed its effectiveness in relaxation; however, most of them focused on changes in single or several stress indicators during biofeedback training [18,19,20,23,24]. The aim of the present study is to investigate alterations and dynamics of distinct stress indicators before, after, as well as during the biofeedback session in virtual reality. We hypothesized that even a single session of VR-based respiratory biofeedback would have positive effects on average heart and respiratory rate, heart rate variability, galvanic skin response values, salivary cortisol levels, and the psychological state of study participants. Since the duration of the relaxation session in the current study was only 12 min, we assumed that this strategy might be a particularly attractive alternative for subjects who are not able to regularly practice strictly scheduled and time-consuming stress management programs. 4. Discussion

The study aimed to investigate the impact of a brief VR-based respiratory-biofeedback-assisted relaxation session on the ANS and HPA axis activity, as well as on subjective assessments of strain, fatigue, and mood. The results indicated significantly decreased strain and fatigue and improved mood after the single relaxation session (effect sizes ranged from small to moderate). Also, there was a significant decrease in cortisol levels from before to after the session, as well as in the cortisol-to-cortisone ratio (moderate and small effect size, respectively). Analysis of percentage change values of salivary steroid hormone concentrations revealed that relaxation session had a large effect size on the percentage change of cortisol level. The majority of physiological biomarkers (i.e., heart rate, respiratory rate, galvanic skin response) also enhanced relaxation state during the last 2 min compared with the first 2 min of the relaxation session. Linear mixed-effects models confirmed the aforementioned results, with the exception of non-significant change in breathing rate during the entire session.

Although it is agreed that slow breathing techniques promote autonomic changes by shifting ANS activity toward a parasympathetic predominance [30,31], results from previous studies examining the impact of these techniques on HR/HRV measures are conflicting [23,24]. Rockstroh et al. [24] investigated the effectiveness of a single-session VR-based HRV biofeedback exercise in a group of healthy young adults. The authors found a significant decrease in HR from mid- to post-relaxation session, without significant changes from pre- to mid-session. In contrast, RMSSD increased from pre- to mid- and decreased from mid- to post-session, with no differences from pre- to post-session values. It should be noted that the aforementioned changes in HR and RMSSD values during VR-based HRV biofeedback exercise were similar to alterations observed in traditional HRV biofeedback treatment with a graphical indicators group and control condition. A recent study conducted by Blum et al. [23] examined the feasibility of a VR-based diaphragmatic breathing biofeedback algorithm. Results showed that subjects assigned to the feedback group had lower mean respiratory rate and higher RMSSD values during a 7 min breathing exercise compared with the participants in the control group without biofeedback. Since both groups experienced the same virtual environment, the results supported the idea that biofeedback has an additional positive effect in learning to voluntarily control breathing rate and thus exert greater parasympathetic activity. Another study [32] investigated the effects of a single 60 min session of HRV-BF with paced breathing (6 breaths/min) training on time- and frequency-domain HRV indices and compared it with the results obtained from subjects who received autogenic training (AT). Analysis revealed higher SDNN, LF, lnLF, and LF/HF values, as well as lower respiratory rate after the HRV-BF training with large effect sizes (ηp2 ranging from 0.32 to 0.36). Moreover, higher LF, lnLF, LF/HF, and a lower breathing rate post-training was observed in the HRV-BF group compared with AT group, while no differences in heart rate variability measures between groups were observed prior to the relaxation sessions. These findings indicate that slowing down the subjects’ breathing rate at approximately 6 breaths/min effectively increases parasympathetic tone and baroreflex gain. However, Van Diest et al. [33] showed that inhalation/exhalation ratio is a more important factor than respiration rate for inducing a self-reported relaxation state. Also, the study suggested that there is a combined effect of breathing rate and inhalation/exhalation ratio, as an increase in respiratory sinus arrhythmia was higher when participants were breathing at 6 breaths/min with a low inspiration/expiration ratio. Since the inhalation/exhalation ratio was not controlled in our study, this might explain the fact that no significant changes in RMSSD and pNN50 values were found in the current study. Thus, it would be interesting to examine the impact of respiratory-biofeedback-assisted relaxation training with controlled respiration rate and inspiration/expiration ratio (for example, an inhalation of 1.5 s and exhalation of 3.5 s) on heart rate variability measures.Ambiguous results are presented in Tinga et al.’s [34] work. Their study examined the effectiveness of respiratory biofeedback in lowering subjective and objective arousal after stress. Respiratory biofeedback was compared to a control feedback placebo condition in which visual feedback unpaired to participants’ breathing was presented, and a control condition in which no feedback was provided at all. The decrease in heart rate and subjective tension was significantly stronger in the biofeedback group compared to the control feedback placebo; however, RMSSD and EEG theta-to-alpha ratio were higher in the control biofeedback group. Taken together, the results indicated that the control feedback placebo was superior to respiratory biofeedback in reducing arousal. The current findings highlight the importance of including a control feedback placebo condition when studying the effectiveness of biofeedback in order to establish the exact additional value of providing biofeedback [34].The present study demonstrated that a VR-based respiratory-biofeedback-assisted relaxation approach reduced sympathetic activation as indicated by decreased skin conductance level at the end of the session. Similar findings were presented in the recent study [19], which aimed to evaluate the effectiveness of respiratory sinus arrhythmia (RSA) biofeedback training in a group of managers with high-level work responsibilities. The authors reported a significant decrease in skin conductance level after five weekly sessions of RSA-BF training and concluded that RSA-BF is an effective technique in reducing physiological arousal. To the best of our knowledge, only one study [35] evaluated the impact of rhythmic breathing on HPA axis activity by measuring morning salivary cortisol concentration before and after biofeedback-based stress management training which lasted for 28 days. Contrary to our study, Lemaire et al. [35] reported no statistically significant alterations in morning salivary cortisol level from pre- to post-training. It should be highlighted that the comparison between studies is not valid due to methodological differences, as we measured salivary glucocorticoid levels before and immediately after the single 12min relaxation session, while in the study conducted by Lemaire et al. [35], cortisol concentration was determined at baseline and after a 28-day relaxation trial.

The main strength of the study is that multiple stress-related biomarkers and indicators were used to evaluate the effectiveness of VR-based respiratory-biofeedback-assisted relaxation session. However, our study has several methodological limitations. It should be noted that there was no control group that did not participate in the VR-based relaxation session. Thus, the effectiveness of relaxation training should be confirmed in future studies with both experimental and control groups included. Since only young healthy volunteers, mainly women, were enrolled in the study, our work should be replicated in more heterogeneous populations. Moreover, data on menstrual cycle phase and sleep duration and quality was unavailable in the current study, and we were not able to include these potential confounding factors in the statistical analysis. Finally, a possible explanation of non-significant alterations in HRV measures during the session is the fairly short duration of the relaxation session, as well as the lack of information about HF component changes which reflect parasympathetic activity of the ANS. Therefore, the actual effect of the VR-based respiratory-biofeedback-assisted technique on subjects’ heart rate variability should be examined using both time- and frequency-domain HRV measures during longer duration training.