Thirty healthy male amateur esports players (age 23.1 ± 3.0 years), who spend an average of 12.3 h per week on esports, were tested. The subjects were recruited through a mailing to the students and employees of the University of Bayreuth and through the esports team of the University of Bayreuth. In total, approximately n = 16.000 people were reached directly. A power analysis according to Hopkins (Hopkins, 2020) was performed, which led to a minimum population size of n = 17. According to Saeidifard et al. (2018) and Amaro-Gahete et al. (2019), the difference in energy expenditure between sitting and standing is around 14%, which is why we calculated the smallest change in energy expenditure required to exclude a purely sedentary activity with +14%. We assumed a typical error of 13.6% based on past studies we have conducted on resting metabolic rate. The number of participants of n = 30 is, therefore, sufficient to prove our hypothesis.

The main inclusion criteria were playing the required game (Counter-Strike: Global Offensive or FIFA 20) for at least 10 h per week with interruptions of <1 month in the last year prior to study participation. The study team set this prerequisite to ensure participants were amateur players with a minimum gaming experience and routine level. Additional inclusion criteria required participants to be in good health, of male gender and between 18 and 40 years of age. Exclusion criteria were regular smoking or medical contraindications for participation in the study. In addition, no coffee, alcohol, or performance-enhancing supplements were to be consumed and no sport was to be practiced on the day of the trial. Table 1 provides the anthropometric and (e)sports activity data for all subjects.

The study was held at the sports medicine laboratory of the University of Bayreuth (Institute of Sports Science). It consisted of one gaming session, split into the pre-gaming, gaming, and post-gaming stage. The pre- and post-gaming phases lasted 10 min, the gaming session at least 30 min and consisted either of Counter-Strike: Global Offensive (CS:GO) or FIFA 20 (FIFA). CS:GO (Valve, Bellevue, US) is a tactical, first-person multiplayer shooter and was played either on the subject's personal device or an ASUS ZenBook UX434F (ASUSTeK COMPUTER INC., Taipei, Taiwan). The soccer simulation video game FIFA 20 (Electronic Arts, Redwood City, US) was performed on PlayStation 4 (Sony, Minato, Tokyo, Japan). Initially on the measuring day, an anthropometric measurement of body composition was performed using bioelectrical impedance analysis (BIA, In-Body 720, Biospace Co. Ltd., Seoul, South Korea), for which the participants were instructed to fast for at least 3 h before they arrived at the laboratory, and esports and sporting activity in hours per week were surveyed via a written questionnaire. Pre-, during, and post-gaming respiratory parameters, EE, cardiovascular parameters, as well as blood and hormonal parameters, were determined.

Respiratory parameters [oxygen uptake (V·O₂), carbon dioxide output (V·CO₂), respiratory exchange ratio (RER) and ventilation (VE)] were monitored continuously (METALYZER® 3B, CORTEX Biophysik GmbH, Leipzig, Germany) in sitting position starting 10 min before, over the entire course of the game, and until 10 min thereafter. The spirometry data were obtained breath-by-breath and were analyzed for every 5 s. Data were calculated as 5-min intervals and as 0.5-, 1-, and 2-min intervals after the start of the gaming session. When referring to these data, index entries, therefore, contain either “pre,” the minutes in the game (e.g., RER_min2), “post” or the number of minutes after the end of the game (e.g., RER_min+10). EE was calculated by converting the measured RER values to their caloric equivalent and multiplying this by the corresponding V·O₂ values.

Cardiovascular parameters [heart rate (HR), cardiac output (Q·) and stroke volume (SV)] were monitored (PhysioFlow Enduro, Manatec Biomedical, Paris, France) continuously in a sitting position starting 10 min before, over the entire course of the game, and until 10 min thereafter. The cardiovascular data were obtained and analyzed simultaneously to the respiratory parameters due to the connected measurement system (MetaSoft® Studio, CORTEX Biophysik GmbH, Leipzig, Germany).

A saliva sample was taken 15 min before and 15 min after the gaming session to determine the cortisol level (Cort). For this purpose, a cotton roll was soaked with saliva in the mouth for 1–2 min (Salivette® Cortisol, SARSTEDT AG & Co., Nümbrecht, Germany). The saliva was then isolated by centrifugation and analyzed using an enzyme-linked immunosorbent assay. Pre-gaming, 5, 10, 20, and 30 min into the gaming phase, as well as 10 min post-gaming capillary blood samples were taken from an hyperemized earlobe. These were used to determine lactate and blood glucose concentrations (BIOSEN S-Line Lab+, EKF-diagnostic GmbH, Barleben, Germany).

All statistical calculations were performed using Graphpad Prism 8.0 (GraphPad Software, US). Data are presented as the mean ± standard deviation (SD). Data were tested for normal distribution via Kolmogorov-Smirnov test. All parameters except cortisol were normally distributed. Possible changes over time were checked by an analysis of variance for repeated measurements (one-way ANOVA) with post-hoc tests (Bonferroni post-hoc test). Differences between groups at identical time points were calculated using unpaired t-tests. The significance of differences in Cortisol levels pre- and post-gaming was analyzed with a non-parametric Wilcoxon matched-pairs signed rank test. The significance level was set at p < 0.05.

CS:GO and FIFA players differed significantly in absolute and relative body fat, visceral fat, and hours of esports activity per week. No significant differences were found in age, height, body mass, lean body mass, and hours of sporting activity per week between CS:GO and FIFA players (Table 1).

Figure 1 shows the course in V·O₂, V·CO₂, RER, and EE before, during, and after the gaming session. V·O₂ and V·CO₂ remained constant over all three phases. Mean V·O₂ during the gaming session was 0.40 ± 0.06 L/min (V·O_2,pre: 0.40 ± 0.07 L/min; V·O_2,post: 0.40 ± 0.06 L/min). The RER showed a significant time effect (p < 0.001) with significant lower values in the post-phase (RER_pre: 0.87 ± 0.03, RER_min+10: 0.82 ± 0.05, p_pre;post < 0.001). In the calculated EE, no significant changes occurred (EE_pre: 1.98 ± 0.30 kcal/min; EE_Gaming: 1.99 ± 0.43 kcal/min; EE_post: 1.96 ± 0.43 kcal/min). Additionally, there were no significant differences in V·O₂, V·CO₂, RER, and EE between CS:GO and FIFA players.

As shown in Table 2, HR and Q· showed a significant time effect (p < 0.001) and decreased significantly toward the end of the game and the post-phase in relation to HR_min0.5 and Q·_min0.5. In stroke volume, no significant changes occurred. Between CS:GO and FIFA players, no significant differences in HR, Q·, and SV were detected.

Blood glucose showed a significant time effect (p < 0.05) with significantly lower values in the post-phase, while lactate remained constant over all three phases (Table 3). Cortisol levels decreased significantly from pre- to post-phase. No significant differences in blood glucose, lactate, and cortisol of CS:GO and FIFA players were found.

Looking at other sports like chess, which are mentally but not physically demanding, studies found an increased sympathetic activation for professional and amateur players (Troubat et al., 2009; Fuentes-García et al., 2019). Competitive, performance-oriented esports shows very similar tendencies: Looking at professional esports players in a tournament situation, there is an increase in HR and trends toward a heightened Q·, which implies a high sympathetic activation (Behnke et al., 2019). For example, Rudolf et al. found a slight increase in HR from 102 to 108 bpm, comparing pre to during the tournament situation. This slight increase, as well as the high HR at baseline, are likely caused by the tournament situation and not by the esports activity itself since, in the same players, a relatively low HR (80 bpm) during training of the identical game could be measured (Rudolf et al., 2016). The stress reaction during esports tournaments is partially reinforced at the hormonal level by the significantly elevated cortisol concentrations found in some studies (Schütz, 2016; Schmidt et al., 2020). However, it was not detected in other trials measuring consistent cortisol levels during competitive esports (Oxford et al., 2010; Chaput et al., 2011; Rudolf et al., 2016; Gray et al., 2018). On the other hand, non-competitive casual gaming (CG) was contrary to these results, as no changes in cardiac activity and no hormonal stress indicators could be detected as a reaction to 15–60 min of gaming (Ballard et al., 2006; Arriaga et al., 2008; Staude-Müller et al., 2008; Siervo et al., 2013, 2018).

These findings lead to the expectation that sympathetic activation as an increase in HR, Q·, SV, and cortisol could also occur in amateur esports, although to a lesser extent than in professional esports. This assumption is also supported by the case study of Haupt et al., in which a mental but no metabolic stress response was triggered in an amateur esports player (Haupt et al., 2021).

In the present study, during the course of the game and just after the end of the game, HR and Q· decreased. This suggests that the playing itself did not trigger a stress reaction. At the hormonal level, this assumption is strengthened by the significant decrease in the cortisol level. A rising cortisol level is often used as an indicator for mental stress situations, as it is needed to mobilize glucose for the skeletal musculature to prepare the body for fight-or-flight reactions (Tozman et al., 2017), which does not happen in this study. Similarly, in other competitive, sedentary situations, such as professional chess, no or only small increases in cortisol levels have been found (Tozman et al., 2017; Mendoza et al., 2020). Although the circadian rhythm could have contributed to this, the observed cortisol reduction was significantly more pronounced than what would typically occur during this period in the absence of external influences (Bailey and Heitkemper, 2001).

In physically stressful situations, blood glucose is needed to provide energy for any potential fight-or-flight reactions. During such increased metabolic demands, the release and breakdown of glucose balance each other out, resulting in approximately constant blood glucose levels (Bamberger et al., 1996). In mentally stressful situations, on the other hand, glucose concentration increases due to the elevated glycogenolysis while glucose consumption remains constant. Therefore, the significant decrease in blood glucose values from pre-gaming to 10 min after the end of the game suggests that there is no stress situation caused by playing. The lactate values, which did not change, align with our assumptions because lactate concentration is known to increase during severe stress situations (Hermann et al., 2019).

Overall, when comparing our data to data in the literature, amateur esports should be distinguished in its acute stress-related, physiological effects from professional esports with enormous sympathetic arousal, in which the players find themselves in a highly competitive as well as pressuring situation and are strongly mentally challenged as well as subject to high performance pressure. Therefore, amateur esports can be better compared with the low sympathetic activities in CG.

In studies on other cognitively oriented sports such as chess, no increased metabolic response could be observed in the professional and amateur levels. However, studies have found an increased stress response in chess players of different performance levels despite the unchanged metabolic response (Troubat et al., 2009; Fuentes-García et al., 2019). At the moment, there is no literature data in the competitive esports area for comparison, but in the CG sector, the same tendencies can be observed in the metabolic reaction, as shown in this study. In chess as well as in CG players, there were no or only minimal increases in respiration rate, V·O₂ and EE (Staude-Müller et al., 2008; Lanningham-Foster et al., 2009; Lyons et al., 2011; Barry et al., 2016). As the only study to date that investigated metabolism and EE in esports, the case study by Haupt et al. likewise did not detect an increase in metabolic response or EE. However, a major stress situation occurred during the gaming session (Haupt et al., 2021). At the metabolic level, these findings are supported by this study, as metabolism, expressed by the parameters V·O₂, V·CO₂, and RER, as well as the EE, showed no acute change due to the 30-min esports intervention.

Overall, the present study's findings suggest that in amateur esports, no acute changes in metabolism or EE occur, while HR and Q· slightly decrease. Consequently, the lower HR and Q· seem not to be coupled to the metabolic reactions, as it is mentally but not physically induced.

Performing esports at an amateur level, players of both games in this study have shown identical physiological responses. However, it is necessary to explicitly distinguish this from the professional area since skill level as well as competition situation could play a major role here and make game-specific investigations necessary.

However, there were differences between the two player groups in body composition and the amount of time spent playing per week. The CS:GO players had a higher body fat percentage than the FIFA players and spent more time playing their respective games per week. Since lean body mass and the exercise activity of both groups did not differ, this could indicate that more gaming hours per week lead to less activity in everyday life. This is supported by a study by DiFrancisco-Donoghue et al. (2020), demonstrating that collegiate esports players are significantly less active than non-esports players and have a higher body fat percentage. However, it is equally conceivable that FIFA as a sports simulation generally attracts players with an already more active lifestyle. All in all, additional research is necessary to make generally applicable statements.

The consequence of this would be that esports cannot provide the same health-promoting functions as traditional sports or presumably do not have any health-promoting effects in general. Due to the lack of any form of physical strain, neither the energy metabolism nor the cardiovascular health can be increased, nor can muscles be built, nor can disease-preventing myokines be released (Wilmore, 2003; Chen et al., 2016). Presumably, esports is a form of purely sedentary behavior, which could be associated with potentially harmful effects, such as higher risks for cardiovascular diseases, type 2 diabetes and increased all-cause mortality rates (Katzmarzyk et al., 2019; Saunders et al., 2020). In order to counteract these possible negative health consequences, deliberate physical activity and exercise routines adapted to these demands should be part of the daily life of amateur esports players. Considering the recommendations of the ACSM's guidelines for exercise testing and prescription 2021, adults should perform 150 min of moderate-intensity aerobic physical activity per week for improved health promotion and disease prevention (American College of Sports Medicine, 2021). However, since esports does not constitute aerobic physical activity, these recommendations cannot be fulfilled by esports and additional physical activity is needed to meet these requirements.

Although the power analysis requirements were met, the number of participants was nonetheless limited to n = 30, making it difficult to provide generalized conclusions. Additionally, the playing time of 30 min was relatively short and thus may only offer a brief insight into, but not a general overview of the gaming situation. With longer game durations more in line with the daily life of professional or amateur players in esports, prolonged exposure to esports could have different effects on metabolism and stress responses and therefore needs to be investigated. Moreover, only acute effects and not long-term effects were investigated by this study. Furthermore, only two games, CS:GO and FIFA, were examined. The results for other games could differ from those of this study and need to be examined individually. Also, the individual game experience of the players was only partially controlled and may have differed between participants. To determine cardiovascular responses, respectively, sympathetic activity more precisely, blood pressure measurements would have been of the highest interest. However, since conventional blood pressure measurements would have disturbed the game's flow, they were not used in this study.

The regular use of caffeinated or alcoholic products, dietary supplements, or pharmacological agents as well as preliminary fatigue or hydration status were not controlled, which could, to some extent, influence the study outcomes. However, this would probably influence all three measurement phases and keep the changes over the course of the measurement largely unaffected. In addition, changes in emotional state before, during and after the 30-min gaming session were not measured and could have influenced the course of the heart rate.

Additionally, only the weekly sporting activity was surveyed via a questionnaire. In order to be able to know the daily physical activity, it would have been necessary to measure the physical activity using an activity tracker. Moreover, it is unclear to what extent these results can be transferred to female amateur players. Further research on gender differences in the psychophysiological response to esports is needed.