Physiology & Behavior 130 (2014) 170–175

Contents lists available at ScienceDirect

Physiology & Behavior journal homepage: www.elsevier.com/locate/phb

The social environment during a post-match video presentation affects the hormonal responses and playing performance in professional male athletes Christian J. Cook a,b,c,d, Blair T. Crewther d,⁎ a

School of Sport, Health and Exercise Sciences, Bangor University, Bangor, UK United Kingdom Sports Council, London, UK c Health and Sport Portfolio, College of Engineering, Swansea University, Swansea, UK d Hamlyn Centre, Imperial College, London, UK b

H I G H L I G H T S • • • •

The social environment during a video assessment might influence athlete hormones and performance. Familiarity and body size are two factors influencing the social environment. Athlete recovery can be influenced by post-match psychological factors. Social interactions could play a broader role in moderating stress reactivity.

a r t i c l e

i n f o

Article history: Received 30 May 2013 Received in revised form 10 March 2014 Accepted 2 April 2014 Available online 12 April 2014 Keywords: Behaviour Stress Neuroendocrine Sport Recovery

a b s t r a c t This study examined the social environment effects during a post-match video presentation on the hormonal responses and match performance in professional male rugby union players. The study participants (n = 12) watched a 1-hour video of mixed content (player mistakes and successes) from a match played 1 day earlier in the presence of; (1) strangers who were bigger (SB), (2) strangers who were smaller (SS), (3) friends who were bigger (FB) and (4) friends who were smaller (FS). The salivary testosterone (T) and cortisol (C) responses to a physical stress test were assessed 3 days later, along with pre-match T levels and match-ranked performance 6–7 days later. All treatments were associated with elevated T responses (% change from baseline) to the stress test with SS N SB and FB N FS. The C stress responses after the SS and SB interventions were both greater than FS and FB. On match-day, the FB approach was linked to higher T concentrations than SB and better ranked performance than FS and SS. The subsequent testing of a population sub-group (n = 8) across a video (V) and a nonvideo (NV) presentation in a neutral social environment produced similar stress-test and performance outcomes, but pre-match T concentrations differed (V N NV). In conclusion, the presence of other males during a post-match video assessment had some influence on the hormonal responses of male athletes and match performance in the week that followed. Thus, the social environment during a post-match assessment could moderate performance and recovery in elite sport and, in a broader context, could be a possible modulator of human stress responses. © 2014 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/3.0/).

1. Introduction Much evidence has been presented on the role of testosterone (T) in regulating or supporting male social behaviours in different species, especially those relating to dominance and aggression [1–4]. The social environment appears to be one important factor influencing male T concentrations and the subsequent expression of dominance behaviours ⁎ Corresponding author at: Hamlyn Centre, Imperial College, South Kensington Campus, London SW7 2AZ, UK. Tel.: +44 20 7594 0814; fax: +44 20 7594 8904. E-mail address: [email protected] (B.T. Crewther).

relating to status-seeking motivation [1,5]. Specifically, it has been suggested that males receive various forms of sensory information during social interactions that can potentially be used to establish and maintain dominance hierarchies through these neuroendocrine signals [6]. Challenges in the social environment can modify human hormonal activity, including difficult family environments, traumatic social events, and competition [1,3,7]. Innocuous, everyday events can also induce a hormonal change. For instance, close proximity sleeping between fathers and their children can decrease the T concentrations of adult males over time [8], as does daytime father–child interactions [9]. Male T concentrations (and associated risk-taking behaviours) can

http://dx.doi.org/10.1016/j.physbeh.2014.04.001 0031-9384/© 2014 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/3.0/).

C.J. Cook, B.T. Crewther / Physiology & Behavior 130 (2014) 170–175

also acutely increase in the presence of females [10] and when interacting with adult women who are potential mates [11], but T levels may actually decrease if the woman is a conjugal partner of a close friend [11]. The presence of friends or strangers during male-to-male social interactions adds to these complexities. As an example, competing (non-physically) against either familiar men or strangers subsequently promoted different T responses in men [12] and male T levels increased when they defeated strangers, but not their friends [11]. Differences in male T concentrations have also been demonstrated when interacting socially with other males simply perceived to be similar (i.e. increasing T) or dissimilar (i.e. lowering T) [13]. Cortisol (C) may jointly work with T to moderate status-seeking behaviours [4]. Several mechanisms may explain this effect including; suppression the hypothalamic–pituitary–gonadal-axis (and T secretion), inhibition of the T actions on target tissue and/or the downregulation of the androgen receptors [14]. Similar to T, C responsiveness can vary in the presence of strangers or friends [12,15]. The acute neuroendocrine responses to social interactions have possible implications for modifying future performance and recovery. Indeed, transient changes in T and C levels have been recently linked to recovery from a competitive sport and/or subsequent match performance [16,17]. Body size is another variable to consider during social interactions. In men, greater height is associated with feelings of greater power [18] and taller men are more likely to acquire power than shorter men [19]. Humans and other animals can also express power through open, expansive postures or poses [20]. For example, men assuming a power pose exhibited rapid T increases whilst those assuming more submissive poses showed decreases in T levels [20]. Power posing also decreased C levels [20], thereby potentially magnifying any behavioural change linked to T. Thus, in some social interactions, males might be attuned to the physical size of other males, possibly with larger males exhibiting or perceived to exhibit more dominant behaviour with associated neuroendocrine responses. A post-match video presentation of an athlete or team performance is a common practice in elite sport. Video presentations can acutely modify male T concentrations [21–24] and thus, could potentially link through to changes in behaviour and short-term (b1–2 h) physical performance [17,21]. In fact, it was recently demonstrated that a psychological strategy involving a post-match video can influence these outcomes up to a week later in professional male athletes [16]. To our knowledge, no studies have examined if the social environment (i.e. presence of other males) during a post-match video presentation can also modify athlete hormones and performance on similar timescales. As the social environment is highly malleable, this would be a worthwhile addition to the field of psychobiology. We examined the social environment effects during a post-match video presentation on the subsequent hormonal responses (i.e. stress test changes, pre-match levels) and match performance in professional male athletes. To modify the social environment the video presentations were completed in the presence of; (1) strangers who were bigger (SB), (2) strangers who were smaller (SS), (3) friends who were bigger (FB) and (4) friends who were smaller (FS). Considering previous work [18–20], we hypothesised that the SS and FS interventions would produce more favourable outcomes (i.e. greater T and lower C stress responses, higher pre-match T concentrations and better match performance) [16,17,21] during the following week when compared to the SB and FB approaches. 2. Methods 2.1. Participants Twelve elite male rugby union players (6 forwards and 6 backs) playing for the same team were recruited for this study. They were all full-time athletes with between 2 and 4 years of training experience

171

in a professional environment, and were considered healthy and injury-free at the time of this study. On a weekly basis, the participants were engaged in numerous training sessions (i.e. 4–5 days per week, 1–2 h in duration) involving physical and skill conditioning, team/ positional preparation, recovery work, and they played in 1 competitive match. Each participant had a full explanation of the protocols and signed informed consent. The experimental procedures were performed with university ethics approval. 2.2. Experimental design Using a cross-over design, participants (n = 12) attended a video review 1 day after a professional rugby union match in the presence of other males (i.e. SS, SB, FS and FB interventions). To isolate the video effect from the social environment, a sub-group (n = 8) was tested a year later to compare the effects of a video (V) and a non-video (NV) intervention in a neutral social environment. The delay in the V and NV treatments was unavoidable due to the competition schedule and athlete availability. As a consequence, the data collected across the 2 experimental blocks were not directly compared. All assessments were carried out the week after each intervention; the T and C responses to a physical stress test (3 days later) and pre-match T and match-ranked performance (6–7 days later). The experimental procedures were implemented within the training schedules of participants to improve adherence and the ecological validity of the study findings. 2.3. Post-match video reviews (Block 1) The SB and SS interventions were implemented first with the FB and FS interventions completed across later games in the same season. Participants reported to the training facility 24 h after each match, where they were randomly assigned to 1 of 2 groups of equal size. Each group completed a different intervention (all 1 h in duration) and the 2 groups crossed over on the subsequent week to complete the corresponding intervention. The mean age, body mass and height of participants were 20.0 ± 0.7 years, 95.7 ± 9.5 kg and 1.85 ± 0.06 m, respectively. Testing in this block was conducted as follows: 1. SB intervention: Both groups saw video footage of the match played the previous day, which consisted of mixed content of player mistakes and successes. The video footage was selected by the coaching staff, based on the specific requirements of this study, and formatted by the team analyst to be played on a large video screen. During the watching of these videos, each group was accompanied by an equal number of male strangers whom they were told were there to simply observe and learn. The group of strangers was also rugby players, but they were chosen to be significantly (P b 0.01) taller (1.91 ± 0.03 m) and heavier (105.3 ± 6.7 kg) than the study population. 2. SS intervention: Both groups watched the same video footage described above, but in the presence of another group of male strangers. This group was also rugby players and chosen to be significantly (P b 0.05) shorter (1.78 ± 0.05 m) and lighter (88.0 ± 6.0 kg) than the assessed group. In both the SB and SS treatments, the unknown males made no comments concerning the video presentations, but they did interact in a social manner. The discussions between these athlete groups were not scripted in anyway, but the strangers were instructed to maintain a friendly conversation at all times. 3. FB intervention: Participants watched selected video footage of the previous match played, at which time they were accompanied by an equal number of male friends known to the study participants for more than two years and who were also rugby players. The group of friends was chosen to be significantly (P b 0.01) taller (1.91 ± 0.02 m) and heavier (115.3 ± 5.2 kg) than the assessed group of players.

172

C.J. Cook, B.T. Crewther / Physiology & Behavior 130 (2014) 170–175

4. FS intervention: Participants watched the same video footage as above in the presence of another group of friends, but this group was selected to be significantly (P b 0.05) lighter (87.0 ± 4.9 kg) than the study population. The group of friends also tended to be shorter (1.81 ± 0.02 m) than the study participants (P = 0.09). Once again, both groups of friends made no comments concerning the video presentations and they interacted with the study participants in a friendly, social manner. 2.4. Post-match video reviews (Block 2) The V and NV interventions were implemented a season later. Again, the participants reported to the training facility 1 day after each match, where they were randomly assigned to 1 of 2 groups of equal size. Each group completed one intervention (both 1 h in duration) and they crossed over the following week. Testing was as follows: 1. V intervention: Participants watched selected video footage of the previous match played, as described in testing block 1, but this intervention was completed in a neutral social environment without the presence of any friends or strangers. 2. NV intervention: Participants sat quietly in the room in front of the video screen (no video was played) in a neutral social environment without any friends or strangers present. To eliminate other possible confounders of the social environment, all of the video reviews described were performed without coach presence. However, being an in-season professional sporting environment there were practicalities that had to be adhered to. The participants maintained their normal training schedules although training volume and intensity were kept relatively consistent to ensure participants were exposed to the same physical stressors across each week of testing. The pre- and post-match procedures were also kept consistent in terms of sleep (N 7 h), the time of waking and nutritional intake. Participants were assessed after a consistent consumption of breakfast between 0800 and 0830 h, which generally comprised of cereals, yoghurt, toast, eggs, fruit juice and/or water. Participants were also instructed to avoid alcohol consumption 2 days before mid-week and game-day testing. 2.5. Mid-week assessments The stress test was completed 3 days after each intervention to assess hormonal reactivity to a physical challenge. The test itself was completed at the team training facility using published protocols [16], being a series of resistance exercises (i.e. 3 sets of power cleans; 3 sets of back squats and 3 sets of bench presses) and each exercise was performed for 5 repetitions using a relative load of 40%, 60% and 70% of individual 1 repetition maximum. Rest periods of 1 and 3 min were used between sets and exercises, respectively. The loading protocols were consistent throughout this study. Exercise duration was approximately 20 min and testing was completed in a group, as part of normal training procedures. Participants were assessed at the same time of day (1100 h) to control for daily variation in hormones and physical performance [25]. A saliva sample was collected from participants 5 min before, and 5 min after the stress test. 2.6. Match-day assessments The rugby union matches were played 6 to 7 days after the postmatch video presentations. These matches were played at different venues on a home (n = 4) and away (n = 2) basis, but they all started at a similar time of day (1430–1500 h). The 4 home games were won and the 2 away games produced a loss and a draw. Approximately 2 h before each match began, participants arrived at the playing venue to begin technical and tactical preparation, followed by a standard warm-up, and a final team debrief 20 min before the match started. A

saliva sample was taken 40 min before kick-off. Player performance was ranked across the matches played in testing block 1 (where 1 = best performance to 4 = worst performance) and block 2 (where 1 = better performance to 2 = worse performance). The match rankings were determined by 2 members of the coaching staff (i.e. head coach and assistant coach), based on several key skills identified for each playing position (e.g. for a forward lineout wins and ruck clears) [16, 17], after individual scores were combined and then averaged. Performance rankings for each participant were conducted only after all matches were played, so that each match was ranked relative to each other.

2.7. Salivary hormone assessment The saliva sampling procedures employed are consistent with other research [16,21,26,27]. Briefly, saliva was collected by passive drool into sterile containers, approximately 2 ml over a timed collection period of 2 min. The samples were subsequently stored at −30° C until assay. No food was taken 90 min before the first saliva sample to reduce the effect of food intake on salivary hormone concentrations [28]. After thawing and centrifugation (2000 rpm × 10 min), the saliva samples were analysed in duplicate for T and C concentrations using commercial kits (Salimetrics LLC, USA) and the manufacturers' guidelines. Pre-match C was not measured in this study. The minimum detection limit for the T assay was 6.1 pg/ml with inter-assay coefficients of variation (CV) of b9.7%. The C assay had a detection limit of 0.12 ng/ml with inter-assay CV of b 8.1%.

2.8. Statistical analyses Before analyses all variables were tested for normality. The stresstest changes in T and C concentrations (% change from pre to post) were assessed within each treatment using paired t-tests [16,17,27]. The analysis of percent (%) changes allowed comparisons between the hormonal markers and across studies using similar outcomes. The changes in T and C were compared across each treatment using a 1way analysis of variance (ANOVA) with repeated measures. The treatment effect on pre-match T concentrations was also examined using a 1-way ANOVA with repeated measures. Where appropriate, post hoc testing was conducted using paired t-tests with Bonferroni corrections. Match-ranked performance was compared using the Friedman test for non-parametric data and post hoc testing was conducted using the Wilcoxon Signed-Rank test with Bonferroni corrections. The effects of the V and NV interventions on the hormonal and performance outcomes were examined using the same statistics described above. Significance for all analyses was set at P ≤ 0.05.

Table 1 Salivary testosterone responses to the physical stress test following the post-match interventions in different social environments (Mean ± SD).

Pre testosterone (pg/ml) Post testosterone (pg/ml) % change

M SD M SD M SD

SS

SB

FS

FB

125 21 188 21 51.8⁎,#,α

126 24 153 28 22.3⁎,α

135 22 146 21 8.6⁎

137 23 165 26 21.4⁎,α

14.2

13.7

5.0

7.5

SS = strangers who were smaller; SB = strangers who were bigger; FS = friends who were smaller; FB = friends who were bigger. ⁎ Significant from baseline P b 0.001. # Significant from SB and FB P b 0.01. α Significant from FS P b 0.05.

C.J. Cook, B.T. Crewther / Physiology & Behavior 130 (2014) 170–175

173

3. Results 3.1. Hormonal stress-test responses following the social environment interventions The salivary T responses to the physical stressor were significantly elevated within each intervention (P b 0.001, Table 1). A significant treatment effect on the % changes in T concentrations was also identified (F3, 33 = 33.91, P b 0.001). The SS approach produced a significantly greater T response (51.8%) than SB (22.3%) and FB (21.4%), whilst the SB, SS and FB responses were all superior to FS (8.6%). In response to the stress test, salivary C was significantly elevated in the SS (16.0%) and SB (26.5%) interventions, whereas the FS (− 5.8%) and FB (− 9.3%) approaches both produced negative responses (P b 0.001, Table 2). Significant treatment effects were also identified for the % changes in C concentrations (F3, 33 = 102.1, P b 0.0001). The C response following SB was significantly greater than SS and both outcomes differed from FS and FB (P b 0.01).

Fig. 1. Pre-match salivary testosterone concentrations following the post-match interventions in different social environments (Mean ± SD). SS = strangers who were smaller; SB = strangers who were bigger; FS = friends who were smaller; FB = friends who were bigger. αSignificant from SB P b 0.01.

4. Discussion 3.2. Pre-match testosterone following the social environment interventions Analysis of the different social environment interventions (Fig. 1) revealed a significant effect on pre-match salivary T concentrations (F3, 33 = 5.34, P = 0.004) with the T response following the FB protocol being significantly higher than the SB intervention. No other significant differences were identified between treatments.

3.3. Match-ranked performance following the social environment interventions A significant treatment effect on match-ranked performance was identified, x2(3) = 27.4, P b 0.001 (Fig. 2). The SB, FS and FB interventions all produced significantly better outcomes (i.e. lower-ranked performance) than SS, with the FB approach also associated with a better performance outcome than FS (P = 0.005).

3.4. Hormonal and performance outcomes following the neutral social environment interventions The NV and V interventions were associated with elevated T (22.6% and 23.9%) and C (6.7% and 3.5%) stress-test responses, respectively (P b 0.01, Table 3). The T responses were no different between treatments, but the C response was marginally higher for NV (P = 0.05). Pre-match T concentrations were found to be significantly higher in the V intervention (P = 0.015, Fig. 3), but match-ranked performance was no different (P = 0.48, Fig. 4) between the NV (1.6 ± 0.5) and V (1.4 ± 0.5) treatments. Playing at home or away venues did not influence the match-day T outcomes (P = 0.604), but all of the home games were won and the away games produced a loss and a draw.

We recently demonstrated the effects of post-match video presentations on the hormonal stress-test responses and competitive match performance in professional male athletes [16]. This study added the dimension of a social environment involving the presence of other males who differed in familiarity and/or body size to the study population. All treatments were associated with elevated T responses to a physical stress test (3 days later) but of varying magnitudes, whereas the C responses varied by magnitude and direction. On match day (6–7 days later), the FB intervention was associated with higher prematch T concentrations and better match-ranked outcomes than the SB, SS and/or FS treatments. The T responses to the stress test differed in magnitude across each intervention (SS N SB and FB N FS), as did C responsiveness (SB N SS N FS and FB). Whilst our original hypotheses regarding these outcomes were partly correct, they were clearly over simplified. We do recognise that any translational effect might additionally depend on the environmental and situational conditions during testing. These outcomes might also reflect exposure to other stressors in the interim period, such as critical feedback and social interactions with coaching staff or more senior players within a team, as well as other life stressors (e.g. emotional, financial). Nevertheless, it does appear that the hormonal stress responses of male athletes are highly malleable and influenced by prior exposure to different psychological and social stressors [16,17], similar to that reported in other species [6,29–31]. Collectively, these data suggest that social interactions can promote some neuroendocrine responses that are linked to future stress reactivity.

Table 2 Salivary cortisol responses to the physical stress test following the post-match interventions in different social environments (mean ± SD).

Pre cortisol (ng/ml) Post cortisol (ng/ml) % change

M SD M SD M SD

SS

SB

FS

FB

2.54 0.42 2.92 0.37 16.0*α 7.1

2.43 0.60 3.06 0.71 26.5*#α 7.1

2.53 0.40 2.39 0.39 −5.8* 2.8

2.64 0.42 2.40 0.43 −9.3* 4.1

SS = strangers who were smaller; SB = strangers who were bigger; FS = friends who were smaller; FB = friends who were bigger.

Fig. 2. Match-ranked performance following the post-match interventions in different social environments (Mean ± SD). SS = strangers who were smaller; SB = strangers who were bigger; FS = friends who were smaller; FB = friends who were bigger. αSignificant # from SB, FS and FB P ≤ 0.008, Significant from FB P = 0.005.

174

C.J. Cook, B.T. Crewther / Physiology & Behavior 130 (2014) 170–175

Table 3 Salivary testosterone and cortisol responses to the physical stress test following the postmatch interventions in neutral social environments (Mean ± SD).

Pre Post % change

M SD M SD M SD

Testosterone (pg/ml)

Cortisol (ng/ml)

NV

V

NV

V

141 29 172 32 22.6⁎

154 27 188 20 23.9⁎

2.91 0.23 3.10 0.29 6.7⁎,#

2.82 0.25 2.92 0.28 3.5⁎

4.0

2.8

8.2

9.9

NV = no video; V = video. ⁎ Significant from baseline P b 0.01. # Significant from V P = 0.05. Fig. 4. Match-ranked performance following the post-match interventions in neutral social environments (mean ± SD).

In general, the presence of strangers (SS and SB interventions) in the post-match video presentations appeared to promote greater hormonal stress responses some days later, which could be linked to a larger neuroendocrine response across each intervention. Studies have reported exaggerated hormonal, physiological or emotional responses to a psychological stressor in the presence of strangers, as opposed to friends or partners [15,32–34], and this could be conceivably likened to the post-match reviewing of athlete performance. Male T responses can also differ when competing non-physically against either friends or within-group coalitions versus strangers or between-group coalitions [11,12,35]. However, due to practical and financial constraints no hormonal data were taken to acutely profile the post-match interventions. This would be a valuable addition in future research. We also hypothesised that bigger males would pose a threat (and a more domineering stimulus) during the video presentations, thereby negatively influencing the match-day outcomes. However, the FB treatment was associated with higher pre-match T levels than SB and achieved better performance outcomes than FS (and SS). Our athlete group by nature of their sport engages in combat situations with other very sizeable males. Thus, it is conceivable that the size of other social members can be seen as a positive challenge, rather than a threat, producing a favourable stress response that may not be seen in other groups. Perhaps bigger males can also be perceived as being supportive and non-threatening in some situations, such as conversing in a friendly manner during the video presentations. Indeed, adults supported by strangers during a psychological challenge demonstrated attenuated hormonal and metabolic responses from those receiving no social support at all [15,34]. The pre-match elevation in T levels might help to explain the relative match success after the FB intervention, given that T has been linked to various behaviours (e.g. risk-taking, motivation, willingness to perform) [10,26,36–38] that are arguably important for competitive sport. In

Fig. 3. Pre-match salivary testosterone concentrations following the post-match interventions in neutral social environments (Mean ± SD). αSignificant from V P b 0.05.

other studies, higher T levels were associated with superior physical performance [38,39] and the video-induced changes in T correlated strongly and positively to individual voluntary strength [21]. Similarly, larger T increases in a soccer match were associated with better selfrated performance [40] and our recent work [41] suggests that a higher T response to a mid-week physical stressor can help to predict better game outcomes in professional rugby league. In the financial domain, T measurements are also related to motivational behaviours and financial performance later in the day [42,43]. Although some treatment differences in the match-day results were noted, these outcomes did not reflect the stress-test results from earlier in the week. This differs somewhat from previous work on elite male athletes [17,21,38,39]. These findings could be explained by the 3month period separating the stranger and friend interventions. Team strategies and playing styles could also influence any hormonal links to match performance, along with motivational factors related to playing at home or away venues [44], although our T results did not support a venue effect. The monitoring of game-day C levels would have provided valuable information on player perception of psychological stress under competitive conditions and additional interactions with T. Little research on humans has examined the body size effect in social interactions on subsequent recovery and performance, but in a social context greater height is associated with feelings of greater power in humans [18] and taller men are more likely to acquire power [19]. Powerful humans can in fact overestimate their height [18], perhaps to exaggerate this effect, and men assuming a power pose exhibit produced rapid T increases and C decreases, whilst the opposite effect was noted with more submissive poses [20]. Speculatively, larger males might exhibit, or at least are perceived to exhibit, threatening or more dominant behaviours towards smaller males and this could promote further hormonal changes to maintain social status [1]. However, in humans, this response is likely to be far more complex than body size per se, for example, a supportive bigger male may be viewed as a powerful ally rather than a threat, and indeed our results tend to add to this complexity. The V and NV interventions conducted in the neutral social settings promoted largely similar stress-test and performance outcomes, so the previous observations around the different treatments may be due to the social environment rather than simply getting together and watching a post-game video review. However, pre-match T concentrations were significantly different (V N NV), which suggests that the video itself may have some small independent effect going forward. Although not compared statistically, the V and NV hormone results were centrally distributed within the other treatment results. Further work incorporating all of these variables into one consecutive study would be useful, but in a professional sport setting, it is hard to achieve from the point of view of compliance. Other study limitations include the small sample size, the video content used in relation to each match played (i.e. showing a mixture of home and away games, with team wins and losses), and possible bias

C.J. Cook, B.T. Crewther / Physiology & Behavior 130 (2014) 170–175

associated with the ranking of matches. It is also important to note that team video sessions are not normally accompanied by non-team members and with coaching staff absent. Moreover, the environmental aspects of each assessment might have influenced athlete hormonal state (e.g. the presence of other team members during stress testing) and/or match performance (e.g. the rank and status of the opposing team). Still, our results highlight new pathways for assessing the physiological impact of psychological and social strategies in sport, and a novel approach for priming athlete performance, as well as suggesting interesting social environment influences on stress reactivity. 5. Conclusions Exposing male athletes to different social environments during a post-match video presentation had some influence on their hormonal stress responses, pre-match T concentrations and match-ranked performance during the following week. Thus, the social environment appears to be an important aspect of performance and recovery in elite sport and, in a broader context, a possible modulator of stress responsiveness in humans. Acknowledgments We would like to thank the athletes and coaching staff that contributed to this study. This project was partly supported by the UK Sports Council and the Engineering and Physical Sciences Research Council UK, as part of the Elite Sport Performance Research in Training (ESPRIT) with Pervasive Sensing Programme [EP/H009744/1]. References [1] Mazur A, Booth A. Testosterone and dominance in men. Behav Brain Sci 1998;21:353–97. [2] Wingfield JC, Hegner RE, Dufty AM, Ball GF. The “challenge hypothesis”: Theoretical implications for patterns of testosterone secretion, mating systems, and breeding strategies. Am Nat 1990;136:829–46. [3] Archer J. Testosterone and human aggression: An evaluation of the challenge hypothesis. Neurosci Biobehav Rev 2006;30:319–45. [4] Mehta PH, Josephs RA. Testosterone and cortisol jointly regulate dominance: Evidence for a dual-hormone hypothesis. Horm Behav 2010;58:898–906. [5] Mehta PH, Josephs RA. Social endocrinology: Hormones and social motivation. In: Dunning D, editor. Social motivation. New York: Psychology Press; 2011. p. 171–89. [6] Chen CC, Fernald RD. Visual information alone changes behavior and physiology during social interactions in a Cichlid fish (Astatotilapia burtoni). PLoS One 2011;6:1–12. [7] Flinn MV, Nepomnaschy PA, Muehlenbein MP, Ponzi D. Evolutionary functions of early social modulation of hypothalamic–pituitary–adrenal axis development in humans. Neurosci Biobehav Rev 2011;35:1611–29. [8] Gettler LT, McKenna JJ, McDade TW, Agustin SS, Kuzawa CW. Does cosleeping contribute to lower testosterone levels in fathers? Evidence from the Philippines. PLoS One 2012;7:e41559. [9] Gettler LT, McDade TW, Feranil AB, Kuzawa CW. Longitudinal evidence that fatherhood decreases testosterone in human males. Proc Natl Acad Sci U S A 2011;108:16194–9. [10] Ronay R, von Hippel W. The presence of an attractive woman elevates testosterone and physical risk taking in young men. Soc Psychol Personal Sci 2010;1:57–64. [11] Flinn MV, Ponzi D, Muehlenbein MP. Hormonal mechanisms for regulation of aggression in human coalitions. Hum Nat 2012;23:68–88. [12] Wagner JD, Flinn MV, England BG. Hormonal response to competition among male coalitions. Evol Hum Behav 2002;23:437–42. [13] DeSoto CM, Hitlan RT, Deol RS, McAdams D. Testosterone fluctuations in young men: The difference between interacting with like and not-like others. Evol Psychol 2009;8:173–88. [14] Liening SH, Josephs RA. It is not just about testosterone: Physiological mediators and moderators of testosterone's behavioral effects. Soc Pers Psychol Compass 2010;4:982–94.

175

[15] Kirschbaum C, Klauer T, Filipp S, Hellhammer DH. Sex-specific effects of social support on cortisol and subjective responses to acute psychological stress. Psychosom Med 1995;57:23–31. [16] Crewther BT, Cook CJ. Effects of different post-match recovery interventions on subsequent athlete hormonal state and game performance. Physiol Behav 2012;106:471–5. [17] Cook CJ, Crewther BT. The effects of different pre-game motivational interventions on athlete free hormonal state and subsequent performance in professional rugby union matches. Physiol Behav 2012;106:683–8. [18] Duguid MM, Goncalo JA. Living large: The powerful overestimate their own height. Psychol Sci 2011;23. http://dx.doi.org/10.1177/0956797611422915. [19] Kortt M, Leigh A. Does size matter in Australia? Econ Rec 2010;86:71–83. [20] Carney DR, Cuddy AJ, Yap AJ. Power posing: brief nonverbal displays affect neuroendocrine levels and risk tolerance. Psychol Sci 2010;21:1363–8. [21] Cook CJ, Crewther BT. Changes in salivary testosterone concentrations and subsequent voluntary squat performance following the presentation of short video clips. Horm Behav 2012;61:17–22. [22] Hellhammer DH, Hubert W, Schürmeyer T. Changes in saliva testosterone after psychological stimulation in men. Psychoneuroendocrinology 1985;10:77–81. [23] Carré JM, Putnam SK. Watching a previous victory produces an increase in testosterone among elite hockey players. Psychoneuroendocrinology 2010;35:475–9. [24] Stoléru SG, Ennaji A, Cournot A, Spira A. LH pulsatile secretion and testosterone blood levels are influenced by sexual arousal in human males. Psychoneuroendocrinology 1993;18:205–18. [25] Sedliak M, Finni T, Cheng S, Kraemer WJ, Häkkinen K. Effect of time-of-day-specific strength training on serum hormone concentrations and isometric strength in men. Chronobiol Int 2007;24:1159–77. [26] Carré JM, McCormick CM. Aggressive behavior and change in salivary testosterone concentrations predict willingness to engage in a competitive task. Horm Behav 2008;54:403–9. [27] Edwards DA, Wetzel K, Wyner DR. Intercollegiate soccer: Saliva cortisol and testosterone are elevated during competition, and testosterone is related to status and social connectedness with teammates. Physiol Behav 2006;87:135–43. [28] Gibson LE, Checkley S, Papadopoulos A, Poon L, Daley S, Wardle J. Increased salivary cortisol reliable induced by a protein-rich midday meal. Psychosom Med 1999;61:214–24. [29] Lürzel S, Kaiser S, Sachser N. Social interaction decreases stress responsiveness during adolescence. Psychoneuroendocrinology 2011;36:1370–7. [30] Lürzel S, Kaiser S, Sachser N. Social interaction, testosterone, and stress responsiveness during adolescence. Physiol Behav 2010;99:40–6. [31] Ziegler TE, Schultz-Darken NJ, Scott JJ, Snowdon CT, Ferris CF. Neuroendocrine response to female ovulatory odors depends upon social condition in male common marmosets, Callithrix jacchus. Horm Behav 2005;47:56–64. [32] Phillips AC, Gallagher S, Carroll D. Social support, social intimacy, and cardiovascular reactions to acute psychological stress. Ann Behav Med 2009;37:38–45. [33] Kissel S. Stress-reducing properties of social stimuli. J Pers Soc Psychol 1965;2:378–84. [34] Christenfeld N, Gerin W, Linden W, Sanders M, Mathur J, Deich JD, et al. Social support effects on cardiovascular reactivity: Is a stranger as effective as a friend? Psychosom Med 1997;59:388–98. [35] Oxford J, Ponzi D, Geary DC. Hormonal responses differ when playing violent video games against an ingroup and outgroup. Evol Hum Behav 2010;31:201–9. [36] Mehta PH, Josephs RA. Testosterone change after losing predicts the decision to compete again. Horm Behav 2006;50:684–92. [37] Aarts H, van Honk J. Testosterone and unconscious positive priming increase human motivation separately. NeuroReport 2009;20:1300–3. [38] Crewther BT, Cook CJ, Gaviglio CM, Kilduff LP, Drawer S. Baseline strength can influence the ability of salivary free testosterone to predict squat and sprinting performance. J Strength Cond Res 2012;26:261–8. [39] Cardinale M, Stone MH. Is testosterone influencing explosive performance? J Strength Cond Res 2006;20:103–7. [40] Trumble BC, Cummings D, von Rueden C, O'Connor KA, Smith EA, Gurven M, et al. Physical competition increases testosterone among Amazonian foragerhorticulturalists: A test of the ‘challenge hypothesis’. Proc R Soc B 2012;279. http://dx.doi.org/10.1098/rspb.2012.0455. [41] Crewther BT, Sanctuary CE, Kilduff LP, Carruthers JS, Gaviglio CM, Cook CJ. The workout responses of salivary-free testosterone and cortisol concentrations and their association with the subsequent competition outcomes in professional rugby league. J Strength Cond Res 2013;27:471–6. [42] Coates JM, Gurnell M, Sarnyai Z. From molecule to market: steroid hormones and financial risk-taking. Philos Trans R Soc Biol Sci 2010;365:1363–8. [43] Coates JM, Herbert J. Endogenous steroids and financial risk taking on a London trading floor. Proc Natl Acad Sci U S A 2008;105:6167–72. [44] Nevill A, Holder RL. Home advantage in sport: An overview of studies on the advantage of playing at home. Sports Med 1999;28:221–36.