Person. individ Dr;ff-Vol. 23,No.& pp. 1031-1039, 1997 0 1997ElsevierScienceLtd. All rights reserved
Pergamon
Printed in Great Britain
PII: SO191-8869(97)00125-6
0191-8869/97%17.00+0.00
INDIVIDUAL TEMPORAL DIFFERENCES IN PRECOMPETITION ANXIETY AND HORMONAL CONCENTRATION Martin Eubank,‘* Dave Collins,’ Geoff Lovell,’ Debra Dorling3*4and Steve Talbot3 ‘Department of Exercise and Sport Science, Crewe+Alsager Faculty, The Manchester Metropolitan University, England, *Department of Sports Studies, Roehampton Institute, England, ‘Department of Biology, St Mary’s University College, England and 4Department of Health and Paramedical Studies, Brunel University College, England (Received 23 January 1997)
Summary-This investigation explored individual differences in hormonal concentration and directional perception of anxiety 24 hrs, 2 hrs and 1 hr before an important competition. Those who perceived their anxiety level as positive to performance (i.e. facilitators) showed an increase in testosterone concentration as time-to-event approached that contrasted markedly with those who perceived their anxiety intensity as negative to performance (i.e. debilitators). The facilitatory group also exhibited a lower cortisol response 2 hrs and 1 hr, relative to 24 hrs before the event, whilst the response of the debihtatory group was elevated as time-to-event approached. Finally, the facilitatory group evidenced a trend towards rapid elevations in all of the catecholamines, whilst the concentration amongst the debihtatory group remained stable as the event approached. The findings add further substance to the veracity of the directional construct in acting as the important distinguishing variable which accounts for significant individual differences. Additionally it addresses the mechanism by which the relationship between anxiety direction and performance may be brought about. Such underpinning biochemical mechanisms offer one possible yet long awaited explanation to aid our understanding of the pre-event anxiety direction response of competitive individuals. 0 1997 Elsevier Science Ltd. All rights reserved Key- Words: anxiety direction, temporal patterning, hormonal concentration
INTRODUCTION Recent research examinations of pre-competitive anxiety (Jones & Cale, 1989; Jones, Swain & Cale, 1991; Krane & Williams, 1987) have indicated that the patterning of multidimensional anxiety as time-to-event approaches is a function of individual difference, with skill level being one such confounding factor. It would appear that elite individuals exhibit a consistently different temporal patterning response than non-elite performers as time-to-critical event approaches. Of particular interest is the work of Jones, Hanton and Swain (1994), who observed elite performers to interpret both cognitive and somatic state anxiety intensity as being more facilitative (i.e. positively perceived anxiety affect) to performance than their non-elite counterparts, despite no individual difference between the two skill groups in anxiety intensity (i.e. the level of anxiety experienced). Despite these positive steps to ‘facilitate’ the measurement of the anxiety response, it is unfortunate that research findings are still only able to, at best, hypothesise about the influential nature of anxiety direction and it’s role as an individual difference variable. The scope of self-report data infers an examination of a hypothetical construct, due in the main to an absence of understanding on the underlying mechanisms combined with, in this instance, different perceptions of anxiety. Ultimately, this understanding may be enhanced if individual differences in the perception of anxiety are mirrored by differences in physiological markers, for example hormonal concentration. Consequently, research protocols would benefit greatly from the utilisation of other alternative measures that index the psychological emotions of performers. Stressful environments inflict a change on the concentration of biochemical substrates such as catecholamines, cortisol and testosterone. The work of Dienstbier (1989) hypothesises that changes *To whom all correspondence should be addressed at: The Manchester Metropolitan University, Crewe+Alsager Faculty, Alsager Campus, Hassall Road, Alsager, Stoke-on-Trent, ST7 2HL, England. Tel: 44-161-247-5573; Fax: 44-161-2476375; e-mail:
[email protected]. 1031
1032
Martin Eubank et ai.
in hormonal concentration act as a much sought-after biochemical mechanism of how individual differences in anxiety perception occur, as “catecholamine and cortisol responsivity is manifest in response to psychological challenge/stress situations” (Dienstbier, 1989; p. 91). Of crucial importance to this mechanism is the dichotomous conceptualisation of stress, where the term stress refers only to perceptions that are emotionally negative. A further component of stress, labelled as challenge, refers to that aspect of the interaction that results in positive emotion. Dienstbier states that catecholamines never seem to be too high for optimal arousal. This serves to indicate that the ideal catecholamine pattern for a positive adaptive response results from a low base rate with a fast and strong increase, leading to subsequent arousal generation in response to perceived challenge. In perceived stress conditions the catecholamine pattern will be depressed and fail to exhibit any such sudden hormonal and arousal generation. In contrast, it is suggested that the ideal pattern of cortisol in conditions of challenge is one where the base rate is low and the response is depressed and delayed, whereas elevated levels of cortisol are induced by conditions of perceived stress. In addition continual temporal elevations in testosterone in response to stressful events mirrors the constructive process of chemical change associated with testosterone build-up. Considering the recent anxiety research (Eubank, Smith & Smethurst, 1995; Jones et al., 1994; Jones & Swain, 1995) that has indicated the potential importance of anxiety affect, it may be hypothesised, in association with Dienstbier’s predictions, that as competition approaches, positive anxiety affect responses and perceptions of positive arousal are associated with lower cortisol concentrations but continual temporal elevations in catecholamine and testosterone concentration. Consequently, this study sought to explore the differences in the directional perception of anxiety and temporal fluctuations in hormonal concentration. Following the distinct nature of the work to date involving the perceived influence of anxiety and the temporal patterning of hormonal concentration, it was hypothesised that: l
Facilitators would exhibit an increase in testosterone concentration, a depressed cortisol response and a rapid increase in catecholamine concentration as time-to-event approached.
l
In contrast, debilitators would show the opposite pattern, namely a decrease in testosterone concentration, an elevated cortisol response and no observable rapid increase in catecholamine concentration as time-to-event approached. METHOD
Subjects
The subject sample consisted of 10 elite male marathon canoeists, aged between 19 and 28yrs (mean+S.D. =23.80+2.86yrs). The events used were two races held in the summer of 1994 for selection to the 1995 world championships, and thus were extremely important to the competitors. Each race was a ‘K2’ marathon race, i.e. with two canoeists per boat carried out over a distance of 42 kilometres (equivalent to a marathon distance of 26 miles). Instrumentation Psychological measures. State anxiety was evaluated by the Competitive State Anxiety Inventory2 (CSAI-2; Martens, Burton, Vealey, Bump & Smith, 1990). The inventory is comprised of 27 items measuring cognitive anxiety, somatic anxiety and self-confidence intensity. Each item is scored on a Likert scale ranging from 1 “not at all” to 4 “very much so”, with intensity scores ranging from a low of 9 to a high of 36. Each intensity item was modified to include a direction scale, with possible direction scores ranging from -3 “very debilitative” to f3 “very facilitative” with a low of -27 to a high of +27, as per Jones and Swain (1992). Hormonal measures. To measure hormone levels, subjects were required to empty their bladders at each data collection period, and not at any other time during the data collection period on the actual day of the race. 15 ml of urine was collected to establish catecholamine concentrations in each subject at each time period. 0.5 mls of 6 mol hydrochloric acid was added to the sample to ensure that a pH of between 1.0 and 3.0 was maintained, thus preserving catecholamines and deterring bacterial growth. The samples were then frozen at -20°C. Analysis required the urine
Individual temporal differences in pre-competition anxiety and hormonal concentration
1033
samples to undergo a ‘cleanup’ procedure using Ion Exchange Chromatography. Following this, catecholamine were assayed using High Performance Liquid Chromatography, with the use of relevant reagents to correct for creatinine, so as to account for any differences due to dehydration. Concentrations for epinephrine, norepinephrine and dopamine were determined. To measure testosterone and cortisol concentration, lOm1 of blood was drawn from the antecubital vein of the left or right arm by venepuncture at each time period. Samples were allowed to clot and were then put on ice. The samples were then centrifuged at 3000 r.p.m. for 10 min, following which the serum was stored at a temperature of -20°C for a period of two months. Cortisol and testosterone were subsequently assayed by the Coat A Count method. To account for the extent of dehydration during the race and the potential effect on cortisol and testosterone concentrations, haematocrit levels were also analysed. Results were subsequently corrected for any small differences in dehydration where necessary. The cortisol sample taken 24 hrs before the race was obtained at a time equivalent to the start time of 2.00 p.m. for the following day’s race. This was done to control for the profound circadian rhythm of cortisol, which is higher ante compared to post midday. Procedures
Ethical approval was obtained for the investigation and an informed consent form was completed by each subject. Hormonal concentrations for each subject were evaluated 2-3 weeks prior to the event to form a baseline measure. This was implemented to evaluate and therefore control for any extraneous variability that may have resulted from between-subject training and nutritional differences. Data collection for the event required each subject to complete the directionally modified CSAI-2 and provide urine and blood samples for subsequent cortisol, catecholamine and testosterone analysis all at one day, 2 hrs (i.e. midday) and 1 hr (i.e. 1.OOp.m.) prior to the start of the race. A control subject was employed who was also a male marathon canoeist of similar age but who was not taking part in the race to obtain hormonal data at equivalent times throughout the testing period. Subjects were ‘re-stuck’ 1 hr prior to the event and were aware that this was required. The investigators were alert to the potential influence of blood drawing on anxiety levels. However, all subjects reported that they were not concerned about the procedure as they had previously experienced blood letting and had every confidence in the experience of the qualified phlebotomist. The subject population was split into a ‘facilitatory’ and ‘debilitatory’ group, founded on the basis of each subjects CSAI-2 direction score and following the procedures employed by Jones et al. (1994). RESULTS
Baseline values
A major strength of the study was its utilisation of a genuinely elite group in a real life ecologically valid environment which is, as Jones and Hardy (1990) suggest, the most attractive location for the study of competitive stress, thus facilitating the study of naturally occurring rather than laboratory manipulated states. However, such strengths limit the potential to control for extraneous variability, and unfortunately the collection of effective baseline data proved to be problematic. Logistically it proved impossible to obtain baselines from three out of the 10 subjects. However, it was considered that the data were in any case flawed due to lack of control over training intensity and patterns, which could have masked any baseline level of hormonal concentration that was being sought. In addition the absence of control over the food intake of the subjects may have added to the confounding effects. Such factors were confirmed by the self-report measures implemented to detect confounding sources of such extraneous variability in the biochemical measures. Whilst on the basis of these problems the baseline data were unfortunately rejected, these same self-report measures demonstrated that subsequent findings were free of these confounds, and include significant hormonal effects, despite the lack of allowance for individual differences. Anxiety/hormone
variable analysis
A 2 x 3 (group x time period) multivariate analysis of variance (MANOVA) was used to examine differences in hormonal concentration which occurred as a function of the approaching time-to-
Martin Eubank et al.
1034
Table I. Mean group state anxiety responses as time-to-event
Anxiety variable Cognitive
intensity
Somatic intensity
W&confidence Intensity Cognitive
direction
Somatic direction
Self-confidence Direction
Time-to-event
variable
Testosterone
Cortisol
Epinephrine
Norepinephrine
Dopamine
17.7k4.96 17.8k4.53 17.2+4.71 11.7k2.42 12.0&-2.83 14.0+4.73 28.8 f4.58 28.0 f 5.52 27.8 k5.08 7.83f2.31 9.60k3.96 9.5Ok3.28 9.0 f 2.46 10.0*3.40 8.83k2.68 15.3k4.80 18.6f4.50 17.2k4.97
24hr 2hr 1 hr 24 hr 2hr 1 hr 24 hr 2hr I hr 24hr 2hr I hr 24 hr 2hr 1 hr 24 hr 2hr 1 hr
Table 2. Mean erouo hormonal Hormone
Facilitatory group Mean & S.D
Time-to-event
resoonses as time-to-event Facilitatory group Mean & SD
24hr 2hr 1 hr 24 hr 2hr 1 hr 24 hr 2hr I hr 24hr 2hr 1 hr 24hr 2 hr 1 hr
15.57k3.36 17.47+2.91 20.87 k 3.38 414k39.9 373 + 77.5 491+52.1 9.68k3.78 7.47 f 3.07 13.8k6.53 25.9 f 5.89 26.0& 12.0 32.1+ 10.0 186+28.6 305 k 92.5 871 f 530
approached Debilitatory group Mean & S.D. 22.551.73 21.2+2.36 23.0k3.92 17.Ok3.46 25.3 f 4.75 23.5 + 4.43 23.8 +4.99 23.3f2.06 22.8 k 2.06 -3.50* 1.0 -1.75kl.72 -4.OOk2.16 -2.25kl.99 - .25 f 2.23 -3.50*1.0 14.5*3.79 13.5+3.61 12.3k3.30
aooroached Debilitatory group Mean & S.D. 19.3k 1.41 17.27+ 1.56 16.67+1.63 441 k 36.9 4665 52.7 642 + 77.3 6.23 f4.99 9.71 f 5.26 10.754.06 32.1 f 7.72 20.1 k7.51 20.0* 10.4 210+60.0 203k46.1 2llk62.9
event between the facilitatory and the debilitatory groups (see Table 2). MANOVA were also carried out on the CSAI-2 subscale data to examine any group or time differences that resulted. In all cases, homogeneity assumptions were tested by use of the Box M test. In all cases, this test was nonsignificant. However, given the relatively small number of subjects, the large variances in hormonal data, and the repeated measures design which was used, the conservative Greenhouse-Geiser epsilon adjustment (Keppel, 1982) was uniformly applied to correct df on all these tests. Effect size and power statistics were also calculated in accordance with the procedure outlined by Cohen (1992). In all cases, significant effects were followed up by Neuman Keuls post-hoc analysis. Cognitive anxiety, somatic anxiety and self-confidence
The data for the anxiety responses of the two group can be seen in Table 1. Support for the use of anxiety direction as an individual difference variable was provided by the CSAI-2 subcomponents, with a consistent group difference being observed. The facilitatory group exhibited a more positive cognitive and somatic state anxiety direction response compared to the debilitatory group, with a significant group main effect in the case of the somatic variable (F(l,lO) adj =4.08, ~(0.05; power = 0.70), which was significantly more positive in the facilitatory compared to the debilitatory group. These differences were apparent without corresponding significant between-group differences in cognitive or somatic anxiety intensity. For self-confidence, the facilitatory group showed a consistently higher intensity response compared to the debilitatory group, and perceived this intensity level to have a more positive influence.
Individual
temporal
15 24 hours
differences
in pre-competition
anxiety and hormonal
concentration
I
I
2 hours
1 hour
1035
Time-to-Event Fig. 1. Testosterone concentration over time as a function of time-to-event.
Testosterone Testosterone concentration exhibited a significant interaction as a function of group and timeto-event (see Fig. 1). Follow-up tests showed the facilitatory group to have a significantly lower testosterone concentration than the debilitatory group 24 hrs before the race, (F( 1.8,15) adj = 4.79, p < 0.05; power = 0.84). There was no significant difference in concentration between the two groups 2 hrs before the race, although 1 hr before the race the facilitatory group was found to have a significantly higher testosterone concentration than the debilitatory group (r;Yl.8,15) adj = 4.52, p < 0.05; power = 0.78). Thus, testosterone concentration in the facilitatory group increased significantly as time-to-event approached where-as it decreased significantly over time in the debilitatory group.
Cortisol No group by time interaction was observed. However, plasma cortisol concentration was found to increase significantly from 2hrs to 1 hr before the event in the debilitatory group (F(2,16) adj = 4.22, p < 0.05; power =0.74) (see Fig. 2). Over time, the facilitatory group exhibited consistently lower levels of cortisol than the debilitatory group. Thus a 50% increase in cortisol concentration was observed from 24 hrs to 1 hr before the race in the debilitatory group, whilst the facilitatory group remained stable. Catecholamines No significant temporal or group differences in catecholamine concentration were observed. Despite these findings however, trends between the 2 hrs and 1 hr measures are more supportive of expected results than those observed between 24 and 2 hrs. Effect size statistics (Cohen, 1992) support the existence of differences in the data that were graphically apparent, and suggest that these changes were statistically meaningful for the three catacholamines measured. For example, the group by time interaction effect size for epinephrine between the 2 hrs and 1 hr assessments at the ~~0.05 level can be interpreted as being large and meaningful (ES= 1.61; power=0.75), supporting the existence of trends in the data that are also consistent with previous research (Dienstbier, 1989) (see Figs 3-5). A high and fast rate of increase in all three of the catecholamines of the facilitatory group was observed 2 hrs before the start of the race, while the concentration in the debilitatory group’s catecholamines remained stable or exhibited a relatively smaller increase. In summary, the data lend support to the hypotheses of the investigation. The facilitatory group
1036
Martin Eubank
et al. Debilitatory
650 -
600 =‘ 9 2
550
-
._8 z z
500 -
E
/
u” s .3
-
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
Facilitatory
group
.
_ A’
__-450
/
/
/
/
/
/
/
group
,A
____-------
350 I 24 hours
I
I
2 hours
1 hour
Time-to-Event Fig. 2. Cortisol concentration over time as a function of time-to-event.
Facilitatory
14 -
group
.
13 -
Debilitatory
group
_ -A
9
8 l’l_ 7 /
/
6 24 hours
/
,
/
/
,
/
/
/
,
,
/
,
,
,
/
/
/
,
/
.
I
I
2 hours
1 how
Time-to-Event Fig. 3. Epinephrine concentration over time as a function of time-to-event.
showed an increase in testosterone concentration as time-to-event approached that contrasted markedly with the significantly decreased levels evidenced by the debilitatory group. The facilitatory group exhibited a lower cortisol response at 2 hrs and 1 hr relative to 24 hrs before the race whilst the response of the debilitatory group was elevated as time-to-event approached. Finally as the event approached, the facilitatory group evidenced a trend towards rapid elevations in all of the catecholamines, whilst the concentration amongst the debilitatory group remained stable. Although the lack of baseline correction, and the usual high between-subject variability in absolute levels on such measures seem to have limited the statistical significance of certain results, trends within the data are clear and fit well with previous if less ecologically valid investigations.
Individual temporal differences in pre-competition anxiety and hormonal concentration Facilitatory
33 \ 31 -‘\
1037
group
. \
\
29 21 II 25 -
\
\
\
\
\
\
\
\
\ \
./ \
23 -
\
\
\
\
\
\
\
21 -
\
\
\
\
Debilitatory
group
.~--------_-___--_-__.
19 17 1 15
I
24 hours
Fig. 4. Norepinephrine
I
I
2 hours
1 hour
Time-to-Event concentration over time as a function
of time-to-event.
Facilitatory
100
24 hours
I
I
2 hours
1 hour
group
Time-to-Event Fig. 5. Dopamine
concentration
over time as a function
of time-to-event.
DISCUSSION The findings lend support to the hypotheses of Dienstbier (1989). As predicted the facilitatory group exhibited a consistently lower cortisol concentration, a sharp increase in catecholamine concentration and a significantly higher testosterone response than the debilitatory group as timeto-event approached. Relative to cortisol and catecholamine concentration, consistent support for Dienstbier’s hypotheses, and therefore for the findings observed in this investigation has been provided by previous research, such as Frankenhauser, Lundberg and Forsman (1980), who utilised a positively challenging choice-reaction task and a negatively stressful vigilance task, and Harte and Eifert (1995) who used a task manipulating environmental condition and attentional focus. Similarly, in terms of testosterone concentration, studies completed under competitive conditions (Leshner,
1038
Martin Eubank et al.
1983; Harris, Cook, Walker, Read & Riad-Fahmy (1989)), are also supportive of the findings observed in this investigation. The underlying mechanism that influences the perceived anxiety response of the individual revolves around the appraisal of stress as having the potential for harm or loss (Lazarus & Folkman, 1984). Traditionally exposure to stress is perceived as having the potential to be negatively harmful and is generally associated with a peripheral physiological arousal response and a state anxiety reaction that is emotionally negative in nature. In such conditions, where the individual experiences distress resulting in a perceived feeling of negative affect, the associated hormonal response is characterised by heightened cortisol and depleted catecholamine levels. A continuation of the stressful experience further increases the ‘helpless’ experience and cortisol concentration continues to rise. Appraisal of threat as potential for challenge produces an anxiety response that is emotionally positive in nature. A low catecholamine base rate, followed by a fast increase in concentration is indicative of positive arousal, a response similar to that observed in the facilitatory group. Dienstbier (1989) suggests that catecholamines are preferentially activated when the organism displays an active response to deal with a challenging situation, to increase the activity and excitability of the whole body. Specifically, the drive associated with a desire to meet the challenge enhances the secretion of the ‘fight’ hormone norepinephrine in response to any stressful encounter. The observed increase in cortisol secretion also serves to initiate a series of metabolic effects that offsets the potential damage of the stressful episode. Hence facilitators place emphasis on the hormonal mechanism as the method of coping with the situation that they find themselves in, which may be described as a ‘fight’ response to challenge. In conclusion, the results of the study clearly add further substance to the veracity of the directional construct in acting as the important distinguishing variable which accounts for individual differences. Additionally however, it has addressed the mechanism by which the relationship between anxiety direction and performance may be brought about. As a result of this underpinning biochemical mechanism, an explanation to aid our understanding of the pre-competitive anxiety direction response of elite and non-elite individuals has been offered, providing an approach that has been long awaited. The focus of a forthcoming study will examine more closely the association of coping and self-control with individual differences in perceived anxiety direction. It may be the case that such differences involve a process where performers have learnt to be facilitators of their anxiety intensity response as a function of becoming elite, through the development of appropriate self-control and coping strategies. Alternatively, it may be the case that facilitation is predispositional and therefore acts as a predictor of elitism. Most plausibly, future investigation may indicate that the direction concept is, at the very least, an interactive product of predispositional trait and transitory state. Acknowledgemenrs-The authors wish to thank Diagnostic Products Corporation, Llanberis and BioRad Diagnostics Group, Hemel Hempstead for providing the hormonal analysis equipment and reagents for this investigation. Thanks are also expressed to Mr Hywel Evans and Mr Paul Wailer of the Clinical Biochemistry department at West Middlesex University Hospital
REFERENCES Cohen, J. (1992). A power primer. Psychological Bulletin, 112(l), 155-159. Dienstbier, R. A. (1989). Arousal and physiological toughness: Implications for mental and physical health. Psychological Review, 96(l), 84100. Eubank, M. R., Smith, N. C. & Smethurst, C. J. (1995). Intensity and direction of multidimensional competitive state anxiety: Relationships to performance in racket sports. Journal of Sport Sciences, 13(l), 52-53. Frankenhauser, M., Lundberg, U. & Forsman, L. (1980). Dissociation between sympathetic-adrenal and pituitary-adrenal responses to an achievement situation characterised by high controllability: Comparison between Type A and Type B males and females. Biological Psychology, IO, 79-91. Harris, B., Cook, N. J., Walker, R. F., Read, G. F. & Riad-Fahmy, D., (1989). Salivary steroids and psychometric parameters in male marathon runners. British Journal of Sports Medicine, 23 (2), 89-93. Harte, J. L. & Eifert, G. H. (1995). The effect of running, environment, and attentional focus on athletes’ catacholamine and cortisol levels and mood. Psychophysiology, 32,49-54. Jones, J. G. & Cale, A. (1989). Precompetition temporal patterning of anxiety and self-confidence in males and females. Journal of Sport Behaviour, 12, 183-195. Jones, J. G., Hanton, S. & Swain, A. B. J. (1994). Intensity and interpretation of anxiety symptoms in elite and non-elite sports performers. Personality and Individual Differences, 17, 651463. Jones, J. G. & Hardy, L. (1990). Stress andperformance in sport. Chichester: John Wiley.
Individual
temporal
differences
in pre-competition
anxiety and hormonal concentration
1039
Jones, J. G. & Swain, A. B. J. (1992). Intensity and direction dimensions of competitive state anxiety and relationships with competitiveness. Perceptual and Motor Skills, 74,467412. Jones, J. G. & Swain, A. B. J. (1995). Predispositions to experience debilitative and facilitative anxiety in elite and nonelite performers. The Sport Psychologist, 9(2),-201-211. Jones. J. G.. Swain. A. B. J. & Gale, A. (1991). Gender differences in nrecomnetition temnoral nattemina and antecedents ofanxiety and s&confidence. J&r& of sport and Exercise Psychology, Ij, 1-15. _ _ Keppel, G. (1982). Design and analysis: A researchers handbook. Englewood Cliffs, NJ: Prentice Hall. Krane, V. & Williams, J. M. (1987). Performance and somatic anxiety, cognitive anxiety and confidence changes prior to competition. Journal of Sport Behaviour, IO, 47-56. Lazarus, R. S. & Folkman, S. (1984). Stress, appraisal and coping. New York: Springer. Leshner, A. I. (1983). The hormonal responses to competition and their behavioural significance. In B. B. Svare (Ed.), Hormones and aggressive behaviour (pp. 393404). New York: Plenum Press. Martens, R., Burton, D., Vealey, R. S., Bump, L. A. & Smith, D. E. (1990). Development and validation of the Competitive State Anxiety Inventory-2 (CSAI-2). In R. Martens, R. S. Vealey and D. Burton (Eds.), Competitive Anxiety in Sport (pp. 117-123). Champaign, IL: Human Kinetics.
