Testosterone, cortisol, and women's competition

https://doi.org/10.1016/S1090-5138(01)00100-3Get rights and content

Abstract

Hormone (testosterone, cortisol)–behavior relationships have been extensively studied among male competitors, and far less so among female competitors. To address this gap, we studied members of a nationally recognized college women's rugby team. Seventeen players (ages 18–22 years) provided saliva samples 24 h before, 20 min prior to, and immediately after five league matches. Subjects self-reported aggressiveness, team bonding, pregame mental state, postgame performance evaluation, and whether the opponent was more or less challenging than expected. Results revealed that both testosterone and cortisol levels increased in anticipation of the matches. Postgame levels of both hormones were higher than pregame levels. The pregame rise in testosterone was associated with team bonding, aggressiveness, and being focused, but was unrelated to perceptions of the opponent's skill. Testosterone change during the game was unrelated to winning or losing, evaluations of personal performance, or perceptions of the opponent's threat. Game changes in cortisol were positively related to player evaluations of whether the opponent was more of a challenge than expected, and negatively related to losing. These results are compared with hormone–behavior patterns found among male competitors and are interpreted within a recent theory of sex differences in response to challenges.

Introduction

This study examined temporal hormonal responses to aggressive competition in elite college women athletes. An extensive body of research has focused on hormonal responses to competition, but the subjects are nearly always male. Few studies have explored parallel relationships in women. This study furthers our understanding of the degree to which women's hormones change in anticipation of and response to aggressive competition, and how hormones are related to performance, cognition, and attitudes among women competitors. Findings permit indirect comparison of hormone–behavior links in female competitors with the patterns found in research on males.

The role of testosterone and cortisol in male competitors has attracted substantial attention from behavioral endocrinologists who have studied tennis, wrestling, basketball, and chess, as well as competitive tasks in laboratory settings. The only carefully controlled study of women's competition to date focused on women and men competing with same-sex partners in a video game (Mazur, Susman, & Edelbrock, 1997). The present study of women's rugby is an ideal parallel to the many male studies. Women's rugby requires the physical contact and tackling that are characteristic of American football, yet without the protective equipment. Two teams of 15 players compete against each other. To forward the ball, players run, pass, or kick for 80 min with no time-outs. The opposing team defends by tackling the ball carriers and attempting to drive the ball to their side. All players are involved in tackling, driving, running, and passing. In this forum, active physical domination by women players is required and rewarded. This is one of the rare natural social ecologies in which women's hormonal responses to highly physical aggressive competition can be studied.

In men, the relationship between testosterone and competition is reciprocal. Males characteristically experience a testosterone increase in anticipation of competition (Booth, Shelly, Mazur, Tharp, & Kittok, 1989). The precompetition rise is likely to make the individual more willing to take risks (Daltzman & Zuckerman, 1980), improve psychomotor function and coordination (Herrmann & Beach, 1976), and increase cognitive performance (Herrmann et al., 1976, Klaiber et al., 1971, Vogel et al., 1971. For a few hours following competition, testosterone is high for winners relative to losers Booth et al., 1989, Elias, 1981, Mazur & Lamb, 1980. The rise in testosterone following a win is associated with positive mood (Booth et al., 1989), and is thought to be important because winners often face challenges from others soon after gaining new status. There are exceptions to this general finding. Two studies found no difference in testosterone levels between winners and losers Gonzalez-Bono et al., 1999, Salvadore et al., 1987) and one found losers to have higher levels than winners (Filaire, Masso, Sagnol, Lac, & Ferrand, 2001).

Both the pregame rise and the response to winning or losing are less likely to occur if the individual regards the event as unimportant, if he feels certain that he will win or lose because the opponent has a very different level of skill Mazur et al., 1992, Mazur & Lamb, 1980, Salvadore et al., 1987, or if he believes the outcome was due to luck or a referee's decision (Gonzalez-Bono et al., 1999). Studies show that these changes also occur in nonphysical competition such as chess matches (Mazur et al., 1992) and contests in reaction time (Gladue, Boechler, & McCaul, 1989). Indeed, active participation in the competition is not necessarily required: testosterone levels increase among spectators watching their favorite sports teams win and decrease for fans of the losing teams (Bernhardt, Dabbs, Fielden, & Lutter, 1989). These findings suggest that the link between competition and testosterone in males operates in a wide variety of competitive situations but is highly contingent on perceptions that gain or loss of status is at stake.

Information about the role of cortisol in competition is not as comprehensive. Some studies suggest that cortisol may play a role in behaviors important in competition, including aggression, arousal, and mobilization of physiological resources to deal with impending threat or challenge. There is anticipatory elevation in cortisol in men and a rise during the competitive event Booth et al., 1989, Elias, 1981. Another aspect of the cortisol–competition link is that top-seeded male tennis players exhibited consistently lower cortisol levels than did less talented players, suggesting that highly successful competitors may have above average ability at managing stress (Booth et al., 1989).

To our knowledge, only one carefully controlled study has examined competition and hormone changes in women. Mazur et al. (1997) obtained saliva samples from 28 men and 32 women before, during, and after competing with same-sex partners in a video game. They found that men's testosterone and cortisol followed the familiar pattern of a pregame rise but that there was no postgame difference in response to winning and losing. Women, on the other hand, experienced no change in testosterone and cortisol production, except for a downward trajectory most likely due to diurnal variation. Overall, women's cortisol level was higher than the men's, which may indicate that they experienced the event as more unfamiliar or challenging than did the men. While this was a carefully controlled study, women seldom play video games, compared to men, and may not have felt that the event challenged their status. The fact that male losers and winners did not differ from one another may also indicate that men did not find the video game sufficiently engaging to invest in the experience as a challenge to their status.

Lack of research notwithstanding, there are reasons to expect differences in the hormone–behavior links between women and men competitors. One reason is that women produce five to seven times less testosterone, a hormone that acts to develop the male brain for aggressive or dominant behavior in many species, than men (Nelson, 2000). Another reason is that while activation of the hypothalamic–pituitary–gonadal axis is the primary source for testosterone in men, in women the majority of testosterone is derived from the peripheral metabolism of dehydroepiandrosterone (DHEA) (Parker, 1991). DHEA is secreted by the adrenal gland in response to activation of the HPA axis in reaction to challenging or threatening events and is associated with changes in cortisol. If HPA activation and subsequent release of DHEA are linked to increases in women's testosterone levels, it is biologically plausible to believe that women have a different pattern of relationships among testosterone, cortisol, and competition than do men.

Still another reason to expect sex differences is that parental investments may predispose women to adopt a different response pattern than men to challenges and stress. Taylor et al. (2000) suggest that women's response to challenges may be more defensive in nature than men's, which they characterize as a “tend-and-befriend” strategy to differentiate it from the “fight-or-flight” response attributed to men. In females, fight may put their offspring in danger, while flight may be compromised by pregnancy or may interfere with offspring care. Tending entails nurturing activities intended to protect and calm offspring and befriending involves creating and maintaining networks that provide resources and protection for self and offspring. As such, Taylor et al. propose that the typical low levels of testosterone in adult females indicate that androgens are unlikely to be the organizing factors that evoke a tend-and-befriend response. We will examine our findings in light of this theory.

The study described here explored several important issues. First, it looked at whether women's testosterone and cortisol show a precompetition rise and, if so, sought to ascertain the source of the increase and if it influenced perceived performance. Second, it investigated whether women's testosterone and cortisol respond to winning and losing or relate to the seriousness of the challenge posed by the opponents. Third, whether top-seeded women (highest ranked athletes) have lower cortisol just prior to the match was examined. Then, a number of issues that have not yet been studied in men were explored. For example, are particular cognitive or emotional experiences associated with precompetition elevations in hormones? How are individual differences in testosterone or cortisol related to social affiliation in the form of team bonding and camaraderie? Do individuals who enjoy the aggressive play and think of themselves as aggressive have higher pregame increases in testosterone?

Section snippets

Participants

Seventeen members of a university women's rugby team (ages 18–22 years) gave informed consent to participate in the research. The university's Institutional Review Board approved the use of human subjects in this study. All subjects were varsity players who had individually participated in the National Under-23 Tournament or other regional all-star competitions. Participants had played rugby for at least one previous season. They played an average of three games each, although the number ranged

Descriptive analyses of hormone change

To minimize between-subject variation, each subject's raw hormone values were divided by the subject's highest measured level. The percentages of maximum value per individual were used in all analyses. Baseline, pregame, and postgame levels of testosterone and cortisol are shown in Table 1. The t tests were used to assess the statistical significance of hormone change over time. All P values are two-tailed. Testosterone increased 24% from baseline to just prior to the game [t(49)=−3.811, P

Conclusion and discussion

The study explored relationships between testosterone and cortisol production in anticipation of and response to aggressive and physical competition among women. Because the context of rugby is similar to that of sports used in many studies of male competition, it provides a, heretofore, unique opportunity to compare the role of hormones in women's and men's competition. Such comparisons are important because although much is known about the role of hormones in male competitors (relative to

Acknowledgements

We thank the outstanding young women athletes who participated in this study for their enthusiastic contribution. We also thank Laura Cousino Klein for feedback on an earlier draft of this manuscript and commentary on the “tend-and-befriend” theory and David R. Johnson for advice on methodological issues. The staff of the Penn State Behavioral Endocrinology Laboratory and the Population Research Institute (which receives core support from NICHD Grant 1-HD28263) deserve recognition for their

References (24)

  • B Gladue et al.

    Hormonal response to competition in human males

    Aggressive Behavior

    (1989)
  • D Granger et al.

    Salivary testosterone determination in studies of child health and development

    Hormones and behavior

    (1999)
  • Cited by (215)

    View all citing articles on Scopus
    View full text