Elsevier

Hormones and Behavior

Volume 126, November 2020, 104866
Hormones and Behavior

A longitudinal investigation of bidirectional and time-dependent interrelationships between testosterone and training motivation in an elite rugby environment

https://doi.org/10.1016/j.yhbeh.2020.104866Get rights and content

Highlights

  • Testosterone can drive and reinforce motivation in sport on different time scales.

  • The testosterone-motivation interplay was modeled in an elite rugby environment.

  • Training motivation was positively related (lagged and non-linear) to testosterone.

  • No time-lagged effect of testosterone on motivation to train was identified.

  • Training motivation decreased and testosterone level increased at match onset.

Abstract

In sport, testosterone has been positioned as a substrate for motivation with both directional and time dependencies. However, evidence is scarce when considering the complexities of competitive sport and no work has explicitly modeled these dependencies. To address these gaps, we investigated the bidirectional and time-dependent interrelationships between testosterone and training motivation in an elite rugby environment. Thirty-six male athletes were monitored across training weeks before and after eight international rugby matches. Pre-breakfast measures of salivary testosterone and training motivation (1–10 rating) were taken on training, competition, and recovery days (up to 40 tests). Using a continuous-time (CT) model, within-person estimates of autoregressive effects (persistence) and cross-lagged effects (relationships) were derived. A stronger, more persistent temporal association was identified for testosterone than for motivation. Cross-lagged effects verified that training motivation was positively related to testosterone at latter time points (p < 0.001). Discrete-time analyses revealed a non-linear association; increasing in strength from a zero-time lag to peak after 2.83 days (standardized effect = 0.25), before dissipation over longer lagged intervals. The testosterone relationship with ensuing training motivation was also positive, but non-significant. Match effects also appeared (p < 0.001) with a predicted decline in training motivation, but a rise in testosterone, at match onset. In summary, a positive within-person association emerged between fluctuations in self-appraised motivation to train and testosterone concentration in an elite rugby environment. The lagged, non-linear nature of this relationship and match predictions on both outcomes support, and extend, theoretical models linking testosterone and competitive behaviors.

Introduction

The steroid hormone testosterone plays a key role in regulating behaviors related to social motivation and dominance. This interplay is often conceptualized within the theoretical framework of the Challenge Hypothesis (Archer, 2006; Wingfield et al., 1990) and the Biosocial Model of Status (Mazur, 1985), where testosterone is thought to regulate behaviors that serve to gain and maintain social status in human or animal competition. Such a perspective is restrictive as testosterone can affect, both consciously and unconsciously, broad-spectrum motivations to act (Aarts and van Honk, 2009). Furthermore, descriptive and experimental studies indicate that many elements of social motivation (e.g., persistence, perceived physical dominance, status-seeking, competitive endurance, fear reduction) are related, positively, to individual changes or differences in testosterone (Casto et al., 2020; Enter et al., 2014; Hermans et al., 2006; Losecaat Vermeer et al., 2020; Welker and Carré, 2015; Welling et al., 2016), and often in the absence of overt human-to-human competition.

In recent years, testosterone has been positioned as a biological substrate for motivation in sport (Wood and Stanton, 2012), which could mediate training gains in muscle size and strength (Cook et al., 2013), competitive performance (Casto and Edwards, 2016), and post-competition recovery (Crewther and Cook, 2012). Direct evidence comes from studies on athletic populations as they train for, and compete in, different sports. Examples include positive associations between testosterone concentration or response and training motivation (Crewther et al., 2016), motivation to win (Salvador et al., 2003; Suay et al., 1999), competitiveness (Crewther and Cook, 2018), and voluntary selection of training workloads (Cook et al., 2013), as a proxy for motivation. Likewise, differences in athlete testosterone concentration or response were found to be positively related to social bonding, status and connectedness (Bateup et al., 2002; Edwards et al., 2006), as possible primers of motivation and interpersonal engagement in a team-sport environment.

Debate still exists regarding the cause-effect nature of the testosterone relationship with motivated behaviors. A high testosterone concentration may, for instance, indicate a predisposition for competitive drive or it could reflect a shift in status motivation to reinforce dominance (Chichinadze et al., 2012; Crewther and Cook, 2018). The notion of reciprocity is consistent with the Biosocial Model of Status (Mazur, 1985), whereby divergent testosterone responses to victories (i.e., rising) and defeats (i.e., falling) promote behaviors that serve to achieve and preserve social status, respectively. Whilst this model has theoretical appeal, not all research on human competition adhere to these outcome-specific hormone responses (Casto and Edwards, 2016). Competitive sport is a complex pursuit comprising of multiple activities (e.g., physical and skill training, competition, recovery days), with each arguably producing differential shifts in testosterone and motivation depending on factors like mood, anticipation of victory (Chichinadze et al., 2012), fatigue state (Schiphof-Godart et al., 2018), and rewards (Vallerand, 2012). It remains to be seen whether the theorized bidirectional testosterone and motivation relationship holds true under these conditions.

To our knowledge, no studies have modeled the time-dependency of any testosterone and motivation interactions. This is a fundamental gap in the literature, as fluctuations in testosterone and dominance behaviors or emotional state do not always covary on the same timescale (Crewther et al., 2016), especially around competition (Shearer et al., 2015; West et al., 2014), with added heterogeneity across individuals (Cook et al., 2018). Testosterone also affects the motivational circuitry via rapid and delayed pathways (Wood and Stanton, 2012), thereby mapping onto future behaviors and performance on timescales spanning several minutes to hours, or even days, later (Booth et al., 1989; Carré et al., 2013; Crewther and Cook, 2012; Mehta and Josephs, 2006; Zilioli and Watson, 2014). To capture these complexities, a more detailed analysis of the time-lagged effects of testosterone and motivation on each other is needed. Addressing these dependencies, both in direction and time, would provide a better etiological understanding of social neuroendocrinology in a sporting context.

A longitudinal, descriptive study was undertaken to investigate the bidirectional and time-dependent interrelationships between testosterone and training motivation in an elite rugby environment. If testosterone does promote competitive motivation, and vice versa, then such interplay could be exaggerated among those individuals who both enjoy competing and possess high motivation to win (i.e., elite athlete). Salivary testosterone and training motivation were measured repeatedly (up to 40 times) on training, competition, and recovery days. The primary aim was to test the within-person effect of testosterone/motivation on motivation/testosterone via time-lagged analyses. Evidence indicates that testosterone and motivational factors positively covary (Casto and Edwards, 2016; Cook et al., 2013; Wood and Stanton, 2012), so we hypothesized that the lagged and reciprocal association between testosterone and training motivation would be positive as well. We then explored how these relationships emerge over time by examining the direction and strength of each association at different time lag intervals. As a secondary aim, we examined the match effect on both variables, but no firm hypotheses were made in this regard.

To assess the time-lagged interplay between study variables, a continuous-time (CT) model was implemented. The CT approach uses differential equations to model data as continuous processes over time (Voelkle et al., 2018), which offers a more realistic framework for investigating hormone and behavior system dynamics. Continuous-time models afford other benefits by (1) permitting the analysis of data from longitudinal designs with differing time intervals between and within individuals, (2) accommodating sampling schedules with different start times and missing data, (3) allowing for effects or associations to be explored across time, and (4) facilitating cross-study comparisons (Driver et al., 2017; Hecht and Voelkle, 2019). Subsequently, CT models are well suited to a growing number of studies utilizing experience sampling, ambulatory, and ecological momentary assessment approaches (de Haan-Rietdijk et al., 2017; Hecht and Voelkle, 2019; Voelkle et al., 2018).

Despite the potential benefits offered by CT modeling within neuroendocrine and behavioral research, some barriers still exist regarding their wider use and interpretation (e.g., understanding of differential calculus, concepts and terms related to time-series modeling, availability of software and code). To overcome these problems, we highlight some key CT concepts and terms, as part of our statistical procedures, before providing a brief working example using a simulated rugby dataset. Next, we present step-by-step guidelines for applying a bivariate CT model to the actual and simulated rugby datasets. Annotated R code for conducting all analyses and plotting results are provided as a supplemental file, including a link to download the simulated dataset and tabled results for this working example. The CT models in this study were implemented using the ctsem (version 3.0.9) package (Driver et al., 2017) in the R (version 3.5.1) programming environment (R Core Team, 2020). This software package is freely available online (https://cran.r-project.org/web/packages/ctsem/index.html).

Section snippets

Participants

The study cohort consisted of 36 professional male rugby players with a mean age of 27.7 (SD = 3.3) years, height of 1.88 (SD = 0.09) m, and body mass of 102.3 (SD = 13.0) kg. These athletes formed part of a training squad preparing for an international rugby series in 2010 and an international tournament in 2011. Three matches were played on consecutive weekends (weeks 1, 2, 3) in the rugby series, whilst the tournament comprised of five matches with intermittent scheduling (weeks 13, 14, 16,

Results

Initial data exploration identified 10 testosterone values (1.6% of the dataset) as outliers, based on a cut-off criteria of ±3SD from the grand mean. Subsequently, these values were winsorized to 3SD from the grand mean before CT modeling. Testosterone had a pooled M = 149.3 and SD = 46.6 pg/mL (minimum = 41.0, maximum = 289.6), whilst training motivation had a pooled M = 6.67 and SD = 1.92 score (minimum = 1, maximum = 10). Population means (95% CI) for testosterone and training motivation

Discussion

This study investigated the interrelatedness between testosterone and training motivation in elite male athletes under ecological conditions, where both processes were characterised by transient fluctuations across each week of competition. In this environment, a stronger and more persistent temporal association on successive time points was identified for testosterone than for motivation. As hypothesized, a positive lagged association between training motivation and subsequent testosterone

Funding

This study was supported by the UK Engineering and Physical Sciences Research Council and UK Sports Council, as part of the Elite Sport Performance Research in Training with Pervasive Sensing Programme [EP/H009744/1], and the Scottish Rugby Union.

Declaration of competing interest

The authors of this paper have no competing interests.

Acknowledgements

We wish to thank the study participants and coaching staff who contributed to this research.

Data availability

The research data is unavailable due to a confidentiality agreement, but a simulated dataset has been created based on the CT modeled results. The simulated dataset can be sourced online and loaded directly into the R programme (see supplementary code provided).

References (53)

  • D. Enter et al.

    Alleviating social avoidance: effects of single dose testosterone administration on approach-avoidance action

    Horm. Behav.

    (2014)
  • D. Enter et al.

    Single dose testosterone administration alleviates gaze avoidance in women with social anxiety disorder

    Psychoneuroendocrinology

    (2016)
  • E.J. Hermans et al.

    A single administration of testosterone reduces fear-potentiated startle in humans

    Biol. Psychiatry

    (2006)
  • A. Mazur et al.

    Testosterone, status and mood in human males

    Horm. Behav.

    (1980)
  • P.H. Mehta et al.

    Testosterone change after losing predicts the decision to compete again

    Horm. Behav.

    (2006)
  • D. Redhead et al.

    On the dynamics of social hierarchy: a longitudinal investigation of the rise and fall of prestige, dominance, and social rank in naturalistic task groups

    Evol. Hum. Behav.

    (2019)
  • A. Salvador et al.

    Anticipatory cortisol, testosterone and psychological responses to judo competition in young men

    Psychoneuroendocrinology

    (2003)
  • F. Suay et al.

    Effects of competition and its outcome on serum testosterone, cortisol and prolactin

    Psychoneuroendocrinology

    (1999)
  • L.L. Welling et al.

    Exogenous testosterone increases men’s perceptions of their own physical dominance

    Psychoneuroendocrinology

    (2016)
  • R.I. Wood et al.

    Testosterone and sport: current perspectives

    Horm. Behav.

    (2012)
  • S. Zilioli et al.

    Testosterone across successive competitions: evidence for a ‘winner effect’ in humans?

    Psychoneuroendocrinology

    (2014)
  • H. Aarts et al.

    Testosterone and unconscious positive priming increase human motivation separately

    NeuroReport

    (2009)
  • G. Atkinson et al.

    Selected issues in the design and analysis of sport performance research

    J. Sports Sci.

    (2001)
  • M.C. Beaven et al.

    Ultradian rhythmicity and induced changes in salivary testosterone

    Eur. J. Appl. Physiol.

    (2010)
  • K.V. Casto et al.

    Testosterone reactivity to competition and competitive endurance in men and women

    Horm. Behav.

    (2020)
  • K. Chichinadze et al.

    Testosterone dynamics during encounter: role of emotional factors

    J. Comp. Physiol. A.

    (2012)
  • Cited by (9)

    View all citing articles on Scopus
    View full text