Research Article: The Effects of Social Support on Strenuous Physical Exercise

Date Published: January 11, 2018

Publisher: Springer International Publishing

Author(s): Arran Davis, Emma Cohen.


In humans, socio-environmental cues play an important role in determining adaptive psychophysiological states and behaviours. In sport and exercise, cues to cohesive groups and close, supportive relationships are ubiquitous, possibly because of their effects on neurobiological mechanisms underlying physical performance. Clinical research has shown that the presence of supportive others can lead to reductions in perceptions of pain, while research from sport and exercise science has shown that pain and physical fatigue occupy ranges on a single spectrum of physical discomfort, which works to regulate outputs during strenuous physical exertion. Given the similar neurobiological underpinnings of pain and fatigue, the involvement of both in the self-regulation of strenuous physical outputs, and the effects of social support on perceptions of pain and neurophysiological stress responses more generally, we hypothesised that perceptions of social support affect outputs during strenuous physical exercise by altering activity in self-regulatory mechanisms involved in perceptions of pain and fatigue. We used a between-subjects experimental design to test this hypothesis, varying participants’ social support and measuring physical outputs and perceptions of physical discomfort and exertion during a series of maximum-effort cycling bouts. Analyses showed that participants in the social support condition produced greater initial outputs and steeper declines over time, compared to controls. This effect was moderated by participant neuroticism; an important predictor of how individuals react to social support. We discuss these findings in terms of proposed causal mechanisms linking supportive, cohesive social environments with self-regulation and physical performance.

Partial Text

Studies across a variety of fields using ethnographic, observational, and experimental methods suggest that team cohesion and the presence of close, supportive relationships can enhance performance during sport and exercise via reductions in physical discomfort and fatigue (Davis et al. 2015; Gotaas 2012; Kraus et al. 2010). However, the effects of the presence of close supportive others – an operationalised form of emotional social support (Heinrichs et al. 2003; Schnall et al. 2008) – on physical performance have not been tested experimentally, leaving it an open question as to whether and how social support affects pain and fatigue and the physical outputs they govern. This study tested the hypothesis that social support influences everyday exercisers’ performance during anaerobic exertion by altering activity in self-regulatory mechanisms involved in perceptions of pain and fatigue.

Firstly, Mann-Whitney U tests (used because of non-normal distributions of condition-wise differences in the outcome variables) were conducted to check the effects of the social support manipulation and to identify potential confounds. A mixed ANOVA was then used to test the hypothesis that exercisers in the companion condition would produce greater total anaerobic outputs over the four 30-s exercise bouts. A secondary mixed ANOVA on peak power during each 30-s bout, another commonly used output variable for Wingate Anaerobic Tests (McArdle et al. 2010), was also conducted as a means of confirming that observed patterns of results were not simply an artifact of analysing total anaerobic outputs as the main outcome variable (methods for this analysis are reported in 1.5 of the Electronic Supplementary Material). Mixed ANOVAs were also used to test the effects of the experimental conditions, exercise bouts, and their interaction on exercisers’ rates of perceived exertion (RPE) and reported physical discomfort, as well as their average heart rates during the 30-s exercise bouts. Finally, individual differences known to alter the receipt of social support were included in a series of mixed ANOVAs in order to identify moderators of the effects of social support on strenuous physical exercise outputs and experiences. Moderator variable scores were tertile split; analyses excluded the middle tertile as previous research suggests that social support differentially influences those scoring relatively high or low on these variables (Park et al. 2013; Cohen et al. 2015). All mixed ANOVAs included exerciser sex and estimated anaerobic fitness as control variables.

Regarding the manipulation checks and potential confounds, Mann-Whitney U tests revealed no significant differences between the two experimental conditions in exercisers’ post-exercise ratings of enjoyment, liking of the experimenters, self-consciousness, comfort and support, anxiety, or whether they felt they could have tried harder during the exercise bouts. There were also no significant between-condition differences in the exerciser-companion relationship, estimates of exercisers’ self-reported anaerobic fitness, personality traits, age, or sex (see Table S1 in Electronic Supplementary Material).

This experiment investigated the influence of social support on performance in a difficult exercise challenge. Due to non-significant between-condition effects on mean performance overall (i.e., all exercise bouts collapsed into a single mean), the null hypothesis of no overall difference in performance between the two conditions cannot be rejected. However, analyses revealed a condition × exercise bout interaction on exerciser outputs, with higher initial outputs and steeper declines over the four exercise bouts in the companion condition, compared to the solo condition.




Leave a Reply

Your email address will not be published.