REF-1650 · Wagstaff, 2014 · Journal of Sport and Exercise Psychology, · Paper · PubMed

Holding back an emotional reaction cost a time trial: 20 competitive athletes told to suppress their response to an upsetting video then rode 10 km slower, with lower mean power, a lower peak heart rate and higher perceived exertion, than after the same video with no instruction or after no video (no magnitudes in the abstract)

Main caveat: One study of 20; the abstract gives directions only, so the size of the effect and its uncertainty are unknown.
limited evidenceMeasured on trained to highly trained athletesRead from the abstract only

Wagstaff asked whether regulating one's emotions draws on a resource that endurance performance also needs.

Twenty competitive athletes completed four laboratory conditions in a single-blind, counterbalanced, within-participant design: familiarisation, a control with no video, and two conditions in which they watched an upsetting video, once told to suppress their emotional reaction and once given no instruction about it. Each condition ended with a 10-km cycling time trial.

After suppressing their reaction, athletes rode the time trial more slowly, produced lower mean power, reached a lower maximum heart rate and rated their exertion higher than in both the non-suppression and the control conditions.

The author concludes that impaired emotional self-regulation affects perceived exertion, pacing and performance, in line with earlier work on the self-regulation of persistence in physical tasks.

Participants

n = 20 competitive athletes; both sexes by the PubMed indexing (the split is not in the abstract).
Age, sport, training volume, fitness and competition level: not in the abstract.

Protocol

Four laboratory sessions: familiarisation, control (no video), suppression, non-suppression; order counterbalanced; single-blind.
Experimental conditions: an upsetting video before the ride; in the suppression condition participants were told to suppress their emotional reaction, in the non-suppression condition they received no self-regulation instruction.
Task: a 10-km cycling time trial, self-paced.
Measures: completion time, mean power, maximum heart rate, perceived exertion.
Not in the abstract: the video, the time between video and ride, the exertion scale, how the order was assigned beyond counterbalancing, and any measure of the emotion felt.

Key Results

OutcomeSuppression vs non-suppression and vs control
10-km timeslower
Mean powerlower
Maximum heart ratelower
Perceived exertionhigher
No values, effect sizes or P values are given in the abstract.
Non-suppression vs control (the upsetting video on its own): not reported in the abstract.

Practical Application

Bottling up an upset right before a hard effort went with a slower, harder-feeling ride. The cost appeared when athletes were told to hide their reaction to an upsetting video, compared with watching it freely and with no video; it showed in time, power and perceived exertion together.
Exertion felt higher while power and peak heart rate were lower. Inference, not tested here: a low heart rate in a poor session on a stressful day reads as a ride that was held back, and does not by itself mean the athlete was fresh.
Inference, not tested here: when an athlete reports a stressful day and is "holding it together", a key session may feel harder than its power says; noting the stress with the session RPE helps read the result later, and no specific adjustment was tested.
Not settled by this design / not from this paper: how large the cost is (no magnitudes in the abstract); whether being upset without suppressing it costs anything; whether other strategies, such as talking it through or reappraisal, avoid the cost; whether it applies to real-life stress or to training rather than a time trial.

Limitations

One study of 20; the abstract gives directions only, so the size of the effect and its uncertainty are unknown.
A laboratory video with an instruction to suppress is a narrow model of real-life stress.
The athletes' sport, level and sex split are not in the abstract; single-blind, with no blinding of the experimenter described.
Whether the emotional manipulation worked (any measure of the emotion felt) is not in the abstract.
Written from the abstract and the PubMed record; full text not read.
How this entry was checked and changed
Why this level

a counterbalanced crossover of 20 (the rubric's B+ band for crossovers of 15 or more) held at B: the abstract gives the direction of every result but no magnitudes, test statistics or variability, and it does not report the plain comparison of the upsetting video without suppression against the control

Content correction 2026-11-07
2026-11-07: added from the abstract and the PubMed record; level B (counterbalanced crossover, n = 20, held below B+ because the abstract gives no magnitudes).