1
Revised Manuscript
The effect of ego depletion or mental fatigue on subsequent physical
endurance performance: a meta-analysis
Louis-Solal Giboin
1
& Wanja Wolff
2,3
1: Sensorimotor Performance Lab, Human Performance Research Centre, University of Konstanz,
Germany
2: University of Konstanz, Department of Sport Science, Sport Psychology, Germany
3: University of Bern, Department of Educational Psychology, Bern, Switzerland
Corresponding author: Louis-Solal Giboin, louis-solal.giboin@uni.konstanz.de
Highlights
- Ego depletion and mental fatigue impair subsequent endurance performance
- The duration of the mental effort task doesn’t predict the magnitude of impairment
- The effect is higher on isolation than whole-body tasks
- The effect is higher when the person-situation fit is low
- This effect should not be seen only through the “fatigue” prism but also as “value”
Abstract
Two independent lines of research propose that exertion of mental effort can impair subsequent
performance due to ego depletion or mental fatigue. In this meta-analysis, we unite these research
fields to facilitate a greater exchange between the two, to summarize the extant literature and to
highlight open questions.
We performed a meta-analysis to quantify the effect of ego-depletion and mental fatigue on
subsequent physical endurance performance (42 independent effect sizes).
We found that ego-depletion or mental fatigue leads to a reduction in subsequent physical endurance
performance (ES = -0.506 [95% CI: -0.649, -0.369]) and that the duration of prior mental effort
exertion did not predict the magnitude of subsequent performance impairment (r = -0.043). Further,
analyses revealed that effects of prior mental exertion are more pronounced in subsequent tasks that
use isolation tasks (e.g., handgrip; ES = -0.719 [-0.946, -0.493]) compared to whole-body endurance
tasks (e.g. cycling; coefficient = 0.338 [0.057, 0.621]) and that the observed reduction in performance
is higher when the person-situation fit is low (ES for high person-situation fit = -0.355 [-0.529, -
0.181], coefficient for low person-situation fit = -0.336 [-0.599, -0.073]).
Taken together, the aggregate of the published literature on ego depletion or mental fatigue indicates
that prior mental exertion is detrimental to subsequent physical endurance performance. However, this
analysis also highlights several open questions regarding the effects’ mechanisms and moderators.
Particularly, the surprising finding that the duration of prior mental exertion seems to be unrelated to
subsequent performance impairment needs to be addressed systematically.
2
Keywords
Self-control; cognitive fatigue; mental effort; motivation, conservation of resources
3
Introduction
To perform at our best, we frequently have to control ourselves and have to consciously employ
mental effort in order to achieve a valued goal. For example, in order to achieve excellent grades in
university, a student has to study hard and employ mental effort to ward off any internal (e.g., task-
induced boredom) or external (desirable behavioural alternatives, proposed by friends) detractors of
successful goal pursuit. The same holds for a cyclist who has to invest mental effort to fight off the
urge of slowing down, although her body is aching. Thus, the effective self-regulation of human
performance frequently hinges on the exertion of mental effort.
Despite its ubiquitous nature (or due to it), arriving at an operational definition of mental effort has
been surprisingly challenging (Shenhav et al., 2017). Here, we follow the approach by Shenhav et al.
(2017, p. 100), who define mental effort in terms of information processing: “Effort is what mediates
between (a) the characteristics of a target task and the subject’s available information-processing
capacity and (b) the fidelity of the information-processing operations actually performed, as reflected
in task performance.” Although it refers to mental effort, this definition can also be intuitively
explained with an example from physical effort. Say, a marathon runner is able to run a marathon in
02:14:00h (i.e., capacity). To qualify for the Olympics, he needs to run it in < 02:19:00h (i.e., task
characteristics). Effort is what mediates between his running capacity and the required qualifying time,
on the one hand, and the marathon time that is ultimately achieved. From an information processing
perspective, tasks require more effort if they require the control of more default (i.e., automatic)
responses (Schneider & Shiffrin, 1977). To illustrate, when a cyclist is aching, the default response
would be to stop. However, to win this response needs to be controlled. Control can then be defined as
the force through which mental effort is exerted (Shenhav et al., 2017).
Despite its instrumentality for achieving goals, people avoid exerting mental effort (Shenhav et al.,
2017) and when mental effort is exerted, this feels aversive (Kool & Botvinick, 2014) and leads to
sensations of fatigue (Wolff, Sieber, Bieleke, & Englert, 2019). Thus, mental effort appears to carry an
intrinsic disutility (Kool & Botvinick, 2018) and effort is only mobilized when the goal is subjectively
worth it (Gendolla & Richter, 2010). In addition, effort mobilization directly corresponds to task
difficulty (Wright, Mlynski, & Carbajal, 2019), implying a restrain to mobilize effort in excess
(Richter, Gendolla, & Wright, 2016). Going back to the example of the marathon runner: If the only
goal is to qualify for the Olympics (i.e., no other incentives like price money or a personal record play
a role), the runner should only run as fast as needed to qualify. Indeed, a large body of research has
shown that people try to conserve energetic resources when it comes to the mobilization of effort
(Richter et al., 2016).
Taken together, people invest mental effort sparingly and treat its’ mobilization as if the capacity for
control is limited (Shenhav et al., 2017). Attesting to this possible limitation, a large body of research
has shown that prior exertion of mental effort impairs subsequent cognitive (Hagger, Wood, Stiff, &
Chatzisarantis, 2010) and physical performance (Van Cutsem, Marcora, et al., 2017). Two largely
independent lines of research in psychology and exercise physiology have postulated theoretical
frameworks that account for this phenomenon (e.g., Marcora et al., 2009; Muraven et al., 1998). As
expected with two independent research fields, what appears to be the same phenomenon is studied
with different experimental paradigms and explained using different explanations (Pattyn, Van
Cutsem, Dessy, & Mairesse, 2018). In the present paper, we want to briefly introduce the two
dominant theoretical models from both fields and highlight similarities in regard to the predictions
they make regarding the effect of mental effort exertion on subsequent physical performance. We will
then quantify the empirical evidence for or against these predictions with a meta-analysis of the
published literature.
Ego depletion
4
In the last two decades, the strength model of self-control (Muraven, Tice, & Baumeister, 1998) has
been by far the most popular psychological model for explaining performance decrements due to the
prior exertion of mental effort. According to the strength model, the capacity to exert mental effort
hinges on a depletable global self-control resource (Hagger et al., 2010). The state of depleted self-
control resources is called ego depletion and supposedly leads to impaired performance in subsequent
self-control demanding tasks. This is because depleted self-control is thought to replenish only slowly
and individuals have to make do with very limited resources (Muraven, Collins, Shiffman, & Paty,
2005). According to the strength model, all self-control processes draw on the same limited resource,
implying that applying self-control in one task (e.g. regulating an emotional response) will affect
performance in a completely unrelated self-control demanding physical task (e.g. handgrip task)
(Muraven et al., 1998). Importantly, self-control can also be conceptualized as a trait (Tangney,
Baumeister, & Boone, 2004) and individuals high in self-control are supposedly less prone to ego
depletion (Muraven et al., 2005; but see also Lindner, Nagy, Ramos Arhuis, & Retelsdorf, 2017). In
ego depletion research, the first task is usually called the primary task or the ego depletion task.
A host of research has reported support for the strength models’ propositions and a meta-analysis of k
= 81 studies has found a medium-to-large effect size of ego depletion on diverse outcome domains
(e.g., impulse control, choice behaviour, volition, cognitive processing, Hagger et al., 2010). Applied
to physical performance, researchers have found detrimental effects of ego depletion on choking under
pressure (Englert & Bertrams, 2012), endurance (Englert & Wolff, 2015), or sprint start performance
(Englert, Persaud, Oudejans, & Bertrams, 2015). For an overview on ego depletion and physical
performance, please see (Englert, 2016; Englert, 2017). However, it is important to note that failures to
replicate the strength models propositions have accumulated in recent years (e.g. Lurquin et al., 2016;
Wolff, Sieber, et al., 2019) and a multi-lab preregistered replication report (RRR) failed to find
evidence for the ego depletion effect (Hagger et al., 2016).
Mental fatigue
In exercise physiology, these performance decrements are primarily explained by a mental fatigue
which is thought to occur after prolonged exertion of mental effort (Marcora, Staiano, & Manning,
2009). More specifically, according to the psychobiological model of endurance performance
(Marcora, 2009; Marcora & Staiano, 2010), perception of effort is the ‘cardinal exercise stopper’
(Staiano, Bosio, Morree, Rampinini, & Marcora, 2018) and perception of effort during a physical task
can be affected by – among others – prior induction of mental fatigue.
Indeed, research from cognitive neuroscience indicates that areas in the prefrontal cortex play a crucial
role in the regulation of effortful control (Shenhav, Botvinick, & Cohen, 2013; Vassena, Holroyd, &
Alexander, 2017). More specifically, the anterior cingulate has been linked to the sensation of effort
(Williamson et al., 2001), the decision to further invest effort and heightened prefrontal cortex
activation has been found when participants anticipate the need to invest mental effort (Vassena,
Gerrits, Demanet, Verguts, & Siugzdaite, 2019). In line with this, prefrontal cortex activation has been
found to increase as a function of the effort participants have to put into an endurance task (Wolff,
Bieleke, et al., 2018; Wolff, Sch, et al., 2019). Interestingly, prior to task failure in an exhausting
cycling task a drop in activation has been frequently reported (Rooks, Thom, McCully, & Dishman,
2010), which indicates that a certain level of activation in the prefrontal cortex is needed to perform an
effortful task (Hosking, Cocker, & Winstanley, 2015). From the perspective of mental fatigue, this
indicates that prior mental exertion leads to an accelerated increase in perception of effort which will
then lead to a premature task termination (Marcora et al., 2009). (Pattyn et al., 2018)In this paradigm,
the first task is often referred to as mental fatigue or cognitive fatigue task.
Research has found support for the notion that mental fatigue affects subsequent performance with a
particular emphasis on endurance performance (Van Cutsem, Marcora, et al., 2017). For example,
Marcora et al. (2009) found a detrimental effect of mental fatigue on subsequent cycling performance.
While the published literature on mental fatigue appears to be rather consistent (but see Vrijkotte et al.,
5
2018), it is important to note that highly cited replication failures on ego depletion stem from pre-
registered studies with very large samples (e.g. Hagger et al., 2016; Lurquin et al., 2016). Such studies
have not yet been conducted in mental fatigue research. We believe it will be an important next step
for mental fatigue research to also use studies with bigger samples to get a better estimate of the
effects' true size (Button et al., 2013).
Ego depletion and mental fatigue: how prior mental exertion affects subsequent performance
Ego depletion and mental fatigue make a strikingly similar proposition: Prior exertion of mental effort
will impair endurance performance on a subsequent task and this impairment can only partly be
compensated by motivation (Baumeister & Vohs, 2007; Van Cutsem, Marcora, et al., 2017). Possibly
the main observable difference between research labelled as ‘mental fatigue research’ or ‘ego
depletion research’ is the duration of the fatiguing (above or equal to 30 min; Van Cutsem, Marcora, et
al., 2017) or depleting task (mostly less than 30 min). As Pattyn and colleagues recently stressed, the
use of different terminology in the different scientific fields has led to independent streams of
research, possibly inducing slower dissemination of ideas and rediscovery of ‘old news’ (Pattyn et al.,
2018). Taking this point into account, the aim of the present paper is not to make a comparison of the
relative explanatory merit between ego-depletion and mental fatigue research but to provide a data-
driven discussion of the published literature in regard to key propositions made by both theories. To
our knowledge, this is the first attempt at a unified synthetization of published literature on the effect
of prior mental exertion on subsequent physical endurance performance. We focus on physical
endurance performance, because this is a domain where both, ego depletion and mental fatigue
researchers, have made sizable contributions to, thereby making the synthetization attempt worth wile.
We hope that the present observations will prove helpful in locating the knowns and unknowns
regarding those two similar but largely independent lines of research and to facilitate a better
understanding for the way physical performance is affected by prior mental exertion.
The present study
In both lines of research, mental effort is exerted to induce a state that is then either labelled ego
depletion or mental fatigue. For clarity and neutrality, we will refer to the ego depletion or the mental
fatigue task as the mental effort task as this describes the actual behaviour (i.e., exertion of mental
effort) and not an expected result of this behaviour (i.e., ego depletion or mental fatigue). As a starting
point, we will assess the very premise of both models: does prior mental effort impair subsequent
endurance performance? Building on this main research question, we will then meta-analyse three
questions:
First, both ego-depletion and mental fatigue researchers propose that the duration of the cognitive
demanding task plays an important role in the alteration of the subsequent endurance task. Regarding
ego depletion, a linear association between duration of the depleting task and the size of the ego
depletion effect is expected (Hagger et al., 2010). Thus, the effect should scale with time, but the
model does not specify a lower limit for the duration mental effort needs to be exerted for an ego
depletion effect to occur. Contrary to this, mental fatigue is only thought to reliably occur if the mental
exertion was at least 30 minutes long (Van Cutsem, Marcora, et al., 2017). Interestingly, we are not
aware of any research that has compared the average duration of mental effort tasks between both
fields. To address this, we will compare field-dependent task durations and then assess if a longer
mental effort task indeed leads to a more severe performance impairment.
Second, different types of physical endurance performance tasks pose different physiological (and
possibly psychological) challenges. Endurance tasks can engage the whole body or only involve a few
specific muscles (i.e., isolation task). Running or cycling are typical examples of a whole-body
endurance task, whereas persistence in a handgrip task is an example of an isolation task. It is
conceivable that prior mental exertion differentially affects performance as a function of task type.
Indeed, single joint tasks and whole body tasks possibly use different motor modules that require
