Sleepy Anger: Restricted Sleep Amplifies Angry Feelings

Abstract

Despite extensive ties between sleep disruption, anger, and aggression, it is unclear whether sleep loss plays a causal role in shaping anger. On one hand, negative affect and distress frequently follow curtailed sleep, suggesting increased anger responses. On the other hand, fatigue and withdrawal also follow, potentially muting anger. To examine these competing possibilities, 142 community residents were randomly assigned to either maintain or restrict their sleep over two days. Before and after, these participants rated their anger and affect throughout a product-rating task alongside aversive noise. Sleep restriction universally intensified anger, reversing adaptation trends in which anger diminished with repeated exposure to noise. Negative affect followed similar patterns and subjective sleepiness mediated most of the experimental effects on anger. These findings highlight important consequences of everyday sleep loss on anger and implicate sleepiness in dysregulation of anger and hedonic adaptation.

Full text

Sleepy anger 1
Running head: SLEEPY ANGER
Sleepy Anger: Restricted Sleep Amplifies Angry Feelings
Zlatan Krizan
Garrett Hisler
Iowa State University
In Press at Journal of Experimental Psychology: General
© 2018, American Psychological Association. This paper is not the copy of record and may not exactly
replicate the final, authoritative version of the article. Please do not copy or cite without authors'
permission. The final article will be available, upon publication, via its DOI: 10.1037/xge0000522
Zlatan Krizan, Ph.D.
Department of Psychology
Iowa State University
(515)294-1975
zkrizan@iastate.edu

Sleepy anger 2
Abstract
Despite extensive ties between sleep disruption, anger, and aggression, it is unclear whether sleep
loss plays a causal role in shaping anger. On one hand, negative affect and distress frequently
follow curtailed sleep, suggesting increased anger responses. On the other hand, fatigue and
withdrawal also follow, potentially muting anger. To examine these competing possibilities, 142
community residents were randomly assigned to either maintain or restrict their sleep over two
days. Before and after, these participants rated their anger and affect throughout a product-rating
task alongside aversive noise. Sleep restriction universally intensified anger, reversing adaptation
trends in which anger diminished with repeated exposure to noise. Negative affect followed
similar patterns and subjective sleepiness mediated most of the experimental effects on anger.
These findings highlight important consequences of everyday sleep loss on anger and implicate
sleepiness in dysregulation of anger and hedonic adaptation.
Keywords: Sleep, Anger, Affect, Sleepiness, Adaptation

Sleepy anger 3
Sleepy Anger: Restricted Sleep Amplifies Angry Feelings
Anger is a basic emotion that confers significant adaptive advantages, yet imposes
perennial costs on society. Anger-prone individuals are at a higher risk for cardiovascular
disease, diabetes, and auto-accidents, while feelings of anger frequently drive unhealthy or
aggressive behaviors whose management drains public resources (Chida & Steptoe, 2009; Staicu
& Cutov, 2010). An important factor that may play a key role in anger is sleep; more angry and
aggressive individuals often report sleep more disturbed in terms of duration or continuity.
However, whether sleep loss plays a causal role in amplifying anger is unclear; most evidence is
correlational and focused on individual differences in anger and sleep, with relatively few
experimental studies (Krizan & Herlache, 2016). As a result, it is essential to test whether and
how sleep loss may causally impact anger.
Sleep and Anger
Sleep. People spend much of their lives sleeping (about one-third of it), typically sleeping
once every night for about eight hours. Sleep is behaviorally defined as a reversible state of
perceptual disengagement from—and unresponsiveness to—the environment. In a human adult it
proceeds in alternating cycles of non-rapid eye movement (NREM), which contains slow-wave
sleep, and rapid eye movement (REM) sleep, which is associated with dreams (Carskadon &
Dement, 2011). Although the full purpose of sleep remains a mystery, it plays a vital role in key
physiological functions such as toxin disposal from the brain, thermoregulation, and
physiological growth and repair (Xie et al., 2013). Moreover, sleep is essential for optimal
psychological functions as it facilitates memory, cognitive control, use of newly learned skills,
and integration of emotional experiences (Killgore, 2013; Walker, 2010). This indispensable role

Sleepy anger 4
of sleep for human functioning is confirmed by epidemiological studies that have linked short
sleep duration to metabolic and cardiovascular disease, alongside variety of psychological and
behavioral problems such as depression and substance abuse (Altman et al., 2012; Bixler, 2009).
The role that sleep plays in modulating anger, however, is not yet clear.
Anger and Aggression. Anger can be defined as “an emotional state marked by
subjective feelings that vary in intensity from annoyance or irritation to intense fury and rage”
(Speilberger, 1988, p. 1). It is one of the basic emotions that appears early in life, presents across
numerous species, and exhibits distinct behavioral and physiological signatures (e.g., furrowed
brow, increased blood pressure, Potegal, Stemmler, & Spielberger, 2010). Anger is aroused
following violations to what “ought” to happen and is intesified by perceived blameworthiness of
others’ actions (Frijda, 1986; Ortony et al., 1988). Although the precise conditions needed for
arousing anger are debated, anger can follow both from non-social aversive experiences such as
heat or pain, as well as from purely cognitive appraisals of others’ hostile intent, perceived harm,
or blame (Anderson & Bushman, 2002; Berkowitz, 2012; Potegal et al., 2010). The importance
of anger also extends to behavior as it often fuels hostile aggression, namely intentionally
harmful actions toward the person viewed as responsible for the perceived harm, unfairness, or
frustration that lead to anger (Anderson & Bushman, 2002).
Sleep as a Predictor of Anger
Does people’s sleep inform us about their anger? Individuals prone to anger, aggression,
and antisocial behavior do seem to have sleep disturbed in either duration or continuity. For
example, children and adolescents with poor sleep are at a higher risk for aggressive behavior
(Aronen, Paavonen, Fjaellberg, Soininen, & Toerroenen, 2000; O’Brien et al., 2011; see Gregory
& Sadeh, 2012 for a review). Critically, these links to aggression extend across healthy adult,

Sleepy anger 5
psychiatric, and offender populations; delinquency, hostility, and intimate-partner violence have
all been linked to disrupted sleep (Ireland & Culpin, 2006: Kamphuis, Dijk, Spreen, & Lancel,
2014; Krizan & Herlache, 2016; Rauer & El-Sheikh, 2012). Although these studies usually did
not assess anger, multiple investigations also point to anger itself as an indicator of both self-
reported and objectively-measured sleep disturbance (Pilcher, Ginter, & Sadowsky, 1997; Shin et
al., 2005). For example, Hisler and Krizan (2017) found that adults with more actigraphically-
assessed sleep disruption (e.g., fragmentation) reported higher chronic anger, especially poorer
anger control.
Does Sleep Loss Cause Anger? While revealing a close link between anger and various
forms of sleep disruption, this evidence does not speak to whether sleep disruption itself is the
cause of higher anger (or aggression). It is likely that individuals who experience sleep problems
also experience anger due to other personality or situational factors responsible for both. For
example, chronic stress, media overuse, or social conflicts could be responsible for feelings of
anger as well as hyper-arousal that interfere with sleep (Barlett, Gentile, Barlett, Eisenmann, &
Walsh, 2011; Caprara, Manzi, & Perugini, 1992). Furthermore, it is likely that anger itself
undermines sleep duration or continuity. Feelings of anger and acts of aggression can impair
sleep through both cognitive activity (i.e. rumination, worry) and physiological arousal (i.e.
sympathetic activity). First, angry rumination and cognitive perseveration can interfere with
sleep initiation and maintenance (Thomsen, Mehlsen, Christensen, & Zachariae 2003; Pillai,
Steenburg, Ciesla, Roth, & Drake, 2014). Second, behaving angrily and inflicting harm can
generate worry about potential consequences (e.g., legal sanctions, retaliation) and induce
physiological arousal which then disrupts sleep (Caprara, Manzi, & Perugini, 1992; Akerstedt,
Kecklund, & Axelsson, 2007).

Sleepy anger 6
Critically, prior experimental investigations into effects of experimentally-curtailed sleep
on anger and aggression have produced inconclusive findings. The few investigations that
examined aggressive behavior in face of provocation did not observe increases in aggression
(Cote et al., 2013; Vohs, Glass, Maddox, & Markman, 2011). However, these studies had very
small sample sizes, induced extreme levels of sleep deprivation rarely experienced in
observational studies, and did not measure anger. Investigations directly assessing anger show
inconsistent results, but have not employed provocation. For example, Dinges and colleagues
(1997) observed substantial increases in tension and fatigue across multiple nights of sleep
restriction, but not an increase in anger. Haack and Mullington (2005) did observe increased
anger following three nights of sleep restricted to 4 hours per night, but these effects were much
smaller than other effects on increased fatigue, reduced sociability, or bodily discomfort. Kahn-
Greene and colleagues (2006) found that individuals were more likely to blame and less willing
to de-escalate a conflict situation when interpreting a cartoon character on a projective measure
following extreme 55 hours of wakefulness, but did not directly measure anger. McMakin and
colleagues (2016) showed that adolescents exhibited more observer-rated conflict behaviors
during social interactions with a peer following sleep loss, but again did not directly assess anger.
The strongest evidence that anger is intensified following lost sleep is from a study of
adolescents, where mild sleep restriction over 6 days resulted in more anger, as well as more
anxiety, fatigue, and confusion (Baum, Desai, Field, Miller, Rausch, & Beebe, 2015).
Anger Without Provocation? Although suggestive, as a whole this evidence does not
firmly establish that common levels of sleep loss (i.e., a few hours a night) causally increase
anger among healthy adults. Critically, the extent to which actual anger was observed is unclear
because emotions were not assessed under conditions of provocation, instigating triggers

Sleepy anger 7
essential for manifestation of frustration or anger (Anderson & Bushman, 2002; Slotter & Finkel,
2011). As suggested by the classic “frustration-aggression” hypothesis, anger is typically driven
by aversive events that impede individuals’ goal progress or threaten desired outcomes (Doobs &
Miller; Berkowitz, 2012). In contrast, the studies failing to document increased anger following
sleep loss were typically confined to contrived computerized protocols and relative social
isolation, which constrains the opportunities for social stresses or provocation typically needed to
instigate anger and aggression (cf., Dinges et al., 1997). As a result, measuring anger under
conditions of provocation is essential both for ensuring that the construct of anger is adequately
assessed and for increasing parallelism between anger assessed within artificial laboratory
conditions (typically devoid of provocation) and emotions as they are experienced in daily life.
How May Sleep Loss Shape Anger? Besides the lack of strong evidence that sleep loss
actually increases anger, it is not clear how sleep disruption may impact anger, namely what
psychological or other consequences of sleep loss may more proximally lead to anger. Losing
sleep leads to a complex, and almost limitless, set of consequences that range from changes in
basic arousal mechanisms and immune-metabolic functions to cognitive disruption and
emotional dysregulation (Killgore, 2010; Walker, 2010). Which of these consequences and
processes are most important for potential changes in anger? Many candidates abound; increased
sensitivity to identification of negative stimuli may bias individuals processing toward more
aversive aspects of stimuli (Walker, 2010), the capacity to exert top-down control of emotional
signals may be impaired via reduced functional connectivity between the amygdala and pre-
frontal cortex (Killgore, 2010), or individuals may be indirectly frustrated by external demands
that interfere with seeking sleep and rest. To make further progress on identifying the
mechanisms that are responsible for how sleep changes shape anger, it is thus also critical to

Sleepy anger 8
examine potential mediating variables that are both sensitive to sleep loss and tied to anger. In
this investigation we focus on sleepiness, given it is a direct consequence of sleep disruption and
an aversive visceral state that may impact ongoing emotions such as anger.
Study Aims
This brief review exposes a need for a strong test of whether mild levels of sleep loss
commonly experienced in daily life cause anger. While suggestive, prior investigations suffer
from important limitations that include lack of provocation, small sample sizes, and neglect of
potential mediators. To address these gaps, we conducted a powerful test on the influence of
everyday doses of sleep loss on anger. First, we sought to induce a moderate amount of sleep
restriction, about 5 to 6 hours overall, which reflects an amount of sleep loss frequently
experienced in everyday life. Second, we examined anger during differing levels of provocation
(i.e., trials with less vs. more aversive noise), critically improving upon prior studies that
assessed anger within austere laboratory protocols that may not have been sufficiently provoking
to arouse anger. Third, we recruited a large sample size (more than 150 individuals) and assessed
anger both before and after the manipulation of sleep, ensuring higher precision and enabling a
direct examination of how sleep restriction impacts changes in anger throughout sleep loss.
Finally, we examined whether subjective sleepiness may account for anger changes due
to sleep loss. We suspected sleepiness could play a key role for the following reasons. First,
feelings of sleepiness are one major consequence of accrued sleep debt; they are extremely
sensitive to sleep loss, form a conscious signal that one is lacking sleep, and are predictive of
objective sleepiness indicators at higher levels (e.g., eye closures, psychomotor slowing)
(Carskadon & Dement, 1977; Akerstedt, Anund, Axelsson & Kecklund, 2013). Second,
subjective sleepiness should have a more proximal influence on emotional responding, unlike

Sleepy anger 9
sleeping itself, as it refers to the state of the organism at the time emotions are experienced (i.e.,
when getting angry during the day) rather than when the sleep itself is disrupted (i.e., during the
night). Consistent with this reasoning, individuals who report higher daytime sleepiness also
report higher anger (Goldstein, Ancoli-Israel, & Shapiro, 2004). Subjective sleepiness may thus
capture both the extent to which accrued sleep debt is interfering with ongoing function as well
as constitute an aversive signal about the current state of the organism that translates into anger.
For example, individuals may rely on their visceral state as information about how well they are
doing, leading to the unpleasant introspective sensations of sleepiness that match resultant
feelings of anger (see Schwarz & Clore, 1983).
Hypotheses. Following prior data and theory, we anticipated that exposure to a more
provoking stimulus (i.e., white noise) will yield more intense anger than will exposure to the less
provoking stimulus (i.e., brown noise, see Anderson & Bushman, 2002). Our main hypothesis
was that sleep restriction will amplify feelings of anger (Krizan & Herlache, 2016). Critically, we
anticipated anger to increase among those undergoing sleep restriction relative to those
maintaining their sleep schedule. We also examined if sleepiness played a role in any anger
differences, namely whether higher subjective sleepiness was responsible for more intense anger
due to sleep restriction. Finally, we examined if any changes in anger extended to changes in
affect more generally or to cognitive evaluations of hostility, and whether they were dependent
on personality differences in anger.
Method
Participants
A power analysis using G*Power 3.1 for a two-factor mixed design targeting a within-
between-subject interaction suggested that 160 participants would be sufficient to ensure .80

Sleepy anger 10
power for detecting even a small interaction effect approximating d = .20 (assuming .60
correlation between repeated observations and .05 level of significance; Faul, Erdfelder, Lang, &
Buchner, 2007). To this end, we recruited 158 participants from a large Midwestern university
and the surrounding community to participate in a study on “sleep restriction and performance in
healthy adults.” Prior to participating in the study, all interested parties were screened over the
telephone; to minimize risks due to participating in the study only individuals without a current
diagnosis of sleep, mental, or a physical illness, who did not work overnight shifts, and who slept
more than six hours a night (on average) were invited to participate. Note that in order to
optimize resources this extensive study involved data collection pertinent to several
investigations; although research procedures are described in full, measurement details less
relevant to the current investigation are only summarized. All relevant measures that were
collected and analyzed are disclosed in this report; data and all of the materials are available
upon request from the first author. All research procedures were approved by the local
institution’s Office for Responsible Research.
Of the 158 participants that completed the experiment, 16 were removed from analyses
for not meeting study eligibility criteria (e.g., reporting an ongoing disorder once arriving to the
laboratory), ignoring study instructions (e.g., not wearing the actiwatch), failing to meet attention
checks during surveys, or reporting the use of caffeine or other psychoactive substances during
the two days outside of the lab. Thus, the final sample size was 142; however, due to computer
data loss issues, sample size varies from 137 to 142 depending on outcome analyzed. Sixty-seven
of these participants were randomly assigned to the sleep restriction condition and the remaining
75 were assigned to the control condition who maintained typical sleep patterns. This sample was
50% male, had an average age of 23.43 (SD = 9.07, Range 18-79), and 68% of participants

Sleepy anger 11
identified as European Americans, 4% as African American, 12% as Asian American, 10% as
Hispanic American, and 6% did not identify with an ethnic or racial category.
Design and Procedures
Baseline Assessment. Participants came to the lab individually. To minimize circadian
influences on any differences between pre- and post-assessments, all participants were scheduled
to arrive for both of their sessions either during the morning (8am-12pm) or the afternoon (1pm-
5pm). Participants were compensated between $20 and $60, depending on their compliance.
More specifically, participants earned $20 if they completed the pre-sleep manipulation session,
but did not continue with the study (this occurred with four participants who are not analyzed).
Participants earned $50 if they completed both pre- and post-sleep manipulation sessions, but did
not follow all study instructions by consuming caffeine, alcohol, drugs, not making all actiwatch
wake and sleep time recordings, or not attempting to adhere to sleep schedule (this occurred with
17 participants and three of these participants were excluded from final data analysis based upon
study exclusion criteria). Participants earned $60 if they completed both pre- and post-sessions
and fully adhered to study (all other participants earned $60).
After arriving to the lab, participants first completed baseline assessments of
computerized and randomly ordered measures of sustained attention, working memory, and
response-control. Participants first completed cognitive tasks, all of which were randomly
ordered to reduce order effects (analyzed in a separate report; none of the cognitive variables
were associated with anger). These included Continuous Performance Test to measure sustained
attention (Rosvold, Mirsky, Sarason, Bransome, & Beck, 1956), Sternberg Memory Test to
assess working memory performance (Sternberg, 1966), and the emotional Go/No-Go task to
capture response-control (Denny & Siemmer, 2012). All these measures were delivered via the

Sleepy anger 12
Inquisit platform and lasted around 30 minutes. Afterwards, participants completed baseline
assessments of the key dependent measures including affect and perceived hostility, again
randomly ordered.
Anger and affect. Anger was measured via participants’ affective responses throughout a
product-rating task across two intensities of aversive noise (the provoking stimulus). To institute
provocation, participants were exposed via headphones to approximately 80 dB computer-
generated brown noise (a low-frequency signal sounding like spraying water) or white noise (a
more aversive “static” signal) throughout a computerized judgment task that required
participants to evaluate common consumer items. Individuals’ irritation and annoyance tend to
maximize between 80 and 90 decibels for many sources of noise (Skagerstrand, Koelber, &
Stenflt, 2017). To de-emphasize anger as the key interest of the study, participants were informed
that the study is about “product perceptions under environmental noise;” product evaluations
were not analyzed. These two forms of noise constituted different intensities of an instigating
trigger expected to give rise to anger. White noise is linked to irritation and frequently used as
provocation in research on aggression, while also a common source of frustration (Bushman &
Anderson, 1998; Cohen & Weinstein, 1981).
Specifically, participants completed 12 trials of a product-rating task (six trials for each
type of noise). During each trial, participants were sequentially presented with pictures of four
common consumer items (e.g. chair, couch, table, bed) and were asked to make consumer-type
evaluations (e.g., “How likely you are you to buy this item?”). While viewing items and making
judgements, participants wore headphones that emitted white or brown noise, depending on the
trial, about a minute. After each trial, the noise over the headphones stopped and participants
reported their affective states on an abbreviated version of the Positive and Negative Affect

Sleepy anger 13
Schedule (chosen based on face validity and high factor loadings, see PANAS-X Manual, Watson
& Clark, 1999). Self-reports of ‘angry,’ ‘hostile,’ and ‘irritable’ were used to create a measure of
anger (

= .88 at pre and post-test). Similarly, self-reports of ‘upset,’ ‘nervous,’ and ‘distressed’
were aggregated to from a measure of negative affect (

= .87 and .69 at pre and post-test,
respectively), while reports of ‘happy,’ enthusiastic,’ and ‘joyful were aggregated to form a
measure of positive affect (

= .96 and .95 at pre and post-test, respectively). Trials of white and
brown noise were randomly ordered for each participant to prevent any anticipation of noise.
Perceived hostility. To evaluate cognitive appraisals of hostility that may occur alongside
anger, participants’ rated their reactions to several hypothetical and unrelated provoking
situations (in MediaLab v2014). These vignettes depicted eight situations with unintentional
harm (e.g., “You notice some kids throwing a baseball in front of your house. Their baseball flies
through your front window breaking all the glass. The kids come to tell you that they are sorry”),
eight ambiguous situations (e.g., “You go to your work and you are in a bad mood. As you walk
in the office, a coworker teases you about something that you are wearing”), and eight highly-
provoking situations (e.g., “As you are walking down the sidewalk you meet a group of three
young teenage males. One walks into you and pushes you forcefully out of the way”, see
Tremblay & Belchevski, 2004; Combs, Penn, Wicher, & Waldheter, 2007). Such vignettes are
frequently used to understand hostile biases in social information processing (e.g., Graham,
Hudley, & Williams, 1992; Bushman & Anderson, 1998). Following each vignette, participants
indicated whether the person “performed the action on purpose”, “how angry would it make you
feel”, and “how much would you blame them for it.” To create an index of hostile interpretation,
responses were first standardized and then aggregated given that scales for each item varied. To
avoid content overlap and preserve equal distribution of scenarios across the testing periods, each

Sleepy anger 14
participant responded to a randomly selected half of the scenarios within each category before
sleep restriction and completed the other half of the scenarios following sleep restriction. The
internal consistency of this hostility index was .89 during both baseline and follow-up
assessments.
Once these tasks were completed, participants also completed demographic, sleep, and
personality questionnaires (analyzed in a separate report). These included a measure of trait
anger that was explored as a potential moderator of the effect sleep restriction may have on anger
(from the Aggression Questionnaire; Buss & Perry, 1991). Afterwards, the experimenter revealed
whether the participant was randomly assigned to the sleep restriction group or the control group
(that maintained their usual sleep schedule) and specified a sleep schedule for each participant to
follow over the next two nights.
Sleep Restriction Manipulation. To create sleep schedules, all participants were given a
printed out time-grid and asked what their normal expected times were for going to bed with the
intention of sleeping and for getting out of bed with the intention of staying awake for the next
two nights. These times were then recorded by the experimenter on the sleep schedule. If the
participant was in the Control condition, the experimenter instructed the participant to adhere to
these expected sleep and wake times. However, if the participant was in the Restriction
condition, the experimenter marked a new target sleep time (two hours later than originally
expected by the participant) and a new wake time (two hours earlier than originally expected by
the participant) for each of the two nights on the sleep schedule. Participants in the restricted
condition were then instructed to adhere to these new sleep- and wake-times. Thus, by
attempting to stay awake two hours later and wake up two hours earlier each day, participants in
the restricted condition, on average, we expected to sleep for a total of eight hours less between

Sleepy anger 15
testing sessions than the control group. Although perfect compliance is rare, this procedure can
effectively yield substantial experimental sleep differences and it enhances ecological validity as
participants sleep at home (Swann Yelland, Redman, & Rajaratnam, 2006). Finally, all
participants were also instructed to take home their sleep schedule and record any abnormal
deviations from their typical routine.
After establishing their schedule, participants received the Actiwatch Spectrum Pro (by
Phillips Respironics) to wear on their non-dominant wrist to record their sleep until return to the
laboratory. They were informed that the device would monitor their sleep and that this data
would be used to determine adherence to the sleep schedule when determining monetary
compensation for study participation. Additionally, buttons on the Actiwatch Spectrum Pro
allowed participants to make two self-reported recordings each day. Participants were instructed
to use this function to mark their sleep and rising times (alongside ratings of sleepiness and sleep
quality, respectively).
Two Nights at Home. During the two nights outside of the laboratory participants were
generally free to do to as they pleased, with a few exceptions. Besides adhering to the stipulated
sleep schedule, all participants were asked to refrain from drinking alcohol and caffeine,
consuming other psychoactive drugs, and taking naps. To help participants in the sleep restricted
condition adhere to their new sleep schedule and to minimize differences in activities during that
period, sleep restricted participants were emailed a link to a two-hour nature documentary they
could watch during that period. In addition, to help ensure that participants woke up at their
designated wake times, they received an automated phone call at the time they were supposed to
wake-up; once they answered the phone they were not to get back into bed. No additional
requirements were imposed on these periods.

Sleepy anger 16
After leaving the lab, adherence to the sleep schedule was continuously monitored via the
Actiwatch Spectrum Pro. It records movement at 30-second epochs to estimate sleep-wake state.
Although sleep-wake state as recorded by the actiwatch highly converges with
polysomnographic recordings, all recordings were inspected by the second author to further
increase accuracy (Marino et al., 2013). If necessary, the sleep-wake recordings were manually
adjusted based upon participant-entered sleep- and wake-times, alongside recorded light or
movement (as in Chow et al., 2016). These recordings were then used to obtain objective
measures of sleep duration (minutes participants slept) and sleep efficiency (time they spent
sleeping relative to their total time in bed).
Follow-up Assessment. Once back at the lab following the two nights at home, the
participants returned the Actiwatch and completed a short, printed questionnaire assessing
sleepiness (via Stanford and Epworth Sleepiness scales) and adherence to at-home instructions
(use of alcohol, drugs, caffeine, or any naps over the past two days). Then, participants again first
completed the randomly ordered cognitive assessments, before completing the randomly ordered
anger, affect, and perceived hostility measures. Once these measures were completed,
participants were compensated, debriefed about the study purpose, and dismissed.
Daytime sleepiness. Sleepiness was assessed with both the Stanford Sleepiness Scale and
the Epworth Sleepiness Scale at the beginning of the follow-up experimental session. The
Stanford Sleepiness Scale assessed the participants’ momentary degree of sleepiness upon their
return to the laboratory; participants picked one statement ranging from “Feeling active, vital,
alert, or wide awake” (1) to “No longer fighting sleep, sleep onset soon; having dream-like
thoughts” (7). This scale is sensitive to actual sleep loss, circadian oscillations, and is often used
to measure subjective sleepiness in the moment (Hoddes, Zarcone, Smythe, Phillips, & Dement,

Sleepy anger 17
1973; Shahid, Shen, & Shapiro, 2010). To evaluate their sleepiness more generally, participants
also completed the Epworth Sleepiness Scales which retrospectively queries participants about
how likely they are to fall asleep across eight common scenarios (e.g. reading, riding in a car,
talking to someone; Johns, 1991). For each situation, participants responded from “Would never
doze” (0) to “High chance of dozing” (3). Responses to these eight situations are then averaged
to assess more a general level of sleepiness (

= .78).
Results
Manipulation Checks
How effectively did participants curtail their sleep at home? Examination of actigraphic
recordings indicated that participants in the control condition slept an average of 6.95 per night
(SD =1.05 hours), whereas those who restricted their sleep only slept 4.55 hours per night (SD =
1.2 hours), yielding an average difference of 2.40 hours per night, t (138) = 12.51, 95% CI: 2.02
to 2.77 hours, p < .001, d = 2.12. On average, sleep-restricted participants slept 4.80 hours less
across the two nights than participants in the control condition, a moderate yet significant sleep
loss. Sleep-restricted participants also exhibited slightly poorer sleep efficiency (M = 82.33 %,
SD = 7.75 %) than control participants (M = 84.69 %, SD = 5.52 %), t (138) = -2.36, 95% CI:
-4.59 % to -.14 %, p = .04, d = .25. Means and standard deviations of all actigraphically derived
sleep variables for the control and sleep restricted groups are provided in Table 2. Sleep loss had
clear effects on participants’ sleepiness; sleep-restricted participants scored 3.20 (SD = 1.13) on
the Stanford Sleepiness Scale (3 = “…responsive but not fully alert”), which was a whole point
higher than those in the control condition, who scored an average of 1.96 (SD = .85; 2 =
“functioning at high levels…”), t (138) = 7.27, 95% CI: .91 to 1.57, p < .001, d = 1.88. These
differences extended to experiences throughout their day; sleep-restricted participants scored

Sleepy anger 18
1.46 (SD = .53) on the Epworth Sleepiness Scale (2 = “moderate chance of dozing”), whereas
those in the control condition scored .95 (SD = .47), more than a half-point lower (1 = “slight
chance of dozing”), t (138) = -6.02, 95% CI: -.67 to -.34, p < .001, d = 1.02. These two
sleepiness measures correlated at .48 and they provide additional confirmation of the sleep-
restriction protocols’ effectiveness. Descriptive statistics and inter-correlations among key
dependent variables appear in Table 1. Finally, to check the extent that differences in the sleep-
wake circadian rhythm may be present between baseline and follow-up testing sessions and
between sleep conditions, we conducted a 2 (time: baseline vs. follow-up) by 2 (sleep: control vs.
restriction) repeated measures ANOVA. There were no substantive effects (all p’s > .26),
suggesting a minimal influence of circadian rhythms on differences across time or between
conditions.
Anger
How did anger change as a function of provocation intensity and sleep? To examine the
former, we submitted participants’ reports of anger to a 2 (time: baseline vs. follow-up) by 2
(provocation: mild vs. strong) by 2 (sleep: control vs. restriction) ANOVA. First, the analysis
revealed an expected effect of provocation intensity, as the average anger during the white-noise
trials (M = 1.67, SD = .75) was higher than anger during the brown-noise trials (M = 1.42, SD = .
61), F (1, 138) = 44.64, 95% CI: .18 to .32, p < .001, d = .37. Second, anger decreased from pre-
(M = 1.61, SD = .78) to post-assessment (M = 1.48, SD = .58), F (1, 138) = 4.07, 95% CI: -.002
to -.23, p = .05, d = -.19, consistent with hedonic adaptation to an aversive stimulus with
repeated exposure.

Sleepy anger 19
Critically, the analysis indicated that the changes in anger were dependent on sleep,
evidenced by a 2-way interaction between time and sleep, F (1, 138) = 8.11, p = .005. To probe
this interaction, we examined changes in anger separately within the control and the sleep
restriction condition (Figure 1). Anger levels were effectively equal at baseline for both sleep
conditions (M = 1.58, SD = .73 vs. M = 1.62, SD = .72, d = .06). Following a hedonic adaptation
pattern, the anger of individuals in the control condition decreased from the initial assessment (M
= 1.59, SD = .74) to the post-manipulation assessment (M = 1.38, SD = .53), t (73) = -2.99, p = .
004, 95% CI: -.34 to -.07, d = .33. In contrast, sleep restricted individuals exhibited a trend
toward increased anger from baseline assessment (M = 1.63, SD = .72) to the post-manipulation
assessment (M = 1.75, SD = .62), t (65) = 1.55, p = .13, 95% CI: -.29 to .04, d = .18. Most
importantly, a simple contrast of anger experienced post-manipulation confirms that individuals
who restricted their sleep reported considerably higher anger (M = 1.75, SD = .62) than those
who did not (M = 1.38, SD = .53), t (139) = 3.7, 95% CI: .19 to .57, p = .004, d = .64. These
differential trajectories of anger as a function of sleep disruption did not vary further as a
function of provocation intensity (F (1, 138) = .48, p = .49).
Finally, including age, gender, or trait anger as co-variates did not change the impact of
sleep restriction on anger. When including change in negative affect between the sessions as a
covariate, the impact of sleep condition on anger was reduced by 48% but not eliminated
(remaining d = .36, p = .03), implicating that sleep loss affected anger uniquely above general
negative affect. Moreover, findings did not change after including sleep efficiency, sleep
fragmentations, and Wake After Sleep Onset as covariates, suggesting that greater anger was due
to the study manipulation of sleep duration and not other concomitant changes in sleep
(measured by actigraphy). Similarly, neither gender (in a mixed ANOVA) nor trait anger (in a

Sleepy anger 20
moderated regression with standardized anger scores) qualified the impact of sleep on anger (all
Fs < 1). Overall, participants experienced more anger during more intense provocation and, more
importantly, those who lost sleep ultimately experienced more anger than those who maintained
their sleep schedule, regardless of provocation intensity or trait anger.
Does Sleepiness Mediate the Impact of Sleep-restriction on Anger? To test whether
the sleepiness of individuals was key to their anger, we repeated the main analysis while
including subjective sleepiness (assessed via the Stanford Sleepiness Scale at the beginning of
the follow-up session) as a co-variate. To the extent current sleepiness is responsible for observed
changes in anger, the reported experimental effects should be muted or eliminated in this
analysis. In this manner, all experimental effects involving sleep restriction failed to reach
conventional standards of significance once sleepiness was taken into account (all F’s < 2, p’s > .
17). To marshal more direct evidence that subjective sleepiness mediated the causal impact of
sleep restriction on higher anger, we tested a mediation model where the sleep-restriction
condition (categorical variable) caused subjective sleepiness (assessed at the beginning of the
follow-up session), which then caused final anger (assessed toward the end of the follow-up
session). Estimating a simple mediation model using the PROCESS v3.0 (Hayes, 2018) with
5,000 bootstrap samples revealed a significant indirect effect via sleepiness of .23 (95% CI: .10, .
39), much larger than the (non-significant) direct effect of sleep restriction on anger (.15, 95%
CI: -.07, .36) and accounting for most (70%) of the total effect that sleep restriction had on anger
(.37, 95% CI: .18, .56, see Figure 2). Importantly, sleepiness still accounts for 50% of the total
effect of sleep restriction on anger after controlling for anger during the baseline assessment,
ruling out the possibility that the observed mediation effect just reflects greater tendencies
towards anger. This result further strengthens the case that sleep is driving changes in anger

Sleepy anger 21
(rather than other reactions to sleep restriction or individual differences in anger) and implicates
the subjective state of sleepiness as particularly important for whether anger manifests.
The Impact of Sleep Restriction on General Affect and Perceived Hostility.
Negative affect. We submitted participants’ reports of negative affect to a 2 (time: pre vs.
post) by 2 (provocation: mild vs. strong) by 2 (sleep: control vs. restriction) ANOVA. As
expected, the analysis revealed an effect of provocation intensity. Average negative affect during
the white-noise trials (M = 1.62, SD = .73) was higher than negative affect during the brown-
noise trials (M = 1.40, SD = .56), F (1, 138) = 36.17, 95% CI: .15 to .29 to .15, p < .001, d = .34.
Additionally, average negative affect decreased from pre- (M = 1.59, SD = .78) to post-
assessment (M = 1.44, SD = .51), F (1, 138) = 8.18, 95% CI: -.24 to -.04, p = .005, d = -.23,
again consistent with hedonic adaptation to an aversive stimulus with repeated exposure.
Critically, a 2-way interaction between time and sleep emerged, demonstrating that
changes in negative affect were dependent on sleep (F (1, 138) = 5.08, p = .03). This interaction
was probed by examining changes in negative affect separately within the control and the sleep
restriction conditions. Baseline negative affect was similar for both sleep conditions (M = 1.60,
SD = .76 vs. M = 1.56, SD = .72, d = .05). Following a hedonic adaptation pattern, the negative
affect of individuals in the control condition decreased from baseline (M = 1.60, SD = .77) to
post-manipulation assessment (M = 1.35, SD = .43), t (73) = -3.69, 95% CI: -.39 to -.12, p < .
001, d = -.36. In contrast, sleep restricted individuals maintained the same levels of negative
affect from baseline (M = 1.57, SD = .72) to the post-manipulation assessment (M = 1.53, SD = .
50), t (65) = -.42, 95% CI: -.17 to -.11, p = .68, d = -.07. Contrasting negative affect experienced
during post-manipulation assessment confirmed that sleep restricted individuals reported
considerably higher negative affect (M = 1.53, SD = .50) than those in the control condition (M =

Sleepy anger 22
1.34, SD = .42), t (139) = 2.47, 95% CI: .04 to .34, p = .02, d = .41. These changes in negative
affect as a function of sleep disruption did depend on provocation intensity (F (1, 138) = .15, p
= .73). Including age, gender, or trait anger as co-variates did not change this pattern of results.
Positive affect. We examined participants’ reports of positive affect in a 2 (time: pre vs.
post) by 2 (provocation: mild vs. strong) by 2 (sleep: control vs. restriction) ANOVA. Again, the
analysis revealed an expected effect of provocation intensity, as the average positive affect
during the brown-noise trials (M = 2.06, SD = 1.02) was slightly higher than during the white-
noise trials (M = 1.98, SD = 1.06), F (1, 138) = 19.91, 95% CI: .05 to .13, p < .001, d = .08.
Positive affect also decreased from pre- (M = 2.20, SD = 1.09) to post-assessment (M = 1.85, SD
= .84), F (1, 138) = 39.29, 95% CI: -.48 to -.25, p < .001, d = -.39.
Moreover, changes in positive affect were dependent on sleep, evidenced by a 2-way
interaction between time and sleep, F (1, 138) = 6.01, p = .001. To probe this interaction,
changes in positive affect were examined separately within the control and the sleep restriction
condition. Baseline positive affect was slightly greater in the control condition than the sleep
restricted condition (M = 2.37, SD = 1.11 vs. M = 2.02, SD = .99, d = .33). The positive affect of
individuals in the control condition slightly decreased from the initial assessment (M = 2.37, SD
= 1.11) to the post-manipulation assessment (M = 2.21, SD = 1.06), t (73) = -2.19, 95% CI: -.30
to -.01, p = .03, d = -.15. Sleep restricted individuals demonstrated several times steeper drop in
positive affect from initial assessment (M = 2.01, SD = 1.00) to the post-manipulation
assessment (M = 1.44, SD = .68), t (65) = -6.16, 95% CI: -.30 to -.01, p < .001, d = -.63.
Comparing positive affect experienced at post-manipulation revealed that individuals who
restricted their sleep reported considerably lower positive affect (M = 1.44, SD = .68) than those
who did not (M = 2.24, SD = 1.07), t (139) = -5.20, 95% CI: -1.10 to -.49, p < .001, d = -.91.

Sleepy anger 23
Once again, differential trajectories of affect as a function of sleep disruption did not vary further
as a function of provocation intensity (F (1, 138) = .02, p = .89). Including age and gender as co-
variates did not change this pattern of results.
Perceived hostility. To examine changes in perceived hostility as a function of sleep
restriction, we first submitted participants’ overall hostile perceptions to a 2 (time: pre vs. post)
by 2 (sleep: control vs. restriction) ANOVA. A trend emerged in which hostile perceptions
increased from pre- (M = 3.40, SD = .58) to post-assessment (M = 3.48, SD = .64), F (1, 135) =
3.12, 95% CI: -.01 to .17, p = .08, d = .13. However, hostile perceptions increased from pre- to
post-assessment equally for both sleep restricted and control participants, F (1, 135) = .12, p = .
73. Including age, gender, or trait anger as co-variates did not change this pattern of results.
Next, we examined whether this pattern of results differed depending on whether the
hostile scenarios were unintentional, ambiguous, or intentional. To do so, we repeated the same
analysis, but only used the unintentional, ambiguous, or intentional scenarios as the outcome.
Results from these analyses did not differ from the results using overall perceived hostility (all
F’s < .13).
Discussion
Despite decades of research and centuries of intuitions, whether everyday sleep loss
actually causes greater frustration or anger has remained unsettled. In response, we examined
how a large group of adults responded to frustrating conditions over time as a function of sleep
restriction. The results revealed that losing even moderate amounts of sleep can make a person
angry. Despite typical tendencies to get somewhat used to irritating conditions, sleep restricted
individuals actually exhibited a trend toward increased anger, reversing normal adaptation. These
patterns dovetailed with broader changes in affect and were directly dependent on subjective

Sleepy anger 24
sleepiness of the individual. Finally, there were minimal effects on cognitive appraisals of
hostility, highlighting the importance of sleep for emotional experiences.
From Sleepiness to Anger
Sleep disruption has been extensively linked to anger and aggression, but identifying
reasons for these associations necessitates more direct investigations. Although there are many
reasons why sleep disruption may be connected to anger and aggression, the present findings
clearly point to sleep loss as an important causal factor in anger. In short, losing sleep will make
a person angrier under irritating conditions (e.g., noise), as well as more anxious and less
enthusiastic. While this result dovetails prior research that has frequently identified emotional
problems following extensive sleep deprivation, it also directly establishes greater anger
following commonly experienced doses of sleep loss. Losing a couple of hours of sleep each
night across two days was sufficient to create relatively large differences in anger during aversive
conditions. In addition, the effect of sleep on anger was not qualified by the intensity of
provocation in this study. Although individuals experienced more anger during more aversive
noise, the “anger penalty” incurred due to lost sleep was similar. Likewise, this dynamic applied
to all individuals regardless of personality differences in chronic anger; individuals both low and
high on trait anger were equally susceptible to the corrosive influence of sleep loss. This pattern
suggests that losing sleep will result in heightened anger across various individuals and contexts,
but the role of provocation intensity deserves further scrutiny as its interaction with sleep may
depend on the nature of provocation (e.g., physical, verbal).
Critically, individuals experienced more anger following sleep loss largely to the extent
that their subjective sleepiness was also intensified, pointing to the importance of sleepiness for

Sleepy anger 25
anger. Because sleep loss is associated with more pain sensitivity, individuals’ introspective
sleepiness could also reflect their awareness of aversive psychosomatic experiences, frustrations
than could have contributed to anger (Finan, Goodin, & Smith, 2013; Schwartz & Clore, 1983).
Although sleepiness was a central marker of the effect that sleep loss had on greater anger, it is
further important to consider underlying neuropsychological processes that constitute this
influence. On one hand, higher anger may have been driven by relatively automatic and early
processes, such as increased sympathetic activity in response to emotion-relevant stimuli
(Franzen, Buysse, Dahl, Thompson, & Sielge, 2009; Zhong, Hilton, Gates, Jelic, Stern, Bartels,
DeMeersam, & Basner, 2005). Similarly, sleep may have lowered thresholds for identifying or
recalling negative stimuli, although as the aversive stimuli in this study (i.e., noise) were
constantly and undeniably present in a single modality, such processes may have played a more
minor role (Gobin, Banks, Fins, & Tartar, 2015; Tempesta, Couyoumdjian, Curcio, Moroni,
Marzano, De Gennaro, & Ferrara, 2010).
On the other hand, mounting evidence implicates sleep loss as undermining top-down
control of emotional responses (Goldstein & Walker, 2014). This is evident by heightened
amygdala activity in response to negative stimuli as well as impaired functional connectivity
between pre-frontal control regions and behavioral impulses that originate in the limbic system
(Anderson & Platten, 2011; Simon et al., 2015; Yoo et al., 2007, Killgore, 2013). In this vein,
sleep (and sleepiness) may have led to impaired inhibition of one’s own anger. Separating early
(orienting, reactivity) from late (re-appraisal, inhibition) processes in the experience and
regulation of anger should be one key avenue for future research aiming to identify brain
pathways responsible for the impact of sleep loss on anger. Note that the state of subjective
sleepiness, an aversive visceral state, may also be the target of regulation alongside anger, with

Sleepy anger 26
more sleepy individuals ending up angrier because of broader impairments in control of
emotional reactions following sleep restriction.
The fact that sleep-restricted individuals experienced more anger also supports the notion
that sleep loss may intensify aggressive responding. Despite limited evidence in direct support of
this hypothesis (Cote et al., 2007, Vohs et al., 2011) , the present experimental findings on anger
embolden the notion that losing sleep can release aggression and call for more precise and
powerful investigations on this score (Krizan & Herlache, 2016). Note that deprivation of REM
sleep in rats can lead to more aggression and that REM sleep is denser during the morning hours
of sleep (see Hicks, Moore, Hayes, Phillips, & Hawkins, 1978; Morden, Conner, Mitchell,
Dement, & Levine, 1967). Given the importance of REM sleep for emotional processing and that
the current study made participants get up two hours earlier than normal (Goldstein & Walker,
2014), sleep architecture may thus be an important factor to consider in future investigations.
Sleep and Hedonic Adaptation
Besides revealing substantial differences in anger following sleep-restriction, the findings
also point to sleep as potentially important for basic adaptation processes. When it comes to
aversive stimuli such as noise, individuals frequently show hedonic adaptation, namely less
negative evaluations of the same repeated stimulus (Frederick & Loewenstein, 1999; Hatfield et
al., 2002; Namba & Kuwano, 1988). A hedonic adaption pattern was suggested for participants in
the control condition, whose ratings of anger and distress decreased from the baseline to the
follow-up assessment. However, sleep restriction seemed to mute or even reverse this effect of
repeated noise exposure, with anger (and negative affect) of sleep-restricted participants
increasing, rather than decreasing. As a result, sleep may have interrupted more basic adaptation

Sleepy anger 27
processes, paralleling negative effects that sleep loss exerts on memory consolidation and
retrieval (Sheth & Khan, 2007). To our knowledge, this is the first study to suggest that sleep
degrades hedonic adaptation to aversive noise. Although only two measures of hedonic
experience separated by two days were not ideal to identify hedonic adaptation (even if they
occurred under identical laboratory conditions), it was not possible to adequately analyze
changes throughout the task across noise-blasts given each individual was assigned to a different
random order of the blasts.
How may sleep shape hedonic adaptation? According to the Hedonic Adaptation to
Postive and Negative Experiences model, both “bottom-up” processes (shifting attention and
awareness of the experience) and “top-down” processes (shifting standards and aspirations
applied to the experience) modulate adaptation trends (Sheldon, Boehm, & Ljubomirsky, 2012).
Whether hedonic adaptation occurs in the short-term (fading discomfort from one’s shoes) or
long-term (a frustrating morning drive to work that becomes neutral), sleep may intensify
immediate emotional reactivity in the short term, or negatively bias recollections of repeated
hedonic experiences in the long term. Because people’s experiences repeat so often and hedonic
adaptation plays a key role in both protecting individuals from adverse events and orienting their
perceptions to most relevant stimuli (Frederick & Loewenstein, 1999), the role of sleep in
hedonic adaptation and more basic learning processes (e.g., habituation) deserves more intense
scrutiny.
Limitations and Future Directions
Despite its strengths, the current study is also limited in important ways. First, at-home
sleep restriction protocols raise the possibility that some other reactions to the experimental

Sleepy anger 28
manipulation, rather than the sleep loss itself, are responsible for experimental differences (in
this case, sleep-restricted participants could have become angrier due to schedule changes that
are negatively impacting their lives). However, the strong role that sleepiness played in
mediating experimental effects increases confidence in internal validity, as it reveals that
individuals became more angry mostly to the extent they were also more sleepy. Along these
lines, more research focusing on mediating factors besides subjective sleepiness is needed. While
subjective perceptions of sleepiness offer insight into how a person is feeling and may partially
drive anger, other factors certainly play important roles in how sleep loss affects anger. For
example, future research should focus on physiological changes that results from sleep loss, such
as decreased neural connectivity between the prefrontal cortex and amygdala (Yoo et al., 2007).
Second, the study examined anger under a relatively mild and non-social provocation.
Although this can be considered a strength insofar it constituted a conservative test, it may not
capture dynamics between sleep loss and more extreme anger observed, for example, during
political or marital arguments. In this vein, anger increased only slightly in the sleep restriction
condition; however, this trend needs to be interpreted in the context of normative adaptation
patterns observed in the control condition where individuals exhibited strong decreases in anger.
Ultimately, future research should examine more intense provoking conditions as well as other
components of anger such as physiology or facial expressions. Third, the findings are also
restricted to a particular dose of sleep loss, namely around 5 hours loss over 2 days. Note that the
extent of sleep loss likely plays a key role in the nature of its effects. Whereas total sleep
deprivation (more than 24 hours awake) ubiquitously impairs vigilance and cognitive control,
smaller doses of sleep restriction have smaller or distinct effects on cognitive function and affect
Killgore, 2013; Van Dongen et al., 2003). In this vein, Hisler & Krizan (2017) found that

Sleepy anger 29
whereas moderate sleepiness predicted more behavioral risk-taking, extreme sleepiness predicted
less of such behavior, a curvilinear relation. Examining how distinct doses of sleep loss (or
fragmentation) disrupt cognitive and affective function should be the central agenda for research
in this area.
Conclusions
Intuitions about frustrations following sleep loss are commonplace, but sleep science has
only recently delved into sleep’s role in anger. In a large group of adults, those who restricted
their sleep saw near-increases in anger during noise, instead of decreases, as did their more well-
rested counterparts. Subjective sleepiness accounted for most of the impact that sleep loss
exerted on anger, pointing to an apparent paradox; one can be both sleepy and angry at the same
time.
Context of the Research
The ideas, hypotheses, and methods were developed by the first author in order to
directly examine prior suggestions that sleep loss actually causes increased anger and aggression.
Despite consistent individual-difference associations between sleep disruption and anger as well
as aggression, prior findings were inconclusive on this point. As a result, an experimental
investigation was necessary in order to separate directions of causal influence which may flow in
both directions. Ongoing research from our team is examining whether the causal impact of sleep
on anger extends to daily life and actual aggressive behavior.

Sleepy anger 30
Acknowledgments
Zlatan Krizan and Garrett Hisler, W112 Lagomarcino Hall, Department of Psychology,
Iowa State University, Ames, IA 50011. This research was supported by the National Science
Foundation Award #1525390 awarded to the first author. We thank Isabella Kane, Sierra Lauber,
Charlotte Moser, Jessica Rissman, Kelly Spoth, and Andrew Weiss for their assistance with data
collection, and W. D. S. Killgore for helpful comments regarding the research. Correspondence
regarding this work should be sent to Zlatan Krizan, W112 Lagomarcino Hall, Department of
Psychology, Iowa State University, Ames, IA 50011, or via e-mail to zkrizan@iastate.edu.
References
Åkerstedt, T., Anund, A., Axelsson, J., & Kecklund, G. (2014). Subjective sleepiness is a
sensitive indicator of insufficient sleep and impaired waking function. Journal of Sleep
Research, 23, 242-254.
Åkerstedt, T., Kecklund, G., & Axelsson, J. (2007). Impaired sleep after bedtime stress and
worries. Biological psychology, 76, 170-173.

Sleepy anger 31
Anderson, C. A., & Bushman, B. J. (2002). Human aggression. Annual Review of Psychology,
53, 27–51.
Aronen, E. T., Paavonen, E. J., Fjallberg, M., Soininen, M., & Torronen, J. B. (2000). Sleep and
psychiatric symptoms in school-age children. Sleep and Psychiatric Symptoms in School-
Age Children. 39, 502–508.
Barlett, N. D., Gentile, D. A., Barlett, C. P., Eisenmann, J. C., & Walsh, D. A. (2012). Sleep as a
mediator of screen time effects on US children's health outcomes: A prospective
study. Journal of Children and Media, 6, 37-50.
Baum, K. T., Desai, A., Field, J., Miller, L. E., Rausch, J., & Beebe, D. W. (2014). Sleep
restriction worsens mood and emotion regulation in adolescents. Journal of Child
Psychology and Psychiatry, 55, 180-190.
Berkowitz, L. (2012). A different view of anger: The cognitive‐neoassociation conception of the
relation of anger to aggression. Aggressive behavior, 38, 322-333.
Bushman, B. J., & Anderson, C. A. (1998). Methodology in the study of aggression: Integrating
experimental and nonexperimental findings. In R. G. Geen & E. Donnerstein
(Eds.), Human aggression: Theories, research, and implications for social policy (pp. 23-
48).
Caprara, G. V., Manzi, J., & Perugini, M. (1992). Investigating guilt in relation to emotionality
and aggression. Personality and Individual Differences, 13, 519-532.
Carskadon, M. A., & Dement, W. C. (1977). Sleepiness and Sleep State on a 90‐Min
Schedule. Psychophysiology, 14, 127-133.

Sleepy anger 32
Chida, Y., & Steptoe, A. (2009). The association of anger and hostility with future coronary heart
disease: A meta-analytic review of prospective evidence. Journal of the American
College of Cardiology, 53, 936-946.
Chow, C. M., Wong, S. N., Shin, M., Maddox, R. G., Feilds, K. L., Paxton, K., ... & Steinbeck,
K. (2016). Defining the rest interval associated with the main sleep period in actigraph
scoring. Nature and Science of Sleep, 8, 321-328.
Cohen, S., & Weinstein, N. (1981). Nonauditory effects of noise on behavior and health. Journal
of Social Issues, 37, 36-70.
Dinges, D. F., Pack, F., Williams, K., Gillen, K. A., Powell, J. W., Ott, G. E., ... & Pack, A. I.
(1997). Cumulative sleepiness, mood disturbance, and psychomotor vigilance
performance decrements during a week of sleep restricted to 4–5 hours per
night. Sleep, 20, 267-277.
Faul, F., Erdfelder, E., Lang, A. G., & Buchner, A. (2007). G*Power 3: A flexible statistical
power analysis program for the social, behavioral, and biomedical sciences. Behavior
Research Methods, 39, 175-191.
Finan, P. H., Goodin, B. R., & Smith, M. T. (2013). The association of sleep and pain: An update
and a path forward. The Journal of Pain, 14, 1539-1552.
Franzen, P. L., Buysse, D. J., Dahl, R. E., Thompson, W., & Siegle, G. J. (2009). Sleep
deprivation alters pupillary reactivity to emotional stimuli in healthy young
adults. Biological Psychology, 80, 300-305.

Sleepy anger 33
Frederick, S. & Loewenstein, G. (1999). Hedonic adaptation. In Kahneman, D., Diener, E.D., &
Schwartz, N. (eds.) Well-being: The foundations of hedonic psychology. (pp. 302-329).
New York: Russell Sage Foundation.
Gobin, C. M., Banks, J. B., Fins, A. I., & Tartar, J. L. (2015). Poor sleep quality is associated
with a negative cognitive bias and decreased sustained attention. Journal of Sleep
Research, 24, 535-542.
Goldstein, A. N., & Walker, M. P. (2014). The role of sleep in emotional brain function. Annual
Review of Clinical Psychology, 10, 679-708.
Simon, E. B., Oren, N., Sharon, H., Kirschner, A., Goldway, N., Okon-Singer, H., ... & Hendler,
T. (2015). Losing neutrality: The neural basis of impaired emotional control without
sleep. Journal of Neuroscience, 35, 13194-13205.
Gregory, A. M., & Sadeh, A. (2012). Sleep, emotional and behavioral difficulties in children and
adolescents. Sleep Medicine Reviews, 16, 129-136.
Haack, M., & Mullington, J. M. (2005). Sustained sleep restriction reduces emotional and
physical well-being. Pain, 119, 56-64.
Hatfield, J., Job, R. S., Hede, A. J., Carter, N. L., Peploe, P., Taylor, R., & Morrell, S. (2002).
Human response to environmental noise: The role of perceived control. International
Journal of Behavioral Medicine, 9, 341-359.
Hayes, A. F. (2018). An introduction to mediation, moderation, and conditional process analysis:
A regression-based approach, second edition. New York, NY: Guilford Press.

Sleepy anger 34
Hisler, G., & Krizan, Z. (2017). Anger tendencies and sleep: poor anger control is associated
with objectively measured sleep disruption. Journal of Research in Personality, 71, 17-
26.
Hisler, G., & Krizan, Z. (2017). Sleepiness and behavioral risk-taking: Do sleepy people take
more or less risk? Behavioral Sleep Medicine, 1-14.Hoddes, E., Zarcone, V., Smythe, H.,
Phillips, R., & Dement, W. C. (1973). Quantification of sleepiness: A new
approach. Psychophysiology, 10, 431-436.
Ireland, J. L., & Culpin, V. (2006). The relationship between sleeping problems and aggression,
anger, and impulsivity in a population of juvenile and young offenders. Journal of
Adolescent Health, 38, 649-655.
Johns, M. W. (1991). A new method for measuring daytime sleepiness: The Epworth sleepiness
scale. Sleep, 14, 540-545.
Kahn-Greene, E. T., Lipizzi, E. L., Conrad, A. K., Kamimori, G. H., & Killgore, W. D. (2006).
Sleep deprivation adversely affects interpersonal responses to frustration. Personality
and Individual Differences, 41, 1433-1443.
Kamphuis, J., Dijk, D. J., Spreen, M., & Lancel, M. (2014). The relation between poor sleep,
impulsivity and aggression in forensic psychiatric patients. Physiology and
Behavior, 123, 168-173.
Killgore, W. D. (2013). Self-reported sleep correlates with prefrontal-amygdala functional
connectivity and emotional functioning. Sleep, 36, 1597-1608.

Sleepy anger 35
Krizan, Z., & Herlache, A. D. (2016). Sleep disruption and aggression: Implications for violence
and its prevention. Psychology of Violence, 6, 542-552.
Marino, M., Li, Y., Rueschman, M. N., Winkelman, J. W., Ellenbogen, J. M., Solet, J. M., ... &
Buxton, O. M. (2013). Measuring sleep: Accuracy, sensitivity, and specificity of wrist
actigraphy compared to polysomnography. Sleep, 36, 1747-1755.
Namba, S., & Kuwano, S. (1988). Measurement of habituation to noise using the method of
continuous judgment by category. Journal of Sound and Vibration, 127, 507-511.
Pilcher, J. J., Ginter, D. R., & Sadowsky, B. (1997). Sleep quality versus sleep quantity:
Relationships between sleep and measures of health, well-being and sleepiness in college
students. Journal of Psychosomatic Research, 42, 583-596.
Pillai, V., Steenburg, L. A., Ciesla, J. A., Roth, T., & Drake, C. L. (2014). A seven day
actigraphy-based study of rumination and sleep disturbance among young adults with
depressive symptoms. Journal of Psychosomatic Research, 77, 70-75.
Potegal, M., Stemmler, G., & Spielberger, C. (Eds.). (2010). International handbook of anger:
Constituent and concomitant biological, psychological, and social processes. Springer
Science & Business Media.
Rauer, A. J., & El-Sheikh, M. (2012). Reciprocal pathways between intimate partner violence
and sleep in men and women. Journal of Family Psychology, 26, 470-477.
Schwarz, N., & Clore, G. L. (1983). Mood, misattribution, and judgments of well-being:
Informative and directive functions of affective states. Journal of Personality and Social
Psychology, 45, 513.

Sleepy anger 36
Shahid, A., Shen, J., & Shapiro, C. M. (2010). Measurements of sleepiness and fatigue. Journal
of Psychosomatic Research, 69, 81-89.
Sheldon, K. M., Boehm, J., & Lyubomirsky, S. (2013). Variety is the spice of happiness: The
hedonic adaptation prevention model. In S. A. David, I. Boniwell, & A. Conley Ayers
(Eds.), Oxford library of psychology. The Oxford handbook of happiness(pp. 901-914).
New York: Oxford University Press.
Sheth, B., & Khan, M. (2007). Sleep affects adaptation. Journal of Vision, 7, 857-857.
Shin, C., Kim, J., Yi, H., Lee, H., Lee, J., & Shin, K. (2005). Relationship between trait-anger
and sleep disturbances in middle-aged men and women. Journal of Psychosomatic
Research, 58, 183-189.
Simon, E. B., Oren, N., Sharon, H., Kirschner, A., Goldway, N., Okon-Singer, H., ... & Hendler,
T. (2015). Losing neutrality: The neural basis of impaired emotional control without
sleep. Journal of Neuroscience, 35, 13194-13205.
Skagerstrand, A., Koebler, S., & Stenfelt, S. (2017). Loudness and annoyance of disturbing
sounds – perception by normal hearing subjects. International Journal of Audiology, 56,
775-783.
Speilberger, C. D. (1988). Manual for the State-Trait Anger Expression Inventory. Odessa, FL:
Psychological Assessment Resources.
Staicu, M., & Cuţov, M. (2010). Anger and health risk behaviors . Journal of Medicine and
Life, 3, 372–375

Sleepy anger 37
Swann, C. E., Yelland, G. W., Redman, J. R., & Rajaratnam, S. M. (2006). Chronic partial sleep
loss increases the facilitatory role of a masked prime in a word recognition task. Journal
of Sleep Research, 15, 23-29.
Tempesta, D., Couyoumdjian, A., Moroni, F., Curcio, G., Moroni, F., Marzano, C., De Gennaro,
L., && Ferrara, M. (2010). Lack of sleep affects the evaluation of emotional
stimuli. Brain Research Bulletin, 82, 104-108.
Van Dongen, H. P., Maislin, G., Mullington, J. M., & Dinges, D. F. (2003). The cumulative cost
of additional wakefulness: Dose-response effects on neurobehavioral functions and sleep
physiology from chronic sleep restriction and total sleep deprivation. Sleep, 26, 117-126.
Watson, D., & Clark, L. A. (1999). The PANAS-X: Manual for the positive and negative affect
schedule-expanded form.
Xie, L., Kang, H., Xu, Q., Chen, M. J., Liao, Y., Thiyagarajan, M., ... & Takano, T. (2013). Sleep
drives metabolite clearance from the adult brain. Science, 342, 373-377.
Yoo, S. S., Gujar, N., Hu, P., Jolesz, F. A., & Walker, M. P. (2007). The human emotional brain
without sleep—a prefrontal amygdala disconnect. Current Biology, 17, R877-R878.
Zhong, X., Hilton, H. J., Gates, G. J., Jelic, S., Stern, Y., Bartels, M. N., ... & Basner, R. C.
(2005). Increased sympathetic and decreased parasympathetic cardiovascular modulation
in normal humans with acute sleep deprivation. Journal of Applied Physiology, 98, 2024-
2032.

Sleepy anger 38

Sleepy anger 39
Table 1. Descriptive statistics and inter-correlations of key dependent variables.
Variable 1 2 3 4 5 6 7 8 9 10 11 12 13 14 M SD
1. Anger - brown noise (pre) - 1.43 .66
2. Anger - brown noise (post) .38* - 1.41 .55
3. Anger - white noise (pre) .74* .25* - 1.78 .90
4. Anger - white noise (post) .43* .88* .57* - 1.55 .60
5. Negative affect - brown noise (pre) .72* .22* .57* .28* - 1.46 .68
6. Negative affect - brown noise (post) .32* .67* .27* .61* .44* - 1.34 .43
7. Negative affect - white noise (pre) .55* .10 .75* .33* .83* .37* - 1.70 .87
8. Negative affect - white noise (post) .32* .42* .62* .64* .50* .69* .71* - 1.53 .58
9. Positive affect - brown noise (pre) .05 -.04 -.13 -.13 .14†.04 .01 -.18* - 2.26 1.05
10. Positive affect - brown noise (post) .06 -.20* -.12 -.30* .12 -.04 .00 -.21* .74* - 1.89 .98
11. Positive affect - white noise (pre) .08 -.04 -.21* -.18* .16†.04 -.05 -.23* .95* .72* - 2.14 1.11
12. Positive affect - white noise (post) .09 -.18* -.15†-.31* .15†-.04 -.02 -.27* .75* .97* .76* - 1.84 1.02
13. Stanford sleepiness scale .06 .40* .09 .42* .08 .21* .07 .20* -.02 -.34* -.05 -.33* - 2.54 1.17
14. Epworth sleepiness scale .01 .30* .06 .34* -.03 .14†-.01 .10 -.01 -.21* -.02 -.20* .50* - 1.19 .56
15. Trait anger .33* .19* .38* .30* .10 .08 .18 .16†-.02 .02 -.05 -.02 .14 .20* 3.26 1.06
Note. *p < .05, †p < .10

Sleepy anger 40
Table 2. Actigraphic sleep characteristics for each sleep condition.
Control Restricted
Mean (SD) M (SD)
Sleep duration 417.46 (63.44) minutes 273.46 (72.75) minutes
Sleep onset latency 21.39 (20.97) minutes 22.60 (24.70) minutes
Sleep efficiency 84.69 (5.52) % 82.33 (7.74) %
Wake after sleep onset 34.23 (13.31) minutes 22.33 (12.03) minutes
Number of wake bouts 33.12 (10.74) 20.16 (9.12)

Sleepy anger 41
PRE POST
1.00
1.10
1.20
1.30
1.40
1.50
1.60
1.70
1.80
1.90
2.00
Control - Low Provocation Control - High Provocation
Restricted - Low Provocation Restricted - High Provocation
Anger (1-5)
Figure 1. The influence of sleep restriction and provocation intensity on anger
.15

Sleepy anger 42
Figure 2. Subjective sleepiness mediates the impact of sleep-restriction on anger
1.2 .18 Anger
Anger
Subjective
Sleepiness
Sleep
Restriction
Indirect Effect = . 23 (95% CI: .10, .39)
Total Effect = .37 (95% CI: .17, .61)
Direct Effect = .15 (-.07, .36)

Sleepy anger 43
Negative affect Positive affect
-0.8
-0.6
-0.4
-0.2
0
0.2
Control Restricted
Within-person standardized mean difference (d)
Figure 3. Changes in general affect from baseline to follow-up as a function of sleep restriction (error bars represent within-person
standard errors)