Content uploaded by Emma Kirwan
Author content
All content in this area was uploaded by Emma Kirwan on Nov 05, 2021
Content may be subject to copyright.
J Sleep Res. 2020;00:e13044. wileyonlinelibrary.com/journal/jsr
|
1 of 10
https://doi.org/10.1111/jsr.13044
© 2020 European Sleep Research Society
1 | INTRODUCTION
Inadequate sleep may have strong adverse consequences for psy-
chosocial health and everyday functioning (Shochat, Cohen-Zion, &
Tzischinsky, 2014), with young people being especially vulnerable to
such consequences (Schwarz et al., 2019). Young people with poorer
quality of sleep report significantly more psychological health prob-
lems than those with good-quality sleep (Gregory et al., 2011; Lund,
Reider, Whiting, & Prichard, 2010; Sin et al., 2017). Symptoms of psy-
chological distress co-occur with unusual sleep experiences; 29.0%
of patients with a parasomnia reported some form of depression
(Vandeputte & de Weerd, 2003), and recent research suggests that a
person diagnosed with a sleep disorder is at 2.3 times greater risk of
developing depression (Byrne, Timmermann, Wray, & Agerbo, 2019).
Sleep disorders may be underdiagnosed and are relatively com-
mon complaints among young adults. In one large national study of
over 50,000 students, rates of insomnia increased from 2011, with
34% of women and 22% of men reporting insomnia (Sivertsen et al.,
2019). A large study from the USA suggested that 36% of students
reported symptoms and were at risk of developing at least one sleep
disorder (Petrov, Lichstein, & Baldwin, 2014). One such potential
parasomnia that has been significantly under-researched, especially
Received:2Januar y2020
|
Revised:2 5February2020
|
Accepted:18March2020
DOI : 10.1111 /js r.130 44
REGULAR RESEARCH PAPER
Exploding head syndrome, chronotype, parasomnias and
mental health in young adults
Emma Kirwan | Donal G. Fortune
Department of Psychology, University of
Limerick, Limerick, Ireland
Correspondence
DonalG.Fortune,Departmentof
Psychology, University of Limerick,
Castletroy, Limerick, Ireland.
Email: donal.fortune@ul.ie
Abstract
Although inadequate sleep among young people is well documented in the literature,
anomalous sleep experiences, such as the parasomnia termed exploding head syn-
drome (EHS), have received little empirical attention. The current study examined
the association of sleep quality, symptoms of psychological distress and other unu-
sual sleep experiences with EHS in a sample of young adults (n = 135, age M = 21.77,
SD=2.08).Wealsoaimedtoaccountforthepossibleeffectofparticipantchrono-
type on sleep experiences. The lifetime prevalence of EHS among participants was
20.0%. Three-quarters (75.6%) of participants reported poor quality sleep according
to the Pittsburgh Sleep Quality Index (PSQI). Univariate analysis showed that partici-
pants with a lifetime prevalence of EHS experienced more symptoms of anxiety and
poorer sleep quality; age, gender and symptoms of depression were not significantly
related to EHS. Parasomnias (OR [95% CI] = 1.62 [1.02–2.57], p = .040) and action-
relatedsleepdisorders (OR [95%CI]=1.87[1.09–3.20],p = .023) were associated
with lifetime experience of EHS in a logistic regression analysis. Chronotype did not
significantly impact mood, sleep quality or presence of EHS. Results suggest that
EHS is more common in young people than previously considered and ought to be
examined in conjunction with the presence of other unusual sleep disorders. This
study provides valuable insight into young peoples' sleep experiences and key factors
associated with EHS.
KEYWORDS
chronotype, distress, exploding head syndrome, parasomnia, sleep disorder
2 of 10
|
KIRWAN ANd FO RTUNE
in young adults, is exploding head syndrome (EHS). EHS is a sensor y
parasomnia characterized by sudden sensations of an explosive loud
soundoccurringinthehead(AmericanAcademyofSleepMedicine,
2014). Although painless, the episodes have an awakening, usually
jarring, effect on the suf ferer and may induce a great deal of distress
(Sharpless, 2014).
Exploding head syndrome is likely to be under-reported and may
be more common than previously thought (Denis, Poerio, Derveeuw,
Badini, & Gregory, 2019; Sharpless, 2015). It is possible that the ap-
parent rarity of EHS is attributable to its unusual symptoms, which
patients might be reluctant to present to healthcare professionals
asthesolecomplaint(Pearce,1989).Only11%ofasampleofyoung
people (n = 304) reported EHS to medical professionals (Sharpless,
2018).The reporting of EHS mayalso beinfluenced by its benign
nature; reassurance regarding the harmlessness of the experience
appears to be more appropriate than medical treatment (Frese,
Summ, & Evers,2014).Fulda et al.(2008)reported 10.8% lifetime
prevalence of EHS among healthy participants (n=180).Morere-
cently, a much higher prevalence of 29.6% was reported (n = 1673)
(age, mean [M] = 34.17 years, standard deviation [SD] = 13.62) (Denis
et al., 2019).
Research suggest s a greater prevalence of EHS among females
(Rozen, 2004; Sharpless, 2014), although no significant difference
has also been reported (Denis et al., 2019; Sharpless, 2015). EHS
was thought to occur primarily in individuals over 50 years of age
(Pearce,1989;Rozen,2004);however,theexperiencemaybefairly
common(18.0%)amongyoungadults(n = 211) (age, M = 19.7 years,
SD = 1.77) (Sharpless, 2015). Given that individuals have reported
experiencing EHS as frequently as seven episodes a night (Sachs &
Svanborg, 1991) or as little as a single episode in a lifetime (Pearce,
1989),itisimportanttoconsiderboththecurrentandlifetimeprev-
alence of EHS.
Exploding head syndrome can demonstrate comorbidity with
other parasomnias, in particular sleep paralysis, with some 36.9% of
patients with sleep paralysis also reporting EHS (Sharpless, 2015).
Psychological factors such as unease and emotional tension may
increase the likelihood of an EHS episode (Sharpless, 2014). Little
systematic research has examined the associations of EHS, other ad-
verse sleep experiences and psychological distress par ameters; how-
ever, Denis et al. (2019) reported that neither anxiety, depression nor
stress predicted EHS in multiple logistic regression models among
two samples. Investigations of sleep must also include awareness
that individuals synchronize (entrain) differently depending on ex-
ternal factors such as light exposure and internal factors such as cir-
cadian responses that produce different chronotypes (Roenneberg,
Pilz, Zerbini, & Winneback, 2019). Chronotype can be considered on
a spectrum ranging from an ex treme early-type ‘morning chrono-
type’ to an extreme late-type ‘evening chronotype’: the remainder
of the population would fall into a category of ‘intermediate chro-
notype’. There is compelling research showing that the late chrono-
type and circadian misalignment (sleep jetlag) are associated with a
range of psychobiological challenges and may represent a transdi-
agnosticrisk factorforsuchchallenges(Taylor&Hasler,2018).We
hypothesize that this may include the experience of EHS for some
individuals. In other disorders, such as migraine for example, there is
nowevidenceforbothearly (Van Oosterhout etal.,2018)andlate
chronotype effects ( Viticchi et al., 2019). In EHS, there are no data
on possible chronotype effects as the role of chronotype in EHS re-
mains to be examined.
The purpose of this study was to examine the relationship be-
tween exploding head syndrome, sleep qualit y, symptoms of psy-
chological distress and other unusual sleep experiences. Considering
that peak sleep schedule ‘lateness’ is reached in young adulthood
(Roenneberg et al., 2004), we also aimed to examine participants'
preference for particular sleep behaviour times or chronot ype,
which may affect reported sleep experiences.
• H1. Individuals who indicate a lifetime prevalence of EHS will have
significantly higher anxiety and depression scores than those who
do not have such a prevalence.
• H2. Late chronotype will show an association with EHS, distress
and poor sleep quality.
• H3. Participants who have poor sleep quality will also have higher
anxiety and depression scores when compared to students with
good quality sleep.
• H4. There will be a significant difference in the sleep-quality
scores between participants who have experienced EHS and
those who have not.
• H5.Finally,weexpectthatlatechronotype,sleepquality,symp-
toms of depression and frequency of other adverse sleep expe-
riences will be associated with lifetime prevalence of EHS in a
logistic regression model.
2 | METHODS
2.1 | Participants and procedure
Approval for the study was obtained from the Education and Health
Sciences Ethics Committee at the University of Limerick.
We aimed to collect data for this study in the months following
students' examinations so as to reduce the potential impact of tran-
sient stress on measures of our key variables.
Students were recruited through a post on the University's
Department of Psychology social media page. The post detailed
that the researchers were seeking par ticipants for an “online sur-
vey about sleep qualit y, mood, sleep schedule and sleep issues”.
Participation was voluntary and those who provided informed
consent completed an online survey using the Unipark Questback
website (https://ww2.unipa rk.de/www). Participants generally com-
pleted the survey within 11 min.
A total of 149 participants took part in this cross-sectional
study. Nine participants were omitted from the final dataset as
data checks suggested that they may not meet the inclusion cri-
teria, which required participants to be (a) over 18 years of age
and (b) an undergraduate or postgraduate student. A further five
|
3 of 10
KIRWAN ANd FO RTUNE
participants were excluded from the final dataset as they re-
ported experiencing ‘severe pain’ during EHS episodes. According
to ICSD-3 diagnostic criteria for EHS (American Academy of
Sleep Medicine, 2014), the experience is not associated with
significant pain complaints. The final sample (n = 135) (age,
M = 21.77 ± SD=2.08years;91.9%undergraduate)included81
females (60.0%) and 54 males (40.0%).
2.2 | Materials
The online survey consisted of scales related to sleep quality, mood,
chronotype and sleep disorders: (a) the Pittsburgh Sleep Qualit y
Index (PSQI), (b) the Patient Health Questionnaire-4 (PHQ-4), (c)
theMunichChronotypeQuestionnaire (MCTQ)and(d)theMunich
Parasomnia Screening. Participants who indicated ongoing or pre-
vious experience(s) of EHS were filtered to a page containing five
items from the EHS Inter view (EHSI) (Sharpless, 2015). The survey
began with four items to address participant demographics, includ-
ing age, student type, gender, and an open-ended item to allow dis-
closure of chronic illness.
2.2.1 | Sleep Quality
ThePittsburghSleepQualityIndex(PSQI;Buysse,Reynolds,Monk,
Berman , & Kupfer, 1989) assessed pa rticipants' sleep quality and
regular sleep habit s over the past month with 19 self-rated items
(total score 0–21, scoring above 5 indicates poorer quality of sleep).
The measure has been well validated in a student sample (Carney,
Edinger,Meyer,Lindman,&Istre,2006)andhasdiagnosticspecific-
ity of 86.5%and sensitivityof89.6%in categorizing sleeperswith
goodandpoorqualityofsleep(Buysseetal.,1989).ThePSQIglobal
sleep quality score has shown good internal consistency across
its seven subscales, as well as good test–retest reliability (r = .85)
(Brown, Buboltz, & Soper, 2002). The measure showed strong reli-
ability in this study (α = .94).
2.2.2 | Psychological distress symptoms
The Patient Health Questionnaire-4 (PHQ-4) (Kroenke, Spitzer,
Williams, & Löwe, 2009) measured symptoms of anxiety and de-
pression over the past 2 weeks. The four-item measure consists
of a two-item anxiety scale (GAD-2) with the essential criteria for
generalized anxiety disorder (score range, 0–6), combined with
a two-item depression scale (PHQ-2) that measured the core di-
agnostic criteria for depressive disorders (score range, 0–6).
Research has suppor ted the validity and reliability of the PHQ-4 as
a brief measure of symptoms of anxiety and depression in young
adults (Khubchandani, Brey, Kotecki, Kleinfelder, & Anderson,
2016). The PHQ-4 had good internal consistency in the current
study (α=.84).
2.2.3 | Chronotype
TheMunichChronotypeQuestionnaire(MCTQ)(Roenneberg,Wirz-
Justice,&Merrow,2003)wasusedtomeasureparticipants'chrono-
type.TheMCTQiswellestablishedasavalidmeasure(Roenneberg
et al., 2019) and focuses on participants' typical sleep schedule over
the past 4 weeks considering work days and work-free days sepa-
rately. “College” was added after the word “work ”, such that the
questionnaire read “work/college” to avoid any confusion, given that
many students work while studying. Chronotype has been defined
as the midpoint between sleep onset and waking up on free days
(mid sleep on free days [MSF]) and was corrected for potential ex-
cessive free-day sleep time resulting from sleep deprivation during
work days (MSFsc; for correction algorithm, see supplement to the
study by Roennebergetal.,2004).Ahigher scoreonthe MSF indi-
cates a later chronotype (evening type) and a lower score indicates
an earlier chronotype (morning type). Participants were grouped into
distinc t categories of chronotype (Antypa et al., 2017) according to
their MSFs c score (early typ es: MSFsc ≤ 3.99; interme diate types:
MSFsc4.00–5.79;latetypes:MSFsc≥5.8).Thediscrepancybetween
work/college and free days served as a measure of “social jetlag”. It
wascomputedbysubtracting theMSW(midpointofsleep on work
days)fromtheMSF.
2.2.4 | Sleep disorder symptoms
Sleep disorder symptoms were measured using the Munich
ParasomniaScreening(MUPS)(Fuldaetal.,2008),a21-itemself-rat-
ing measure that assessed the lifetime prevalence and current fre-
quency of various nocturnal behaviours and sleep issues. Responses
ranged from (1), “no, never”, to (7), “very frequently - every or nearly
every night”, and also included a response that assessed lifetime
prevalence, (2), “previously observed years ago, but not now”. The
MUPShasshowngoodvalidityinassessingthelifetimeprevalence
of sleep disorder symptoms: sensitivity of the measure was above
90% for all but two of the items, where the answers given by pa-
tients inthe MUPSreflected responses in adetailedclinicalinter-
view. Specif icity was ab ove 80% for all item s (Fulda et al., 20 08).
The MUPS showed good internal consistencyin the current study
(α=.84).
2.2.5 | Exploding head syndrome
AsingleitemfromtheMUPSmeasuredlifetimeprevalenceofEHS
and was worded as follows: When falling asleep or waking, perceiv-
ing a loud bang, a sound similar to a bang (e.g., door bang), or having
the sensation of an “explosion in the head”. This item has been previ-
ouslyusedtoassesstheprevalenceofEHS(Denisetal.,2019;Fulda
etal.,2008).Participantswhoindicatedeverhavingexperiencedan
EHS episode (i.e., from (2) “was observed many years ago, but not
now” to (7) “very frequently - every or nearly every night ”) were
4 of 10
|
KIRWAN ANd FO RTUNE
directed to a subscale comprising five items from the EHS Interview
(EHSI), with the aim of facilitating further understanding of the ex-
perience. Participants were asked: Using this scale how would you
rate these sounds/sensations? Responses ranged from (0) “not at all”
to (9) “very severe”. The items enquired about how loud, how fright-
ening, how painful and how distressing the experience(s) is/are, as
well as how severely it disrupts the participants’ sleep. The subscale
had good internal consistency in this study (α=.89).
2.3 | Statistical analyses
The distribution of data was examined by the use of Q-Q plots and
histograms fitted to a normal curve. Examination of the plots sug-
gested th at independent variables , except depression score (PHQ-4),
approximated a normal distribution. Given the dat a for the depres-
sion variable showed evidence of a positive skew, a log transforma-
tion was undertaken to normalize the data and render it suitable for
use in parametric analyses.
The EHS variable was dichotomized into EHS absent and pres-
ent, due to its non-normal distribution, with EHS present indicating
that the participant had experienced one or more episode(s) in their
lifetime. Participants who selec ted (2) “behaviour was observed
years ago but not anymore” to (7) “very frequently - every or almost
every night)” were categorized as having lifetime prevalence of EHS.
Subsequently, this variable ser ved as the outcome variable in the bi-
nary logistic regression analysis.
In order to assess if lifetime prevalence and current frequency
of unusual sleep experiences were associated with a lifetime prev-
alenceofEHSinalogisticregressionanalysis,itemsfromtheMUPS
were factor analysed with the primar y aim of uncovering the under-
lying constructs upon which the questionnaire items were based in
our sample, such that any derived higher-order constructs might be
examined in the analysis. Thus, factor analysis was used to examine
whethertheexperiencesunderlyingMUPSrepresentedasingledi-
mension of experience or whether there were empirically different
experiences of MUPS thatmayhave importanceforunderstanding
EHS.
Relationships between EHS and sleep quality, social jetlag, and
symptoms of anxiety and depression were examined by an inde-
pendent samples t-test. Relationships among predictor variables
were examined by Pearson's correlation. Consequently, a logis-
tic regression analysis predicting prevalence of EHS from sleep
quality, symptoms of anxiety and unusual sleep experiences was
conducted.
3 | RESULTS
3.1 | Sleep quality
Over two-thirds (67.4%) of participant s rated their own sleep qual-
ity as “fairly good”, 9.6% rated their sleep as “very good” and 23.0%
of participants rated their own sleep as “fairly bad”. No participants
rated their own sleep quality as “very bad”. However, three-quarters
(75.6%, n = 102) of participants qualified as having poor quality of
sleep according to their scores on the PSQI, suggesting poor aware-
ness of sleep norms.
The mean sleep-quality score for the sample was 6.83
(SD = 2.07); higher PSQI scores are indicative of poorer quality
of sleep. The sleep-quality score was related to anxiety (r = .4 9,
p < .001) and depression (r = .37, p < .001), as well as the overall
distress score (r = .50, p < .0 01). Participants with a score >5 on the
PSQI were classified as poor-quality sleepers in accordance with the
cut-off scores provided by the developers of the measure (Buysse
etal.,1989)andwerecomparedtoparticipantsscoringbelowthis
cut-off score (good sleep quality). Participants classified as having
poor sleep quality (M = 2.39, SD=1.68)significantlydifferedinanx-
iety scores (PHQ-4) when compared to participants with good sleep
quality (M = 1.00, SD = 1.00), t(92.87)=−5.78,p < .0 01, d=−1.20.
Results also confirmed a significant difference in depression scores
(PHQ-4), where participants with poor sleep quality (M = 1.50,
SD = 1.43) had higher scores for symptoms of depression than stu-
dents with good sleep quality (M = 0.70, SD=0.85),t(133)=−3.05,
p = .003, d=−0.53.
TABLE 1 Lifetime prevalence of sleep disorders in the sample
(N = 135)
Sleep disorder Present , n (%) Absent, n (%)
Hypnic jerks 109(80.7%) 2 6 ( 19. 3% )
Nightmares 113(83.7%) 22 (16.3 %)
Sleep talking 97 (71.9 %) 38(28.1%)
Nocturnal leg cramps 73 (54 .1%) 62 (45.9%)
Rhythmic feet movements 59 (43.7%) 76 (56. 3% )
Hypnagogic hallucinations 63 (46.7%) 72 (53.3%)
Sleep terrors 53 (39.3%) 82(60.7%)
Nocturnal eating 53 (39.3%) 82(60.7%)
Sleep-related bruxism 48(35.6%) 87(64.4%)
Confusional arousals 47(34.8%) 88(65.2%)
Periodic leg movements 46 (34.1%) 89(65.9%)
Sleep-related groaning 41 (30.4%) 94 ( 69. 6% )
Sleep walking 33 (24.4%) 102 (75 .6%)
Rhythmic movement disorder 32 (23.7%) 103 (76. 3% )
REMsleepbehaviourdisorder 32 (23.7%) 103 (76. 3% )
Sleep paralysis 31 (23.0%) 10 4 (77. 0%)
Exploding head syndrome 27 (20.0%) 108(80.0%)
Violent behaviour 23 (17.0%) 112(83.0%)
Sleep enuresis 23 (17.0%) 113(83.0%)
Sleep-related abnormal
swallowing
18(13.3%) 117(86.7%)
Sleep-related eating 3 (2.2%) 132(97.8%)
Note: Present indicates a lifetime prevalence of the sleep disorder
symptom.REM,rapideyemovement.
|
5 of 10
KIRWAN ANd FO RTUNE
3.2 | Chronotype
MidsleeponfreedaysandMSFscshowedastrongpositiverelation-
ship (r=.87,p<.001).TheMSFscsuggestedthatthemajorityofpar-
ticipantsalignedwith an intermediatechronotype (57.8%, n = 78),
whereas 20% (n = 27) had an early chronotype and 22.2% (n = 30)
were of a late chronotyp e. Although late types were more likely to be
younger (F(2, 132) = 3.55, p = .031), they were not significantly more
likely to report EHS (χ2(1) = 0.312, p=.859).Furthermore,therewas
no significant effect of chronotype on gender (χ2(1)=2.98,p = .225).
Late types were also not significantly more anxious (F(2,132)=2.78,
p = .065) or depressed (F(2, 132) = 1.15, p = .317), did not report
greater frequency of sleep disorder symptoms (F(2, 132) = 1.54,
p = .165) and did not rate their sleep significantly more poorly on
the PSQI than early or intermediate types (F(2, 132) = 2.27, p = .107).
Similarly, presence of EHS was not significantly related to social jet-
lag (t(133)=−1.55,p = .121).
3.3 | Prevalence of sleep disorders
Almost all (99.3%) participants reported lifetime prevalence of one
ormoresleepdisordersmeasured intheMUPS; 47.4%ofthesam-
ple indicated lifetime prevalence of seven or more sleep disorders.
Lifetime prevalence rates for all sleep disorders measured in the
MUPSaredisplayedinTable1.Weconsideredthattheparticularly
high prevalence rate of sleep disorders may have been due to the
inclusion of participants who reported previous, non-current life-
time prevalence. Thus, a second analysis was conducted examining
only current or recent frequency of sleep disorders. This resulted in
aslightreductionofprevalencerates;97.8%ofstudentsindicatedat
least one current or recent sleep disorder symptom, whereas 27.6%
of participants indicated seven or more current sleep disorder symp-
toms (Table S1).
3.4 | Factor analysis of sleep disorders
The MUPS questionnaire was factor analysed in order to firstly
examine the underlying constructs upon which the questionnaire
items are based in this student sample, and secondly to see if it
was possible to reduce the variables to a smaller number of key as-
sessment dimensions that may be more meaningfully used in fur-
ther analysis. Given that correlation was expected among factors, a
Direct Oblimin rotation was performed. Bar tlett's' test of sphericity
confirmed the significance of all the correlations (χ2(190) = 643.03,
p<.0 01)andtheKaiser-Meyer-Olkinmeasureofs amplingadequacy
indicated adequate strength of the relationships among the vari-
ables (KM O = 0.74). Initiall y, the analysi s yielded six fac tors with
an eigenvalue greater than one. However, the eigenvalue one rule
(Kaiser's criterion) often overestimates the number of factors due to
well-recognised sampling effects (Cliff, 1988). This wasconfirmed
bya screeplot(FigureS1).Aseriesof factoranalysesresultedina
three-factor solution, which explained 39.6% of the variance. Two
items, “wetting oneself during sleep” and “sleep walking”, were elimi-
nated because they did not contribute to the factor structure. The
obtained pattern matrix is displayed in Table 2.
The first factor, labelled parasomnias, had an eigenvalue of 4.47
and accounted for 22.4% of the variance in the data. The second fac-
tor, labelled sleep-related movement disorders, had an eigenvalue
of 1.95 and accounted for a further 9.7% of the variance and factor
three, action-related sleep disorders, had an eigenvalue of 1. 50 and
accounted for a further 7.5% of the total variance.
3.5 | Frequency and qualitative aspects of EHS
One-fifth (20%, n = 27) of participants reported experiencing at least
one episode of EHS in their lifetime, whereas 15.6% of participants
reportedcurrentormorefrequentEHS.Figure1displaysthedistri-
bution of prevalence rates of EHS.
Participants generally reported that their EHS experiences were
mildly to moderately loud (M = 4.56, SD = 2.10), moderately fright-
ening (M = 5.15, SD = 2.66), not at all painful (M = 1.96, SD = 1.29),
mildly to moderately distressing (M = 4.56, SD = 2.4 4) and mildly
disruptive to sleep (M=3.89,SD = 2.17). Relationships bet ween the
TABLE 2 PatternmatrixofvariablesfromtheMunich
ParasomniaScreening(MUPS)(N = 135)
Scale items
Factor
1 2 3
Hypnagogic hallucinations 0.739
Sleep paralysis 0.646
Sleep-related abnormal
swallowing
0.508
Nightmares 0.478
Violent behaviour 0.429
REMsleepbehaviourdisorder 0.415
Sleep terrors 0.378
Periodic leg movements −0.721
Rhythmic feet movements −0.711
Hypnic jerks −0.571
Sleep-related bruxism −0.524
Rhythmic movement disorder −0.496
Nocturnal leg cramps −0.387
Sleep-related groaning 0.767
Sleep-related eating 0.642
Sleep talking 0.594
Confusional arousals 0.562
Nocturnal eating 0.528
Note: Factor1waslabelled‘‘parasomnias’’,Factor2waslabelled‘‘sleep-
relatedmovementdisorders’’andFactor3waslabelled‘‘action-related
sleepdisorders’’.REM,rapideyemovement.
6 of 10
|
KIRWAN ANd FO RTUNE
predictor variables and the qualitative aspec ts of EHS were exam-
ined using Spearman's rank-order correlations. Only anxiety was sig-
nificantly related to severit y of sleep disruption, indicating a weak
positive relationship (rs = .40, p = .038). (See Appendix S1 forthe
figures showing the distribution of responses for qualitative aspec ts
of EHS.)
3.6 | Univariate analysis of EHS associations
Gender (χ2(1)=0.008,p = .930), age (t(30.24)=−1.61,p=.118)and
depression score (t(133) = −1.94, p = .055) were not significantly
related to EHS. Six participants detailed ongoing illness, includ-
ing migraine, asthma, endometriosis and eczema; this was also not
significantly related to the presence of EHS (χ2(1)=0.68, p = .411).
Furthermore,chronotypewasnotassociatedwithEHS;latechrono-
types were not significantly more likely to report EHS (χ2(1)=0.83,
p = .659) than early or intermediate types.
Participants who reported experiencing EHS differed signifi-
cantly in anxiety score (t(133) = −3.03, p = .003, d = − 0.53) and
sleep-quality score (t(133)=−3.42,p = .001, d=−0.59)tothosewho
did not. The means and standard deviations for the key variables in
the study are shown in Table 3.
3.7 | Logistic regression analysis predicting lifetime
prevalence of EHS
A logistic regression analysis was conducted to predict lifetime prev-
alence of EHS using the variables that were significantly associated
with this experience in the univariate analyses above, as well as three
sleep disorder dimensions resulting from the factor analysis. Given
that anxi ety was related to seve rity of sleep disr uption associated w ith
EHS (EHS subscale) and that the sample had particularly poor-quality
sleep according to the PSQI, these variables were entered into the
first block of the regression model. Sleep disorders were entered into
the second block and binary coded EHS was the dependent variable.
The results indicated that the model reliably distinguished be-
tween EHS present and absent (chi-squared = 24.67, p < .001 with
df = 5) and explained 26.4% (Nagelkerke's R2) of the variance in EHS.
Predictionsuccess overallwas83.0% (97.2%for absentand25.9%
for present). The Wald criterion demonstrated that both parasom-
nias (OR [95% CI] = 1.62 [1.02–2.57], p = .040) and action-related
sleep disorders (OR[95%CI] = 1.87[1.09–3.20],p = .023) made a
significant contribution to prediction of EHS. Sleep-related move-
ment disorders, sleep quality and symptoms of anxiety were not
significant predictors for lifetime prevalence of EHS in the final
model (Table 4). The Exp(B) value indicated that as parasomnias are
raised by one unit (frequency of experience increases) the odds ratio
suggests that individuals are 1.62 times more likely to experience
EHS at least once in a lifetime. Similarly, results suggest that as the
frequency of experiences of action-related sleep disorders is raised
byoneunit,participantsare1.87timesmorelikelytohavelifetime
prevalence of EHS.
4 | DISCUSSION
This study found that two key dimensions of sleep disorder symp-
toms, labelled ‘parasomnias’ and ‘action-related sleep disorders’, were
related to the lifetime prevalence of EHS in a sample of young adults.
We believe this research to be among the first systematic studies on
associations of EHS, and at the time of writing it is the first to assess
the possible influence of chronotype on EHS in young adults.
The prevalence of EHS in our study is almost double that re-
ported involving healthy participants (10.8%) using the same
measure (Fulda et al., 2008), and higher than a previous esti-
mate among young adults (18.0%) (Sharpless, 2015). Contrary
to suggestions that EHS occurs primarily in older adults (Pearce,
1989;Rozen, 2004), 20.0%of our sample of young people (age,
FIGURE 1 Histogram showing the
distribution of frequency of exploding
head syndrome in the student sample
108 (80%)
6(4.4%)13 (9.6%)
4(3.0%)4(3.0%)
00
0
20
40
60
80
100
120
Behaviour
never
observed
by me or
others
Was
observed
years ago
but not
anymore
Very
seldom
Seldom SometimesFrequentlyVery
frequently
Response frequency
Frequency of exploding head syndrome
|
7 of 10
KIRWAN ANd FO RTUNE
M = 21.77, SD=2.08years)reportedtheexperience.Wereport
no significant difference in the prevalence of EHS among males
(20.4%)andfemales(19.8%).
Echoing previous work, sleep quality was associated with symp-
toms of both anxiety and depression. Par ticipants with poorer
sleep quality had significantly higher anxiety and depression scores
Variable Possible range
Overall EHS present EHS absent
Mean SD Mean SD Mean SD
Age (years) 18–34± 21.77 2.08 22.56 3.06 21.57 1.72
Sleep quality
(PSQI)
0–2 1 6.83 2.07 8.00 2.06 6.54 1.97
Anxiety (GAD-2) 0–6 2.05 1.65 2.89 1.67 1.84 1.59
Depression
(PHQ-2)
0–6 1.30 1.36 1.74 1.56 1 .19 1.29
Social jetlag
(MCTQ)
0–4:35± 1:14 0:53 1:28 1:02 1:10 0:50
Hypnic jerks 1–7 3.87 1.71 3.96 1.51 3.84 1.77
Rhythmic feet
movements
1–7 2.42 1.90 2.22 1.60 2 .47 1.99
Rhythmic
movement
disorder
1–7 1.63 1.29 1.60 1 .15 1.64 1.33
Hypnagogic
hallucinations
1–7 2.33 1.64 3. 26 1.89 2.09 1.49
Periodic leg
movements
1–7 2.07 1 .74 2.48 1.82 1.97 1.71
Nocturnal leg
cramps
1–7 2.37 1.49 2.41 1.50 2.36 1.49
Sleep-related
bruxism
1–7 2.17 1.87 2.63 2.34 2.06 1.73
Sleep talking 1–7 2.90 1. 57 3.44 1 .74 2 .76 1.50
Sleep-related
swallowing
1–7 1.37 1.09 1.56 1.16 1.32 1.08
Sleep-related
groaning
1–7 1.87 1.50 2.48 1. 76 1.71 1.40
Sleep enuresis 1–7 1 .19 0.47 1.26 0.45 1.18 0. 47
Nightmares 1–7 3.67 1.63 4.56 1.40 3.44 1.61
Sleep terrors 1–7 2.00 1.45 2.89 1.87 1.78 1 .24
Nocturnal eating 1–7 2.04 1.48 2.33 1.4 4 1.96 1.49
Sleep-related
eating
1–7 1.04 0.32 1 .19 0.68 1.01 0.10
Confusional
arousals
1–7 1.89 1.41 2.85 1.85 1.65 1.17
Sleep paralysis 1–7 1.53 1.12 1.96 1.43 1.42 1.00
Sleepwalking 1–7 1.36 0.73 1.49 0.58 1.33 0.76
REMsleep
behaviour
disorder
1–7 1.46 0.92 1.85 1.26 1.36 0.79
Violent
behaviour
1–7 1.34 0.85 1.85 1.26 1.30 0.79
Note.: ±: actual r ange. Values represent hr:min.
Abbreviations: EHS, exploding head syndrome; GAD-2, Generalized Anxiety Disorder-2 (a higher
scoreindicatesmoresymptomsofanxiety);MCTQ,MunichChronotypeQuestionnaire;PHQ-2,
Patient Health Questionnaire-2 (a higher score indicates more symptoms of depression); PSQI,
Pittsburgh Sleep Quality Index (a higher score indicates poorer quality of sleep).
TABLE 3 Descriptive statistics for all
variables included in the study
8 of 10
|
KIRWAN ANd FO RTUNE
compared to those with good sleep quality. Previous work has also
linked sleep quality with experiences of nocturnal disorders (Veldi,
Aluoja, & Vasar, 2005).In line with this well-est ablished finding, our
univariate analysis confirmed that participants who have experi-
enced EHS specifically have significantly poorer sleep quality com-
pared to those who have not experienced EHS. Recent research
(Denis et al., 2019) reported that insomnia symptoms were associ-
ated with EHS in a multiple predictor model; however, in our logistic
regression model poor sleep quality did not signif icantly predic t EHS.
It is worth noting this may be due to the use of different measures;
Denis et al. (2019) aimed to measure insomnia specifically, whereas
we obtained a measure of overall sleep quality.
We anticipated that the late chronotype might have an important
relationship with EHS, mood and sleep quality. However, chronotype
was not significantly associated with these key variab les in our study.
It has previously been demonstrated that a late chronotype is asso-
ciated with poor sleep and meta-analytic data show a small effect
size in relation to mood disorders (Au & Reece, 2017). It is difficult
to reconcile this particular finding with the absence of a relation-
ship between chronotype, self-reported sleep quality or distress in
our study. However, we have shown that participants showed poor
awareness of their poor sleep quality, which may account for the
absence of a significant relationship between the chronotype data
and self-reported sleep qualit y. It may also be the case that current
students' distress may be subsyndromal and as such may present
differently to formal mood disorders, with a likely effect on any po-
tential relationships between mood and chronotype.
We found that symptoms of anxiety were related to EHS in uni-
variate analyses. A lthough links between sleep disturbance and symp-
toms of psychological distress are strong (sleep disturbances have
been shown to be both risk factors for and a symptom of psycho-
logical distress) (Byrneetal.,2019;Nutt,Wilson,&Paterson,2008;
Perlis, Giles, Buysse, Tu, & Kupfer, 1997), we did not find a significant
relationship between symptoms of depression and EHS in our sample.
A logistic regression model showed that two dimensions of sleep
disorders labelled ‘‘parasomnias’’ and ‘‘action-related sleep disor-
ders’’, were significantly associated with the lifetime prevalence of
EHS among our sample. Our results are similar to those of Denis
et al. (2019), who reported sleep paralysis, nightmares and other
potentially distressing sleep experiences were associated with
the presence of EHS. Despite being significantly related to EHS in
univariate analyses, symptoms of anxiety were not significantly re-
lated to EHS in a logistic regression model. It remains a possibility
that both symptoms of anxiety and sleep quality have an indirec t
relationship to EHS. We considered that chronotype, particularly the
late chronotype, may impact the relationship; however, we did not
findanysignificantroleforchronotypeinthecurrentsample.Future
studies using longitudinal designs should examine this possibility
further in order to understand the associations of EHS.
4.1 | Implications of the research
Our findings suggest that EHS is related to and ought to be con-
sidered in conjunction with the presence of other anomalous sleep
experiences such as sleep paralysis, hypnagogic hallucinations and
nightmares. Effective treatment options for EHS are few, with reas-
surance and education deemed sufficient in most cases (Ganguly,
Mridha, Khan, &Rison,2013). In order to facilitate more targeted
interventions, we suggest it is necessary for EHS to be more fre-
quently included in clinical assessments alongside anomalous sleep
experiences.
An impor tant observation from this study is that students ap-
pear to be largely unaware of their poor sleep habits; the frequency
of poor-quality sleepers (75.6%) exceeds previous estimates (Becker
et al., 2018; Lund et al., 2010), suggesting that inad equate sleep
is present at high levels in our current sample. However, less than
one-quarter (23.0%) of students recognized their own sleep as being
inadequate. It may be helpful for universities and similar institutions
to acknowledge that students' sleep habits are a significant issue
that may warrant inter vention or at the very least, education.
5 | LIMITATIONS OF THE STUDY
There are some limitations of our study. The prevalence of sleep
disorder symptoms in our study was high. The study was adver-
tised to students through a brief social media post with a link to
the survey; although limited information was provided in the adver-
tisement, it contained the phrases “sleep quality”, “sleep schedule”
and “sleep issues”. Thus, we cannot rule out that the sample may
have had higher propor tions of young people with unusual sleep
experiences or poorer quality of sleep than the general student
population.
The use of self-report dat a must be cautiously interpreted.
However, substantial evidence supports the use of self-report esti-
matesofsleepandchronotype,andtheMUPSasanassessmentof
lifetimeprevalenceofEHShasbeenwellvalidated.Furtherresearch
may benefit from the use of adjunctive clinical inter views for a more
comprehensive measure.
Due to the cross-sectional design of this study, it remains un-
clear whether factors such as other adverse sleep experiences are
the cause of EHS, or conversely, that sleep disruption as a result of
EHS may influence other sleep disorders. Although our research is of
TABLE 4 Predictors of lifetime prevalence of exploding head
syndrome
Independent variable
Odds
ratio
Confidence
interval (95%)
Sig
p
Sleep-related movement
disorders
1.31 0.81–2 .11 .266
Parasomnias 1.62 1.0 2–2 . 57 .040
Action-related sleep disorders 1.87 1.09–3.20 .023
Sleep quality 1.29 0. 9 3–1.69 .068
Anxiety 1.1 2 0.80–1.57 . 515
|
9 of 10
KIRWAN ANd FO RTUNE
a similar size to a number of published studies in this area, the sample
size could be considered limited and therefore replication in a larger
international sample would be helpful.
6 | CONCLUSION
In sum, our f indings suggest the following. (a) EHS may be more prev-
alent than previously thought in younger people. (b) Chronotype is
not significantly associated with EHS. (c) Poor sleep quality is pre-
sent among a large number of young people, is poorly recognized
by young people and is associated with symptoms of anxiet y and
depression. (d) Symptoms of sleep disorder are common in college
students. (e) Participant s with a lifetime prevalence of EHS have
significantly poorer quality of sleep than those who do not. (f) Two
dimensions of unusual sleep experiences, parasomnias and action-
related sleep disorders, are significantly associated with the experi-
ence of EHS.
CONFLICT OF INTEREST
No conflicts of interest declared.
ORCID
Emma Kirwan https://orcid.org/0000-0001-8536-023X
Donal G. Fortune https://orcid.org/0000-0003-0916-3247
REFERENCES
AmericanAcademyofSleepMedicine(2014).International classification
of sleep disorders (3rd ed.). Darien, IL: American Academy of Sleep
Medicine.
Anty pa, N., Verkuil, B. , Molendijk, M ., Schoevers, R ., Penninx, B. W.,
&Van Der Does, W. (2017). Associations between chronotypes and
psychological vulnerability factors of depression. Chronobiology
International, 34(8), 1125–1135. https://doi.org/10.1080/07420
528.2017.1345932
Au, J., &Reece, J. (2017). The relationship between chronotype and de-
pressive symptoms: A meta-analysis. Journal of Affective Disorders,
218, 93–104. https://doi.org/10.1016/j.jad.2017.04.021
Becker, S. P.,Jarrett, M. A., Luebbe, A.M., Garner,A. A.,Burns,G. L.,
&Kofler, M. J. (2018) . Sleep in a large, mu lti-universit y sample of
college students: Sleep problem prevalence, sex differences, and
mental health correlates. Sleep Health, 4(2), 174–181. https://doi.
org/10.1016/j.sleh.2018.01.001
Brown, F.C.,Buboltz,W.C.Jr,&Soper,B. (2002).Relationship ofsleep
hygiene awareness, sleep hygiene practices, and sleep qualit y in
university students. Behavioral Medicine, 28(1), 33–38. ht tps://doi.
org/10.1080/08964280209596396
Buysse,D.J.,Reynolds,C.F.III,Monk,T.H.,Berman,S.R.,&Kupfer,D.
J.(1989).ThePittsburghSleepQualityIndex:A newinstrumentfor
psychiatric practice and research. Psychiatry Research, 28(2), 193–
213.https://doi.org /10.1016/0165-1781(89)90 047-4
Byrne , E. M., Timmer man, A., Wray, N. R ., &Agerb o, E. (2019). Sleep
disorders and risk of incident depression: A population case-control
study. Twin Research and Human Genetics, 22, 140 –146. https://doi .
org /10.1017/thg.2019.22
Carne y, C. E., Ed inger, J. D., Meyer, B., Lindm an, L., &Ist re, T. (2006).
Daily activities and sleep qualit y in college students. Chronobiology
International, 23 (3), 623–637.https://doi.org/10.1080/0742052060
0650695
Cliff, N. (1988). The eigenvalues-greater-than-onerule andthereliabil-
ity of components. Psychological Bulletin, 103(2), 276. https://doi.
org /10.1037/0033-2909.103.2.276
Denis,D.,Poerio,G.L.,Derveeuw,S.,Badini,I.,&Gregory,A.M.(2019).
Associations between exploding head syndrome and measures of
sleep quality and experiences, dissociation, and well-being. Sleep,
42(2), 1–11. https://doi.org/10.1093/sleep /zsy216
Frese, A ., Summ, O., &Ever s, S. (2014). Expl oding head s yndrome: Si x
new cases and review of the liter ature. Cephalalgia, 34(10),823–827.
htt ps://doi.org /10.1177/03331 02414 53 6059
Fulda, S. , Hornyak, M ., Müller, K., Ce rny,L ., Beitin ger,P. A., &Wett er,
T.C.(2008).DevelopmentandvalidationoftheMunichParasomnia
Screening (MUPS). Somnologie-Schlafforschung und Schlafmedizin,
12(1),56–65.https://doi.org/10.1007/s11818-008-0336-x
Ganguly,G., Mridha,B.,Khan,A.,&Rison,R.A .(2013).Explodinghead
syndrome: A case report. Case Reports in Neurology, 5(1), 14–17.
htt ps://doi.org /10.1159/00034 6595
Gregory,A. M., Buysse, D. J.,Willis, T.A., Rijsdijk, F.V.,Maughan, B.,
Rowe, R., … Eley, T. C. (2011). Associations between sleep quality
and anxiety and depression symptoms in a sample of young adult
twins and sibling s. Journal of Psychosomatic Research, 71(4), 250–255.
https://doi.org/10.1016/j.jpsyc hores.2011.03.011
Khubchandani, J., Brey, R., Kotecki, J., Kleinfelder, J., &Anderson,
J. (2016). The psychometr ic properties of PHQ-4 depression
and anxiety screening sc ale among college student s. Archives of
PsychiatricNursing, 30(4), 457–462. https://doi.org/10.1016/j.
apnu.2016.01.014
Kroenke, K., Spitzer, R. L., Williams, J. B., &Löwe, B. (20 09). An ul-
tra-brief screening scale for anxiety and depression: The PHQ–4.
Psychosomatics, 50(6), 613–621. https://doi.org/10.1016/S0033
-3182(09)70864-3
Lund, H. G., Reider, B. D., Whiting, A. B., &Prichard, J. R. (2010). Sleep
patterns and predictors of distur bed sleep in a large population
of college students. Journal of AdolescentHealth, 46(2), 124–132.
https://doi.org/10.1016/j.jadoh ealth.2009.06.016
Nutt,D.,Wilson,S.,&Paterson,L.(2008).Sleepdisordersascoresymp-
toms of depression. Dialogues in ClinicalNeuroscience, 10(3), 329.
Pearce, J. M. (1989).Clinicalfeatures of the exploding headsyndrome.
Journal of Neurolog y, Neurosurgery & Psychiatry, 52(7), 907–910.
https://doi.org/10.1136/jnnp.52.7.907
Perlis,M.L.,Giles,D.E.,Buysse,D.J.,Tu,X.,&Kupfer,D.J.(1997).Self-
reported sleep disturbance as a prodromal s ymptom in recurrent de-
pression. Journal of AffectiveDisorders, 42(2–3), 209–212. https://doi.
org/10.1016/S0165 -0327(96)01411 -5
Petrov,M.E.,Lichstein,K.L.,&Baldwin,C.M.(2014).Prevalenceofsleep
disorders by sex and ethnicity among older adolescents and emerg-
ing adult s: Relations to daytime functioning, working memory and
mental health. Journal of Adolescence, 37(5), 587–597. https://doi.
org/10.1016/j.adole scence.2014.04.007
Roenneberg, T., Kuehnle, T., Pramstaller, P. P., Ricken, J., Havel, M.,
Guth,A.,&Merrow,M.(2004).Amarkerfortheendofadolescence.
Current Biology, 14(24), R1038–R1039. https://doi.org/10.1016/j.
cub.2004.11.039
Roenneberg, T., Pilz, L . K., Zerbini, G ., &Winnebeck, E. C. (2019).
Chronot ype and social jetlag: A (self-) critical review. Biology, 8(3), 54.
https://doi.org/10.3390/biology8030054
Roenneberg, T., Wirz-Justice, A., &Merrow, M. (2003). Life between
clocks: Daily temp oral pat terns of human chronotypes. Journal of
BiologicalRhythms, 18(1), 80–90. https://doi.org/10.1177/07487
30402 239679
Rozen, T. D. (2004). Short-lasting headache syndromes and treatment
options. Current Pain and Headache Reports, 8(4),268–273.https://
doi.org/10.1007/s1191 6-004-0006-z
10 of 10
|
KIRWAN ANd FO RTUNE
Sachs, C ., &Svanborg, E. (1991). The exploding head syndrome:
Polysomnographic recordings and therapeutic suggestions. Sleep,
14(3), 263–266. https://doi.org/10.1093/sleep /14.3.263
Schwarz,J.,Axelsson,J.,Gerhardsson,A.,Tamm,S.,Fischer,H.,Kecklund,
G., & Åkerstedt, T. (2019). Mood im pairment is stro nger in young
than in older adults after sleep deprivation. Journal of Sleep Research,
28(4),e12801.https://doi.org/10.1111/jsr.12801
Sharpl ess, B. A. (2014). Ex ploding head sy ndrome. Sleep MedicineReviews,
18(6),489–493.https://doi.org/10.1016/j.smrv.2014.03.001
Sharpless, B. A . (2015). Exploding head syndrome is common in col-
lege students. Journal of Sleep Research, 24(4), 447–449. https://doi.
org /10.1111/jsr.12292
Sharpless,B.A.(2018).Characteristicsymptomsandassociatedfeatures
of exploding head syndrome in undergraduates. Cephalalgia, 38(3),
595–599.https://doi.org/10.1177/0333102417702128
Shochat,T.,Cohen-Zion,M.,&Tzischinsky,O. (2014).Functionalconse-
quences of inadequate sleep in adolescents: A systematic review.
Sleep MedicineReviews, 18(1), 75–87. https://doi.org/10.1016/j.
smrv.2013.03.005
Sin, N. L ., Almeida, D. M., Crain, T. L., Kossek, E . E., Berkman, L . F.,
&Buxton, O. M. (2017). Bidirectional, temporal associations of
sleep with positive events, affect, and st ressor s in daily life across
a week. Annals of Behavioral Medicine, 51(3), 402–415. https://doi.
org/10.1007/s12160-016-9864-y
Sivertsen, B., Vedaa, Ø., Harvey, A. G., Glozier, N., Pallesen, S., Aarø, L.
E.,…Hysing,M.(2019).Sleeppatternsandinsomniainyoungadults:
A national survey of Norwegian university students. Journal of Sleep
Research, 28(2), e12790. https://doi.org/10.1111/jsr.12790
Taylor, B. J., &Hasler, B. P. (2018). Chronotype and mental health:
Recent advances. Current Psychiatry Reports, 20(8), 59. https://doi.
org/10.1007/s11920-018-0925-8
Van Oosterhout, W. P. J., Van Someren, E. J. W., Schoonman, G. G.,
Louter,M.A.,Lammers,G.J.,Ferrari,M.D.,&Terwindt,G.M.(2018).
Chronotypes and circadian timing in migraine. Cephalalgia, 38(4),
617–625.htt ps://doi.org/10.1177/0333102417698953
Vandeputte, M., &deWeerd, A. (2003). Sleep disorders and depres-
sive feelings: A global survey with the B eck depression scale.
Sleep Medicine, 4(4), 343–345. https://doi.org/10.1016/S1389
-9457(03)00059 -5
Veldi,M.,Aluoja,A .,&Vasar,V.(2005).Sleepqualit yandmorecommon
sleep-related problems in medical students. Sleep Medicine, 6(3),
269–275. https://doi.org/10.1016/j.sleep.2004.12.003
Viticchi,G.,Falsetti,L.,Paolucci,M.,Altamura,C.,Buratti,L.,Salvemini,
S., …Silvestrini, M. (2019). Influence of chronotype on migraine
characteristics. Neurological Sciences, 40(9),1841–1848. https://doi.
org/10.1007/s10072-019-03886-4
SUPPORTING INFORMATION
Additional suppor ting information may be found online in the
Supporting Information section.
How to cite this article:KirwanE,FortuneDG.Exploding
head syndrome, chronotype, parasomnias and mental health
in young adults. J Sleep Res. 2020;00:e13044. h t tp s : //d o i.
org /10.1111/j sr.13 044
