93

British Journal of Health Psychology (2016), 21, 93113

2015 The British Psychological Society
www.wileyonlinelibrary.com

The role of emotion dysregulation in insomnia:

Longitudinal findings from a large community
sample
Markus Jansson-Fr
ojmark1,2*, Annika Norell-Clarke1,3 and
Steven J. Linton1
1

Center for Health and Medical Psychology (CHAMP), School of Law, Psychology and

Social Work, Orebro

University, Sweden
2
Department of Psychology, Stockholm University, Sweden
3
Centre for Research on Child and Adolescent Mental Health, Karlstad University,
Sweden
Objectives. The purpose of this longitudinal investigation was to examine the
association between emotion regulation and future insomnia (incidence and persistence).
Design. A longitudinal study in the general population.
Methods. A survey was sent out to 5,000 individuals in the community. To those who
returned the baseline questionnaire (n = 2,333), two follow-up surveys, 6 and 18 months
later, were sent out and then completed by 1,887 and 1,795 individuals, respectively. The
survey contained information about demographic factors, insomnia symptomatology, the
Difficulties in Emotion Regulation Scale, anxiety, and depression.
Results. The findings suggested that emotion regulation at baseline was not associated
with the incidence or persistence of insomnia. Overall, the effect sizes were very small to
medium. When examining changes in emotion regulation over time, a different pattern
emerged. Partial support was established for the notion that decreases in emotion
regulation were related to incident and persistent insomnia, as a decrease in emotion
regulation was associated with a higher likelihood of future insomnia. Yet, the effect sizes
were very small to small.
Conclusion. This study does partly point towards a longitudinal association between
emotion dysregulation and insomnia. This might have implications for the conceptualization and management of insomnia as well as for future research.

Statement of contribution
What is already known on this subject?
 Previous research has indicated that emotion dysregulation might be enhanced in patients with
insomnia.
 A number of limitations have however hindered progress in understanding how emotion
dysregulation is related to insomnia, such as limited research on the topic and relying solely on
cross-sectional data.

*Correspondence should be addressed to Markus Jansson-Frojmark, Department of Psychology, Stockholm University, 106 91
Stockholm, Sweden (email: markus.jansson-frojmark@psychology.su.se).
DOI:10.1111/bjhp.12147

94

Markus Jansson-Frojmark et al.

What does this study add?

 The current investigation showed that emotion dysregulation is a risk factor for the development of
incident and persistent insomnia.
 This study also shows that increased emotion dysregulation over time heightens the risk of incident
and persistent insomnia.

Insomnia as a diagnosis is a chronic difficulty that involves problems getting to sleep,

maintaining sleep, or waking in the morning not feeling restored, along with daytime
impairment (Edinger et al., 2004), a condition reported by 6% of the general population
(Ohayon, 2002). The consequences for those with insomnia are severe and include
functional impairment, work absenteeism, impaired concentration and memory, and
increased use of medical services (Roth & Ancoli-Israel, 1999). Mounting evidence
suggests that sleep and emotions may be closely related. Research shows, for example,
that sleep physiology may be involved in the modulation, regulation, and preparation of
emotional brain processes (Walker, 2009). There is also evidence that daytime emotional
stress has a negative impact on sleep physiology, particularly REM sleep and arousals, but
also affecting dreams (Vandekerckhove & Cluydts, 2010). Because sleep seems to be
critical for emotion regulation and because sleep loss has serious repercussions for nextday emotional functioning (Walker, 2009), it could be argued that insomnia and emotions
are also importantly linked.
One consistent finding is that those with insomnia are emotionally distressed during
the day, such as displaying heightened anxiety, dysphoria, and depression (Edinger, Stout,
& Hoelscher, 1988). Another, related finding is that there is a high comorbidity between
insomnia and psychiatric disorders, particularly with anxiety and mood disorders, both of
which can be characterized as having a strong emotional component (Harvey, 2008). The
question is, however, how the association between insomnia and emotions might be
viewed theoretically. One early theoretical account was based on the findings that
internalizing problems were more typical than externalizing difficulties among those with
insomnia, and Kales, Caldwell, Preston, Healey, and Kales (1976) proposed the
internalization of conflicts model of insomnia, in which the tendency to internalize
conflicts results in elevated emotional arousal, which influences physiology and the ability
to sleep. Other, more current, aetiological theories of insomnia view heightened
emotional arousal to be a stable characteristic of patients with insomnia (Espie, 2002;
Perlis, Giles, Mendelson, Bootzin, & Wyatt, 1997; Riemann et al., 2010). Perlis et al.
(1997) have proposed that insomnia is a result of enhanced cortical hyperarousal,
experienced subjectively by patients as cognitive hyperarousal. Espie (2002) has instead
posited that affect dysregulation mediates the effect of cognitive and autonomic
hyperarousal on sleep. Riemann et al. (2010) proposed alternatively that a bottom-up
process may be involved in the aetiology of insomnia: A genetically determined sleep
wake dysfunction regulating neural circuitries, in conjunction with precipitating
stressors, may lead to sleep disruption as well as to cognitive and emotional problems.
In sum, there are several notions about how insomnia and emotions might be linked.
Another account, with potential heuristic value, has posited that emotion dysregulation is a core phenomenon and might also maintain psychopathology. Emotion
regulation has been conceptualized as the processes by which individuals influence
which emotions they have, when they have them, and how they experience and express
these emotions (Gross, 1998). In all, emotion regulatory processes are viewed as resulting
in a critical outcome for human beings and other organisms; they decrease, increase, or

Emotion dysregulation and insomnia

95

maintain both negative and positive emotions. Based on conceptual and empirical work,
Gratz and Roemer (2004) have suggested that emotion regulation may be viewed as
involving four principal domains: (1) being aware of and understanding emotions, (2)
accepting emotions, (3) controlling impulsive behaviours and behaving in agreement
with desired goals when experiencing negative emotions, and (4) using contextually
appropriate regulation strategies flexibly to modulate emotional responses as desired to
meet goals and situational demands. Gratz and Roemer (2004) also postulate that the
relative absence of any or all of these four abilities would indicate the presence of emotion
dysregulation. In relation to clinical psychology, emotional dysregulation is involved in
the majority of the common psychiatric disorders (e.g., anxiety disorders) and in all of the
personality disorders (American Psychiatric Association, 2000; Gross, 1998). There is also
evidence that various emotion regulatory processes might maintain psychiatric disorders
(Aldao, Nolen-Hoeksema, & Schweizer, 2010).
Based on the literature encompassing a wide range of psychiatric disorders, the question
that we pose in this study is whether emotion dysregulation is associated with the incidence
and persistence of insomnia. Although several studies and theories have pinpointed the link
between insomnia and emotions in general, very little empirical work has looked into
whether emotion dysregulation plays a part in insomnia (Gruber, Eidelman, & Harvey, 2008;
Violani, Battagliese, Pisanti, & Lombardo, 2012; Waters, Adams, Binks, & Varnado, 1993). An
early study in this area found evidence that people with insomnia are more emotionally
reactive to and take longer to recover from stressors during the day, suggesting that
emotional reactivity might result in distress (Waters et al., 1993). In another investigation,
self-report scales indexing rumination, worry, and negative automatic thoughts were used
(Gruber et al., 2008). The results demonstrated that both rumination and worry, with small
to moderate effect sizes, were enhanced in those with insomnia, relative to a non-clinical
control group, indicating that rumination and worry might maintain insomnia. In a
longitudinal study on non-insomnia participants, emotion dysregulation (i.e., reappraisal
and suppression) was examined in relation to insomnia status 2 months after the initial
baseline assessment (Violani et al., 2012). The findings showed that when negative affect
and negative life events were controlled for, neither reappraisal nor suppression was
associated with future insomnia. In all, previous studies provide mixed findings; while two
investigations indicate that emotion regulation might be impaired in those with insomnia, a
third longitudinal study using covariates in the analyses does not show a significant link
between emotion regulation strategies and insomnia.
In summary, it appears that emotion dysregulation might play a part in insomnia. A
number of limitations have, however, hindered progress in understanding how emotion
dysregulation is related to insomnia. First, the number of investigations on the topic is very
limited. Second, given the small sample sizes in the previous studies, low statistical power
is an obvious limitation. Third, previous investigations have investigated self-identifying
individuals with insomnia or no insomnia. However, as such individuals might differ from
the general population (Ancoli-Israel & Roth, 1999; Morin, LeBlanc, Daley, Gregoire, &
Merette, 2006), the generalizability of findings based on those who self-identify and those
who do not is unknown. Fourth, although cross-sectional studies are important as a first
step when examining links between emotion regulation and insomnia (Gruber et al.,
2008), a longitudinal study is a methodologically stronger option, as it is possible to draw
more robust conclusions from such investigations concerning temporal sequences. Fifth,
no study, to our knowledge, has investigated whether emotion dysregulation plays a part
in the long-term persistence as well as incidence or remission of insomnia. Studying all of
these questions is important not only to provide information about what might drive

96

Markus Jansson-Frojmark et al.

persistent insomnia but also to identify possible aetiological factors. The present
investigation was thus designed to examine the longitudinal association between emotion
regulation and the incidence as well as the persistence of insomnia in a large sample from
the general population, using standard diagnostic criteria to define insomnia (Edinger
et al., 2004) and employing screening for other sleep disorders. This study is important as
it is designed to identify risk factors for incident and persistent insomnia, which, in turn,
might inform theory and provide insights into possible preventative interventions and
therapeutic options.
The purpose of this study was to examine the longitudinal association between
emotion regulation and insomnia. More specifically, the investigation had five aims. The
two-first aims focused on the incidence of insomnia by exploring the association between
emotion regulation at baseline and the incidence of insomnia, and by examining the
association between changes in emotion regulation over time and the incidence of
insomnia, respectively. The three last aims focused on the persistence of insomnia.
Specifically, the third aim was to examine the relationship between emotion regulation at
baseline and persistence of insomnia, relative to those with persistent normal sleep, while
the fourth aim was to examine the relationship between emotion regulation at baseline
and persistence/remission of insomnia, among those with insomnia at baseline. The fifth
and last aim was to investigate the association between changes in emotion regulation
over time and persistence/remission of insomnia. Based on the previous research in
insomnia and other forms of psychopathology, our hypothesis was that emotion
regulation and deteriorations in emotion regulation would be associated with increased
risk for persistent insomnia. No formal hypothesis was formulated concerning the
association between emotion regulation and incident insomnia. In all the analyses, sociodemographic parameters and anxiety and depression were statistically controlled as
potential confounding factors (Violani et al., 2012).

Methods
Overview of the study
This research is part of the Prospective Investigation on Psychological Processes for
Insomnia (PIPPI) study and was approved by the Regional Ethics Board in Uppsala,
Sweden. A survey consisting of sleep-related questionnaires was mailed to the general
public on three occasions over 18 months.

Procedure
The baseline questionnaire was mailed to the study participants. No response on the
survey or the response promoting efforts within 2 weeks after was considered a
declination to participate. Respondents at Time 1 were sent a questionnaire at Time 2, and
at Time 3. To increase the response rates at the three assessment points, a number of steps
were taken in line with a recent Cochrane review (Edwards et al., 2007). We sent a prenotification letter, included a small incentive with the questionnaires (a pen or a lottery
ticket), sent an information letter describing the project including its aim and importance,
used a multicoloured and user-friendly questionnaire, used closed questions and placed
relevant and easy questions first, attached a prepaid return envelope, sent reminders to
non-responders (including a new questionnaire), and provided an assurance of
confidentiality.

106

Markus Jansson-Frojmark et al.

Table 6. Emotion regulation strategies at baseline between those with persistent insomnia and those
with remission from insomnia: Descriptive statistics and logistic regression analyses
M (SD)

B (SE)

Insomnia at T1 and T2 (n = 150) Remission: I at T1 and NS at T2 (n = 62)

Unadjusted analysis
Emotion regulation (T1)
REM: 24.0 (7.2)
.01 (0.02)
I: 24.4 (7.1)
Adjusted analysis
Emotion regulation (T1)
REM: 24.0 (7.2)
.03 (0.03)
I: 24.4 (7.1)
Gender: Female (T1)
REM: 59.7%
.40 (0.33)
I: 66.0%
Civil status: Living alone,
REM: 25.8%
.14 (0.36)
divorced or widowed (T1)
I: 32.0%
Ethnicity: Not born in Sweden (T1)
REM: 9.7%
.11 (0.56)
I: 9.7%
Anxiety (T1)
REM: 4.8 (3.6)
.06 (0.06)
I: 5.9 (3.4)
Depression (T1)
REM: 4.0 (3.4)
.10 (0.06)
I: 5.3 (3.5)
Insomnia at T1 and T3 (n = 127) Remission: I at T1 and NS at T3 (n = 62)
Unadjusted analysis
Emotion regulation (T1)
REM: 23.9 (7.2)
.02 (0.02)
I: 24.9 (7.5)
Adjusted analysis
Emotion regulation (T1)
REM: 23.9 (7.2)
.01 (0.03)
I: 24.9 (7.5)
Gender: Female (T1)
REM: 51.6%
.84 (0.34)**
I: 68.5%
Civil status: Living alone,
REM: 21.0%
.50 (0.38)
divorced or widowed (T1)
I: 33.1%
Ethnicity: Not born in Sweden (T1) REM: 12.9%
.44 (0.57)
I: 8.1%
Anxiety (T1)
REM: 4.9 (3.6)
.05 (0.05)
I: 6.1 (3.7)
Depression (T1)
REM: 4.2 (3.8)
.08 (0.06)
I: 5.5 (3.5)

R2

OR

95% CI

1.01

0.971.05

.01

0.98

0.931.03

.07**

1.50

0.782.88

1.15

0.572.31

1.11

0.373.33

1.06

0.951.19

1.10

0.991.23

1.02

0.981.06

.01

1.00

0.951.05

.11

2.32

1.194.53

1.65

0.783.49

0.65

0.211.96

1.05

0.941.17

1.08

0.971.20

Note. CI = confidence interval; I = insomnia; M = mean; NS = normal sleep; REM = remission;

SD = standard deviation.
**Significant at the .01 level.

Emotion regulation and persistence/remission of insomnia

The fourth aim was to investigate the association between emotion regulation at baseline
and persistence/remission of insomnia, among those with insomnia at baseline. The
descriptive statistics for the DERS at baseline across the two insomnia groups related to
persistence/remission (persistent insomnia and remitted insomnia) are displayed in
Table 6. As can be seen in the table, emotion regulation (d = 0.06 for Time 1Time 2 and
d = 0.14 for Time 1Time 3) was not significantly related to persistent insomnia, relative
to remitted insomnia, neither in unadjusted nor in adjusted analyses. In one of the adjusted

98

Markus Jansson-Frojmark et al.

consisted of significantly more women, more divorced, widowed or single people, and
more people born outside Sweden.

Measures
The following demographic parameters were assessed: Age, gender, civil status, level of
education, vocational status, and ethnicity. For all the measures described in the section
below, the participants were asked to respond based on the past month.

Night-time symptoms
To assess night-time symptoms, research diagnostic criteria were used to determine
symptoms that are typical for insomnia (Edinger et al., 2004). The participants were asked
to complete the following categorical questions based on the previous month: Sleep onset
latency (SOL; <15, 1630, 3160, and >60 min), wake time after sleep onset (same
alternatives as for SOL), early morning awakening (same alternatives as for SOL), total
sleep time (<4, 45, 56, 67, 78, 89, 910, and >10 hr), sleep restoration (completely
[1], a lot [2], somewhat [3], a little [4], not at all [5]), and sleep quality (very good [1], quite
good [2], neither good nor poor [3], quite poor [4], very poor [5]). To determine sleep
disturbance, the participants were asked to complete the following question: Sleep
disturbance during the past month (yes or no). In addition, the Insomnia Severity Index
(ISI; score range 028 points), a 7-item scale, was also included (Bastien, Vallieres,
& Morin, 2001).

Daytime symptoms
To determine daytime symptoms, research diagnostic criteria were used to assess
symptoms that are typical for insomnia (Edinger et al., 2004). The participants were
asked to report on the degree of sleep-related impairment during the previous month:
Fatigue/malaise, impairment in attention, concentration, or memory, social dysfunction, vocational dysfunction, mood disturbance, irritability, daytime sleepiness,
reduction in motivation, energy, or initiative, proneness for errors or accidents at
work or while driving, tension headaches, gastrointestinal symptoms, and concerns or
worries about sleep (Edinger et al., 2004). The response alternatives for these
indications of daytime impairment were as follows: Not at all (1), somewhat (2), quite
much (3), and a lot (4). The response alternatives for the two functional domains were
as follows: No negative consequences (1), small negative consequences (2), marked
negative consequences (3), large negative consequences (4), and very large negative
consequences (5).

Emotion regulation
To index emotion regulation, the Difficulties in Emotion Regulation Scale (DERS) was
used (Gratz & Roemer, 2004). The DERS is a multidimensional index of emotion
regulation and consists of six subscales: Non-acceptance of emotional responses (nonacceptance), difficulties engaging in goal-directed behaviour (goals), impulse control
difficulties (impulse), lack of emotional awareness (awareness), limited access to
emotion regulation strategies (strategies), and lack of emotional clarity (clarity).
Participants are asked to indicate how often the items apply to themselves, with

Emotion dysregulation and insomnia

99

responses ranging from 1 to 5, where 1 is almost never (010%), 2 sometimes (1135%),

3 about half the time (3665%), 4 most of the time (6690%), and 5 almost always (91
100%). The DERS items were recoded so that higher scores in every case indicated
greater difficulties in emotion regulation (i.e., greater emotion dysregulation). The
internal consistency of the DERS and its subscales is high (a > .80). Analyses over time
have shown that the testretest reliability of the DERS and its subscales is adequate or
good. Considering that the entire survey consisted of 243 items, the DERS was shortened
to reduce the risk of a low response rate from 36 to 12 items based on the original paper
(Gratz & Roemer, 2004). To include all six of the subscales, the two items that loaded the
highest onto each subscale (.671.00) were used in the current survey. The following 12
items from the DERS were used to index emotion regulation (with the number from the
original DERS scale, its factor loading and factor in parenthesis): When Im upset, I feel
ashamed with myself for feeling that way (item 25; .76; non-acceptance), When Im
upset, I feel guilty for feeling that way (item 29; .91; non-acceptance), When Im upset,
I have difficulty focusing on other things (item 22; .88; goals), When Im upset, I have
difficulty concentrating (item 30; .88; goals), When Im upset, I have difficulty
controlling my behaviours (item 31; .79; impulse), When Im upset, I lose control over
my behaviours (item 37; 1.00; impulse), I pay attention to how I feel (item 3, reversed;
.67; awareness), I am attentive to my feelings (item 7, reversed; .74; awareness), When
Im upset, I believe that I will remain that way for a long time (item 19; .79; strategies),
When Im upset, I believe that Ill end up feeling very depressed (item 20; .86;
strategies), I have no idea how I am feeling (item 5; .71; clarity), and I have difficulty
making sense out of my feelings (item 6; .81; clarity). In an unpublished, pilot study
prior to the start of the project on 26 individuals (54% women, mean age: 27.2 years), a
high correlation (r = .96) between the 36-item and the 12-item DERS versions was
documented, indicating that the short DERS version that was used in this study is an
appropriate index of emotion regulation. The internal consistency of the total DERS
using this studys sample was high at a = .84.

Covariates: Socio-demographic parameters, anxiety, and depression

In this study, both unadjusted (using emotion regulation only) and adjusted analyses
(emotion regulation and covariates) were used. For two reasons, five covariates were
used in this study. First, socio-demographic parameters, anxiety, and depression have
previously been shown to be associated with insomnia (Ohayon, 2002). In the
current data set, we have established that three socio-demographic factors (i.e.,
gender, civil status, and ethnicity) as well as anxiety and depression are significantly
related to insomnia. Second, as emotion regulation might overlap with the five
covariates, the use of unadjusted and adjusted analyses is enabling for determining
how much of the prediction that emotion regulation explains is due to its overlap
with the covariates.
To assess socio-demographic parameters, gender (malefemale), civil status (cohabitant, married or having a partnerliving alone, divorced, or widowed), and ethnicity (born
in Swedennot born in Sweden) were assessed in the survey. The Hospital Anxiety and
Depression Scale (HADS) was used to assess anxiety and depression (Bjelland, Dahl, Haug,
& Neckelmann, 2002; Zigmond & Snaith, 1983). The HADS is a self-rating scale with 14
questions in which the severity of anxiety (HADS-A) and depression (HADS-D) is rated on
4-point scales (score range 021). Based on the current sample, the internal consistency
for the total scale was high at .90 (anxiety: .85, depression: .84).

100

Markus Jansson-Frojmark et al.

Sleep disorders
The SLEEP-50 was used to assess six DSM-IV-TR sleep disorders: Apnoea, narcolepsy,
restless legs/periodic limb movement disorder, circadian rhythm disorder, and sleep
walking (Spoormaker, Verbeek, van den Bout, & Klip, 2005). The instrument has high
internal consistency, testretest correlation ranging between .65 and .89, and a factor
structure that matches the DSM-IV-TR sleep disorders. The sensitivity and specificity
scores have been found to be reasonable for the sleep disorders (sensitivity: .671.00;
specificity: .691.00). The agreement between clinical diagnoses and classification
derived from the SLEEP-50 is substantial (j = .77) (Spoormaker et al., 2005). The
participants were asked to rate to what extent the items have been applicable during the
past month (1 = not at all, 4 = very much).

Sleep group classification

The participants were classified in distinct groups according to their sleep patterns,
daytime impairment, and evidence of sleep disorders other than insomnia. The
classification used an algorithm based on a combination of insomnia diagnostic
criteria from Research Diagnostic Criteria for insomnia (Edinger et al., 2004),
established quantitative criteria for insomnia (Lichstein, Durrence, Taylor, Bush, &
Riedel, 2003), and screening for sleep disorders other than insomnia (Spoormaker
et al., 2005).

Insomnia
A proxy for the DSM-IV-TR definition of insomnia was used. Participants had to confirm a
sleep disturbance during the last month and report initial, middle, or late insomnia
(>30 min awake involuntarily at any stage during an estimated average night) (Lichstein
et al., 2003). They also had to report some daytime impairment (a response of three or
more on one or more on the daytime symptoms). In addition, those classified as having
insomnia could not meet criteria for apnoea, narcolepsy, restless legs syndrome/periodic
limb movement disorder, circadian rhythm disorder, or sleepwalking as assessed with the
SLEEP-50. Using data from this study, we investigated the concordance between our
insomnia definition and two validated ISI cut-offs (Morin, Belleville, Belanger, & Ivers,
2011). The ISI measures subjective insomnia severity and has widely been used as a
measure of change after treatment. Score below eight is a cut-off for subthreshold
insomnia in both community and clinical samples, and the cut-off at 10, marking insomnia,
has been validated in clinical samples of patients with insomnia. The two ISI cut-offs that
have been validated have both shown nearly 98% correct classification rates differentiating clinical insomnia patients and normal sleepers. The concordance between the ISI
cut-offs and the insomnia classification in this study was high. The cut-off at eight points
correctly classified 99.4% of those with insomnia, and the ISI cut-off at 10 points correctly
classified 89.4% (Jansson-Fr
ojmark et al., 2012). The high concordance with an
established insomnia measure indicates that our insomnia definition captures the
insomnia construct to a high degree.

Normal sleep
The participants in this group had to deny a sleep disturbance during the last month and
not fulfil criteria for any sleep disorder as assessed with the SLEEP-50.

Emotion dysregulation and insomnia

101

Remission
The participants in this group fulfilled the criteria for insomnia at baseline but, at
subsequent assessments, met criteria for the normal sleep classification and did not fulfil
criteria for any sleep disorder as assessed with the SLEEP-50.

Incidence
The participants in this group did not meet criteria for insomnia at baseline but, at
subsequent assessments, fulfilled criteria for insomnia and did not fulfil criteria for any
sleep disorder as assessed with the SLEEP-50.

Statistical analyses
The data were first summarized to provide descriptive statistics. An analysis of outliers was
conducted employing various forms of detection (e.g., univariate and multivariate) and
outlier profiling (Hair, Anderson, Tatham, & Black, 1998).
To examine the differences on emotion regulation between the groups, multivariate
analyses were used in all analyses. As the indicator variables may be related to each other, a
model was determined using multivariate logistic regression analysis, that is including all
the independent variables in one step. As the indicator variables, the DERS and the five
covariates (i.e., gender, civil status, ethnicity, anxiety, and depression) were used. For
aims 2 and 5, changes in the DERS over time were used as indicator variables, calculated as
standardized residual change scores. The standardized residual is a way of expressing the
score at Time 2 as larger or as smaller than the score predicted linearly by Time 1 score
(Cronbach & Furby, 1970). This procedure is recommended rather than using raw change
scores, as standardized residuals take into account the score at Time 1, at the same time
adjusting for possible random errors of measurement (Stekette & Chambless, 1992). The
standardized residuals were computed by converting the raw scores to Z scores and were
further computed as follows: Z2 (Z1 9 r12) in which Z2 is the follow-up score, Z1 the
baseline score, and r12 the correlation between both ratings. For aims 1, 3, and 4, the DERS
scores at baseline were employed as indicator variables. As the outcome variable, group
status was used. Group status was categorized in different ways, depending on the aims
investigated: For aims 1 and 2, the groups were no insomnia (0) and insomnia (1); for aim
3, the groups were persistent normal sleep (coded as 0) and persistent insomnia (1); and
for aims 4 and 5, the groups were remitted insomnia (0) and persistent insomnia (1). All
the continuous independent variables were transformed into z-scores to enable
comparisons between odds ratios and confidence intervals. A two-tailed p-value smaller
than .05 was considered statistically significant. Nagelkerke R2 is also reported for the
logistic regression analyses. Between-group effect sizes were also calculated (baseline or
standardized residual change score for the group coded as 1 minus baseline or
standardized residual change score for the group coded as 0/pooled standard deviation).
Cohen (1988) proposed a threefold classification of between-group effect sizes: Small
(.20.49), medium (.50.79), and large (.80 and above).

Results
Descriptive statistics
Of the 2,333 baseline responders, 309 participants reported other sleep disorders and
were therefore omitted from the analyses. Of the 309 participants with a sleep disorder

102

Markus Jansson-Frojmark et al.

Table 2. Correlations between the study variables

1 Insomnia
2 Gender
3 Civil status
4 Ethnicity
5 Emotion regulation
6 Anxiety
7 Depression

.07*
.10**
.05*
.22**
.39**
.35**

.01
.01
.15**
.15**
.03

.01
.05
.08*
.08*

.08**
.07*
.09**

.46**
.36**

.61**

Note. Pearsons correlation coefficient and eta statistics were used. Insomnia: No (0) and yes (1); gender:
Male (0) and female (1); civil status: Cohabitant, married or having a partner (0) and living alone, divorced
or widowed (1); and ethnicity: Born in Sweden (0) and not born in Sweden (1).
*Significant at the .05 level.
**Significant at the .01 level.

other than insomnia, apnoea was fulfilled by 44.6%, narcolepsy by 2.9%, restless legs/
periodic limb movement disorder by 46.6%, circadian rhythm disorder by 23.1%, and
sleep walking by 0.2%. In all, 2,024 participants were included in this studys analyses. Of
those, 322 individuals (15.9%) fulfilled the criteria for insomnia at baseline. Over time,
4.95.6% met the criteria for incident insomnia. Further, 6.27.4% fulfilled the criteria for
persistent insomnia (i.e., insomnia at T1 and T2 or T1 and T3). Also, 2.93.3% met the
criteria for remitted insomnia over time. The above-mentioned rates for insomnia
resemble to a high degree a previous longitudinal population study (Jansson-Fr
ojmark &
Linton, 2008).
Correlations between the study variables are available in Table 2.

Emotion regulation and incidence of insomnia

The first aim focused on the incidence of insomnia by exploring the association
between emotion regulation at baseline and the incidence of insomnia. The
descriptive statistics for the DERS at baseline across the two insomnia groups related
to incidence (insomnia and no insomnia) are displayed in Table 3. As can be seen in
the table, higher (more dysfunctional) emotion regulation levels at baseline were
associated with increased risk for incident insomnia in one of two unadjusted analyses
(OR = 1.04); the effect sizes between the two groups on emotion regulation were
below the cut-off for being considered a small effect size (d = 0.26 for Time 1Time 2
and d = 0.17 for Time 1Time 3). In adjusted analyses, emotion regulation was not
associated with the incidence of insomnia. In both adjusted analyses, only baseline
depression was a significant risk factor for incident insomnia (d = 0.62 for Time 1
Time 2 and d = 0.55 for Time 1Time 3). The adjusted models explained 6% of the
variance in incident insomnia status.

Changes in emotion regulation and incidence of insomnia

The second aim aimed at examining the association between changes in emotion
regulation over time and the incidence of insomnia. The descriptive statistics for the DERS
change scores across the two insomnia groups concerning incidence (insomnia and no

Emotion dysregulation and insomnia

103

Table 3. Emotion regulation strategies at baseline between those who develop insomnia and those who
do not develop insomnia at subsequent time points: Descriptive statistics and logistic regression analyses
M (SD)

B (SE)

Not insomnia at T1 (n = 1,385): I at T2 (n = 68) NI at T2 (n = 1,317)

Unadjusted analysis
Emotion regulation (T1)
NI: 20.7 (5.8)
.04 (0.02)*
I: 22.3 (6.5)
Adjusted analysis
Emotion regulation (T1)
NI: 20.7 (5.8)
.01 (0.02)
I: 22.3 (6.5)
Gender: Female (T1)
NI: 53.2%
.23 (0.27)
I: 59.7%
Civil status: Living alone,
NI: 18.3%
.22 (0.32)
divorced or widowed (T1)
I: 20.9%
Ethnicity: Not born in Sweden (T1)
NI: 6.5%
.27 (0.56)
I: 6.3%
Anxiety (T1)
NI: 2.4 (2.5)
.05 (0.06)
I: 3.7 (3.4)
Depression (T1)
NI: 2.2 (2.4)
.15 (0.05)**
I: 3.9 (3.1)
Not insomnia at T1 (n = 1,339): I at T3 (n = 75) NI at T3 (n = 1,264)
Unadjusted analysis
Emotion regulation (T1)
NI: 20.8 (5.6)
.03 (0.02)
I: 21.8 (6.4)
Adjusted analysis
Emotion regulation (T1)
NI: 20.8 (5.6)
.01 (0.02)
I: 21.8 (6.4)
Gender: Female (T1)
NI: 54.0%
.22 (0.26)
I: 52.1%
Civil status: Living alone,
NI: 16.7%
.35 (0.31)
divorced or widowed (T1)
I: 22.5%
Ethnicity: Not born in Sweden (T1)
NI: 5.6%
.06 (0.56)
I: 5.8%
Anxiety (T1)
NI: 2.3 (2.4)
.11 (0.06)
I: 3.6 (3.1)
Depression (T1)
NI: 2.1 (2.4)
.13 (0.05)*
I: 3.6 (3.1)

R2

OR

95% CI

1.04

1.011.08

.01

1.01

0.971.06

.06**

1.26

0.742.16

1.25

0.672.34

0.77

0.262.29

1.05

0.951.17

1.17

1.061.29

1.03

0.991.07

.01

1.00

0.951.04

.06**

0.81

0.481.35

1.42

0.782.61

0.94

0.322.80

1.11

1.001.24

1.13

1.031.25

Note. CI = confidence interval; I = insomnia; M = mean; NI = not insomnia; OR = odds ratio;

SD = standard deviation.
*Significant at the .05 level.
**Significant at the .01 level.

insomnia) are displayed in Table 4. As can be seen in the table, a larger change (towards
more dysfunctional levels) in emotion regulation was related to increased risk for incident
insomnia in both unadjusted analyses (OR = 1.12); the effect sizes between the two
groups on changes in emotion regulation were small (d = 0.46 for Time 1Time 2 and
d = 0.44 for Time 1Time 3). In both adjusted analyses, a larger change (towards more
dysfunctional levels) in emotion regulation and baseline depression were significant risk
factors for incident insomnia. The adjusted models explained 10% of the variance in
incident insomnia status.

104

Markus Jansson-Frojmark et al.

Table 4. Changes in emotion regulation strategies between those who develop insomnia and those who
do not develop insomnia at subsequent time points: Descriptive statistics and logistic regression analyses
M (SD)
Not insomnia at T1 (n = 1,385): I at T2 (n = 68) NI at T2 (n
Unadjusted analysis
Emotion regulation (T1T2)
NI: 0.31 (5.08)
I: 1.92 (4.69)
Adjusted analysis
Emotion regulation (T1T2)
NI: 0.31 (5.08)
I: 1.92 (4.69)
Gender: Female (T1)
NI: 53.2%
I: 59.7%
Civil status: Living alone,
NI: 18.3%
divorced or widowed (T1)
I: 20.9%
Ethnicity: Not born in Sweden (T1) NI: 6.5%
I: 6.3%
Anxiety (T1)
NI: 2.4 (2.5)
I: 3.7 (3.4)
Depression (T1)
NI: 2.2 (2.4)
I: 3.9 (3.1)
Not insomnia at T1 (n = 1,339): I at T3 (n = 75) NI at T3 (n
Unadjusted analysis
Emotion regulation (T1T3)
NI: 0.32 (5.13)
I: 2.13 (6.01)
Adjusted analysis
Emotion regulation (T1T3)
NI: 0.32 (5.13)
I: 2.13 (6.01)
Gender: Female (T1)
NI: 54.0%
I: 52.1%
Civil status: Living alone,
NI: 16.7%
divorced or widowed (T1)
I: 22.5%
Ethnicity: Not born in Sweden (T1) NI: 5.6%
I: 5.8%
Anxiety (T1)
NI: 2.3 (2.4)
I: 3.6 (3.1)
Depression (T1)
NI: 2.1 (2.4)
I: 3.6 (3.1)

OR

95% CI

R2

.11 (0.02)**

1.12

1.071.18

.05**

.11 (0.03)**

1.11

1.061.17

.10**

.21 (0.28)

1.24

0.712.14

.24 (0.33)

1.28

0.672.41

.21 (0.55)

0.81

0.282.40

.02 (0.05)

1.02

0.921.13

.17 (0.05)*

1.19

1.071.31

.12 (0.02)**

1.12

1.071.17

.06**

.10 (0.02)**

1.11

1.061.16

.10**

.22 (0.27)

0.80

0.471.36

.30 (0.32)

1.35

0.722.51

.08 (0.55)

0.92

0.312.71

.07 (0.06)

1.08

0.961.20

.14 (0.05)*

1.15

1.041.26

B (SE)
= 1,317)

= 1,264)

Note. CI = confidence interval; I = insomnia; M = mean; NI = not insomnia; OR = odds ratio;

SD = standard deviation.
Standardized residuals were used as change scores in the analyses.
*Significant at the .05 level.
**Significant at the .01 level.

Emotion regulation and stability in insomnia

The third aim focused on examining the relationship between emotion regulation at
baseline and persistence of insomnia. The descriptive statistics for the DERS across the
two sleep groups (persistent insomnia and persistent normal sleep) are displayed in
Table 5. As can be seen in the table, higher (more dysfunctional) emotion regulation levels
at baseline (OR = 1.091.11) were associated with increased risk for persistent insomnia

Emotion dysregulation and insomnia

105

Table 5. Emotion regulation strategies at baseline between those with persistent insomnia and those
with persistent normal sleep: Descriptive statistics and logistic regression analyses
M (SD)

B (SE)

Insomnia at T1 and T2 (n = 150) Normal sleep at T1 and T2 (n = 1,037)

Unadjusted analysis
Emotion regulation (T1)
NS: 20.6 (5.7)
.09 (0.01)**
I: 24.4 (7.1)
Adjusted analysis
Emotion regulation (T1)
NS: 20.6 (5.7)
.01 (0.02)
I: 24.4 (7.1)
Gender: Female (T1)
NS: 52.0%
.59 (0.22)**
I: 66.0%
Civil status: Living alone,
NS: 18.4%
.53 (0.24)*
divorced or widowed (T1)
I: 32.0%
Ethnicity: Not born in Sweden (T1)
NS: 5.5%
.25 (0.40)
I: 9.7%
Anxiety (T1)
NS: 2.3 (2.5)
.25 (0.04)**
I: 5.9 (3.4)
Depression (T1)
NS: 2.1 (2.4)
.19 (0.04)**
I: 5.3 (3.5)
Insomnia at T1 and T3 (n = 127) Normal sleep at T1 and T3 (n = 980)
Unadjusted analysis
Emotion regulation (T1)
NS: 20.5 (5.5)
.11 (0.01)**
I: 24.9 (7.5)
Adjusted analysis
Emotion regulation (T1)
NS: 20.5 (5.5)
.01 (0.02)
I: 24.9 (7.5)
Gender: Female (T1)
NS: 52.4%
.64 (0.24)**
I: 68.5%
Civil status: Living alone,
NS: 17.1%
.70 (0.26)**
divorced or widowed (T1)
I: 33.1%
Ethnicity: Not born in Sweden (T1) NS: 5.3%
.05 (0.49)
I: 8.1%
Anxiety (T1)
NS: 2.2 (2.4)
.27 (0.04)**
I: 6.1 (3.7)
Depression (T1)
NS: 2.0 (2.3)
.20 (0.04)**
I: 5.5 (3.5)

OR

95% CI

R2

1.09

1.071.12

.07**

1.00

0.961.03

.31**

1.81

1.182.77

1.71

1.082.70

1.29

0.592.81

1.28

1.191.39

1.20

1.121.30

1.11

1.081.14

.09**

1.01

0.971.05

.35**

1.90

1.183.07

2.01

1.213.32

0.95

0.372.45

1.31

1.201.42

1.22

1.131.33

Note. CI = confidence interval; I = insomnia; M = mean; NS = normal sleep; SD = standard deviation.

*Significant at the .05 level.
**Significant at the .01 level.

in both unadjusted analyses; the effect sizes between the two groups were medium
(d = 0.59 for Time 1Time 2 and d = 0.68 for Time 1Time 3). In adjusted analyses,
gender, civil status, baseline anxiety (d = 1.22 for Time 1Time 2 and d = 1.28 for Time
1Time 3), and baseline depression (d = 1.08 for Time 1Time 2 and d = 1.21 for Time 1
Time 3) were significant risk factors, with large effect sizes, for persistent insomnia in both
analyses. However, emotion regulation was not significantly associated with persistent
insomnia in adjusted analyses. The adjusted models explained 3135% of the variance in
persistent insomnia status.

106

Markus Jansson-Frojmark et al.

Table 6. Emotion regulation strategies at baseline between those with persistent insomnia and those
with remission from insomnia: Descriptive statistics and logistic regression analyses
M (SD)

B (SE)

Insomnia at T1 and T2 (n = 150) Remission: I at T1 and NS at T2 (n = 62)

Unadjusted analysis
Emotion regulation (T1)
REM: 24.0 (7.2)
.01 (0.02)
I: 24.4 (7.1)
Adjusted analysis
Emotion regulation (T1)
REM: 24.0 (7.2)
.03 (0.03)
I: 24.4 (7.1)
Gender: Female (T1)
REM: 59.7%
.40 (0.33)
I: 66.0%
Civil status: Living alone,
REM: 25.8%
.14 (0.36)
divorced or widowed (T1)
I: 32.0%
Ethnicity: Not born in Sweden (T1)
REM: 9.7%
.11 (0.56)
I: 9.7%
Anxiety (T1)
REM: 4.8 (3.6)
.06 (0.06)
I: 5.9 (3.4)
Depression (T1)
REM: 4.0 (3.4)
.10 (0.06)
I: 5.3 (3.5)
Insomnia at T1 and T3 (n = 127) Remission: I at T1 and NS at T3 (n = 62)
Unadjusted analysis
Emotion regulation (T1)
REM: 23.9 (7.2)
.02 (0.02)
I: 24.9 (7.5)
Adjusted analysis
Emotion regulation (T1)
REM: 23.9 (7.2)
.01 (0.03)
I: 24.9 (7.5)
Gender: Female (T1)
REM: 51.6%
.84 (0.34)**
I: 68.5%
Civil status: Living alone,
REM: 21.0%
.50 (0.38)
divorced or widowed (T1)
I: 33.1%
Ethnicity: Not born in Sweden (T1) REM: 12.9%
.44 (0.57)
I: 8.1%
Anxiety (T1)
REM: 4.9 (3.6)
.05 (0.05)
I: 6.1 (3.7)
Depression (T1)
REM: 4.2 (3.8)
.08 (0.06)
I: 5.5 (3.5)

R2

OR

95% CI

1.01

0.971.05

.01

0.98

0.931.03

.07**

1.50

0.782.88

1.15

0.572.31

1.11

0.373.33

1.06

0.951.19

1.10

0.991.23

1.02

0.981.06

.01

1.00

0.951.05

.11

2.32

1.194.53

1.65

0.783.49

0.65

0.211.96

1.05

0.941.17

1.08

0.971.20

Note. CI = confidence interval; I = insomnia; M = mean; NS = normal sleep; REM = remission;

SD = standard deviation.
**Significant at the .01 level.

Emotion regulation and persistence/remission of insomnia

The fourth aim was to investigate the association between emotion regulation at baseline
and persistence/remission of insomnia, among those with insomnia at baseline. The
descriptive statistics for the DERS at baseline across the two insomnia groups related to
persistence/remission (persistent insomnia and remitted insomnia) are displayed in
Table 6. As can be seen in the table, emotion regulation (d = 0.06 for Time 1Time 2 and
d = 0.14 for Time 1Time 3) was not significantly related to persistent insomnia, relative
to remitted insomnia, neither in unadjusted nor in adjusted analyses. In one of the adjusted

Emotion dysregulation and insomnia

107

analyses, gender was a significant predictor, with women having an increased risk for
persistent insomnia. The two models explained 711% of the variance in persistent
insomnia status.

Changes in emotion regulation and persistence/remission of insomnia

The fifth aim was to explore the relationship between changes in emotion regulation over
time and persistence/remission of insomnia, among those with insomnia at baseline. The
descriptive statistics for the DERS change scores across the two insomnia groups
concerning persistence/remission (persistent insomnia and remitted insomnia) are
displayed in Table 7. As can be seen in the table, a larger change (towards more
dysfunctional levels) in emotion regulation (OR = 1.09) was related to increased risk for
persistent insomnia in one of two unadjusted analyses; the effect sizes between two
groups on changes in emotion regulation were very small to small (d = 0.06 for Time 1
Time 2 and d = 0.30 for Time 1Time 3). In adjusted analyses, baseline depression
(d = 0.38) was a significant predictor for the first model (T1T2) and changes in emotion
regulation and gender were significant predictors in the second model (T1T3). The two
models explained 919% of the variance in persistent insomnia status.

Discussion
The current results suggest that emotion regulation is not associated with the incidence
and persistence of insomnia. However, partial support, with small effect sizes, emerged
for the notion that decreases in emotion regulation were related to incident and persistent
insomnia. It is also important to underscore that these findings emerged while controlling
for previously established risk factors (e.g., gender, anxiety, and depression) (Ohayon,
2002).
Emotion regulation and insomnia incidence
The two-first purposes focused on the incidence of insomnia. The first aim was to examine
the association between emotion regulation at baseline and the incidence of insomnia. In
adjusted analyses, emotion regulation was not associated with the incidence of insomnia;
note also that the effect sizes were below the cut-off for being considered a small effect
size.
The second aim was to investigate the association between changes in emotion
regulation over time and the incidence of insomnia. A decrease in emotion regulation
(more dysfunctional levels) was associated with incident insomnia in adjusted analyses.
We would, however, like to emphasize that the effect sizes were small.
Taken together, the current findings indicate that emotion dysregulation is only partly
evident for those who develop insomnia, compared to those who do not. To our
knowledge, these findings are the first in the research literature to establish a partial, small
link between emotion dysregulation and the incidence of insomnia.

Emotion regulation and insomnia persistence

The three last purposes of this study focused on the persistence of insomnia. The third aim
of this study was to examine the relationship between emotion regulation at baseline and
persistence of insomnia. In adjusted analyses, emotion regulation was not associated with

108

Markus Jansson-Frojmark et al.

Table 7. Changes in emotion regulation strategies between those with persistent insomnia and those
with remission from insomnia: Descriptive statistics and logistic regression analyses
M (SD)

B (SE)

Insomnia at T1 and T2 (n = 150) Remission: I at T1 and NS at T2 (n = 62)

Unadjusted analysis
Emotion regulation (T1T2)
REM: 0.62 (5.78)
.03 (0.03)
I: 0.29 (5.37)
Adjusted analysis
Emotion regulation (T1T2)
REM: 0.62 (5.78)
.02 (0.04)
I: 0.29 (5.37)
Gender: Female (T1)
REM: 59.7%
.46 (0.35)
I: 66.0%
Civil status: Living alone,
REM: 25.8%
.30 (0.38)
divorced or widowed (T1)
I: 32.0%
Ethnicity: Not born in Sweden (T1)
REM: 9.7%
.25 (0.58)
I: 9.7%
Anxiety (T1)
REM: 4.8 (3.6)
.03 (0.06)
I: 5.9 (3.4)
Depression (T1)
REM: 4.0 (3.4)
.12 (0.06)*
I: 5.3 (3.5)
Insomnia at T1 and T3 (n = 127) Remission: I at T1 and NS at T3 (n = 62)
Unadjusted analysis
Emotion regulation (T1T3)
REM: 1.52 (6.12)
.08 (0.03)**
I: 0.38 (6.75)
Adjusted analysis
Emotion regulation (T1T3)
REM: 1.52 (6.12)
.11 (0.04)**
I: 0.38 (6.75)
Gender: Female (T1)
REM: 51.6%
.98 (0.36)**
I: 68.5%
Civil status: Living alone,
REM: 21.0%
.62 (0.42)
divorced or widowed (T1)
I: 33.1%
Ethnicity: Not born in Sweden (T1) REM: 12.9%
.90 (0.60)
I: 8.1%
Anxiety (T1)
REM: 4.9 (3.6)
.04 (0.05)
I: 6.1 (3.7)
Depression (T1)
REM: 4.2 (3.8)
.08 (0.06)
I: 5.5 (3.5)

OR

95% CI

R2

1.03 0.961.09 .01

1.02 0.961.10 .09

1.58 0.803.10
1.34 0.652.80
0.78 0.252.44
1.03 0.921.15
1.13 1.011.27

1.09 1.021.16 .05

1.11 1.041.19 .19**

2.66 1.325.38
1.87 0.834.22
0.41 0.131.34
1.04 0.931.15
1.08 0.971.21

Note. Standardized residuals were used as change scores in the analyses.

CI = confidence interval; I = insomnia; M = mean; NS = normal sleep; REM = remission; SD = standard deviation.
*Significant at the .05 level.
**Significant at the .01 level.

the persistence of insomnia, relative to persistent normal sleep. Although the effect sizes
were similar to earlier studies, this is not in line with the previous research (Gruber et al.,
2008; Waters et al., 1993). This might possibly be due to the covariates included in the
current study.
The fourth aim was to examine the relationship between emotion regulation at
baseline and the persistence/remission of insomnia. Emotion regulation was not

Emotion dysregulation and insomnia

109

associated with the persistence of insomnia, relative to remitted insomnia; the effect sizes
were below the cut-off for being considered a small effect size.
The fifth aim was to investigate the association between changes in emotion regulation
over time and persistence/remission of insomnia. A decrease in emotion regulation (more
dysfunctional levels) was related to persistent insomnia, relative to remitted insomnia, in
one of the adjusted analyses; overall, the effect sizes were very small to small.
In all, it appears as if emotion dysregulation is only partly characteristic for those with
persistent insomnia, compared to both normal sleepers and those whose insomnia remits
over time. Overall, it is also important to underscore the very small to medium effect sizes.
Our findings are only partly in line with the previous research showing that various
emotion regulatory processes are enhanced and might maintain several common
psychiatric disorders (Aldao et al., 2010). A notable difference between our and
previous findings (Aldao et al., 2010) is however that earlier studies have mainly used
disorder-specific measures for emotion regulatory strategies (e.g., rumination in the
context of mood disorders), while the current study employed a more generic index of
emotion regulation. This may have some impact. The notion about a critical difference
in the measures used (disorder-specific vs. generic) is supported by strong previous
evidence suggesting that insomnia-specific processes are involved as emotion dysregulation processes in insomnia (Baglioni, Lombardo, et al., 2010; Baglioni, Spiegelhalder, Lombardo, & Riemann, 2010; Harvey, Watkins, Mansell, & Shafran, 2004; Kaplan,
Talbot, & Harvey, 2009), and non-existent to moderate associations between generic
emotion regulation and insomnia in this and previous studies (Gruber et al., 2008;
Violani et al., 2012). A second point that might have implications for the results is that
the items in the DERS demand a relatively high level of effort and insight into processes
that occur, at least to some part, outside everyday consciousness. Self-report
instruments with simpler items or objective, laboratory measures could therefore be
an option for future research.
Although the current findings suggested that there is an association, albeit a weak one,
between changes in emotion regulation and insomnia over time, this conclusion can be
interpreted in several ways. Based on an extensive literature (Aldao et al., 2010), one
interpretation is that emotion regulation processes might maintain or drive insomnia. In
our view, it is also possible to think of the current results in the light of a top-down process;
it could, for example, be argued that enhanced cortical hyperarousal is causing both
insomnia and emotion dysregulation (Perlis et al., 1997). Recent findings show support
for the top-down process framework; in one study, it was shown that individuals with
insomnia, relative to good sleepers, showed increases in amygdala activation when trying
to downregulate their emotional response with cognitive reappraisal (interpreting the
meaning of a stimulus to feel less negative), thereby providing evidence that there might
be a dysfunction in the neural circuitry underlying emotion regulation among those with
insomnia (Franzen, Siegle, Jones, & Buysse, 2013). The findings can, however, also be
interpreted in terms of a bottom-up process framework (Riemann et al., 2010), in which
an inborn tendency for sleepwake dysfunction causes sleep disruption as well as
emotional difficulties. Our findings might finally be interpreted using a tripartite
conceptual model often used to explain the development of insomnia: The so called 3
P model (Spielman, 1986; Spielman & Glovinsky, 1991). This model hypothesizes that
predisposing conditions, precipitating circumstances, and perpetuating factors all, at
different times and in different ways, are involved in the development of insomnia.
According to this model, emotion dysregulation might be thought of as both a
precipitating circumstance that initiates insomnia and a perpetuating factor that

110

Markus Jansson-Frojmark et al.

maintains insomnia, that is an underlying vulnerability factor. The current results imply
that deteriorations in emotion regulation might increase the risk for both the incidence
and persistence of insomnia. It should however be remembered that the sample in this
study consisted of participants who were categorized as individuals with insomnia based
on their survey responses. An important area for future research is therefore to, with both
cross-sectional and prospective designs, explore the association between emotion
regulation and insomnia in clinical samples, consisting of individuals diagnosed with
insomnia. This is important not only because it would then be able to see whether it would
be possible to replicate the current findings, but also because it is likely that in samples
with diagnosed insomnia the levels of emotion dysregulation can be expected to be even
higher, relative to the sample in this study. An important, related area is also to examine
emotion dysregulation among those with insomnia comorbid with psychiatric disorders, a
patient subgroup for whom it is likely that they have more elevated dysregulation, relative
to those without such comorbidity. In terms of prevention and intervention, the current
findings might suggest that improving emotion regulation strategies might prevent
incident insomnia and decrease persistent insomnia. Although this needs to be tested
empirically, it is possible that interventions aimed at decreasing emotion dysregulation
(Barlow et al., 2011; Harvey, Sharpley, Ree, Stinson, & Clark, 2007; Ong, Shapiro,
& Manber, 2008) might improve current treatments for insomnia, such as cognitive
behavioural therapy (Morin, 2006). It is also possible that, as hypothesized by Baglioni,
Lombardo, et al. (2010), Baglioni, Spiegelhalder, et al. (2010), elevated negative
emotionality or emotion regulation difficulties might modulate the link between insomnia
and psychiatric conditions, and, if so, additional interventions targeting emotional
processes could enhance treatment outcomes.
A number of methodological limitations should be underscored. First, there was a
moderate response rate at baseline. Although the response rate might have influenced the
results, several studies have emerged showing that high non-response rate does not
necessarily have a significant impact on the results when examining associations between
variables rather than estimates of a single population parameter (Curtin, Presserm, &
Singer, 2000; Keeter, Kennedy, Dimock, Best, & Craighill, 2006). Second, because we had
to limit the number of items in the survey as a whole to reduce the risk of high nonresponse, the current study only employed 12 of the original 36 DERS items. Although we
chose the 12 items which loaded the highest on the subscales and that pilot work
suggested a very high correlation between the abbreviated and the full DERS, it is possible
that we did not fully capture emotion regulation as a construct. A third limitation is that we
used self-report scales as indexes for emotion regulation, sleep, anxiety, and depression.
We decided to use the DERS primarily due to its acceptable psychometric properties, but
whether individuals with and without insomnia are able to accurately report on processes
that are not fully within everyday consciousness is an area for future research. Although
we used self-report items to assess sleep, we note that we used a screening instrument to
assess sleep disorders as well as research diagnostic criteria for insomnia that underscore
the condition as a subjective complaint. Regarding insomnia, anxiety, and depression, it is
important to emphasize that the self-report format that was used does not necessarily
make the findings generalizable to clinical diagnoses. A fourth limitation is that we have no
information on whether or not the participants fulfilled the criteria for insomnia, normal
sleep, or remission in between the three time points. As we only obtained information
from the participants concerning the past month, it is possible that important changes in
sleep status occurred in between time points. A related limitation is also that we did not
assess the history of insomnia in the sample. A fifth limitation with the prospective method

Emotion dysregulation and insomnia

111

is that although a timeline can be established and it can be confirmed that emotion
regulation and insomnia are associated over time, a causal link cannot nevertheless be
made. Possible third factors could have effects on both emotion regulation and insomnia
severity, or maybe emotion regulation is a product of insomnia and would thus reduce as
insomnia remitted. Including more relevant covariates in a longitudinal design or studying
emotion regulation and insomnia experimentally could be a future option.
The current findings showed that emotion dysregulation was not associated with
increased risk for the incidence and persistence of insomnia. However, partial support
emerged for the notion that increases in emotion dysregulation were related to incident
and persistent insomnia. It is possible that this has implications for how we should view
insomnia as a clinical disorder and how it should be treated. The results might pave the
way for future research investigating temporal links between emotion regulatory
strategies and insomnia, and subsequent studies examining whether treatment targeting
emotion dysregulation might add to the improvements noted for current insomnia
treatments.

Acknowledgements
We would like to express our gratitude to the Swedish Council for Working Life and Social
Research for financial support and to Allison G. Harvey and Lars-Gunnar Lundh for study
design.

References
Aldao, A., Nolen-Hoeksema, S., & Schweizer, S. (2010). Emotion-regulation strategies across
psychopathology: A meta-analytic review. Clinical Psychology Review, 30, 217237.
doi:10.1016/j.cpr.2009.11.004
American Psychiatric Association (2000). Diagnostic and statistical manual of mental disorders
(4th ed.). Washington, DC: Author.
Ancoli-Israel, S., & Roth, T. (1999). Characteristics of insomnia in the United States: Results of the
1991 National Sleep Foundation Survey. Sleep, 22, 347353.
Baglioni, C., Lombardo, C., Bux, E., Hansen, S., Salveta, C., Biello, S., . . . Espie, C. A. (2010).
Psychophysiological reactivity to sleep-related emotional stimuli in primary insomnia.
Behaviour Research and Therapy, 48, 467475. doi:10.1016/j.brat.2010.01.008
Baglioni, C., Spiegelhalder, K., Lombardo, C., & Riemann, D. (2010). Sleep and emotions: A focus on
insomnia. Sleep Medicine Reviews, 14, 227238. doi:10.1016/j.smrv.2009.10.007
Barlow, D. H., Farchione, T. J., Fairholme, C. P., Ellard, K. K., Boisseau, C. L., Allen, L. B., &
Ehrenreich-May, J. (2011). The unified protocol for transdiagnostic treatment of emotional
disorders: Therapist guide. New York, NY: Oxford University Press.
Bastien, C., Vallieres, A., & Morin, C. (2001). Validation of the Insomnia Severity Index as an
outcome measure for insomnia research. Sleep Medicine, 2, 297307. doi:10.1016/S13899457(00)00065-4
Bjelland, I., Dahl, A. A., Haug, T. T., & Neckelmann, D. (2002). The validity of the Hospital Anxiety
and Depression Scale: An updated literature review. Journal of Psychosomatic Research, 52,
6977. doi:10.1016/S0022-3999(01)00296-3
Cohen, J. (1988). Statistical power for the behavioral sciences. Hillsdale, NJ: Erbaum.
Cronbach, L. J., & Furby, L. (1970). How should we measure change Or should we?
Psychological Bulletin, 74, 6880. doi:10.1037/h0029382
Curtin, R., Presserm, S., & Singer, E. (2000). The effects of response rate changes on the index of
consumer sentiment. Public Opinion Quarterly, 64, 413428. doi:10.1086/318638

112

Markus Jansson-Frojmark et al.

Edinger, J. D., Bonnet, M. H., Bootzin, R. R., Doghramji, K., Dorsey, S. M., Espie, C. A., . . .
Stepanski, E. J. (2004). Derivation of research diagnostic criteria for insomnia: Report of an
American Academy of Sleep Medicine Work Group. Sleep, 27, 15671596. doi:10.1097/
00006842-198801000-00010
Edinger, J. D., Stout, A. L., & Hoelscher, T. J. (1988). Cluster analysis of insomniacs MMPI
profiles: Relation of subtypes to sleep history and treatment outcome. Psychosomatic Medicine,
50, 7787.
Edwards, R. I., Clarke, M., DiGuiseppi, C., Pratap, S., Wentz, R., Kwan, I., . . . Pratap, S. (2007).
Methods to increase response rates to postal questionnaires (review). Cochrane Database of
System Review, 2, 1190. doi:10.1002/14651858.MR000008.pub3
Espie, C. A. (2002). Insomnia: Conceptual issues in the development, persistence, and treatment of
sleep disorder in adults. Annual Review of Psychology, 53, 215243. doi:10.1146/
annurev.psych.53.100901.135243
Franzen, P. L., Siegle, G. J., Jones, N. P., & Buysse, D. J. (2013). Elevated amygdala activation during
voluntary emotion regulation in primary insomnia. Sleep, 36, 560.
Gratz, K. L., & Roemer, L. (2004). Multidimensional assessment of emotion regulation and
dysregulation: Development, factor structure, and initial validation of the Difficulties in Emotion
Regulation Scale. Journal of Psychopathology and Behavioral Assessment, 26, 4154.
doi:10.1023/B:JOBA.0000007455.08539.94
Gross, J. J. (1998). The emerging field of emotion regulation: An integrative review. Review of
General Psychology, 2, 271299. doi:10.1037/1089-2680.2.3.271
Gruber, J., Eidelman, P., & Harvey, A. G. (2008). Transdiagnostic emotion regulation processes in
bipolar disorder and insomnia. Behaviour Research and Therapy, 46, 10961100. doi:10.1016/
j.brat.2008.05.004
Hair, J. F., Anderson, R. E., Tatham, R. L., & Black, W. C. (1998). Multivariate data analysis (5th ed.).
Upper Saddle River, NJ: Prentice Hall.
Harvey, A. G. (2008). Insomnia, psychiatric disorders, and the transdiagnostic perspective. Current
Directions in Psychological Science, 17, 299303. doi:10.1111/j.1467-8721.2008.00594.x
Harvey, A. G., Sharpley, A. L., Ree, M. J., Stinson, K., & Clark, D. M. (2007). An open trial of cognitive
therapy for chronic insomnia. Behaviour Research and Therapy, 45, 24912501. doi:10.1016/
j.brat.2007.04.007
Harvey, A. G., Watkins, E., Mansell, W., & Shafran, R. (2004). Cognitive behavioural processes
across psychological disorders: A transdiagnostic approach to research and treatment. New
York, NY: Oxford University Press.
Jansson-Fr
ojmark, M., Harvey, A. G., Norell-Clarke, A., & Linton, S. J. (2012). Associations between
psychological factors and nighttime/daytime symtomatology in insomnia. Cognitive Behaviour
Therapy, 41, 273287. doi:10.1080/16506073.2012.672454
Jansson-Fr
ojmark, M., & Linton, S. J. (2008). The course of insomnia over one year: A longitudinal
study in the general population in Sweden. Sleep, 31, 881886.
Kales, A., Caldwell, A. B., Preston, T. A., Healey, S., & Kales, J. D. (1976). Personality patterns in
insomnia: Theoretical implications. Archives of General Psychiatry, 33, 11281134.
doi:10.1001/archpsyc.1976.01770090118013
Kaplan, K. A., Talbot, L. S., & Harvey, A. G. (2009). Cognitive mechanisms in chronic insomnia:
Processes and prospects. Sleep Medicine Clinics, 4, 541548. doi:10.1016/j.jsmc.2009.07.007
Keeter, S., Kennedy, C., Dimock, M., Best, J., & Craighill, P. (2006). Gauging the impact of growing
nonresponse on estimates from a national RDD telephone survey. Public Opinion Quarterly,
70, 759779. doi:10.1093/poq/nfl035
Lichstein, K. L., Durrence, H. H., Taylor, D. J., Bush, A. J., & Riedel, B. W. (2003). Quantitative criteria
for insomnia. Behaviour Research and Therapy, 41, 427445. doi:10.1016/S0005-7967(02)
00023-2
Morin, C. M. (2006). Cognitive-behavioral therapy for insomnia. Sleep Medicine Clinics, 1, 375386.
doi:10.1016/j.jsmc.2006.06.008

Emotion dysregulation and insomnia

113

Morin, C. M., Belleville, G., Belanger, L., & Ivers, H. (2011). The Insomnia Severity Index:
Psychometric indicators to detect insomnia cases and evaluate treatment response. Sleep, 34,
601608.
Morin, C. M., LeBlanc, M., Daley, M., Gregoire, J. P., & Merette, C. (2006). Epidemiology of insomnia:
Prevalence, self-help treatments, consultations, and determinants of help-seeking behaviors.
Sleep Medicine, 7, 123130. doi:10.1016/j.sleep.2005.08.008
Ohayon, M. M. (2002). Epidemiology of insomnia: What we know and what we still need to learn.
Sleep Medicine Reviews, 6, 97111. doi:10.1053/smrv.2002.0186
Ong, J. C., Shapiro, S. L., & Manber, R. (2008). Combining mindfulness meditation with cognitivebehavior therapy for insomnia: A treatment-development study. Behavior Therapy, 39,
171182. doi:10.1016/j.beth.2007.07.002
Perlis, M. L., Giles, D. E., Mendelson, W. B., Bootzin, R. R., & Wyatt, J. K. (1997). Psychophysiological
insomnia: The behavioural model and a neurocognitive perspective. Journal of Sleep Research,
6, 179188. doi:10.1046/j.1365-2869.1997.00045.x
Riemann, D., Spiegelhalder, K., Feige, B., Voderholzer, U., Berger, M., Perlis, M., & Nissen, C.
(2010). The hyperarousal model of insomnia: A review of the concept and its evidence. Sleep
Medicine Reviews, 14, 1931. doi:10.1016/j.smrv.2009.04.002
Roth, T., & Ancoli-Israel, S. (1999). Daytime consequences and correlates of insomnia in the United
States: Results of the 1991 National Sleep Foundation Survey. II. Sleep, 22, 354358.
Spielman, A. J. (1986). Assessment of insomnia. Clinical Psychology Review, 6, 1125. doi:10.1016/
0272-7358(86)90015-2
Spielman, A. J., & Glovinsky, P. (1991). The varied nature of insomnia. In P. Hauri (Ed.), Case studies
in insomnia (pp. 115). New York, NY: Plenum Press.
Spoormaker, V. I., Verbeek, I., van den Bout, J., & Klip, C. (2005). Initial validation of the SLEEP-50
questionnaire. Behavioral Sleep Medicine, 3, 227246. doi:10.1207/s15402010bsm0304_4
Stekette, G., & Chambless, D. L. (1992). Methodological issues in prediction of treatment outcome.
Clinical Psychology Review, 12, 387400. doi:10.1016/0272-7358(92)90123-P
Vandekerckhove, M., & Cluydts, R. (2010). The emotional brain and sleep: An intimate relationship.
Sleep Medicine Reviews, 14, 219226. doi:10.1016/j.smrv.2010.01.002
Violani, C., Battagliese, G., Pisanti, R., & Lombardo, C. (2012). Emotion regulation, negative affect
and insomnia. Journal of Sleep Research, 21, 169170.
Walker, M. P. (2009). The role of sleep in cognition and emotion. Annals of the New York Academy
of Science, 1156, 168197. doi:10.1111/j.1749-6632.2009.04416.x
Waters, W. F., Adams, S. G., Binks, P., & Varnado, P. (1993). Attention, stress and negative emotion in
persistent sleep-onset and sleep-maintenance insomnia. Sleep, 16, 128136.
Zigmond, A. S., & Snaith, R. P. (1983). The Hospital Anxiety and Depression Scale. Acta Psychiatrica
Scandinavica, 67, 361370. doi:10.1111/j.1600-0447.1983.tb09716.x
Received 16 October 2014; revised version received 5 June 2015