Haraldmerckelbach.nl2

Journal of Abnormal Psychology
Sleep Normalization and Decrease in Dissociative
Experiences: Evaluation in an Inpatient Sample
Dalena van der Kloet, Timo Giesbrecht, Steven Jay Lynn, Harald Merckelbach, and André de
Online First Publication, August 15, 2011. doi: 10.1037/a0024781 van der Kloet, D., Giesbrecht, T., Lynn, S. J., Merckelbach, H., & de Zutter, A. (2011, August 15). Sleep Normalization and Decrease in Dissociative Experiences: Evaluation in an Inpatient Sample. Journal of Abnormal Psychology. Advance online publication. doi: 2011 American Psychological Association Sleep Normalization and Decrease in Dissociative Experiences: We conducted a longitudinal study to investigate the relation between sleep experiences and dissociativesymptoms in a mixed inpatient sample at a private clinic evaluated on arrival and at discharge 6 to 8weeks later. Using hierarchical regression analyses and structural equation modeling, we found a linkbetween sleep experiences and dissociative symptoms and determined that specifically decreases innarcoleptic experiences rather than insomnia accompany a reduction in dissociative symptoms. Althoughsleep improvements were associated with a general reduction in psychopathology, this reduction couldnot fully account for the substantial and specific effect that we found for dissociation. Our findings areconsistent with Watson’s (2001) hypothesis that disruptions in the sleep–wake cycle lead to intrusions ofsleep phenomena into waking consciousness, resulting in dissociative experiences. Accordingly, sleephygiene may contribute to the treatment or prevention of dissociative symptoms.
Keywords: dissociative experiences, unusual sleep experiences, sleep hygiene For decades, clinicians and researchers have studied dissociative concluded that the extant research provides strong support for a symptoms (e.g., depersonalization, derealization, memory lapses, link between dissociative experiences and a labile sleep–wake absorption) in a systematic fashion. However, a consensus about cycle that is evident across a range of phenomena, including their genesis remains elusive. Because of their dream-like charac- waking dreams, nightmares, and hypnagogic and hypnopompic ter, some authors have recently pointed to and discerned a possible hallucinations. Studies that have offered evidence for a link be- link between dissociative symptoms and sleep (Watson, 2001, tween dissociative experiences and sleep disturbances relied on clinical and nonclinical samples, and, with only one exception Relying on two large nonclinical samples, Watson (2001) found (Hartman, Crisp, Sedgwick, & Borrow, 2001), yielded correlations that dissociation, as measured by two validated dissociation scales in the range of .30 –.55. Similarly, Agargun and colleagues (2003a) (E. M. Bernstein & Putnam, 1986), correlates with unusual sleep- tested an undergraduate sample and found that chronic nightmare and dream-related experiences (Watson, 2001, 2003b). Based on sufferers scored higher on dissociation, compared with controls.
these findings, Watson argued that dissociation, schizotypy, and These authors also reported an increased prevalence of nightmare certain sleep experiences map onto a common domain that encom- disorder among patients with dissociative identity disorder (Agar- passes unusual perceptions and cognitions. Watson (2001) referred gun et al., 2003b). The link between unusual sleep experiences and to the continuity in unusual perceptions and cognitions across the dissociative tendencies is further illustrated by dissociative phe- day and night as “cross-state continuity.” nomena (e.g., out-of-body experiences), which often accompany Several laboratories have replicated Watson’s finding of an hypnagogic and hypnopompic hallucinations (Girard & Cheyne, association between unusual sleep experiences and dissociation (e.g., Fassler, Knox, & Lynn, 2006; Giesbrecht & Merckelbach, Disruptions in sleep patterns figure prominently in mood and 2004, 2006; Soffer-Dudek & Shahar, 2009). In a review of 19 anxiety disorders, schizophrenia, and borderline personality disor- studies, van der Kloet, Merckelbach, Giesbrecht, and Lynn (2011) der (Benca, Obermeyer, Thisted, & Gillin, 1992; Morin & Ware,1996), with fairly specific associations among discrete sleep com-plaints and forms of psychopathology (Koffel & Watson, 2009a).
For example, insomnia and tiredness appear to be primarily asso- Dalena van der Kloet, Timo Giesbrecht, Harald Merckelbach, and ciated with depression and anxiety, whereas unusual sleep expe- Andre´ de Zutter, Department of Clinical Psychological Science, Maastricht riences (e.g., hypnagogic hallucinations) appear to be primarily University, Maastricht, the Netherlands; Steven Jay Lynn, Psychology related to dissociative symptoms (Koffel & Watson, 2009a). Ac- Department, State University of New York, Binghamton.
cording to factor analytic research, dissociation and schizotypy are Correspondence concerning this article should be addressed to Dalena more strongly correlated with unusual sleep experiences (i.e., as van der Kloet, Department of Clinical Psychological Science, Faculty of measured by the Iowa Sleep Experiences Survey) than they are Psychology and Neuroscience, Maastricht University, PO Box 616, 6200MD Maastricht, The Netherlands. E-mail: Dalena.vanderkloet@ correlated with mood and anxiety (Koffel & Watson, 2009b).
Koffel and Watson (2009b) aptly concluded that “unusual sleep experiences (e.g., nightmares, vivid dreams, narcolepsy symp- study, predominantly anxiolytics and antidepressants. Participants toms) are associated with symptoms of dissociation in both clinical suffered from alcohol dependence (15%); medication dependence, and nonclinical samples” (p. 557).
especially sleep medication (18%); physical complaints (33%); Although researchers have consistently found a robust correla- depression (72%); anxiety (15%); burnout (i.e., mixed anxiety and tion between dissociative and sleep experiences, studies have depression symptoms that fail to reach diagnosis threshold; 5%); generally relied on cross-sectional designs. To arrive at meaningful attention-deficit/hyperactivity disorder (2%); psychotic symptoms causal conclusions, Giesbrecht, Smeets, Leppink, Jelicic, and (2%); identity problems (1%); difficulty stopping smoking (1%); Merckelbach (2007) deprived 25 healthy volunteers of 1 night of and other complaints (12%). A comparison on the primary out- sleep and determined that sleep loss engenders a substantial in- come measures of patients using medication and patients not using crease in dissociative symptoms. An important finding was that medication yielded no significant differences, justifying the inclu- this increase could not be attributed to mood or response bias.
sion of medication-using participants in the analyses.
Researchers have not, as yet, examined whether promoting healthy sleeping reduces dissociative symptoms. If such results Procedure
were secured in a clinical sample, it would generalize previousfindings and have important implications for understanding and Participants were informed that the scales were part of routine treating dissociative symptoms. The current study represents the diagnostic testing and that data would be used for study purposes, first prospective study to evaluate the hypothesis that improving after which they gave written informed consent. Participants com- sleep results in a decrease in dissociative experiences and the first pleted questionnaires in a set order during their first days at the prospective test of the hypothesis that unusual sleep experiences clinic (baseline), and 195 participants completed the same mea- are associated with dissociation, whereas insomnia is more reliably sures the day before discharge (follow-up). At baseline and follow- associated with anxiety and depression (see Koffel & Watson, up, computerized measures were completed via the user-friendly software program EMIUM (Janssen, 2008). Participants received We evaluated a mixed sample of inpatients treated for 6 to 8 instructions from a psychologist who explained how to use the weeks in a private clinic that emphasizes sleep hygiene as a core program and who was available to answer questions.
treatment component; we anticipated that participants with the After the baseline measures, patients received therapy as usual greatest sleep improvement at retest would display the strongest (TAU) for 6 to 8 weeks. TAU comprised individual and group decrease in dissociative experiences. Moreover, we evaluated therapy and included cognitive– behavioral therapy, mindfulness, whether anxiety and depression could account for the amelioration daily fitness exercises, and creative work. It is important to note that of dissociative experiences pre–posttreatment: Previous studies the clinic actively encouraged sleep hygiene practices and rules, have revealed a connection between dissociation and anxiety and considered to promote good sleep regulation (Costa e Silva, 2006).
depression (Giesbrecht, Merckelbach, van Oorsouw, & Simeon, Patients were awakened in the morning, denied access to their room 2010; Sierra et al., 2002), as well as a connection between both during the day to preclude napping, and returned to their rooms at the anxiety and arousal, and arousal and sleep (Giesbrecht et al., same time every night. Patients had no access to alcoholic beverages, 2010). Finally, we included a measure of childhood traumatic and caffeine in the evening and night was not permitted. Fitness experience. We anticipated that this measure would be related to activities were restricted to the morning, and participants had access to dissociation levels (e.g., Gast, Rodewald, Nickel, & Emrich, relaxing activities such as massages and sauna in the evening. After- 2001), but that it would remain stable over time.
noons included healthy outdoor activities. Staff, therapists, and pa-tients were naive with respect to the hypotheses under study.
Measures
Participants
Dissociative Experiences Scale (DES; Cronbach’s alpha
Participants were 266 inpatients (132 men, 113 women, 21 not baseline ؍ .94, follow-up ؍ .94; E. M. Bernstein & Putnam,
recorded; mean age: 44.2 years, SD ϭ 11.5; range: 18 –74 years) 1986; Dutch version: Boon & Draijer, 1995).
admitted for 6 to 8 weeks to U-Center in Epen, Netherlands.
self-report scale that requires participants to indicate on 100-mm U-Center is a private clinic with an eclectic treatment approach.
visual analogue scales (anchors: 0 ϭ never; 100 ϭ always) to what Seventy-one participants did not complete treatment as a result of extent they experience 28 dissociative experiences in daily life. Van leaving the clinic prematurely on a voluntary basis or being re- IJzendoorn and Schuengel (1996) provide meta-analytic evidence for ferred to other clinics (e.g., university hospital) because of somatic the sound psychometric properties of the DES. Following the three- or psychiatric complications. Completers versus noncompleters factor solution proposed by Carlson et al. (1991), in addition to the did not differ on any of the baseline measures (e.g., Dissociative DES total score, we calculated subscale scores for amnesia, absorp- Experiences Scale, Beck Anxiety Inventory, Brief Symptom In- tion and imaginative involvement, and depersonalization/derealiza- ventory, Childhood Trauma Questionnaire, and Beck Depression tion. Furthermore, we examined the subset of eight DES items that Inventory—II; ps Ͼ .05) or with respect to age, gender, use of constitute the so-called DES-Taxon (DES-T; Waller, Putnam, & medication, or diagnosis ( ps Ͼ .05).
Carlson, 1996), which tap more pathological symptoms of dissocia- A psychologist and resident psychiatrist collaborated to deter- tion (e.g., depersonalization and amnesia). Following Waller et al.
mine diagnoses on the basis of test scores, clinical interviews, (1996), we created a dichotomous measure of taxon membership information from (medical) records and intake, and collateral versus nontaxon membership; patients with a taxon probability ex- information. Part of the sample (38%) used medication during the ceeding 0.90 were assigned to the DES-T taxon group. Psychometric shortcomings notwithstanding (Watson, 2003a), the DES-T has been Clinical Interview for DSM Disorders is good (r ϭ .83; Sprinkle et considered a useful measure in the dissociation field (Simeon, Knu- telska, Nelson, Guralnik, & Schmeidler, 2003).
Childhood Trauma Questionnaire (CTQ; Cronbach’s alpha
SLEEP-50 (Cronbach’s alpha baseline ؍ .84, follow-up ؍
baseline ؍ .90, follow-up ؍ .52; D. P. Bernstein et al., 2003).
.93; Spoormaker, Verbeek, van den Bout, & Klip, 2005).
The CTQ is a widely used self-report scale of traumatic childhood Sleep experiences were assessed with subscales of the 50-item Dutch events, such as emotional, physical, and sexual abuse, and emotional version of the SLEEP-50, which index sleep complaints and sleep and physical neglect. In the present study, we employed the 25-item disorders listed in the Diagnostic and Statistical Manual of Mental short form scored on 5-point scales anchored 1 (never) and 5 (very Disorders (4th ed., text rev.; American Psychiatric Association, often). The Dutch version of the CTQ possesses satisfactory psycho- 2000): sleep apnea (Cronbach’s alpha baseline: .58; follow-up: .69; metric properties (Thombs, Bernstein, Lobbestael, & Arntz, 2009).
change score: .52), insomnia (Cronbach’s alpha baseline: .87; follow-up: .88; change score: .76), restless legs (Cronbach’s alpha baseline: .71; follow-up: .81; change score: .45), circadian rhythm sleep disor-der (Cronbach’s alpha baseline: .56; follow-up: .55; change score: Individual Differences Measures
.27), sleepwalking (Cronbach’s alpha baseline: .65; follow-up: .84; Statistical analyses were performed using SPSS 18.0 software.
change score: .30), nightmares (Cronbach’s alpha baseline: .84; fol- Table 1 shows mean scores of all measures at baseline and follow- low-up: .90; change score: .93), factors influencing sleep (Cronbach’s up. With exception of the CTQ (i.e., self-reported childhood alpha baseline: .81; follow-up: .82; change score: .53), the impact of trauma experiences), paired-samples t tests revealed significant sleep complaints on daily functioning (Cronbach’s alpha baseline: decreases across the two time points for all measures. This pattern .66; follow-up: .70; change score: .72), and narcolepsy (Cronbach’s supports our hypothesis that TAU would lead to an improvement alpha baseline: .51; follow-up: .73; change score: .61). Each item is in sleep quality as measured by the SLEEP-50 subscales, as well scored on a 4-point Likert scale ranging from 0 (not at all) to 3 (very as a general decrease in psychopathology as measured by the DES, much). Spoormaker et al. (2005) have demonstrated adequate test– BSI, BAI, and BDI–II. For instance, of the completers, 46 of 195 retest reliability for the SLEEP-50 total score (r ϭ .78). The participants (24%) displayed dissociation levels exceeding the SLEEP-50 Narcolepsy subscale covers unusual sleep phenomena, clinical cutoff for dissociative disorders (i.e., Ͼ30; Bernstein- including hypnagogic imagery and excessive daytime sleepiness, that Carlson & Putnam, 1993) at baseline. This number was reduced to overlap with the Iowa Sleep Experiences Survey (ISES) General 24 (12%) at follow-up (Fisher’s exact p ϭ .005). Table 2 presents subscale (Koffel & Watson, 2009b). The use of the SLEEP-50 pro- the Pearson product–moment correlations between all psychopa- vides the opportunity to test predictions derived from Koffel and thology and dissociation measures at baseline and follow-up, as Watson (2009b) pertaining to unusual experiences versus insomnia.
well as the correlations among change scores.
Brief Symptom Inventory (BSI; Cronbach’s alpha base-
line ؍ .97, follow-up ؍ .97; Boulet & Boss, 1991).
Correlations Between Change Scores of SLEEP-50
53-item BSI assesses general symptoms and complaints experi- Subscales, DES, and Psychopathology Composite
enced by people with psychiatric problems. Although the BSIcomprises nine subscales, analyses were based on the total score.
Given their high intercorrelations, BSI, BAI, and BDI–II were Items are scored on a 5-point Likert scale (anchors: 0 ϭ not at all, collapsed into one psychopathology composite by standardizing 4 ϭ extremely). The Dutch version of the BSI has good convergent the baseline, follow-up, and change scores and summing the stan- and divergent validity and has proven to be a useful outcome dardized values. Table 3 displays the Pearson product–moment measure for therapy efficiency (de Beurs & Zitman, 2006).
correlations between the change scores of the SLEEP-50 sub- Beck Anxiety Inventory (BAI; Cronbach’s alpha baseline ؍
scales, change scores of DES, and psychopathology composite.
Next, we tested whether the correlations between dissociation and .93, follow-up ؍ .92; De Ayala, Vonderharr-Carlson, & Doy-
sleep factors were different from the correlations with psychopa- oung, 2005).
The BAI is a 21-item widely used self-report mea- thology. Whereas differences between the correlations of sure of anxiety symptoms. Each item is scored on a 4-point Likert SLEEP-50 and DES and SLEEP-50 and the psychopathology scale (anchors: 0 ϭ not at all bothered by this symptom, 3 ϭ severely composite (BSI, BAI, BDI–II) did not reach significance for most bothered by this symptom). The range of total scores is 0 to 63, with subscales of the SLEEP-50, the correlation between psychopathol- higher scores indicating more anxiety symptoms. The BAI has high ogy and the Insomnia subscale was significantly greater than the internal consistency (Cronbach’s alpha ϭ .93) and modest test–retest correlation between the DES and the Insomnia subscale.
reliability (r ϭ .66; Kelett, Beail, Newman, & Frankis, 2003).
Beck Depression Inventory—II (BDI–II; Cronbach’s alpha
Modeling Mood as Mediator of the Dissociation–Sleep
baseline ؍ .92, follow-up ؍ .93; Sprinkle et al., 2002; Dutch
Connection
version: Van der Does, 2002).
measure of depressive symptoms comprising 21 items. Each item We determined whether the decrease in dissociative symptoms is scored on a 4-point Likert scale ranging from 0 (not at all at follow-up was mediated by a reduction in psychopathology.
bothered by this symptom) to 3 (severely bothered by this symp- According to this hypothesis, the relationship between dissociation tom). The range of total scores is 0 to 63, with higher scores and sleep scores should be eliminated when general psychopathol- reflecting more depressive symptoms. The BDI has high test–retest ogy is statistically controlled. We subjected change scores to a reliability (r ϭ .96), and convergent validity with the Structured hierarchical multiple regression analysis with dissociation (DES) Table 1Mean Scores at Baseline and Follow-Up and t Statistics of Inpatient Sample (n ϭ 195) DES ϭ Dissociative Experiences Scale; PLMD ϭ periodic leg movement disorder; BSI ϭ Brief Symptom Inventory; BAI ϭ Beck’s Anxiety Inventory; BDI–II ϭ Beck’s Depression Inventory—II; CTQ ϭChildhood Trauma Questionnaire.
ء p Ͻ .05 (two-tailed). ءء p Ͻ .01 (two-tailed).
as the dependent variable and the Narcolepsy, Insomnia, Night- remained significant. We repeated this approach for all three DES mares, and Daily Functioning subscales of the SLEEP-50, the subscales. The results are summarized in Table 5. Decrease in psychopathology composite, and self-reported trauma (CTQ) as narcoleptic symptoms was a significant predictor in explaining the predictors. Only SLEEP-50 subscales with change scores that had decrease in absorption, amnesia, and depersonalization. Again, this Cronbach’s alphas exceeding .60 were included. Because change effect was partially mediated by a decrease in general psychopa- scores tend to have lowered reliability, we chose a lower bound of thology (i.e., psychopathology composite).
acceptability than the commonly recommended Cronbach’s al- Finally, we conducted a logistic regression analysis with DES-T pha ϭ .80. The analysis consisted of the following steps: First, we membership probability as dependent variable. We found a signif- entered the SLEEP-50 subscales. Next, we entered other predictors icant decrease in membership from baseline (n ϭ 48, 24.61%) to (i.e., psychopathology composite and CTQ). Following this, we follow-up (n ϭ 19, 9.74%), Pearson ␹2 ϭ 35.67, p Ͻ .001. Change removed nonsignificant predictors by means of backward elimi- scores on SLEEP-50 subscales and change in general psychopa- nation. We present a hierarchical decomposition in Table 4. Nei- thology were entered as predictors. Improvements in narcolepsy ther changes in insomnia, nightmares, or daily functioning scores explained most of the decrease in DES-T membership probability could account for the decrease in dissociative symptoms at follow- (B ϭ 0.37, SE ϭ 0.13, p Ͻ .01), with another part of the change up. Moreover, we employed a bootstrapping methodology in taxon membership being explained by the general psychopa- (Preacher & Hayes, 2004) to determine whether the decline in thology composite (B ϭ 0.18, SE ϭ 0.08, p Ͻ .05).
dissociative symptoms was (partially) mediated by a decrease ingeneral psychopathology. We used 10,000 bootstrap resamples of Childhood Trauma, Improvement in Sleep, and
the data with replacement and found a significant mediation effect Reduction in Dissociation
(bootstrap coefficient ϭ .38, SE ϭ .14), with a 95% confidenceinterval of .15 to .71 (significance indicated by the 95% confidence We hypothesized that sleep improvement would lead to a re- interval not crossing zero). Thus, part of the decrease in dissoci- duction in dissociative symptoms. However, because of shortcom- ation was explained by a decrease in narcoleptic symptoms due to ings associated with change scores (Peter, Churchill, & Brown, a reduction in general psychopathology. However, the effects of 1993), we tested three theoretically motivated mediation models sleep improvement on dissociation were only partly mediated and using structural equation models (see Cole & Maxwell, 2003): Table 2Pearson Product–Moment Correlations Between Dissociation, Psychopathology, and Childhood Trauma at Baseline, Follow-Up, andChange Scores Baseline: n ϭ 256; follow-up: n ϭ 201. Numbers in italics display the correlations of the measures between baseline and follow-up. DES ϭ Dissociative Experiences Scale; BSI ϭ Brief Symptom Inventory; BAI ϭ Beck’s Anxiety Inventory; BDI–II ϭ Beck’s Depression Inventory—II; CTQ ϭChildhood Trauma Questionnaire.
ء p Ͻ .05 (two-tailed). ءء p Ͻ .01 (two-tailed).
Model 1: no mediation, decrease in narcoleptic symptoms leads values of .08 or less indicate adequate fit. Table 6 gives the fit directly to a decrease in dissociation; Model 2: partial mediation, indices for all three models. As can be seen, both the partial and decrease in narcoleptic symptoms leads to decreases in both dis- the full mediation model (Models 2 and 3) fulfilled all criteria for sociation and general psychopathology, but there is also a direct acceptable fit (see also Figure 1). However, Akaike’s information effect of decrease in narcoleptic symptoms on dissociation; and criterion (AIC), the Browne–Cudeck criterion (BCC), the Bayes- Model 3: full mediation, the effect of decrease in narcoleptic ian information criterion (BIC), and the parsimony comparative fit symptoms on dissociation is fully accounted for by a reduction in index (PCFI) indicated superior fit for Model 2. That is, AIC, psychopathology. The analyses were conducted with AMOS 17 BCC, and BIC values were all numerically smaller for Model 2.
Unfortunately, the AIC, BCC, and BIC do not lend themselves to In all three models, we used the psychopathology composite as statistical testing (Barrett, 2007). Fortunately, Model 1 and 3 are a latent variable consisting of BAI, BSI, and BDI–II. The follow- nested in Model 2. Specifically, Model 1 restricts Model 2’s ing fit indices were used: the Bentler–Bonett normed fit index connection between psychopathology at T1 (follow-up time point) (NFI), the comparative fit index (CFI), the goodness-of-fit index and DES at T1 to zero, and Model 3 restricts sleep at T1 to DES (GFI), and the root mean square error of approximation (RMSEA).
at T1 to zero. Therefore, we tested the differences between models We assumed in line with Finch and West (1997) that the fit is by means of a chi-square test (see, e.g., Schreiber, Nora, Stage, acceptable if NFI, CFI, and GFI are .90 or greater, and RMSEA Barlow, & King, 2006). This test showed that Model 2 is statisti- Table 3Pearson Product–Moment Correlations Between SLEEP-50 Subscales, Dissociation, andPsychopathology Composite (All Change Scores; n ϭ 195) and Differences BetweenCorrelation Coefficients Impact of Sleep Complaints on Daily Functioning DES ϭ Dissociative Experiences Scale; PLMD ϭ periodic leg movement disorder.
ء p Ͻ .05 (two-tailed). ءء p Ͻ .01 (two-tailed).
cally superior to Models 1 and 3 ( ps Ͻ 0.01). It is interesting that, psychopathology and dissociation proved to be nonsignificant, in this model, self-reports of childhood trauma at baseline (CTQ) contributed longitudinally to narcolepsy at follow-up. Finally, amodel we assessed with sleep as a potential mediator between Discussion
Our research replicates and extends previous findings and pro- vides important insights regarding the relation between sleep and Summary of Hierarchical Multiple Regression Analysis on the dissociation. More specifically, in a mixed inpatient sample, we Dissociative Experiences Scale (All Change Scores; n ϭ 195) replicated research showing a robust link between sleep experi-ences and dissociation (Giesbrecht & Merckelbach, 2004, 2006; Soffer-Dudek & Shahar, 2011; Watson, 2001). Our findings are in line with the Giesbrecht et al. (2007) study in which sleep depri- vation promoted dissociative experiences, an outcome entirely consistent with the hypothesis that disruptions in the sleep–wake cycle lead to intrusions of sleep phenomena into waking con- sciousness, resulting in dissociative experiences (Watson, 2001).
Because disruptions in circadian rhythms exert detrimental effects on attentional control and memory, they may contribute to the attention deficits that are typically found in patients with dissocia- tive disorders (Guralnik, Giesbrecht, Knutelska, Sirroff, & Simeon, 2007; Guralnik, Schmeidler, & Simeon, 2000).
An important result of this study was that, using a longitudinal design, we demonstrated that improvements in sleep quality and, more specifically, decreases in narcoleptic/unusual sleep symp- toms accompany a reduction in dissociative experiences, including DES total scores, the three DES subscales, and the DES-T taxon membership. It is interesting that, at baseline assessment, 24% of the patients who completed treatment exceeded the clinical cutoff for dissociative disorders (i.e., Ͼ30; Bernstein-Carlson & Putnam, 1993); however, only 12% of the “completers” met this cutoff at follow-up. Similarly, when taxon probability scores, indicative of more serious dissociative pathology, were considered, 24.61% of participants met the criterion for taxon membership at baselineversus only 9.74% at the completion of therapy.
Psychopathology ϭ Psychopathology composite, consisting of to- tal change scores on Beck’s Anxiety Inventory, Brief Symptom Inventory, Improvements in sleep were associated with a general reduction and Beck’s Depression Inventory; CTQ ϭ Childhood Trauma Question- in psychopathological symptoms. However, this reduction could naire; Insomnia ϭ SLEEP-50 Insomnia subscale; Narcolepsy ϭ SLEEP-50 not account for the substantial and specific beneficial effect of the Narcolepsy subscale; Nightmares ϭ SLEEP-50 Nightmares subscale; decrease in narcoleptic symptoms on dissociation. Although struc- Daily functioning ϭ SLEEP-50 Impact of Sleep Complaints on Daily tural equation modeling revealed that narcoleptic symptoms are associated with both decreases in dissociation and general psycho- Table 5Summary of Hierarchical Multiple Regression Analyses on the Dissociative Experiences ScaleSubscales (All Change Scores; n ϭ 195) Psychopathology ϭ Psychopathology composite, consisting of total change scores on Beck’s Anxiety Inventory, Brief Symptom Inventory, and Beck’s Depression Inventory; CTQ ϭ Childhood Trauma Questionnaire;Insomnia ϭ SLEEP-50 Insomnia subscale; Narcolepsy ϭ SLEEP-50 Narcolepsy subscale; Nightmares ϭ SLEEP-50Nightmares subscale; Daily functioning ϭ SLEEP-50 Impact of Sleep Complaints on Daily Functioning subscale.
Table 6Fit Indices of the Nonmediation Model (1), the Partial Mediation Model (2), and the FullMediation Model (3) NFI ϭ Bentler–Bonett normed fit index; CFI ϭ comparative fit index; PCFI ϭ parsimony comparative fit index; GFI ϭ goodness-of-fit index; RMSEA ϭ root mean square error of approximation; AIC ϭ Akaike’sinformation criterion; BCC ϭ Browne–Cudeck criterion; BIC ϭ Bayesian information criterion.
pathology, we also found tentative evidence for a specific link symptoms over and above narcoleptic symptoms, which are a between narcoleptic symptoms and dissociation. Multiple regres- typical manifestation of unusual sleep experiences. In contrast, sion analyses converged on the conclusion that changes in disso- relative to changes in dissociation, changes in the general psycho- ciation, as indexed by DES and taxon probability scores, could not pathology composite were more strongly related to changes in be fully accounted for by global changes in psychopathology.
insomnia from pre- to posttreatment.
Indeed, a substantial part of the decrease in dissociation was Even though decreases in dissociation after treatment could not uniquely explained by improvement in sleep and specifically by be fully accounted for in terms of reductions in global psychopa- the decrease in narcoleptic symptoms. Furthermore, we found thology, we did find that the association between sleep and psy- support for Koffel and Watson’s (2009a) contention that insomnia chopathology was not specific to dissociation. That is, other mea- appears to be associated with depression and anxiety, as measured sures of psychopathology were, like dissociation, associated with by the psychopathology composite in our study. Changes in un- sleep, a finding that is in keeping with the literature (e.g., Benca et usual sleep experiences and narcolepsy (e.g., vivid dreams, hyp- al., 1992). Indeed, the measure of psychopathology was correlated nopompic and hypnagogic hallucinations) were associated with not only with insomnia but with narcolepsy symptoms as well; the dissociative symptoms as well as anxiety/depression. However, the latter correlation was comparable to the correlation between dis- regression analyses demonstrated that the other SLEEP-50 sub- sociation and narcolepsy. However, when we controlled for scales were unable to explain further variance in dissociative changes in general psychopathology, narcoleptic experiences The partial mediation model. Structural equation modeling revealed that a partial mediation model best described the data. Reduction of dissociative symptoms was predicted by decrease in narcoleptic symptomsdirectly, as well as indirectly via improvement in general psychopathology. Note that self-reports of trauma atbaseline only contributed longitudinally in the model as an influence on narcolepsy at follow-up. T0 ϭ Timepoint baseline; T1 ϭ time point follow-up; Narcolepsy ϭ Narcolepsy subscale of SLEEP-50; DES ϭ Disso-ciative Experiences Scale; BAI ϭ Beck’s Anxiety Inventory; BSI ϭ Brief Symptom Inventory; BDI ϭ Beck’sDepression Inventory—II; CTQ ϭ Childhood Trauma Questionnaire.
emerged as a predictor of absorption, amnesia, and depersonaliza- pression). Future research should examine which patients benefit tion. Moreover, the fact that changes in narcoleptic symptoms most from sleep hygiene programs to explore treatment options were the most prominent predictor of changes in DES-T probabil- and to ascertain the possible role of sleep difficulties in diverse ity scores—associated with serious dissociative pathology— suggests that sleep may indeed affect serious dissociative pathol- Before closing, a number of caveats in interpreting our findings merit mention. Although we employed a prospective longitudinal A widely held notion about the etiology of dissociative symp- design, our findings may have been influenced by as yet un- toms is that they serve a defensive function in that they help the specified confounding variables. Accordingly, it is necessary to individual to cope with traumatic memories (e.g., Gershuny & interpret the direction of the relation between sleep problems and Thayer, 1999). One shortcoming of this conceptualization is that it dissociation with caution. Measuring variables of interest over remains silent as to how trauma contributes to dissociation. In three or more points in time would allow researchers to more contrast, the sleep– dissociation approach we evaluated suggests finely assess the temporal and causal links between dissociation that traumatic experiences or the sequelae of trauma disrupt sleep, and sleep problems. Nonetheless, the mere fact that we obtained a which contributes to or exacerbates dissociation. As in previous substantial connection between sleep problems and psychopathol- work (e.g., Gast et al., 2001), we found an association between ogy—notably dissociative experiences—is clinically relevant and self-reported trauma and dissociation. Finding a correlation be- theoretically meaningful in terms of the sleep– dissociation hy- tween self-reported trauma and dissociation does not constitute pothesis (van der Kloet et al., 2011; Watson, 2001, 2003b).
proof of a relationship between objectively documented trauma One could argue that the decline in narcoleptic symptoms, and dissociation and in no way implies a causal relationship dissociative experiences, and general psychopathological com- (Giesbrecht et al., 2010). Although sleep did not emerge as a plaints reflect report bias. However, given the fact that CTQ scores potential mediator between psychopathology and dissociation, remained stable, a report bias related to global demands for “pos- highlighting the specificity of our partial mediation model, our itive reporting” over time is unlikely an adequate alternative ac- structural equation modeling findings are consistent with a causal count of our results. Our finding that narcoleptic symptoms pre- model in which trauma fuels sleep disturbances that in turn pro- dicted dissociation even when psychopathology was statistically mote dissociation. Thus, we found childhood trauma (CTQ) at controlled also argues against a reporting bias interpretation.
baseline to contribute longitudinally to narcoleptic symptoms at Moreover, staff, therapists, and patients were naive with respect to follow-up and thus indirectly to dissociation. Our findings suggest the hypotheses, further reducing sources of potential bias. Still, our a role for trauma—mediated by sleep disturbances—in the genesis research does not permit determination of which of the multifac- of dissociation, possibly by hampering recovery through its impact eted treatment components (e.g., sleep hygiene vs. cognitive– on sleep. If future studies replicate this pattern, it would provide a behavioral therapy) were responsible for symptom reduction and possible basis for a rapprochement between the posttraumatic and sleep improvement. Because our study sampled participants at sociocognitive model of dissociation, which holds that social and only two time points, causal statements about the link between cognitive variables shape dissociative symptoms (Lilienfeld et al., sleep and dissociation remain speculative. We suggest that future studies (a) administer objective measures of dissociation and sleep Our findings suggest that research on dissociation might benefit on multiple occasions, (b) dismantle complex treatments, and (c) from the literature on the origins of and treatment options fornarcoleptic symptoms (e.g., Scammell, 2003). Indeed, there exists control for expectancies and motivation to identify efficacious an urgent need for fresh treatment ideas, as studies have found treatment components and mechanisms of sleep hygiene. Further- dissociative disorders to be recalcitrant to standard therapeutic more, future studies should use the ISES (Watson, 2001) in addi- approaches, including cognitive– behavioral therapy and pharma- tion to the SLEEP-50, as the ISES specifically taps unusual sleep cological interventions such as fluoxetine (Lilienfeld, 2007). Our experiences. We also recommend that future studies include psy- findings also highlight the importance of sleep problems in clinical chometrically sound measures of quality of life (Krystal, Thakur, settings. The prevalence of sleep problems is often underestimated & Roth, 2008). Although we found that daily functioning was (Stores, 2007). Whereas they seldom are the primary focus of associated with dissociation and other indices of psychopathology, therapy, sleep problems are known to exert a decided impact on the measure most closely associated with quality of life was based psychological well-being (Kumar, 2008), quality of life (Costa e on a single subscale of our sleep measure.
Silva, 2006), and mental (Wu et al., 2008) and physical stability In closing, our study replicated and extended previous research and implicates sleep hygiene as a means of treating or preventing Because sleep abnormalities are concomitants of various psy- dissociative symptoms, as well as symptoms of psychopathology chopathological conditions (Benca et al., 1992; Morin & Ware, more broadly. Studies in which sleep hygiene variables and treat- 1996), it is not surprising that we found sleep problems to be ment components of sleep hygiene programs are manipulated, and prominently present in a heterogeneous inpatient sample. An im- dissociative and other symptoms of psychopathology are moni- portant finding was that we determined that a decline in psycho- tored over multiple time points, would be a next logical step.
pathological symptoms accompanied improvements in sleep.
Ultimately, this line of research holds tremendous promise to However, because our mixed patient sample was quite heteroge- contribute to our understanding of psychopathology in general and neous, our findings are not specific to a particular diagnostic entity dissociation in particular and to the development of effective and may not be necessarily generalizable to more discrete mani- treatment interventions for people with a broad range of psycho- festations of well-diagnosed psychopathology (e.g., anxiety, de- References
Gershuny, B. S., & Thayer, J. F. (1999). Relations among psychological trauma, dissociative phenomena, and trauma-related distress: A review Agargun, M. Y., Kara, H., Ozer, O. A., Selvi, Y., Kiran, U., & Ozer, B.
and integration. Clinical Psychology Review, 19, 631– 657. doi:10.1016/ (2003a). Clinical importance of nightmare disorder in patients with dissociative disorders. Psychiatry and Clinical Neurosciences, 57, 575– Giesbrecht, T., & Merckelbach, H. (2004). Subjective sleep experiences 579. doi:10.1046/j.1440-1819.2003.01169.x are related to dissociation. Personality and Individual Differences, 37, Agargun, M. Y., Kara, H., Ozer, O. A., Selvi, Y., Kiran, U., & Kiran, S.
1341–1345. doi:10.1016/j.paid.2004.01.004 (2003b). Nightmares and dissociative experiences: The key role of Giesbrecht, T., & Merckelbach, H. (2006). Dreaming to reduce fantasy? childhood traumatic events. Psychiatry and Clinical Neurosciences, 57, Fantasy proneness, dissociation, and subjective sleep experiences. Per- 139 –145. doi:10.1046/j.1440-1819.2003.01093.x sonality and Individual Differences, 41, 697–706. doi:10.1016/ American Psychiatric Association. (2000). Diagnostic and statistical man- ual of mental disorders (4th ed., text rev.). Washington, DC: Author.
Giesbrecht, T., Merckelbach, H., van Oorsouw, K., & Simeon, D. (2010).
Arbuckle, J. L. (2008). AMOS 17 user’s guide. Chicago: SPSS.
Skin conductance and memory fragmentation after exposure to an emo- Barrett, P. (2007). Structural equation modelling: Adjudging model fit.
tional film clip in depersonalization disorder. Psychiatry Research, 117, Personality and Individual Differences, 42, 815– 824. doi:10.1016/ 342–349. doi:10.1016/j.psychres.2010.03.010 Giesbrecht, T., Smeets, T., Leppink, J., Jelicic, M., & Merckelbach, H.
Benca, R. M., Obermeyer, W. H., Thisted, R. A., & Gillin, J. C. (1992).
(2007). Acute dissociation after 1 night of sleep loss. Journal of Abnor- Sleep and psychiatric disorders—A meta-analysis. Archives of General mal Psychology, 116, 599 – 606. doi:10.1037/0021-843X.116.3.599 Girard, T. A., & Cheyne, J. A. (2004). Individual differences in laterali- Bernstein, D. P., Stein, J. A., Newcomb, M. D., Walker, E., Pogge, D., sation of hallucinations associated with sleep paralysis. Laterality, 9, Ahluvalia, T., . . . Zule, W. (2003). Development and validation of a brief screening version of the Childhood Trauma Questionnaire. Child Guralnik, O., Giesbrecht, T., Knutelska, M., Sirroff, B., & Simeon, D.
Abuse & Neglect, 27, 169 –190. doi:10.1016/S0145-2134(02)00541-0 (2007). Cognitive functioning in depersonalization disorder. Journal of Bernstein, E. M., & Putnam, F. W. (1986). Development, reliability and Nervous and Mental Disease, 195, 983–988. doi:10.1097/ validity of a dissociation scale. Journal of Nervous and Mental Disease, 174, 727–735. doi:10.1097/00005053-198612000-00004 Guralnik, O., Schmeidler, J., & Simeon, D. (2000). Feeling unreal: Cog- Bernstein-Carlson, E., & Putnam, F. W. (1993). An update on the Disso- nitive processes in depersonalization. The American Journal of Psychi- ciative Experiences Scale. Dissociation, 6, 19 –27.
Boon, S., & Draijer, N. (1995). Screening en diagnostiek van dissociatieve Hartman, D., Crisp, A. H., Sedgwick, P., & Borrow, S. (2001). Is there a stoornissen [Screening and diagnostics of dissociative disorders]. Lisse: dissociative process in sleepwalking and night terrors? Postgraduate Medical Journal, 77, 244 –249. doi:10.1136/pmj.77.906.244 Boulet, J., & Boss, M. W. (1991). Reliability and validity of the Brief Janssen, R. (2008). EMIUM: Effect monitoring via Internet Universiteit Symptom Inventory. Journal of Consulting and Clinical Psychology, 3, Maastricht (Version 1.3) [Computer software]. Research Institute of Experimental Psychopathology, Maastricht University, Maastricht, Carlson, E. B., Putnam, F. W., Ross, C. A., Anderson, G., Clark, P., Torem, M., . . . Braun, B. G. (1991). Factor analysis of the Dissociative Expe- Kalia, M. (2006). Neurobiology of sleep. Metabolism: Clinical and Ex- riences Scale: A multicenter study. In B. G. Braun & E. B. Carlson(Eds.), Proceedings of the Eighth International Conference on Multiple perimental, 55, 2– 6. doi:10.1016/j.metabol.2006.07.005 Personality and Dissociative States (p. 16). Chicago: Rush.
Kelett, S., Beail, N., Newman, D. W., & Frankis, P. (2003). Utility of the Cole, D. A., & Maxwell, S. E. (2003). Testing mediational models with Brief Symptom Inventory in the assessment of psychological distress.
longitudinal data: Questions and tips in the use of structural equation Journal of Applied Research in Intellectual Disabilities, 16, 127–134.
modeling. Journal of Abnormal Psychology, 112, 558 –577. doi: Koffel, E., & Watson, D. (2009a). The two-factor structure of sleep Costa e Silva, J. A. (2006). Sleep disorders in psychiatry. Metabolism: complaints and its relation to depression and anxiety. Journal of Abnor- Clinical and Experimental, 55, 40 – 44. doi:10.1016/j.metabol mal Psychology, 118, 183–194. doi:10.1037/a0013945 Koffel, E., & Watson, D. (2009b). Unusual sleep experiences, dissociation, De Ayala, R. J., Vonderharr-Carlson, D. J., & Doyoung, K. (2005).
and schizotypy: Evidence for a common domain. Clinical Psychology Assessing the reliability of the Beck’s Anxiety Inventory scores. Edu- Review, 29, 548 –559. doi:10.1016/j.cpr.2009.06.004 cation and Psychological Measurement, 65, 742–756. doi:10.1177/ Krystal, A. D., Thakur, M., & Roth, T. (2008). Sleep disturbance in psychiatric disorders: Effects on function and quality of life in mood de Beurs, E., & Zitman, F. G. (2006). De Brief Symptom Inventory (BSI): disorders, alcoholism, and schizophrenia. Annals of Clinical Psychiatry, De betrouwbaarheid en validiteit van een handzaam alternatief voor de 20, 39 – 46. doi:10.1080/10401230701844661 SCL-90. [The Brief Symptom Inventory (BSI): The reliability and Kumar, V. M. (2008). Sleep and sleep disorders. Indian Journal of Chest validity of a brief alternative of the SCL-90]. Maandblad Geestelijke Diseases and Allied Sciences, 50, 129 –135.
Volksgezondheid, 61, 120 –141.
Lilienfeld, S. O. (2007). Psychological treatments that cause harm. Per- Fassler, O., Knox, J., & Lynn, S. J. (2006). The Iowa Sleep Experiences spectives on Psychological Science, 2, 53–70. doi:10.1111/j.1745- Survey: Hypnotizability, absorption, and dissociation. Personality and Individual Differences, 41, 675– 684. doi:10.1016/j.paid.2006.03.007 Lilienfeld, S., Lynn, S. J., Kirsch, I., Chaves, J., Sarbin, T., Ganaway, G., Finch, J. F., & West, S. G. (1997). The investigation of personality & Powell, R. (1999). Dissociative identity disorder and the sociocogni- structure: Statistical models. Journal of Research in Personality, 31, tive model: Recalling the lessons of the past. Psychological Bulletin, 125, 507–523. doi:10.1037/0033-2909.125.5.507 Gast, U., Rodewald, F., Nickel, V., & Emrich, H. M. (2001). Prevalence of Morin, C. M., & Ware, J. C. (1996). Sleep and psychopathology. Applied dissociative disorders among psychiatric inpatients in a German univer- and Preventive Psychology, 5, 211–224. doi:10.1016/S0962- sity clinic. Journal of Nervous and Mental Disease, 189, 249 –257.
Peter, J. P., Churchill, G. A., & Brown, T. J. (1993). Caution in the use of difference scores in consumer research. Journal of Consumer Research, Thombs, B. D., Bernstein, D. P., Lobbestael, J., & Arntz, A. (2009). A 19, 655– 662. doi:10.1086/209329 validation study of the Dutch Childhood Trauma Questionnaire—Short Preacher, A. J., & Hayes, K. S. (2004). SPSS and SAS procedures for Form: Factor structure, reliability, and known-groups validity. Child estimating indirect effects in simple mediation models. Behavior Re- Abuse & Neglect, 33, 518 –523. doi:10.1016/j.chiabu.2009.03.001 search Methods, Instruments, & Computers, 36, 717–731. doi:10.3758/ van der Does, A. J. W. (2002). BDI–II-NL. Handleiding. De Nederlandse versie van de Beck Depression Inventory—2nd edition [BDI–II-NL: The Scammell, T. E. (2003). The neurobiology, diagnosis, and treatment of Dutch version of the Beck Depression Inventory—2nd edition]. Lisse: narcolepsy. Annals of Neurology, 59, 154 –166. doi:10.1002/ana.10444 Schreiber, J. B., Nora, A., Stage, F. K., Barlow, E. A., & King, J. (2006).
van der Kloet, D., Merckelbach, H., Giesbrecht, T., & Lynn, S. J. (2011).
Reporting structural equation modeling and confirmatory factor analysis Dissociation, sleep, and memory: A heuristic integrative model. Manu- results: A review. Journal of Educational Research, 99, 323–337. doi: van IJzendoorn, M. H., & Schuengel, C. (1996). The measurement of Sierra, M. S., Senior, C., Dalton, J., McDonough, M., Bond, A., Philips, dissociation in normal and clinical populations: Meta-analytic validation M. L., . . . David, A. S. (2002). Autonomic response in depersonalization of the Dissociative Experience Scale (DES). Clinical Psychology Re- disorder. Archives of General Psychiatry, 59, 833– 838. doi:10.1001/ view, 16, 365–382. doi:10.1016/0272-7358(96)00006-2 Waller, N. G., Putnam, F. W., & Carlson, E. B. (1996). Types of dissoci- Simeon, D., Knutelska, M., Nelson, D., Guralnik, O., & Schmeidler, J.
ation and dissociative types: A taxometric analysis of dissociative ex- (2003). Examination of the pathological dissociation taxon in deperson- periences. Psychological Methods, 1, 300 –321. doi:10.1037/1082- alization disorder. Journal of Nervous and Mental Disease, 191, 738 – 744. doi:10.1097/01.nmd.0000095126.21206.3e Watson, D. (2001). Dissociations of the night: Individual differences in Soffer-Dudek, N., & Shahar, G. (2009). What are sleep-related experi- sleep-related experiences and their relation to dissociation and schizo- ences? Associations with transliminality, psychological distress, and life typy. Journal of Abnormal Psychology, 110, 536 –535. doi:10.1037/ stress. Consciousness and Cognition, 18, 891–904. doi:10.1016/ Watson, D. (2003a). Investigating the construct validity of the dissociative Soffer-Dudek, N., & Shahar, G. (2011, March 28). Daily stress interacts taxon: Stability analyses of normal and pathological dissociation. Jour- with trait dissociation to predict sleep-related experiences in young nal of Abnormal Psychology, 112, 298 –305. doi:10.1037/0021- adults. Journal of Abnormal Psychology. Advance online publication.
Watson, D. (2003b). To dream, perchance to remember: Individual differ- Spoormaker, V. I., Verbeek, I., van den Bout, J., & Klip, C. (2005). Initial ences in dream recall. Personality and Individual Differences, 34, 1271– validation of the SLEEP-50 questionnaire. Behavioral Sleep Medicine, 3, 227–246. doi:10.1207/s15402010bsm0304_4 Wu, J. C., Gillin, J. C., Bruchsbaum, M. S., Schachat, C., Darnal, L. A., Sprinkle, S. D., Lurie, D., Insko, S. L., Atkinson, G., Jones, G. L., Logan, Keator, D. B., . . . Bunney, W. E. (2008). Sleep deprivation PET corre- A. R., & Bissada, N. N. (2002). Criterion validity, severity cut scores, lations of Hamilton symptom improvement ratings with changes in and test–retest reliability of the Beck Depression Inventory—II in a relative glucose metabolism in patients with depression. Journal of university counselling sample. Journal of Counseling Psychology, 49, Affective Disorders, 107, 181–186. doi:10.1016/j.jad.2007.07.030 381–385. doi:10.1037/0022-0167.49.3.381 Stores, G. (2007). Clinical diagnosis and misdiagnosis of sleep disorders.
Journal of Neurology, Neurosurgery & Psychiatry, 78, 1293–1297.

Source: http://haraldmerckelbach.nl/artikelen_engels/2011/Sleep%20Normalization%20Decreases%20Dissociative.%20Experiences%20In%20Inpatients.pdf