Increased sleep spindle dynamics
suggest overconsolidation in PTSD
A.C. van der Heijden
10178996
Brein & Cognitie
L.M. Talamini
W.F. Hofman
24-‐09-‐2016
Index
1. Introduction
page 3-‐4
2. Materials & Methods
2.1 Participants
page 4-‐5
2.2 Polysomnography and general procedure
page 5
2.3 Data analysis
2.3A Sigma power analysis
page 5
2.3B Sigma fluctuation detection
page 6-‐7
2.3C Sigma fluctuation density &
page 7
average of sigma fluctuation characteristics
2.3D Distribution of peak amplitude density in N2
page 7
density in N2
3. Results
3.1 Sigma power analysis
page 8
3.2 Density of sigma fluctuations &
page 9-‐10
average of sigma fluctuation characteristics
3.3 Distribution of peak amplitude density in N2
page 10
4. Discussion
page 11
5. References
page 12-‐13
Abstract
PTSD is a significant health problem with an estimated prevalence of 8% in the general population. Key
symptoms of the disorder are aversive memory intrusions and overgeneralization of the traumatic event, through flashbacks and nightmares. This may be interpreted as a form of memory disturbance. In fact, the symptoms are reminiscent of runaway consolidation, in which impaired leads to the formation of super-strong, overgeneralized memory representations that get reactivated inappropriately. Sleep disturbances are another key symptom of PTSD. Interestingly, sleep has an important role in memory consolidation. In particular, sleep spindles in different cortical areas reflect the reprocessing and consolidation of specific memory traces. Given their strong relationship with memory reprocessing during sleep and the reported memory and sleep alterations in PTSD, sleep spindles may play a role in the aetiology of PTSD. Moreover, thalamic dysfunction has been reported in PTSD patients, which may be linked to abnormalities in spindle generation. In our project, absolute sigma power, and several sigma fluctuations characteristics were analysed and compared between traumatised police officers and military personnel with PTSD (N=14) and without PTSD (N=14). Given the lack of a physiologically motivated delineation of spindling phenomena, a detection method with minimal assumptions regarding spindle “anatomy” was used to obtain an unbiased representation of all present sigma fluctuations. Waxing/waning couplets with an amplitude over five microvolt were considered. For each detected waxing/waning couplet, several variables were computed (e.g. duration, amplitude). PTSD was associated with increased spindle activity, apparent in increased NREM sigma power, increased mean peak amplitude of sigma fluctuations and a shift in the distribution of sigma fluctuation amplitudes towards a larger amplitude relative to controls. Similar trends, though less pronounced, are apparent considering other spindle parameters, such as duration and wax and wane energy (amplitude*duration). The increased spindle activity in PTSD relative to controls may reflect overconsolidation in the disorder.
Interestingly, unaltered wax and wane speed in PTSD suggest that cell recruitment and desynchronization in spindle generation processes function similarly to those of controls, this indicates that an increase in PTSD spindle activity may be related to higher activity in the spindle generating mechanism rather than a dysfunction of it. The minimal assumption analyses revealed details regarding sigma fluctuation abnormalities in PTSD that would have been missed by analysing only heuristically detected spindles. In conclusion, if replicable in a study without SSRI medication, the increased sleep spindle dynamics in PTSD may form part of the mechanism through which the profound sleep disturbance in this disorder contributes to emotional memory problems.
1. Introduction
Post-traumatic stress disorder (PTSD) is a psychiatric disorder that entails a person's inability to recover from the stressful effects of experiencing a traumatic event (Kessler et. al., 1995). In Europe, 1.9% of the general population suffers from PTSD (Alonso et al., 2004). Key symptoms of the disorder are aversive memory intrusions and overgeneralization of the traumatic event, through flashbacks and nightmares. This may be interpreted as a form of memory disturbance. In fact, the memory
disturbance symptoms are reminiscent of runaway consolidation, in which super-strong,
overgeneralized memory representations are formed that get reactivated inappropriately (Meeter & Murre, 2003).
Besides memory disturbances, PTSD patients suffer from severe sleep problems. Sleep problems are the most prevalent PTSD symptoms, roughly 70% of PTSD patients have co-occurring sleep disorders (Ohayon & Shapiro, 2000). The sleep problems typically include insomnia, nightmares, distressed awakenings, nocturnal panic attacks and sleep terrors (Germain, 2013). Interestingly, sleep has been shown to have an important function in memory consolidation (Stickgold, 2005).
More specific, sleep-dependent memory consolidation is thought to be supported by sleep spindles (Cox et al., 2014, Fogel & Smith, 2011; Kaestner et al., 2011; Mednick et al., 2013). Sleep spindles are waxing and waning thalamocortical oscillations in the sigma frequency band (11-16 Hz), which can be divided into slow (11-13 Hz), and fast (14-16 Hz) spindles (Astori et al., 2013). These oscillations in the sigma frequency band exist of three phases: the waxing, middle and waning phase. Distinct neuronal activity underlies each phase of the oscillation (Astori et al., 2013). Sleep spindles have been
suggested to reflect the reprocessing of specific memory traces (Cox et al., 2014). Given their strong relationship with memory reprocessing during sleep, the reported memory-, sleep disturbances,
fluctuations in the sigma frequency band may offer new insights in the aetiology of PTSD.
A small number of previous studies suggests possible alterations in PTSD sigma activity. First, a dysfunction in the thalamus, the spindle generating structure, has been reported in PTSD patients (Lanius et al., 2006). This thalamic dysfunction in PTSD may indicate altered sigma activity as a consequence. Second, a reduction of GABA type A receptor [11]Flumazenil binding in veterans with PTSD in comparison to veterans without PTSD has been found throughout the cortex, hippocampus and thalamus (Geuze et al., 2008). The GABA A receptor is essential in spindle production, as blocking GABA-A receptors lead to complete disappearance of in vitro spindle oscillations in thalamocortical slice preparations (Jacobsen et al., 2001), whereas GABA-A agonists increased spindle density (Mednick et al., 2013). Third, a previous study from our group demonstrated an increased frontal relative sigma power in PTSD patients relative to trauma controls (Talamini et al., unpublished manuscript).
To gain insight in PTSD sleep spindle activity, absolute sigma power and several sigma fluctuation characteristics, such as duration, peak amplitude and density were examined in war veterans and former police officers. All of them experienced a traumatic event but not all suffered from PTSD. Based on the increase in relative sigma power in PTSD compared to controls that was previously found by our group (Talamini et al., unpublished manuscript), an increased spindle activity in PTSD patients was expected relative to trauma controls. No specific expectations were held of how this increase in sigma activity reflected in absolute sigma power and sigma fluctuation characteristics.
A difficulty in previous spindle research is that many different detection methods were used, which held different a priori criteria for defining sleep spindles. These differences in a priori criteria possibly underlay divergent results in previous spindle research. Moreover, these a priori criteria were lacking evidence of physiological delineation, and often included high amplitude thresholds. As a
consequence, parts of the spindle activity possibly may have been missed in previous research. In collaboration with PHI International (Amsterdam, The Netherlands), a minimal assumption method was developed for this project, to obtain an unbiased representation of all sigma fluctuations. This detection method not only included density measures, as commonly used in traditional sleep spindle research, but also examined waxing and waning characteristics of the sigma fluctuations. Since this project has been the first to analyse spindle activity in PTSD, an extensive set of sigma fluctuation characteristics were analysed.
Full polysomnography was recorded for all participants during one night. Only the EEG signal was analysed for this study. Absolute NREM sigma power, sigma fluctuation density and mean values per sigma fluctuation characteristic in N2 were determined and compared between groups. In addition, a distribution of peak amplitude density in N2 was compared between PTSD and trauma controls.
Material & Methods 2.1 Participants
A total of 28 participants were recruited and measured during one night at the Arq Psychotrauma centres in The Netherlands, which offers third line, highly specialised health care for trauma-related disorders. 14 of the participants were diagnosed with PTSD and 14 were trauma controls. Trauma control participants did experience a traumatic event in their lives, as defined by DSM-IV (APA, 1994), but did not suffer from PTSD. PTSD participants scored >50 points on the Clinician Adminsistered PTSD Scale (CAPS) (Blake et. al.,1995). Groups were matched on age, gender, educational level and profession, see table 1 for the sociodemographic details of the participants. Participants were asked to refrain from medication and were excluded if they suffered from acute suicidality, psychotic behaviour, drug abuse, a prior history of neurological or sleep disorders, an atypical sleep pattern with <6 hours of sleep per night or a sleep window outside 10 pm and 10 am. Six of the participants in the PTSD group were on medication during the experiment that might have had an effect on spindling, but for medical ethical reasons could not be interrupted. Five of the PTSD participants were using serotonin reuptake inhibitors (SSRIs) (Paroxetine, Venlafaxine, Setraline, Citalopram) and one participant was using the
hypnotic temazepam. The participant which used temazepam was excluded for statistical analysis, since temazepam is known to increase spindle amplitude and density (Plante et al., 2015).
2.2 Polysomnography and general procedure
For polysomnographic recording, subjects were given the opportunity to sleep undisturbedly for nine hours during a lights-off period starting between 11 and 12 PM, depending on habitual sleep times. PSG was performed using ambulatory 16-channel Porti amplifiers (TMS-i) and Galaxy sleep analysis software (PHI-international). PSG consisted of an EEG recording (F3, F4, C4, O2, referenced to the average of linked mastoids), two EOG electrodes monitoring eye-movements, and two electrodes for submental EMG. Further sensors were ECG monitoring heart rate, plethysmography monitoring blood oxygenation, tibial EMG to detect leg movements, probes measuring nasal airflow, and piezo
respiratory bands for thoracic and abdominal respiratory effort to monitor breathing and sleep apnea. All signals were sampled at a rate of 512Hz
2.3 Data analysis 2.3a Sigma power
The sigma frequency content of the EEG was analysed using a fast Fourier transform-based Spectral analysis (4 s Hamming window, 50 percent overlap, 11-16 Hz, 0.25 Hz bin size). This was done on each electrode (F3, F4, C4, O2) for light sleep (N1+N2) and deep sleep separately. Sleep stages were scored visually on the C4 derivation, according to the standard AASM criteria (AASM)
Table 1. Sociodemographic details of participants in the PTSD group and in the Trauma control Group
Characteristic PTSD group (n=14) Trauma control group (n=14)
Professional background Police 10 (71%) 10 (71%) Veteran 4 (29%) 4 (29%) Mean age (SD) 46.2 (8.4) 44.4 (8.6) Gender (n, %) Male 13 (93%) 13 (93%) Female 1 (7%) 1 (7%) Educational level (n, %)
Lower or middle vocational
education 11 (79%) 11 (79%)
Higher vocational education
3 (21%) 3 (21%)
Test scores
Mean CAPS score (SD)
82.5 (12.3) 5.4 (4.3)
Mean IES-R score (SD)
52.8 (11.8) 5.3 (6.5)
Mean HADS Depression (SD) 13.9 (3.5) 2.0 (2.9)
Mean HADS Anxiety (SD) 14.5 (3.4) 2.5 (2.2)
Note: CAPS=clinical-administered PTSD scale. IES-R=Impact of Event Scale – Revisited. HADS= Hospital Anxiety and Depression scale.
Table from: Talamini et al., (2015). Quantitative electroencephalography in traumatized police officers and veterans with and without post-traumatic stress disorder during rapid eye movement (REM) and non-REM sleep, Department Psychology Brain & Cognition University of Amsterdam - unpublished manuscript.
consensus report, Iber et al., 2007). Sigma power was computed per 30s epoch. Absolute NREM sigma power was compared between PTSD and control using a two-tailed Mann-Whitney test.
2.3B Sigma fluctuation detection
An algorithm has been created for this experiment to analyse sigma fluctuations, in collaboration with PHI international (Amsterdam, The Netherlands). Based on the results of absolute sigma
power, the electrode for sigma fluctuation analyses was chosen. Sleep EEG data of electrode F4-M2 was filtered in the sigma range (11-16 Hz) using a Finite Impulse Response(FIR) and bandpass filter that were built into the algorithm. With a moving window of 0.2 seconds per sample, the standard deviation of the filtered signal was calculated. The result of this calculation was a sigma envelope with a sample rate of 512 Hz. The envelope represents the burst-like shape of the signal, as shown in figure 1C. To define peaks and troughs, the slope was calculated with a sample rate of 5 Hz. The sample rate of 5 Hz was obtained by summing the previously described standard deviations over 0.2s by taking the root mean square. A peak was defined if the slope changed from positive to negative, a trough vice versa. A low hysteresis of 0.2 µV was used to avoid low amplitude background noise in the envelope, as well as to include a broad range of sigma fluctuations for analyses (see figure 1). All sigma fluctuations with an amplitude between 5 and 35 μV were analysed. Twelve parameters were calculated for each individual sigma fluctuation: duration, peak amplitude, waxing amplitude, waning amplitude, wax wane amplitude, waxing duration, waning duration, waxing energy, waning energy, wax wane energy, waxing speed, waning speed. See table 2 for a description of each sigma fluctuation characteristic
.
Fig. 1
Influence of hysteresis value on the detection of sigma fluctuations
1A. Raw EEG signal of electrode C3 (0.3-35 Hz). 1B. EEG signal of electrode C3 filtered in sigma frequency (11-16 Hz). 1C. Calculated sigma envelope of the signal in 1B. 1D The number of boxes represents the number of sigma fluctuations. A small hysteresis value gives a large number of sigma fluctuations, a large hysteresis vice versa.
Table 2 Definitions of the analysed sigma fluctuation characteristics
The sigma fluctuation characteristics were calculated for each fluctuation in the sigma envelope which had a peak amplitude between 5-35 µV.
Sigma fluctuation characteristics
Parameter Unit Definition Amplitude parameters
Peak amplitude µV Absolute level of the peak of the sigma fluctuation
Waxing amplitude µV Amplitude of the waxing of the sigma fluctuation. Calculated as the difference
between the absolute level of the peak and the absolute level of the trough prior to the peak
Waning amplitude µV Amplitude of the waning of the sigma envelope. Calculated as the difference
between the absolute level of the peak and the absolute level of the trough after the peak.
Mean wax and wane amplitude
µV Mean of the waxing and waning amplitude
Duration parameters
Duration s Duration of a sigma fluctuation, from trough to peak to trough
Waxing duration s Duration of a wax, duration of the sigma fluctuation from trough to peak
Waning duration s Duration of a wane, duration of the sigma fluctuation from peak to trough
Energy parameters
Waxing energy µVs Waxing amplitude * waxing duration
Wave down energy µVs Waning amplitude * waning duration
Wax and wane energy µVs Waxing amplitude * waxing duration + waning amplitude * waning duration
Speed parameters
Waxing speed µV/s Waxing amplitude / waxing duration
Waning speed µV/s Waning amplitude / waning duration
2.3C Density of sigma fluctuations and averages per sigma fluctuation characteristic in N2
As N2 is typically characterised by spindling, in contrary to N1, and relative more time is spent in this sleep stage in comparison to N3, analyses of N2 sigma fluctuations were supposed to give a robust representation. Moreover, deep sleep could possibly be lowered in PTSD as a consequence of the sleep disturbances in the disorder, which in turn could decrease the power of statistical analyses on sigma fluctuations in this sleep stage. For that reason, N2 sigma fluctuation density, the total number of sigma fluctuations divided by the time spent in sleep stage, was compared between PTSD and trauma controls, using a two-tailed Mann-Whitney test. Next, mean values were calculated per participant for each sigma fluctuation characteristic and compared between groups using a two-tailed Mann Whitney-U test.
2.3D Distribution of peak amplitude density in N2
A frequency distribution displayed all peak amplitudes between 5-35 µV and their corresponding densities in N2. Due to time limitations, only a frequency distribution for peak amplitude N2 was made. Bin sizes for the frequency distributions were calculated with the Matlab algorithm, calcnbins, that returns the ideal bin size for a distribution. Within this algorithm the Freedman Diaconis rule is used (h=2*(IQR(x)/n1/3 ), where “h” is the ideal bin size and “IQR” the inter quartile range of the data). The Freedman-Diaconis rule is less sensitive to outliers in the data and is considered to be more suitable for heavy-tailed distributions in comparison to other binning methods. Ideal bin size was determined per participant for peak amplitude in N2. Next, the average of the individual ideal bin sizes over the two groups was calculated for this parameter. The nonparametric two sample Kolmogorov-Smirnov was used to compare the peak amplitude density distributions in N2 between PTSD and controls.
3. Results
3.1 Sigma power (F3, F4, C4, O2)
Sigma power analyses revealed a spatial power shift, which highlighted that increases in NREM sigma power in PTSD relative controls were most pronounced at frontal electrodes (F4: N1 U=44 p=0.038*, N2
U=44, p=0.038*, NREM U=42 p=0.029
*
, F3: N1 U=44, p=0.038*
, N2 U=67, p=0.369, NREM U=51 p=0.086), lesspronounced centrally (C4: N3 U=43.5, p=0.035
*
, N1 U=74, p=0.590, N2 U=70 p=0.457, NREM U=76 p=0.663), whereas occipitally no differences between groups were present (O2, N1 U=84 p=0.980, N2 U=84p=0.980, N3 U=64.5, p=0.304, NREM U=87 p=0.980) (see fig. 2). The increased sigma power in PTSD
relative to controls may represent an increase in spindling activity. No alterations in PTSD in frontal N3 sigma power were present (F3: U=71.5, n1=13 n2=13, p=0.504, F4: U=76.5 n1=13 n2=13 p=0.687
.
Results of3.2 Density of sigma fluctuations and average sigma fluctuation characteristics in N2
Analysing characteristics of fluctuations in the sigma envelope, revealed a higher mean and maximum peak amplitude in PTSD than controls (U=50, n1=13, n2=14, p=0.029, max. PTSD 34.7 µV, max. controls
32.6 µV), while local peak-to trough fluctuations in the sigma envelope were more shallow in PTSD
relative to controls (waxing amplitude:U=37, p=0.005, waning amplitude: U=37, p=0.005). This indicates an inflated envelope and increased sigma activity in PTSD in contrast to controls. Similar trends, although less pronounced, were demonstrated by decreases in duration and energy characteristics of the sigma fluctuations in PTSD relative to controls (Duration U=63, p=0.112, waxing duration U=62, p=0.103, waning
duration U=64, p=0.124, waxing energy U=80, p=0.421, waning energy p=0.597, wax wane energy U=85,
p=0.567). Interestingly, the wax and wane speed did not differ between groups (Waxing speed:U=95,
p=0.901, waning speed: U=99, p=0.982). This suggests that increases in the spindle activity were not due
to long stretched sigma fluctuations in PTSD, but are more likely due to an increase in sigma fluctuation density underlying the envelope. The density of sigma fluctuations in the sigma envelope
A
B
C
D
Fig. 2 Absolute sigma power of PTSD versus controls at electrodes F3, F4, C4 and O2
A shift towards an increase in frontal PTSD power relative to controls is represented in the above standing figures. * p<0.05
*
*
*
itself was unaltered in PTSD (U=81, n1=13, n2=14, p=0.44), however, this does not mean that sigma fluctuation density underlying the envelope does not differ between PTSD and controls, since the window over which the envelope is calculated is several fluctuations long and may have smoothed out individual sigma fluctuations. See table 3 for a summary of the means for the twelve sigma fluctuation parameters between PTSD and controls spindle in N2 and figure 3 for a visual representation of how these means reflect in a PTSD versus control sigma fluctuation.
Table 3 Summary of results for twelve sigma fluctuation parameters in N2
For reach sigma fluctuation parameter in N2 at electrode F4, unities, mean values, standard deviations and p-values are displayed
Sigma fluctuation N2 Parameter PTSD Control p-value Amplitude parameters Peak amplitude µV 7.06 ± 1.93 6.69 ± 1.64 p=0.029* Waxing amplitude µV 1.97 ± 1.84 2.10 ± 1.78 p=0.005** Waning amplitude µV 1.96 ± 1.83 2.10 ± 1.77 p=0.005**
Wax wane amplitude µV 1.97 ± 1.37 2.10 ± 1.33 p=0.005**
Duration parameters Duration s 0.21 ± 0.15 0.22 ± 0.15 p=0.112 Waxing duration s 0.10 ± 0.10 0.11 ± 0.10 p=0.103 Waning duration s 0.10 ± 0.10 0.11 ± 0.10 p=0.124 Energy parameters Waxing energy µVs 0.35 ± 0.61 0.37 ± 0.57 p=0.421 Waning energy µVs 0.35 ± 0.60 0.36 ±0.56 p=0.597
Wax wane energy µVs 0.71 ± 0.91 0.73 ± 0.86 p=0.567
Speed parameters
Waxing speed µVs 21.74 ± 12.03 21.74 ± 12.31 p=0.901
Fig. 3 Visual representation of the sigma fluctuation characteristics found in PTSD versus controls Fictive numbers on the x- and y-axes have been used for easy interpretations, no negative numbers were present in the real data. Decreases in PTSD waxing amplitude & waning amplitude (p<0.05), duration (p=0.112) and energy (amplitude * duration) (p>0.05) characteristics were shown in PTSD relative to controls. Waxing and
waning speed between groups was equal (p>0.05).
3.3 Distribution of peak amplitude density in N2
Peak amplitude density in N2 was differently distributed in PTSD and controls, PTSD patients showed a shift towards higher amplitudes in contrast to controls (two-sample KS-test, p=0.025) (see fig 4). PTSD patients demonstrated a higher density than controls in all amplitudes larger than 7 μV. Between 5-7 μV, amplitude densities did not differ between groups. In this range, 60% of the PTSD sigma fluctuations and 64,7% of the controls sigma fluctuations were present. The high proportion of low amplitude sigma fluctuations in both groups underlines the relevance of a minimal assumption approach, as a method with high amplitude thresholds would have missed a majority of this peak amplitude density distribution in N2.
Fig. 4 Increased density in PTSD across N2 peak amplitude distribution relative to controls -‐4 -‐3 -‐2 -‐1 0 1 2 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 Log D ensity (Counts/m in) Peak amplitude (µV) PTSD CTRL P<0.05
Densities were log transformed for visualisation purposes.
Discussion
PTSD is associated with increased spindle activity, apparent in increased NREM sigma power,
increased mean peak amplitude of sigma fluctuations and a shift in the distribution of sigma fluctuation amplitudes towards a larger amplitude. Similar trends that support an increased spindle activity in PTSD, though less pronounced, were apparent in other sigma fluctuation characteristic, such as duration and wax and wane energy. Interestingly, no alterations in waxing and waning speed were found between PTSD patients and controls, this suggests that gradual cell recruitment and
desynchronization in thalamic-reticular nucleus loops of spindle generation are unaltered in PTSD. An increase in PTSD spindle activity may therefore be expressed in an increased sigma fluctuation density in PTSD relative to controls, reflecting a higher activity in the spindle generating mechanism rather than a dysfunction it.
Increases in PTSD spindle activity support the concept of overconsolidation of traumatic memories in PTSD, since sleep spindles are known to reflect memory reprocessing (Cox et al., 2014). In addition, the frontal location at which an increased PTSD sigma power has been found relative to controls, may also serve as evidence for over consolidation in PTSD, since slow spindles typically occur frontally (11-13 Hz) (Astori et al., 2013), and they are thought to support neural processes for long-term memory representations in the neocortex (Mölle et al., 2011).
Besides the previously described link with memory consolidation, a positive relationship between sigma power and subjective hyperarousal has been found in PTSD patients, but not in trauma controls, trauma free subjects or other frequency bands (Woodward et al., 2000). An explanation for this finding may be that hyperarousal is related to the overconsolidation of the traumatic memory, as the stronger the emotional memory of the traumatic event is overconsolidated, the more severe the subjective hyperarousal may be. Spindle activity might underlie these both symptoms. Future studies could investigate whether a relationship exists between sigma power or spindle activity and
nightmares in PTSD, as nightmares induce a form of subjective arousal and are part of the memory disturbances in the disorder.
In this project, the minimal assumption method has been applied for the first time. It revealed details regarding sigma fluctuation abnormalities in PTSD that would have been missed by analysing only heuristically detected spindles. The minimal assumption method revealed that PTSD and controls have a distinct amplitude density distribution across 5-35μV in N2, parts of this representation would likely have been missed by using heuristic spindle detection methods that use high fixed amplitude thresholds.
Of note, as the method is still under development, spindle activity, energy- and speed parameters were rough estimations of the actual sigma activity. For obtaining more precise estimations of sigma fluctuation speed, further fine-tuning of the algorithm would be necessary. Also, central right power differences between PTSD and trauma controls in during deep sleep should be interpreted carefully, as the profound sleep disturbances in PTSD may have led to a reduction of data points in this sleep stage. Finally, the SSRI medication taken by some of the subjects with PTSD might have influenced sigma fluctuation parameters. The latter possible influence is currently under investigation in our lab.
Concluding, if replicable in a study without SSRI medication, the sigma fluctuation abnormalities in PTSD may form part of the mechanism through which the profound sleep disturbance in this disorder contributes to emotional memory problems. Sigma fluctuations may from a fruitful future research
References
Astori, S., Wimmer, R. D. & Lüthi, A. (2013). Maniupulating sleep spindles – expanding views on sleep, memory and disease. Trends in Neuroscience, 36(12), 738-748.
doi:10.1016/tins2013.10.001
Alonso, J., Angermeyer, M.C., Bernert, S., Brffaerts, R., Brugha, R., de Girolamo, G. et al., 2004. Prevalence of mental disorders in Europe: results from the European Study of the Epidemiology of Mental Disorders (ESEMeD) project. Acta Psychiatrica Scandinavica Supplementum, 2004a(420), 21-27. doi: 10.1002/jts.20357
Andrillon, T., Nir, Y., Staba, R. J., Ferrarelli, F., Cirelli, C., Tononi, G. et al. (2011). Sleep spindles in humans: Insights from intracranial EEG and unit recordings. Journal of Neuroscience, 31(49), 17821-17834. doi: 10.1523/JNEUROSCI.2604-11.2011
Cox, R., Hofman, W.F., de Boer, M. & Talamini, L.M. (2014). Local sleep spindle modulations in relation to specific memory cues. Neuroimage, 99, 103-110. doi:10.1016
Fogel, S. M. & Smith, C. T. (2011). The function of the sleep spindle: A physiological index of intelligence and a mechanism for sleep-dependent memory consolidation. Neuroscience and
Biobehavioral Reviews, 35(2011), 1154-1165. doi:10.1016/j.neubiorev.2010.12.003
Geuze, E., Berckel, B. M. N., Lammertsma, A. A., Boellaard, R., Kloet, C. S., Vermetten, E. et al. (2008). Reduced GABAA benzodiazepine receptor binding in veterans with post-traumatic stress disorder. Molecular Psychiatry, 13, 74-83.
Germain, A., Hall, M., Shear, M. K., Nofzinger, E. A. & Buysse, D. J. (2006). Ecological study of sleep disruption in PTSD: a pilot study. Annals of the New York Academy of Sciences, 1071, 438–441. 10.1176/appi.ajp.2012.12040432
Jacobsen, R. B., Ulrich, D. & Huguenard, J. R. (2001). GABA-B and NMDA receptors contribute to spindle-like oscillations in rat thalamus in vitro. Journal of Neurophysiology, 86, 1365-1375.
Kessler, R., Sonnega, A., Bromet, E., Hughes, M., & Nelson, C. (1995). Posttraumatic stress disorder in the National Comorbidity Survey. Archives of General Psychiatry, 52(12), 1048–1060.
Lanius, R. A., Bluhm, R. & Lanius, U. (2006). A review of neuroimaging studies in PTSD:
Heterogeneity of response to symptom provocation. Journal of Psychiatric Research, 40, 709-729.
Mednick, S. C., McDevitt, E. A., Walsh, J. K. Wamsley, E., Paulus, M., Kanady, J. C. et al. (2013). The critical role of sleep spindles in hippocampal dependent memory: a pharmacology study. The
Journal of Neuroscience, 13(10), 4494-4504. doi:10.1523/JNEUROSCI.3127-12.2013
Meeter, M. (2003). Control of consolidation in neural networks: Avoiding runaway effects. Connection
Science, 15(1), 45-61. doi:10.108/0954009031000149591
Mölle, M., Bergmann, T. O., Marshall, L. & Born, J. (2011). Fast and slow spindles during the sleep slow oscillation: Disparate coalescence and engagement in memory processing. Sleep, 34(10), 1411-1421.
Ohayon, M. M., & Shapiro, C. M. (2000). Sleep disturbances and psychiatric disorders associated with posttraumatic stress disorder in the general population. Comprehensive Psychiatry, 41(6), 469– 478. doi: 10.1053/comp.2000.16568
Plante, D.T., Goldstein, M.R., Cook, J.D., Smith, R.,Rriedner, B.A., Ferrarelli, F. et al. (2015). Effects of oral temazepam on sleep spindles during NREM eye movement: A high-density EEG investigation.
European Neuropsychopharmicalogy, 10(25), 1600-1610. doi: 10.1016/j.jad.2012.06.016
Stickgold, R. (2005). Sleep-dependent memory consolidation. Nature, 437(7063), 1272-1278. Doi: 10.1038/81756
Woodward, S. H., Murgburg, M. M. & Bliwise, D. L. (2000). PTSD-related hyperarousal assessed during sleep. Psychology & Behaviour, 48, 1081-1087.
