1068 Using Sleep to Avoid Inpatient PTSD Treatment

Introduction

During inpatient psychiatric treatment, mattress actigraphy can be used to track nightly sleep in a zero-burden fashion to examine its relationships with other markers of treatment progress. We report here on associations between actigraphic sleep parameters and treatment markers in combat veteran engaged in inpatient treatment for posttraumatic stress disorder. We focus first on sleep scheduling, the best-validated outcome available from actigraphic data.

Methods

140 combat veteran inpatients provided between 5 and 164 nights of mattress actigraphy over the course of their inpatient psychiatric hospitalizations. The sample was characterized by multiple traumas and a high prevalence of comorbid mood and substance use disorders. Approximately one-half of the sample had undergone objective screening for OSA and for this subsample AHI < 20 was used as an inclusion criterion. Unscreened participants were required to have a BMI < 35. Inclusion also required mean actigraphic SE > 60%. The PTSD Checklist and Combat Exposure Scale were obtained at admission, from which a trauma load index was created by adding z-scores obtained from standardizing these scales.

Results

In a multiple regression analysis, after accounting for secular change in LOS, higher trauma loads (p < 0.001) and earlier bed times (p = 0.001) were both significantly associated with shorter lengths of stay (LOS; F(3,137) = 10.4, p < 0.001; adj. R-squared = 0.17). Trauma load and bed time did not interact (p = 0.7).

Conclusion

Avoidance is a criterial symptom of PTSD. Early bed times and early exits from treatment may both reflect the impact of syndromal avoidance on treatment, limiting its dosage and benefit. If so, prescribing later bed times may have utility in inpatient treatment programming.

Support

Award W81XWH-15-2-0005 from the Department of Defense, Military Operational Medicine Research Program and the Department of Veterans Affairs. This material is the result of work supported with resources and the use of facilities at the Palo Alto VA Medical Center.

Laboratory sleep correlates of nightmare complaint in PTSD inpatients

BACKGROUND

Nightmares are rare in the sleep laboratory, even in patients with posttraumatic stress disorder for whom nightmare complaints are diagnostic. Nevertheless, it is possible that laboratory conditions do not preclude the observation of telltales-nightmare-related modifications of tonic sleep-given sufficiently large samples.

METHODS

Sixty-three unmedicated, nonapneic Vietnam combat veterans undergoing inpatient treatment for posttraumatic stress disorder underwent polysomnographic testing and assessment of nightmare complaint.

RESULTS

Trauma-related nightmare complaint, but not non-trauma-related complaint, was associated with increased wake-after-sleep-onset in the sleep laboratory. No relationships between nightmare complaint and rapid eye movement sleep architecture were observed.

CONCLUSIONS

Increased wake-after-sleep-onset was specifically associated with trauma-related nightmare complaint, confirming data from other quarters suggesting they are both phenomenologically and functionally distinct from normal dreaming.

PTSD-related hyperarousal assessed during sleep

Posttraumatic stress disorder is widely understood to include "persistent symptoms of increased arousal." This presumption has rarely been tested under conditions in which effects of anticipatory anxiety could be ruled out. In this study, heart rate and electroencephalogram spectral power were assessed during sleep, a state free of most sources of artifact contaminating indices of tonic arousal. Fifty-six unmedicated nonapneic Vietnam combat-related inpatients with posttraumatic stress disorder (PTSD) and 14 controls spent 3 or more nights in the sleep laboratory during which their electrocardiograms and electroencephalograms were continuously recorded. Heart rate and electroencephalogram spectral power were quantified continuously off-line and averaged by sleep stage over all postadaptational nights. Sleep heart rate exhibited no group differences and no covariation with the severity of subjective hyperarousal reported by PTSD patients. PTSD patients exhibited a trend toward reduced low-frequency electroencephalogram spectral power during nonrapid-eye-movement (NREM) sleep. This reduction was significant during slow-wave sleep in those subjects producing scoreable slow-wave sleep. The relationship of rapid-eye-movement (REM) beta-band power to NREM beta-band power was different in PTSD patients and controls, with the patients exhibiting more beta in REM versus NREM sleep than controls. In patients, NREM sleep sigma-band electroencephalogram spectral power exhibited a positive correlation with subjective hyperarousal. Finally, a novel and surprisingly strong inverse correlation between REM-NREM sleep heart rate difference and REM percent of sleep was observed in PTSD patients only. In summary, peripheral and central measures of tonic arousal during sleep demonstrated contrastive relations to PTSD diagnostic and symptom status. The data suggest that more consideration should be directed to mechanisms of central arousal in PTSD.

Heart rate during group flooding therapy for PTSD

The objective of this study was to assess, via heart rate, the arousal levels of participants in group trauma reexposure therapy for posttraumatic stress disorder, and so to better understand this common mode of treatment, particularly in regards to its presumed curative factor, extinction. Six Vietnam combat-related PTSD inpatients participated twice weekly in group trauma reexposure therapy during which their electrocardiograms were recorded. Heart rate was quantified continuously off-line. Heart rates of participants not directly engaged in imaginal reexposure to their personal combat traumas consistently exhibited mild linear declines from the beginnings to the ends of the approximately 2.5 hour sessions. Participants actively engaged in personal combat trauma reexposure exhibited higher whole-session heart rates. Most also exhibited more specific elevation extending over the later portions of therapy sessions during which intensive reexposure usually occurred. Surprisingly, no patients exhibited focal increases in heart rate concurrent with periods of most intensive traumatic incident review as judged from videotape. Administering imaginal reexposure in a group context does not preclude substantial physiological (sympathetic) arousal, as is preconditional for extinction. Under conditions in which the actively engaged reexposure patient is reliably identified, group trauma reexposure therapy may not provide an opportunity for "vicarious" flooding in nonengaged participants.

N400 evidence of abnormal responses to speech in Alzheimer's disease

The status of semantic priming in Alzheimer's disease (AD) was examined using the speech elicited N400 component of the event-related brain potential (ERP). Speech was naturally paced, with 1 s of silence before the final word. In the semantic task, subjects attended to the meaning of the sentences for a subsequent memory test. In the phonemic monitoring task, they counted the words beginning with the letter 'p'. The effects of age were assessed by comparing young and elderly, and the effects of disease by comparing elderly and AD subjects. In healthy young and elderly subjects, N400s were large to semantically unprimed words and small to semantically primed words. In AD subjects, N400s were large to primed words, reflecting a failure of the sentence stem to prime the final word, and probably an impairment in semantic knowledge. The N400 priming effect was not smaller during the phonemic than semantic task in any group, suggesting that the semantic qualities of speech are processed even when subjects are attending to phonemic qualities. N400 latency was delayed with age and further delayed with dementia.

First night effects in post-traumatic stress disorder inpatients

Recent data have suggested that first night effects are attenuated in inpatient depressive subjects. We examined first night effects in 80 inpatients hospitalized for post-traumatic stress disorder (PTSD) as well as non-hospitalized PTSD sufferers and non-ill control subjects. PTSD inpatients exhibited attenuated first night effects compared to non-hospitalized PTSD sufferers and non-trauma-exposed controls. Non-ill combat-exposed subjects also exhibited small first night effects. Within the inpatient sample, severity indices of PTSD, depression and anxiety failed to account for variance in first night effects. These data demonstrate attenuation of first night effects in a new inpatient population and suggest their statistical independence vis-a-vis a range of relevant symptoms. Both the attenuation of first night effects in PTSD inpatients and their accentuation in PTSD outpatients may be indicative of enhanced sensitivity to the sleep environment. Conversely, the trend to small first night effects in non-ill combat-exposed subjects may reflect a dimension of their apparent resistance to traumatic stress.

Sleep and depression in combat-related PTSD inpatients

The sleep of 27 unmedicated Vietnam combat-related posttraumatic stress disorder (PTSD) inpatients was monitored for 3 nights. Depressive comorbidity was considered both as a diagnostic category using DMS-III-R criteria, and as a continuous variable using the Beck Depression Inventory (BDI). Data collected included sleep architecture features that have discriminated unipolar depressives from controls in many prior studies, rapid eye movement (REM) sleep latency, and slow-wave sleep time, as well as two additional indices that have sometimes discriminated depressives from controls in waking studies-baseline heart rate and facial electromyography. Structured Clinical Interview for the DSM-III-R (SCID)-diagnosed PTSD+major depressive disorder (MDD) patients failed to exhibit shorter REM latencies, greater REM percents of sleep, or greater REM densities than PTSD-MDD patients, but did exhibit less slow wave sleep. PTSD+MDD patients also exhibited less facial (mentalis) electromyographic activity. REM densities and baseline heart rates were equivocal. REM density, baseline heart rate, and mentalis electromyography all correlated with the BDI, the former two positively, the last, negatively. In summary, SCID-diagnosed PTSD+MDD patients failed to exhibit the classic REM sleep architectural modifications associated with unipolar depression, despite the fact that several other psychophysiologic indices of dysphoria were detectable in their sleep.

Subjective versus objective sleep in Vietnam combat veterans hospitalized for PTSD

Twenty-five Vietnam combat veterans with chronic severe posttraumatic stress disorder (PTSD) completed a sleep self-report questionnaire on admission to an inpatient treatment program. Between 1 and 2 months later each spent 3 or more nights in the sleep laboratory. When self-report and laboratory findings were compared, significant relationships were observed between sleep schedule items such as time-to-bed/time-out-of-bed and polysomnographic measures of sleep. In contrast, global ratings of sleep quality were generally unrelated to polysomnographic measures. These findings may have implications for survey research assessing sleep quality in traumatized populations.

N4 to spoken sentences in young and older subjects

Event-related potentials (ERPs) to spoken sentences were recorded from young and elderly subjects. Sentences ended in semantically typical or atypical words. Subjects listened to the sentences during two attention conditions. In a semantic attention task, subjects listened to the meanings of the sentences for a subsequent memory test. In a phonemic attention task, subjects counted words beginning with the letter "p". Following these tasks, the sentence-final words were presented in isolation. A conventional auditory oddball paradigm was also presented. The N4 components elicited by semantically atypical sentence-final words were slightly attenuated and delayed in the older group. N4 was not influenced by the attention manipulation. Young but not older subjects' N4s were attenuated by repetition. When the sentence-final words were presented a third time in isolation, the N4s elicited were similar in amplitude and latency to those elicited in sentence context, but were not attenuated by repetition. In young and older subjects, the peak amplitude of the sentence-elicited N4 difference wave form was highly correlated with the number of correct recognitions of typical words, but not atypical sentence-final words. In young but not older subjects, sentence-elicited N4 amplitude was correlated with the amplitude of the P3 elicited in the auditory oddball paradigm.

Probing the time-course of the auditory oddball P3 with secondary reaction time

An elementary neural model of the P3 is proposed in which the P3 is held to manifest a brief, widely-distributed, inhibitory event. A preliminary and indirect test of the model is described using secondary-task methodology. Manual reaction times were measured to probe clicks delivered during the presumed time-course of an auditory oddball P3. We observed that reaction times to probes presented after oddball stimuli were significantly slowed as compared to reaction times to probes presented after standards. The latency of maximum reaction time slowing corresponded generally to the latency of the P3. The latency of maximum reaction time slowing did not respond to a manipulation varying the latency of the P3. Thus, some of the obtained results were consistent with the P3-inhibition hypothesis, whereas others were not. Secondary-task methodology may provide a valuable new approach to understanding the late event-related potentials.