Research Article: Changes in sleep architecture in German Armed Forces personnel with posttraumatic stress disorder compared with depressed and healthy control subjects

Date Published: April 17, 2019

Publisher: Public Library of Science

Author(s): Laura Haberland, Helge Höllmer, Holger Schulz, Kai Spiegelhalder, Robert Gorzka, Soraya Seedat.


This study compares the sleep architecture of patients with posttraumatic stress disorder (PTSD) with that of both patients with depression and subjects with no mental disorder.

45 German armed forces personnel with PTSD, 72 German armed forces personnel with depression and 24 healthy control subjects underwent 24-hour polysomnography. The effects of group membership, medication and the statistical interaction of group and medication were analysed for the following variables: sleep onset latency, REM sleep latency, slow-wave sleep and REM sleep percentages.

Sleep onset latency was significantly prolonged in both the PTSD and the depression group. Moreover, psychotropic medication was associated with significantly prolonged REM sleep latency.

The impact on sleep onset latency is of special clinical relevance in that according to preliminary studies, it is of major importance for subjective sleep quality. In contrast to the other parameters, an increase in sleep onset latency results in a subjective reduction in sleep quality which can lead to hyperarousal and increased preoccupation with sleep, which in turn may lead to dysfunctional sleep patterns.

Partial Text

Non-organic sleep disorders play an important part in mental illness. Numerous studies and systematic reviews have shown that pathological changes in sleep architecture play a major role as secondary symptoms in affective disorders [1], addiction [2] and schizophrenia [3]. It is especially after experiencing traumatic events that dyssomnias in the form of sleep-onset and sleep-maintenance insomnia as well as parasomnias such as nightmares are generally observed [4]. Studies of war veterans [5] and interviews of Holocaust survivors [6] suggest that such sleep disorders may persist for years after the traumatic event.

An analysis of covariance was performed to test hypothesis 1. The PTSD group was tested against the healthy and depressed groups. The age variable was used as a covariate. Table 1 shows significant differences of medium effect size between the three study groups in terms of sleep onset latency. Post-hoc analyses with Sidak adjustment yielded higher values for subjects of the PTSD group (M = 51.3, SD = 37.8) than for subjects of the healthy group (M = 20.1, SD = 15.9, p < .001). The same was observed in a comparison of the depressed group (M = 38.9, SD = 34.0) with the healthy group (p = .046). The PTSD group was not significantly different from the depressed group in terms of sleep onset latency (p = .113). With a small effect size, the REM sleep percentage of sleep period time was greater in the PTSD group than in the control group. The variance percentage explained by the age covariate was significant only for the variable of slow-wave sleep (p = .040). The analysis results for the sleep onset latency variable show that this parameter is noticeably prolonged in subjects with PTSD as well as depression (differences with a small effect size, but not significant). While the literature suggests that this parameter is unspecific and pathologically prolonged in many mental disorders [3,34,35,36] it is nevertheless of special clinical relevance. Of all parameters analysed in this study, sleep onset latency is associated most closely with subjective sleep quality. Many studies show that an increase in slow-wave sleep (e.g. induced by medication) did not result in a major subjective increase in sleep quality [1], while a reduction in sleep onset latency and consolidation of sleep architecture, including a reduction in nighttime waking, did. According to Riemann et al. [17], prolonged sleep onset latency may lead to hyperarousal and increased preoccupation with sleep, which in turn may lead to dysfunctional sleep patterns. Questionnaires on sleep behaviour and subjective sleep quality have shown that PTSD patients in particular are strongly affected and that this may even correlate with the severity of PTSD symptoms [30]. This would suggest that, taking into account hyperarousal theory, disturbed sleep may not only be a posttraumatic symptom, but also a comorbidity that maintains the disorder or even a risk factor directly involved in its pathogenesis [18,19]. In summary, the following conclusions can be drawn from this study:   Source: