Spatially specific changes in EEG spectral power in post traumatic stress disorder during REM and NREM sleep.

Objectives: Sleep problems are a core feature in PTSD. However, a robust objective measure for the sleep disturbance has yet to be found. The current study assessed whether the spatial distribution of EEG spectral power in PTSD would provide such a measure.

Methods: EEG power in F3, F4, C4 and O2 was calculated (FFT) for non-REM and REM sleep, in PTSD patients and trauma-exposed controls. In addition, a full polysomnographical evaluation was performed, including sleep staging, assessment of respiratory function, limb movements and heart rate.

Results: A large power shift was observed in NREM sleep of PTSD patients relative to controls, with a reduction in slow oscillation power and increased activity in higher frequency bands (delta to gamma). REM sleep showed a substantial power shift in the opposite direction. There was a distinct spatial pattern, with NREM sleep abnormalities being most prominent in the right frontal region and REM sleeprelated changes most eminent occipitally. These pronounced power spectral changes occurred in the context of severe subjective sleep problems, as well as changes in sleep macrostructure, including reduced sleep efficiency, increased awakening, trends towards increased N1 and decreased SWS, and increased REM latency. Furthermore, patients showed increased leg movements.

Conclusion: Large changes were found in EEG spectral topography during both major sleep states in PTSD. Less pronounced changes were shown in sleep macrostructure. Spectral analysis thus appears much more powerful in revealing PTSD-related sleep abnormalities than traditional sleep staging. The observed abnormalities provide a candidate biomarker for sleep disturbances in PTSD.