0932 Comparison of Language and Somatic Experiences Between Reports of Trauma & Trauma-Related Dreams

Kimberly Hickey,Augustus Kram Mendelsohn,Edward Pace-Schott
DOI: https://doi.org/10.1093/sleep/zsae067.0932
IF: 6.313
2024-04-20
SLEEP
Abstract:Abstract Introduction Trauma-related nightmares (TRNs) are common sequela of psychological trauma. We explored linguistic and somatic-experience differences between self-reports of a trauma and those of nightmares related to the trauma. Methods Seventeen participants with varying severity of PTSD symptoms reporting recurring TRNs (mean age 27, 14 females) recalled traumatic experiences and nightmares related to that trauma. Trauma reports were written by participants, while nightmare reports were transcribed from audio recordings made as soon as the nightmares were recalled. Following reports, participants indicated co-occurring somatic experiences by choosing from a predetermined list. Choices were later grouped into cardiovascular, respiratory, interoceptive, and tension categories. Linguistic content was measured using the Linguistic Inquiry and Word Count (LIWC) program and words were grouped into positive emotion (positive tone, prosocial, power, affiliation), negative emotion (negative tone, anxiety, anger, sadness, conflict), and somatosensory (visual, auditory, feeling) categories. Shapiro-Wilk tests showed non-normally distributed data therefore nonparametric tests (Wilcoxon Signed-Rank Test, Spearman Correlation) were used. Because trauma reports had significantly higher word counts than TRNs (p=0.0495), LIWC data were normalized for word count. Total and symptom-cluster severities were assessed using the PTSD check list for DSM-5 (PCL-5). Results We found significantly more somatic experiences of interoception (p=0.0084) and tension (p=0.024) in trauma versus nightmare reports but no differences in cardiovascular or respiratory experiences. No differences between report type were found for LIWC categories. The PCL-5 Intrusion (cluster B) PTSD symptoms were associated with cardiovascular (rho=0.592, p=0.0156) and respiratory (rho=0.619, p=0.0109) experiences in trauma reports and interoception (rho=0.718, p=0.0033) and tension (rho=0.556, p=0.0224) experiences in nightmare reports. No other symptom clusters, nor PCL-5 total score, were correlated with other somatic-experience categories. No LIWC categories in either report type were correlated with symptom clusters or total PCL-5 score. Conclusion More somatic experiences of interoception and tension were recalled from traumas than TRNs. Because the brain is relatively deafferented from sensory input during dreaming, we expected, but did not find, state differences in other somatic experiences. Word categories in narratives also did not show state differences. Only the intrusion symptoms of PTSD predicted bodily sensations in trauma as well as TRN reports. Support (if any) R21MH128619
neurosciences,clinical neurology
What problem does this paper attempt to address?